TechRefManualMar2000 HEC HMS

US Army Corps
of Engineers
Hydrologic Engineering Center
Hydrologic Modeling System

HEC-HMS
Technical Reference Manual
March 2000
Approved for Public Release Distribution Unlimited CPD-74B

Form Approved
REPORT DOCUMENTATION PAGE OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington,
VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188), Washington DC 20503.
1. AGENCY USE ONLY (Leave Blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED
March 2000 Computer Software Technical Reference Manual
4. TITLE AND SUBTITLE 5. FUNDING NUMBERS
Hydrologic Modeling System HEC-HMS U.S. Army Corps of Engineers R&D
Technical Reference Manual Work Unit #32444
6. AUTHOR(S)
Arlen D. Feldman, Editor
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION

U.S. Army Corps of Engineers REPORT NUMBER
Hydrologic Engineering Center, HEC CPD-74B

609 Second St.
Davis, CA 95616-4687
9. SPONSORING / MONITOGING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING / MONITORING
HQ U.S. Army Corps of Engineers AGENCY REPORT NUMBER
441 G St., NW
Washington, DC 20314-1000
11. SUPPLEMENTARY NOTES
12A. DISTRIBUTION / AVAILABILITY STATEMENT 12B. DISTRIBUTION CODE

Distribution is unlimited.
13. ABSTRACT (Maximum 200 words)

The Hydrologic Modeling System (HEC-HMS) is designed to simulate the precipitation-runoff processes of dendritic
watershed systems. It supercedes HEC-1 and provides a similar variety of options but represents a significant advancement in
terms of both computer science and hydrologic engineering. The program includes a variety of mathematical models for
simulating precipitation, evapotranspiration, infiltration, excess precipitation transformation, baseflow, and open channel
routing. A versatile parameter estimation option is also included to assist in calibrating the various models.
This technical reference manual describes the mathematical models that are included as part of the computer program,
including equation derivations and solution algorithms where appropriate. In addition, the manual provides information and
guidance regarding how and when to use the various models. Guidance for estimating a model's parameters with measured
field data, empirical relationships, and the parameter estimation option is included.
14. SUBJECT TERMS 15. NUMBER OF PAGES

Hydrology, watershed, precipitation runoff, river routing, flood control, water supply, computer 145
simulation. 16. PRICE CODE
17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF
OF REPORT OF THIS PAGE OF ABSTRACT ABSTRACT
Unclassified Unclassified Unclassified Unlimited
NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89) USAPPC V1.00
Prescribed by ANSI Std. Z39-18 298-102
Hydrologic Modeling System
HEC-HMS
Technical Reference Manual
March 2000
US Army Corps of Engineers

Institute for Water Resources
Hydrologic Engineering Center
609 Second Street
Davis, CA 95616-4687 USA
Phone (530) 756-1104

Fax (530) 756-8250
Email hms@usace.army.mil
Hydrologic Modeling System HEC-HMS, User's Manual
Public Domain 2000. This Hydrologic Engineering Center, HEC, documentation and software was
developed at U.S. Federal Government expense and is therefore in the public domain. Components
of the HEC software were developed using proprietary commercial productivity software. HEC has
executed agreements with the commercial software providers that enable the HEC software to be
copied and distributed without restriction. The agreements require that the components of the
commercial software used in the HEC applications software be only the runtime form bound in the
executable. The commercial portions of the HEC software may not be used for any other
applications. Further, the agreements require that users accept the provisions of the copyright
restrictions of the commercial software components as noted in the following paragraphs.
Copyright 1995 Galaxy Runtime Components by Visix Software, Inc. All Rights Reserved.
Copyright 1995 AgX Toolmaster Runtime Components by Advanced Visual Systems, Inc. All Rights
Reserved.
BY USING OR COPYING THIS SOFTWARE, USER AGREES TO ABIDE BY THE COPYRIGHT

LAW AND ALL OTHER APPLICABLE LAWS OF THE U.S., INCLUDING, BUT NOT LIMITED TO,
EXPORT CONTROL LAWS, AND THE TERMS OF THE FOLLOWING LICENSE AGREEMENT.
THE HYDROLOGIC ENGINEERING CENTER SHALL HAVE THE RIGHT TO TERMINATE THIS
LICENSE IMMEDIATELY BY WRITTEN NOTICE UPON USER'S BREACH OF, OR
NONCOMPLIANCE WITH, ANY OF ITS TERMS. USER MAY BE HELD LEGALLY RESPONSIBLE
FOR ANY COPYRIGHT INFRINGEMENT THAT IS CAUSED OR ENCOURAGED BY USER'S
FAILURE TO ABIDE BY THE TERMS OF THIS LICENSE.
The above information and the full License Agreement are presented for acceptance when the
software is first used. That information is also provided in Appendix F of this manual for reference.
ii
Table of Contents
INTRODUCTION ............................................................................................................... 1
WHAT'S IN THIS MANUAL? ............................................................................................................... 1
PROGRAM OVERVIEW ..................................................................................................................... 1
OTHER PROGRAM REFERENCES...................................................................................................... 2
ORGANIZATION OF THIS MANUAL ..................................................................................................... 2
REFERENCES ................................................................................................................................. 3
PRIMER ON MODELS....................................................................................................... 4
WHAT IS A MODEL? ........................................................................................................................ 4
MODEL CLASSIFICATION.................................................................................................................. 4
CONSTITUENTS OF A MODEL ........................................................................................................... 6
State Variables......................................................................................................................... 7
Parameters .............................................................................................................................. 7
Boundary Conditions................................................................................................................ 7
Initial Conditions....................................................................................................................... 7
Solving the Constituents .......................................................................................................... 7
MODELS AND COMPUTER PROGRAMS .............................................................................................. 8
Model ....................................................................................................................................... 8
Program ................................................................................................................................... 8
Input ......................................................................................................................................... 8
REFERENCES ................................................................................................................................. 9
PROGRAM COMPONENTS ........................................................................................... 10
RUNOFF PROCESSES.................................................................................................................... 10
REPRESENTATION OF THE RUNOFF PROCESS ................................................................................ 11
SYNOPSIS OF INCLUDED MODELS .................................................................................................. 12
PROGRAM SETUP AND APPLICATION .............................................................................................. 14
REFERENCES ............................................................................................................................... 15
DESCRIBING PRECIPITATION ...................................................................................... 16
FIELD-MONITORED HISTORICAL PRECIPITATION ............................................................................. 16
Precipitation Measurement .................................................................................................... 16
Runoff-Computation Requirements ....................................................................................... 17
Mean-Areal Precipitation Depth Computation .......................................................................18
Temporal Distribution of Precipitation .................................................................................... 19
Inverse-Distance-Squared Method ........................................................................................ 20
RADAR "OBSERVATIONS" OF HISTORICAL PRECIPITATION ............................................................... 21
Radar Data ............................................................................................................................. 23
Computations with Radar-Measured Precipitation ................................................................24
HYPOTHETICAL STORMS ............................................................................................................... 24
Standards-Based Design Concepts ....................................................................................... 24
Frequency-Based Hypothetical Storm ................................................................................... 25
The Standard Project Storm .................................................................................................. 28
User-Defined Hypothetical-Storm Distribution .......................................................................30
Storm Selection ...................................................................................................................... 31
Risk-Based Design Concepts ................................................................................................ 32
EVAPORATION AND TRANSPIRATION............................................................................................... 32
REFERENCES ............................................................................................................................... 32
COMPUTING RUNOFF VOLUMES............................................................................... 35
BASIC CONCEPTS ......................................................................................................................... 35
INITIAL AND CONSTANT LOSS MODEL ............................................................................................. 36
Basic Concepts and Equations .............................................................................................. 36
Estimating Initial Loss and Constant Rate .............................................................................36
iii
Table of Contents
DEFICIT AND CONSTANT LOSS MODEL........................................................................................... 37

SCS CURVE NUMBER LOSS MODEL .............................................................................................. 37
Basic Concepts and Equations ............................................................................................. 37
Estimating CN........................................................................................................................ 38
Gridded SCS Curve Number Loss Model ............................................................................. 38
GREEN AND AMPT LOSS MODEL ................................................................................................... 39
Estimating Model Parameters ............................................................................................... 39
SOIL MOISTURE ACCOUNTING LOSS MODEL .................................................................................. 40
Storage Component .............................................................................................................. 41
Flow Component ................................................................................................................... 43
Order of Model Computations ............................................................................................... 46
Estimating Model Parameters ............................................................................................... 46
APPLICABILITY AND LIMITATIONS OF THE RUNOFF-VOLUME MODELS ............................................... 46
REFERENCES ............................................................................................................................... 48
MODELING DIRECT RUNOFF ...................................................................................... 50
BASIC CONCEPTS OF THE UNIT HYDROGRAPH MODEL ................................................................... 50
USER-SPECIFIED UNIT HYDROGRAPH ........................................................................................... 51
Estimating the Model Parameters ......................................................................................... 51
Application of the User-Specified UH .................................................................................... 52
PARAMETRIC AND SYNTHETIC UNIT HYDROGRAPHS ....................................................................... 52
What is a Parametric UH? ..................................................................................................... 52
What is a Synthetic UH?........................................................................................................ 52
SNYDER UNIT HYDROGRAPH MODEL ............................................................................................. 53
SCS UNIT HYDROGRAPH MODEL .................................................................................................. 55
CLARK UNIT HYDROGRAPH MODEL ............................................................................................... 57
MODCLARK MODEL ...................................................................................................................... 59
Setting Up and Using the ModClark Model ........................................................................... 60
KINEMATIC WAVE MODEL ............................................................................................................. 60
Setting Up and Using the Kinematic Wave Model ................................................................. 66
APPLICABILITY AND LIMITATIONS OF DIRECT RUNOFF MODELS ....................................................... 67
REFERENCES ............................................................................................................................... 68
MODELING BASEFLOW ............................................................................................... 70
BASIC CONCEPTS AND IMPLEMENTATION ....................................................................................... 70
Constant , Monthly-Varying Baseflow ................................................................................... 70
Exponential Recession Model ............................................................................................... 70
Linear Reservoir Model ......................................................................................................... 72
ESTIMATING BASEFLOW MODEL PARAMETERS............................................................................... 72
Constant , Monthly-Varying Baseflow ................................................................................... 72
Exponential Recession Model ............................................................................................... 72
Linear Reservoir Model ......................................................................................................... 73
REFERENCES ............................................................................................................................... 73
MODELING CHANNEL FLOW ..................................................................................... 74
OPEN-CHANNEL FLOW EQUATIONS AND SOLUTION TECHNIQUES.................................................... 74
iv
Table of Contents
Basic Equations of Open-Channel Flow ................................................................................ 74

Approximations ...................................................................................................................... 75
Solution Methods ................................................................................................................... 76
Parameters, Initial Conditions, and Boundary Conditions .....................................................76
MODIFIED PULS MODEL ................................................................................................................ 77
Defining the Storage-Outflow Relationship ............................................................................78
Estimating Other Model Parameters ...................................................................................... 79
MUSKINGUM MODEL ..................................................................................................................... 80
Estimating the Model Parameters .......................................................................................... 81
LAG MODEL .................................................................................................................................. 83
Basic Concept ........................................................................................................................ 83
Estimating the Lag ................................................................................................................. 84
KINEMATIC WAVE MODEL.............................................................................................................. 84
Setting Up the Model ............................................................................................................. 84
MUSKINGUM-CUNGE MODEL ......................................................................................................... 85
Setting Up the Model and Estimating Parameters .................................................................87
APPLICABILITY AND LIMITATIONS OF ROUTING MODELS................................................................... 88
MODELING CONFLUENCES AS JUNCTIONS ...................................................................................... 90
Setting Up a Confluence Model ............................................................................................. 91
Limitations of the Confluence Model ...................................................................................... 91
MODELING BIFURCATIONS AS DIVERSIONS ..................................................................................... 91
Setting Up a Bifurcation Model .............................................................................................. 92
Limitations of the Bifurcation Model ....................................................................................... 92
REFERENCES ............................................................................................................................... 92
CALIBRATING THE MODELS........................................................................................ 94
WHAT IS CALIBRATION? ................................................................................................................ 94
SUMMARY OF THE CALIBRATION PROCEDURE ................................................................................ 94
GOODNESS-OF-FIT INDICES .......................................................................................................... 97
SEARCH METHODS ..................................................................................................................... 100
Univariate-Gradient Algorithm ............................................................................................. 100
Nelder and Mead Algorithm ................................................................................................. 102
CONSTRAINTS ON THE SEARCH ................................................................................................... 104
REFERENCES ............................................................................................................................. 106
MODELING WATER-CONTROL FACILITIES.............................................................107
DIVERSION MODELING ................................................................................................................ 107
Basic Concepts and Equations ............................................................................................ 107
Setting Up a Diversion Model .............................................................................................. 108
Return Flow from a Diversion .............................................................................................. 108
Applicability and Limitations of the Diversion Model ............................................................109
RESERVOIR AND DETENTION MODELING ...................................................................................... 109
Basic Concepts and Equations ............................................................................................ 110
Setting Up a Reservoir Model .............................................................................................. 111
Applicability and Limitations of the Detention Model ...........................................................113
REFERENCES ............................................................................................................................. 113
v
Table of Contents
CN TABLES .................................................................................................................... 114
SOIL MOISTURE ACCOUNTING MODEL DETAILS ................................................. 119

TIME INTERVAL SELECTION ......................................................................................................... 119
GLOSSARY ..................................................................................................................... 121
TERM DEFINITIONS ..................................................................................................................... 121
vi
Preface
PREFACE
The HEC-HMS program was developed at the Hydrologic Engineering Center (HEC)
of the US Army Corps of Engineers. HEC-HMS is a component of the HEC Next
Generation Software Development Project. This project is under the guidance of
Darryl Davis, Director, HEC. Arlen Feldman manages the HEC-HMS project.
The program was developed by a team of HEC staff and consultants. Elisabeth Pray,
HEC, developed the majority of the graphical user interface and integrated the
various components to produce the finished program. Paul Ely, contractor, developed
the computation engine and hydrologic algorithm library. William Scharffenberg,
HEC, contributed to graphical user interface design, managed testing, and wrote the
program users manual. Thomas Evans, HEC, developed the algorithms for storing
gridded data. Richard Raichle, contractor, developed the soil moisture accounting
graphical user interface. Shannon Newbold, contractor, developed the meteorological
model graphical user interface. Todd Bennett, HEC, provided technical evaluations
that led to the design of the soil moisture accounting loss method. Jessica Thomas,
HEC, conducted testing and prepared validation documents.
The program continues to benefit from many individuals who contributed to previous
versions. John Peters, HEC, managed the development team until his retirement in
1998. During that time he designed the graphical user interface, managed
development, and wrote the first version of the user's manual. Arthur Pabst, HEC,
and Tony Slocum, consultant, provided essential input to the object-oriented design
of the program. Slocum also wrote the code for the schematic representation of the
basin model. William Charley, HEC, developed the design for the computation
engine. David Ford Consulting Engineers provided recommendations for the scope
and content of the optimization manager. Troy Nicolini, HEC, led the Version 1.0 beta
testing team and managed the maiden release. Several students from the University
of California, Davis, working as temporary employees at HEC, provided excellent
assistance to the software development, testing, and documentation: Ken Sheppard,
Jake Gusman, and Dan Easton.
David Ford Consulting Engineers wrote original drafts of this manual, supplementing
material provided by HEC with new, original text and figures. HEC staff reviewed and
modified the drafts to produce the final manual.
vii
Chapter 1 Introduction
CHAPTER 1
Introduction
What's in this Manual?

This document is the technical reference manual for the Hydrologic Modeling System
(HEC-HMS). The program is a product of the US Army Corps of Engineers' research
and development program, and is produced by the Hydrologic Engineering Center
(HEC). The program simulates precipitation-runoff and routing processes, both
natural and controlled. The program is the successor to and replacement for the
Flood Hydrograph Package HEC-1 (USACE, 1998) and for various specialized
versions of HEC-1. The program improves upon the capabilities of HEC-1 and
provides additional capabilities for distributed modeling and continuous simulation.
This technical reference manual describes the mathematical models that are included
as part of the program. In addition, the manual provides information and guidance
regarding how and when to use the models and how to estimate a models
parameters.
The presentation of the models is aimed at an engineer or scientist who has studied
hydrology in a university-level course. Thus, examples of common models are not
provided; such information may be found by consulting available texts and journals.
On the other hand, examples of the computations for the new or uncommon models
within the program are included.
Program Overview
For precipitation-runoff-routing simulation, the program provides the following
components:
Precipitation-specification options which can describe an observed

(historical) precipitation event, a frequency-based hypothetical
precipitation event, or a event that represents the upper limit of
precipitation possible at a given location.
Loss models which can estimate the volume of runoff, given the
precipitation and properties of the watershed.
Direct runoff models that can account for overland flow, storage and
energy losses as water runs off a watershed and into the stream
channels.
Hydrologic routing models that account for storage and energy flux as
water moves through stream channels.
Models of naturally occurring confluences and bifurcations.
Models of water-control measures, including diversions and storage

facilities.
1
These models are similar to those included in HEC-1. In addition to these, the
program includes:
A distributed runoff model for use with distributed precipitation data, such
as the data available from weather radar.
A continuous soil-moisture-accounting model used to simulate the long-

term response of a watershed to wetting and drying.
The program also includes:
An automatic calibration package that can estimate certain model

parameters and initial conditions, given observations of
hydrometeorological conditions.
Links to a database management system that permits data storage, retrieval and
connectivity with other analysis tools available from HEC and other sources.
Other Program References

Two references are available in addition to this manual:
The Hydrologic Modeling System HEC-HMS Users Manual (USACE,

1998b) describes how to use the computer program. While the user's
manual identifies the models that are included in the program, its focus is
the program's user interface. Thus, the user's manual provides a
description of how to use the interface to provide data, to specify model
parameters, to execute the program, and to review the results. It
provides examples of all of these tasks.
The HEC-HMS on-line help system is a component of the computer

program. It is essentially an electronic version of the user's manual, but
it also includes some material from this reference manual. Because it is
in electronic form, text searches for keywords, and jumping from topic to
topic using hyperlinks are possible.
The users manual and the HEC-HMS program are available on the Hydrologic
Engineering Centers web site. The address is www.hec.usace.army.mil.
Organization of this Manual

Table 1 shows how this manual is organized. Chapters 4-8 and 10 present the
equations of the models, define the terms of the equations, and explain the solution
algorithms used in the program. In addition, parameters of the models and methods
for estimating the parameter are described.
Because of the importance of model calibration, Chapter 9 describes the automated

calibration feature of the program in detail. This can be used to estimate model
parameters with measured precipitation and streamflow.
2
Table 1. Summary of contents of HEC-HMS Technical Reference Manual.

Chapter Topic Description of Contents
1 Introduction Provides an overview of HEC-HMS and
the technical reference manual
2 Primer on precipitation- Defines terms used throughout the
runoff-routing simulation manual and describes basic concepts
models and components of models
3 HEC-HMS components Describes how HEC-HMS represents
the runoff process and identifies the
models that are included in the program
4 Describing precipitation for Identifies each type of precipitation
modeling with HEC-HMS event that may be analyzed with HEC-
HMS, describes the format of the data
for each, and presents the precipitation
processing algorithms
5 Computing runoff volumes Summarizes the models that are
with HEC-HMS included for estimating runoff volume,
given precipitation
6 Modeling direct runoff with Summarizes the models available in
HEC-HMS HEC-HMS for computing runoff
hydrographs, given runoff volume
7 Modeling baseflow with Describes the HEC-HMS model of sub-
HEC-HMS surface flow
8 Modeling channel flow with Describes the alternative models of
HEC-HMS open channel flow that are available
and provides guidance for usage
9 Calibrating the HEC-HMS Describes how HEC-HMS may be
models calibrated with historical
precipitation and runoff data
10 Modeling water-control Describes the HEC-HMS models of
measures diversion and detention
Appx A CN tables Tables of parameters for SCS loss
model
Appx B SMA model details More information about the HEC-HMS
soil-moisture accounting model
Appx C Glossary Briefly defines important terms
References
US Army Corps of Engineers, USACE (1998). HEC-1 flood hydrograph package
users manual. Hydrologic Engineering Center, Davis, CA.
USACE (2000). HEC-HMS hydrologic modeling system users manual. Hydrologic

Engineering Center, Davis, CA.
3
Chapter 2 Primer on Models
CHAPTER 2
Primer on Models
This chapter explains basic concepts of modeling and the most important properties
of models. It also defines essential terms used throughout this technical reference
manual.
What is a Model?
Hydrologic engineers are called upon to provide information for activities for a variety
of water resource studies:
Planning and designing new hydraulic-conveyance and water-control

facilities.
Operating and/or evaluating existing hydraulic-conveyance and water-

control facilities.
Preparing for and responding to floods.
Regulating floodplain activities.
In rare cases, the record of historical flow, stage or precipitation satisfies the
information need. More commonly, watershed runoff must be predicted to provide
the information. For example, a flood-damage reduction study may require an
estimate of the increased volume of runoff for proposed changes to land use in a
watershed. However, no record will be available to provide this information because
the change has not yet taken place. Similarly, a forecast of reservoir inflow may be
needed to determine releases if a tropical storm alters its course and moves over a
watershed. Waiting to observe the flow is not acceptable. The alternative is to use a
model to provide the information.
A model relates something unknown (the output) to something known (the input). In
the case of the models that are included in the program, the known input is
precipitation and the unknown output is runoff, or the known input is upstream flow
and the unknown output is downstream flow.
Model Classification
Models take a variety of forms. Physical models are reduced-dimension
representations of real world systems. A physical model of a watershed, such as the
model constructed in the lab at Colorado State University, is a large surface with
overhead sprinkling devices that simulate the precipitation input. The surface can be
altered to simulate various land uses, soil types, surface slopes, and so on; and the
rainfall rate can be controlled. The runoff can be measured, as the system is closed.
A more common application of a physical model is simulation of open channel flow.
The Corps of Engineers Waterways Experiment Station has constructed many such
models and used these to provide information for answering questions about flow in
complex hydraulic systems.
4
Researchers also have developed analog models that represent the flow of water
with the flow of electricity in a circuit. With those models, the input is controlled by
adjusting the amperage, and the output is measured with a voltmeter. Historically,
analog models have been used to calculate subsurface flow.
The program includes models in a third categorymathematical models. In this

manual, that term defines an equation or a set of equations that represents the
response of a hydrologic system component to a change in hydrometeorological
conditions. Table 2 shows some other definitions of mathematical models; each of
these applies to the models included in the program.
Table 2. What is a mathematical model?

a quantitative expression of a process or phenomenon one is observing, analyzing, or
predicting (Overton and Meadows, 1976)
simplified systems that are used to represent real-life systems and may be substitutes of
the real systems for certain purposes. The models express formalized concepts of the real
systems (Diskin, 1970)
a symbolic, usually mathematical representation of an idealized situation that has the
important structural properties of the real system. A theoretical model includes a set of
general laws or theoretical principles and a set of statements of empirical circumstances.
An empirical model omits the general laws and is in reality a representation of the data
(Woolhiser and Brakensiek, 1982)
idealized representationsThey consist of mathematical relationships that state a theory
or hypothesis (Meta Systems, 1971)
Mathematical models, including those that are included in the program, can be
classified using the criteria shown in Table 3. These focus on the mechanics of the
model: how it deals with time, how it addresses randomness, and so on. While
knowledge of this classification is not necessary to use the program, it is helpful in
deciding which of the models to use for various applications. For example, if the goal
is to create a model for predicting runoff from an ungaged watershed, the fitted-
parameter models included in the program that require unavailable data are a poor
choice. For long-term runoff forecasting, use a continuous model, rather than a
single-event model; the former will account for system changes between rainfall
events, while the latter will not.
5
Table 3. Categorization of mathematical models (from Ford and Hamilton, 1996)

Category Description
Event or continuous This distinction applies primarily to models of watershed-runoff
processes. An event model simulates a single storm. The
duration of the storm may range from a few hours to a few
days. A continuous model simulates a longer period, predicting
watershed response both during and between precipitation
events. Most of the models included in HEC-HMS are event
models.
Lumped or distributed A distributed model is one in which the spatial (geographic)
variations of characteristics and processes are considered
explicitly, while in a lumped model, these spatial variations are
averaged or ignored. HEC-HMS includes primarily lumped
models. The ModClark model is an exception.
Empirical (system This distinction focuses on the knowledge base upon which the
theoretic) or conceptual mathematical models are built. A conceptual model is built upon
a base of knowledge of the pertinent physical, chemical, and
biological processes that act on the input to produce the output.
An empirical model, on the other hand, is built upon observation
of input and output, without seeking to represent explicitly the
process of conversion. HEC-HMS includes both empirical and
conceptual models. For example, Snyders unit hydrograph
(UH) model is empirical: the model is fitted with observed
precipitation and runoff. The kinematic-wave runoff model is
conceptual: it is based upon fundamental principles of shallow
free-surface flow.
Deterministic or If all input, parameters, and processes in a model are
stochastic considered free of random variation and known with certainty,
then the model is a deterministic model. If instead the model
describes the random variation and incorporates the description
in the predictions of output, the model is a stochastic model. All
models included in HEC-HMS are deterministic.
Measured-parameter or This distinction is critical in selecting models for application
fitted-parameter when observations of input and output are unavailable. A
measured-parameter model is one in which model parameters
can be determined from system properties, either by direct
measurement or by indirect methods that are based upon the
measurements. A fitted-parameter model, on the other hand,
includes parameters that cannot be measured. Instead, the
parameters must be found by fitting the model with observed
values of the input and the output. HEC-HMS includes both
measured-parameter models and fitted-parameter models. For
example, the baseflow model of Chapter 7 is empirical, so its
parameters cannot be measured. Instead, for a selected
watershed, the baseflow-model parameters are found by
calibration, as described in Chapter 9. On the other hand, the
Green and Ampt loss model of Chapter 5 has parameters that
are based upon soil characteristics that can be sampled.
Constituents of a Model
The mathematical models that are included in the program describe how a watershed
responds to precipitation falling on it or to upstream water flowing into it. While the
equations and the solution procedures vary, all the models have the various
components in common.
6
State Variables
These terms in the model's equations represent the state of the hydrologic system at
a particular time and location. For example, the deficit and constant-rate loss model
that is described in Chapter 5 tracks the mean volume of water in natural storage in
the watershed. This volume is represented by a state variable in the deficit and
constant-rate loss model's equations. Likewise, in the detention model of Chapter
10, the pond storage at any time is a state variable; the variable describes the state
of the engineered storage system.
Parameters
These are numerical measures of the properties of the real-world system. They
control the relationship of the system input to system output. An example of this is
the constant rate that is a constituent of the runoff-volume-accounting model
described in Chapter 5. This rate, a single number specified when using the model,
represents complex properties of the real-world soil system. If the number increases,
the computed runoff volume will decrease. If the number decreases, the runoff
volume will increase.
Parameters can be considered the "tuning knobs" of a model. The parameter values
are adjusted so that the model accurately predicts the physical system response. For
example, the Snyder unit hydrograph model has two parameters, the basin lag, tp,
and peaking coefficient, Cp. The values of these parameters can be adjusted to "fit"
the model to a particular physical system. Adjusting the values is referred to as
calibration. Calibration is discussed in Chapter 9.
Parameters may have obvious physical significance, or they may be purely empirical.
For example, the Muskingum-Cunge channel model includes the channel slope, a
physically significant, measurable parameter. On the other hand, the Snyder unit
hydrograph model has a peaking coefficient, Cp. This parameter has no direct
relationship to any physical property; it can only be estimated by calibration.
Boundary Conditions
These are the values of the system inputthe forces that act on the hydrologic
system and cause it to change. The most common boundary condition in the
program is precipitation; applying this boundary condition causes runoff from a
watershed. Another example is the upstream (inflow) flow hydrograph to a channel
reach; this is the boundary condition for a routing model.
Initial Conditions
All models included in the program are unsteady-flow models; that is, they describe
changes in flow over time. They do so by solving, in some form, differential
equations that describe a component of the hydrologic system. For example, the
routing models that are described in Chapter 8 solve the differential equations that
describe, in one dimension, the flow of water in an open channel.
Solving the Constituents

The solution of any differential equation is a report of how much the output changes
with respect to changes in the input, the parameters, and other critical variables in
the modeled process. For example, the solution of the routing equations will tell us
the value of Q/t, the rate of change of flow with respect to time. But in using the
models for planning, designing, operating, responding, or regulating, the flow values
7
at various times are needed, not just the rate of change. Given an initial value of
flow, Q at some time t, in addition to the rate of change, then the required values are
computed using the following equation in a recursive fashion:
Qt = Qt t + (Q / t ) (1)
In this equation, Qt -t is the initial condition; the known value with which the
computations start.
The initial conditions must be specified to use any of the models included in the
program. With the volume-computation models, the initial conditions represent the
initial state of soil moisture in the watershed. With the runoff models, the initial
conditions represent the runoff at the start of the storm being analyzed. With the
routing models, initial conditions represent the flows in the channel at the start of the
storm. Moreover, with the models of detention storage, the initial condition is the
state of storage at the beginning of the runoff event.
Models and Computer Programs

For clarity, this manual makes a distinction between a mathematical model, a
computer program and the input to a computer program. These terms are used as
follows:
Model
As noted above, the term model means the equations that represent the behavior of
hydrologic system components. For example, the combination of the continuity and
momentum equations together form a model of open-channel flow for routing.
Program
If the equations of a mathematical model are too numerous or too complex to solve
with pencil, paper, and calculator, they are translated into computer code and an
appropriate equation solver (an algorithm) is used. The result is a computer
program. Thus, HEC-HMS is a computer program that includes a variety of models.
Programs may be classified broadly as those developed for a specific set of

parameters, boundary conditions or initial conditions, and those that are data-driven.
Programs in the first category are "hard wired" to represent the system of interest.
To change parameters, boundary conditions or initial conditions, the program code
must be changed and recompiled. HEC-HMS is in the second category of
programsthose that require no such changes. Instead, these program are tailored
to the system of interest through changes to data in a database or changes to
parameters, boundary conditions, or initial conditions in the input.
Input
When the equations of a mathematical model are solved with site-specific conditions
and parameters, the model simulates the processes and predicts what will happen
within a particular watershed or hydrologic system. In this manual, this is referred to
as an application of the model. In using a program to solve the equations of the
model, input to the program is necessary. With HEC-1, the predecessor to HEC-
HMS, the input is an ASCII text file. This text file includes codes that specify which
models and parameters, initial conditions, and boundary conditions to use. With
HEC-HMS, the same or similar information is supplied by completing forms in the
8
graphical user interface. The input may also include data from an HEC-DSS
database (USACE, 1995).
References
Diskin, M.H. (1970). "Research approach to watershed modeling, definition of terms."
ARS and SCS watershed modeling workshop, Tucson, AZ.
Ford, D.T., and Hamilton, D. (1996). "Computer models for water-excess

management." Larry W. Mays ed., Water resources handbook, McGraw-Hill, NY.
Meta Systems (1971). Systems analysis in water resources planning. Water

Information Center, NY.
Overton, D.E., and Meadows, M.E. (1976). Stormwater modeling. Academic Press,
NY.
USACE (1995). HEC-DSS user's guide and utility manuals. Hydrologic Engineering
Center, Davis, CA.
Woolhiser, D.A, and Brakensiek, D.L. (1982). "Hydrologic system synthesis."

Hydrologic modeling of small watersheds, American Society of Agricultural
Engineers, St. Joseph, MO.
9
Chapter 3 Program Components
CHAPTER 3
Program Components
This chapter describes how the models included in the program conceptually
represent watershed behavior. It also identifies and categorizes these models.
Runoff Processes
Figure 1 is a systems diagram of the watershed runoff process, at a scale that is
consistent with the scale modeled well with the program. The processes illustrated
begin with precipitation. (Currently precipitation is limited to analysis of runoff from
rainfall. Subsequent versions will provide capability to analyze snowmelt also.) In
the simple conceptualization shown, the precipitation can fall on the watershed's
vegetation, land surface, and water bodies (streams and lakes).
Precipitation
evaporation
evaporation evaporation
transpiration
stemflow & Land

Vegetation Water body
throughfall surface flood
infiltration overland flow

capillary rise
Stream
Soil interflow
channel
baseflow
percolation capillary rise
recharge
Groundwater
aquifer
Watershed
discharge
Figure 1. Systems diagram of the runoff process at local scale (after Ward,
1975).
In the natural hydrologic system, much of the water that falls as precipitation returns
to the atmosphere through evaporation from vegetation, land surfaces, and water
bodies and through transpiration from vegetation. During a storm event, this
evaporation and transpiration is limited.
Some precipitation on vegetation falls through the leaves or runs down stems,
branches and trunks to the land surface, where it joins the precipitation that fell
directly onto the surface. There, the water may pond, and depending upon the soil
type, ground cover, antecedent moisture and other watershed properties, a portion
may infiltrate. This infiltrated water is stored temporarily in the upper, partially
saturated layers of soil. From there, it rises to the surface again by capillary action,
moves horizontally as interflow just beneath the surface, or it percolates vertically to
10
the groundwater aquifer beneath the watershed. The interflow eventually moves into
the stream channel. Water in the aquifer moves slowly, but eventually, some returns
to the channels as baseflow.
Water that does not pond or infiltrate moves by overland flow to a stream channel.
The stream channel is the combination point for the overland flow, the precipitation
that falls directly on water bodies in the watershed, and the interflow and baseflow.
Thus, resultant streamflow is the total watershed outflow.
Representation of the Runoff Process

The appropriate representation of the system shown in Figure 1 depends upon the
information needs of a hydrologic-engineering study. For some analyses, a detailed
accounting of the movement and storage of water through all components of the
system is required. For example, to estimate changes due to watershed land use
changes, it may be appropriated to use a long record of precipitation to construct a
corresponding long record of runoff, which can be statistically analyzed. In that case,
evapotranspiration, infiltration, percolation, and other movement and storage should
be tracked over a long period. To do so, a detailed accounting model is required.
The program includes such a model.
On the other hand, such a detailed accounting is not necessary for many of the
reasons for conducting a water resources study. For example, if the goal of a study
is to determine the area inundated by a storm of selected risk, a detailed accounting
and reporting of the amount of water stored in the upper soil layers is not needed.
Instead, the model need only compute and report the peak, or the volume, or the
hydrograph of watershed runoff. In this and similar cases, the "view" of the
hydrologic process can be somewhat simpler. Then, as illustrated in Figure 2, only
those components necessary to predict runoff are represented in detail, and the other
components are omitted or lumped. For example, in a common application, detailed
accounting of movement of water within the soil can be omitted. In this "reductionist"
mode, the program is configured to include models of infiltration from the land
surface, but it does not model storage and movement of water vertically within the
soil layer. It implicitly combines the near surface flow and overland flow and models
this as direct runoff. It does not include a detailed model of interflow or flow in the
groundwater aquifer, instead representing only the combined outflow as baseflow.
11
Precipitation
Evapo
transpiration
Land
Water body
surface
infiltration
overland flow
& interflow Stream
Soil
channel
baseflow
Groundwater
aquifer
Watershed
discharge
Figure 2. Typical representation of watershed runoff.
Synopsis of Included Models

The program uses a separate model to represent each component of the runoff
process that is illustrated in Figure 2, including:
Models that compute runoff volume.
Models of direct runoff, including overland flow and interflow.
Models of baseflow.
Models of channel flow.
The models that compute runoff volume are listed in Table 4. Refer to Table 3 for
definitions of the categorizations. These models address questions about the volume
of precipitation that falls on the watershed: How much infiltrates on pervious
surfaces? How much runs off of the impervious surfaces? When does it run off?
The models of direct runoff are listed in Table 5. These models describe what
happens as water that has not infiltrated or been stored on the watershed moves
over or just beneath the watershed surface. Table 6 lists the models of baseflow.
These simulate the slow subsurface drainage of water from the system into the
channels.
12
Table 4. Runoff-volume models.

Model Categorization
Initial and constant-rate event, lumped, empirical, fitted parameter
SCS curve number (CN) event, lumped, empirical, fitted parameter
Gridded SCS CN event, distributed, empirical, fitted parameter
Green and Ampt event, distributed, empirical, fitted parameter
Deficit and constant rate continuous, lumped, empirical, fitted parameter
Soil moisture accounting (SMA) continuous, lumped, empirical, fitted parameter
Gridded SMA continuous, distributed, empirical, fitted
parameter
Table 5. Direct-runoff models.

User-specified unit hydrograph (UH) event, lumped, empirical, fitted parameter
Clarks UH event, lumped, empirical, fitted parameter
Snyders UH event, lumped, empirical, fitted parameter
SCS UH event, lumped, empirical, fitted parameter
ModClark event, distributed, empirical, fitted parameter
Kinematic wave event, lumped, conceptual, measured parameter
User-specified unit hydrograph (UH) event, lumped, empirical, fitted parameter
Table 6. Baseflow models.

Constant monthly event, lumped, empirical, fitted parameter
Exponential recession event, lumped, empirical, fitted parameter
Linear reservoir event, lumped, empirical, fitted parameter
The choices for modeling channel flow with HEC-HMS are listed in Table 7. These
so-called routing models simulate one-dimensional open channel flow.
Table 7. Routing models.

Kinematic wave event, lumped, conceptual, measured parameter
Lag event, lumped, empirical, fitted parameter
Modified Puls event, lumped, empirical, fitted parameter
Muskingum event, lumped, empirical, fitted parameter
Muskingum-Cunge Standard Section event, lumped, quasi-conceptual, measured
parameter
Muskingum-Cunge 8-point Section event, lumped, quasi-conceptual, measured
parameter
Confluence continuous, conceptual, measured
parameter
Bifurcation continuous, conceptual, measured
13
parameter
In addition to the models of runoff and channel processes, models are included for
simulating a water control structure such as a diversion or a reservoir/detention pond.
Those models are described in Chapter 10.
Program Setup and Application

To analyze a hydrologic system, the program user must complete the following steps:
1. Start a new project.
2. Create gage data.
3. Enter basin model data.
4. Enter precipitation model data.
5. Enter control specifications.
6. Create and execute a "run" (an application of the program.)
7. View the results.
8. Exit the program.
To complete step 3, the user must select the models that will be used for the
analysis. This requires a volume model from Table 4, a direct-runoff model from
Table 5, and a baseflow model from Table 6. For routing computations, a routing
model from Table 7 must be selected. For each model, the user must specify the
initial conditions and the model parameters.
For step 4, the user must select the appropriate form of precipitation, the boundary
condition for a rainfall-runoff model. To make this selection properly, the user must
answer the question: Does historical observed rainfall provide the necessary
information, or is an event with specified frequency needed? These alternatives are
described in more detail in Chapter 4 of this document.
14
Figure 3. Example input screen for runoff-model parameters.
The data-entry steps, program execution, and result visualization are easy. The user
indicates model choices and specifies initial conditions and parameters using a
graphical user interface (GUI). With this GUI, a user can start a project; draw on the
screen a schematic of the watershed; fill in forms to specify basin-model information,
precipitation-model information, and control specifications (as illustrated in Figure 3);
run the models; and view the results. The HEC-HMS user's manual (USACE, 2000)
and the on-line help system provide additional details about this.
References
USACE (1998). HEC-HMS users manual. Hydrologic Engineering Center, Davis, CA.
Ward R.C. (1975). Principles of hydrology. McGraw-Hill Book Company (UK) Limited,
London.
15
Chapter 4 Describing Precipitation
CHAPTER 4
Describing Precipitation
In the view of watershed hydrology, as illustrated by Figure 2, the response of a
watershed is driven by precipitation that falls on the watershed and
evapotranspiration from the watershed. The precipitation may be observed rainfall
from a historical event, it may be a frequency-based hypothetical rainfall event, or it
may be an event that represents the upper limit of precipitation possible at a given
location. (In future versions of the program, the precipitation may also be snowmelt.)
Historical precipitation data are useful for calibration and verification of model
parameters, for real-time forecasting, and for evaluating the performance of proposed
designs or regulations. Data from the second and third categoriescommonly
referred to as hypothetical or design stormsare useful if performance must be
tested with events that are outside the range of observations or if the risk of flooding
must be described. Similarly, the evapotranspiration data used may be observed
values from a historical record, or they may be hypothetical values. This chapter
describes methods of specifying and analyzing historical or hypothetical-storm
precipitation and evapotranspiration.
Field-Monitored Historical Precipitation

Precipitation Measurement
Each of the precipitation measuring devices described in Table 8 captures rainfall or
snowfall in a storage container that is open to the atmosphere. The depth of the
collected water is then observed, manually or automatically, and from those
observations, the depth of precipitation at the location of the gage is obtained.
Table 8. Precipitation field monitoring options (WMO, 1994)

Option Categorization
Manual (also referred to as non- This gage is read by a human observer. An example is
recording, totalizer, or shown in Figure 4. Often such gages are read daily,
accumulator gage) so detailed information about the short-term temporal
distribution of the rainfall is not available.
Automatic hydrometeorological This type of gage observes and records precipitation
observation station automatically. An example is a weighing gage with a
strip-chart data logger. With this gage, the temporal
distribution is known, as a continuous time record is
available. In the HEC-HMS User's Manual, a gage at
which the temporal distribution is known is referred to
as a recording gage.
Telemetering This type of gage observes and transmits precipitation
hydrometeorological observation depth automatically, but does not store it locally. An
station example is an ALERT system tipping bucket raingage
with UHF radio transmitter. Telemetering gages are
typically recording gages. Figure 5 is an example of
such a gage.
Telemetering automatic This type of gage observes, records, and transmits
hydrometeorological observation automatically. It is a recording gage.
station
16
Runoff-Computation Requirements
Chapter 6 provides details of the models for computing direct runoff from
precipitation: the alternatives are various forms of the unit-hydrograph model and the
kinematic-wave model. Inherent in models of both types is an assumption that the
precipitation is distributed uniformly over the watershed for a given time duration.
This, in turn, requires specifying the properties of this uniform rainfall. In the
program, these properties include (1) the total depth of the watershed precipitation,
and (2) the temporal distribution of that precipitation.
Figure 4. Manual precipitation gage.
Figure 5. Telemetering precipitation observation gage.
17
Mean-Areal Precipitation Depth Computation

The required watershed precipitation depth can be inferred from the depths at gages
using an averaging scheme. Thus:

w p ( t )
i i
PMAP = i t
(2)
w i
i
where PMAP = total storm mean areal precipitation (MAP) depth over the watershed;
pi(t) = precipitation depth measured at time t at gage i; and wi = weighting factor
assigned to gage/observation i. If gage i is not a recording device, only the quantity
pi(t), the total storm precipitation at gage i, will be available and used in the
computation.
Common methods for determining the gage weighting factors for MAP depth
computation include:
Arithmetic mean. This method assigns a weight to each gage equal to

the reciprocal of the total number of gages used for the MAP
computation. Gages in or adjacent to the watershed can be selected.
Thiessen polygon. This is an area-based weighting scheme, based

upon an assumption that the precipitation depth at any point within a
watershed is the same as the precipitation depth at the nearest gage in
or near the watershed. Thus, it assigns a weight to each gage in
proportion to the area of the watershed that is closest to that gage.
As illustrated in Figure 6(a), the gage nearest each point in the

watershed may be found graphically by connecting the gages, and
constructing perpendicular bisecting lines; these form the boundaries of
polygons surrounding each gage. The area within each polygon is
nearest the enclosed gage, so the weight assigned to the gage is the
fraction of the total area that the polygon represents.
Details and examples of the procedure are presented in Chow,

Maidment, and Mays (1988), Linsley, Koehler, and Paulus (1982), and
most hydrology texts.
Isohyetal. This too is an area-based weighting scheme. Contour lines

of equal precipitation are estimated from the point measurements, as
illustrated by Figure 6(b). This allows a user to exercise judgment and
knowledge of a basin while constructing the contour map. MAP is
estimated by finding the average precipitation depth between each pair
of contours (rather than precipitation at individual gages), and weighting
these depths by the fraction of total area enclosed by the pair of
contours.
Again, details and examples of the procedure are presented in most

hydrology texts.
18
0 .5 0 " 1.00 "

B 0 .5 5 " 2.00 "
3 .0 0 "
B
4 .0 0 "
C 0 .5 0 " C
4.21 "
A A
0 .87 "
1 .0 0 "
4 .0 0 "
2 .0 0 " 3.00 "

E E
D
2 .3 3 " 2.00 "
D
1 .8 9 "
2.00 "
(a) (b)
Figure 6. Illustration of MAP depth computation schemes.
Temporal Distribution of Precipitation

To compute a hydrograph, which represents flow variations with time, information
about the MAP variations with time must be provided. To do so, a precipitation
pattern with ordinates, ppattern(t) is defined and the temporal distribution of the MAP
depth is computed as (from Equation 2):

p pattern (t )
p MAP (t ) = P (3)
p MAP
pattern (t )
t
in which pMAP(t) = the watershed MAP at time t. As with total storm depth, the pattern
can be inferred from gage observations with a weighting scheme:
p pattern (t ) =
w (t ) p (t )
i i
(4)
w (t ) i
in which pi(t) = precipitation measured at gage i at time t, and wi(t) = weighting factor
assigned to gage i at time t. In this computation, only recording gages are used.
If a single recording gage is used in Equation 3, the resulting MAP hyetograph will
have the same relative distribution as the observed hyetograph. For example, if the
gage recorded 10% of the total precipitation in 30 minutes, the MAP hyetograph will
have 10% of the MAP in the same 30-minute period.
On the other hand, if two or more gages are used, the pattern will be an average of
that observed at those gages. Consequently, if the temporal distribution at those
gages is significantly different, as it might be with a moving storm, the average
pattern may obscure information about the precipitation on the watershed. This is
illustrated by the temporal distributions shown in Figure 7. Here, hyetographs of
rainfall at two gages are shown. At gage A, rain fell at a uniform rate of 10 mm/hr
from 0000 hours until 0200 hours. No rain was measured at gage A after 0200. At
gage B, no rain was observed until 0200, and then rainfall at a uniform rate of 10
mm/hr was observed until 0400. The likely pattern is that the storm moved across
the watershed from gage A to gage B. If these gage data are used with Equations 4-
2 and 4-3 to compute an average pattern, weighting each gage equally, the result is
19
a uniform rate of 5 mm/hr from 0000 until 0400. This may fail to represent well the
average temporal pattern. A better scheme might be to use one of the gages as a
pattern for the watershed average.
10
Rainfall rate
at Gage A,
mm/hr
0
0000 0200 0400
10
Rainfall rate
at Gage B,
mm/hr
0
0000 0200 0400
Figure 7. Illustration of the hazard of averaging rainfall temporal distributions.
Inverse-Distance-Squared Method
As an alternative to separately defining the total MAP depth and combining this with
a pattern temporal distribution to derive the MAP hyetograph, one can select a
scheme that computes the MAP hyetograph directly. This so-called inverse-distance-
squared weighting method computes P(t), the watershed precipitation at time t, by
dynamically applying a weighting scheme to precipitation measured at watershed
precipitation gages at time t.
The scheme relies on the notion of nodes that are positioned within a watershed
such that they provide adequate spatial resolution of precipitation in the watershed.
The program computes the precipitation hyetograph for each node using gages near
that node. To select these gages, hypothetical north-south and east-west axes are
constructed through each node and the nearest gage is found in each quadrant
defined by the axes. This is illustrated in Figure 8. Weights are computed and
assigned to the gages in inverse proportion to the square of the distance from the
node to the gage. For example, in Figure 8, the weight for the gage C in the
northeastern quadrant of the grid is computed as:
1
d C2
wC = (5)
1 1 1 1
2
+ 2 + 2 + 2
dC d D d E d A
in which wC = weight assigned to gage C; dC = distance from node to gage C; dD =

distance from node to gage D in southeastern quadrant; dE = distance from node to
gage E in southwestern quadrant; and dF = distance from node to gage F in
northwestern quadrant of grid. Weights for gages D, E and A are computed similarly.
With the weights thus computed, the node hyetograph ordinate at time t is computed
as:
20
p node (t ) = w A p A (t )+ wC p C (t ) + wD p D (t ) + wE p E (t ) (6)
This computation is repeated for all times t.
Note that gage B in Figure 8 is not used in this example, as it is not nearest to the
node in the northwestern quadrant. However, for any time that the precipitation
ordinate is missing for gage A, the data from gage B will be used. In general terms,
the nearest gage in the quadrant with data (including a zero value) will be used to
compute the MAP.
No rth
C
A
W est East
E
D
S o uth
Figure 8. Illustration of inverse-distance-squared scheme.
Radar "Observations" of Historical Precipitation

Figure 9 shows a typical (but very simple) situation. Runoff is to be predicted for the
watershed shown. Rainfall depths are measured at reporting gages A and B near
the watershed.
Figure 9. MAP can be computed as a weighted-average of depths at gages A

and B.
21
From the gaged data, one might estimate MAP as a weighted average of the depths
observed. The weights assigned might depend, for example, on how far the gage is
from one or more user-specified index points in the watershed. In this example, if an
index point at the centroid of the watershed is selected, then the weights will be
approximately equal, so the MAP will equal the arithmetic average of the depths
observed at gages A and B.
The MAP estimated from the gage network in this manner is a good representation of
rainfall on a subwatershed if the raingage network is adequately dense in the vicinity
of the storm. The gages near the storm must also be in operation, and must not be
subject to inadvertent inconsistencies (Curtis and Burnash, 1996).
The National Weather Service provides guidelines on the density of a raingage

network. These suggest that the minimum number of raingages, N, for a local flood
warning network is:
N = A 0.33 (7)
in which A = area in square miles. However, even with this network of more than the
minimum number of gages, not all storms may be adequately gaged. Precipitation
gages such as those illustrated in Figure 4 and Figure 1 are typically 8-12 in (20-30
2
cm) in diameter. Thus, in a one sq-mi (2.6 km ) watershed, the catch surface of the
th
gage thus represents a sample of precipitation on approximately 1/100,000,000 of
the total watershed area. With this small sample size, isolated storms may not be
measured well if the storm cells are located over areas in which "holes" exist in the
gage network or if the precipitation is not truly uniform over the watershed.
The impact of these "holes" is illustrated by Figure 10. Figure 10(a) shows the
watershed from Figure 9, but with a storm superimposed. In this case, observations
at gages A and B would not represent well the rainfall because of the areal
distribution of the rainfall field. The "true" MAP likely would exceed the MAP
computed as an average of the observations. In that case, the runoff would be
under-predicted. Similarly, the gage observations do not represent well the true
rainfall in the case shown in Figure 10(b). There, the storm cell is over gage A, but
because of the location of the gage, it is not a good sampler of rainfall for this
watershed. Thus, in the second case the runoff might be over-predicted.
A A
B B
(a) (b)
Figure 10. Lack of coverage can complicate MAP estimation.
One potential solution to the problem of holes in the rainfall observations is to

increase the number of gages in the network. But even as the number of gages is
22
increased, one cannot be assured of measuring adequately the rainfall for all storm
events. Unless the distance between gages is less than the principal dimension of a
typical storm cell, the rainfall on a watershed may well be misestimated.
A second solution is use of rainfall depth estimates from weather radar.
Radar Data
The WMO Guide to hydrological practices (1994) explains that
Radar permits the observation of the location and movement of areas of

precipitation, and certain types of radar equipment can yield estimates of rainfall
rates over areas within range of the radar.
Weather radar data are available from National Weather Service (NWS) Weather
Surveillance Radar Doppler units (WSR-88D) throughout the US. Each of these units
provides coverage of a 230-km-radius circular area. The WSR-88D radar transmits
an S-band signal that is reflected when it encounters a raindrop or another obstacle
in the atmosphere. The power of the reflected signal, which is commonly expressed
in terms of reflectivity, is measured at the transmitter during 360 azimuthal scans,
centered at the radar unit. Over a 5- to 10-minute period, successive scans are
made with 0.5 increments in elevation. The reflectivity observations from these
scans are integrated over time and space to yield estimates of particle size and
density in an atmospheric column over a particular location. To simplify data
management, display and analysis, the NWS digitizes and reports reflectivity for cells
in a Hydrologic Rainfall Analysis Project (HRAP) grid. Cells of the grid are
approximately 4 km by 4 km.
Given the reflectivity, the rainfall rate for each of the HRAP cells can be inferred
because the power of the reflected signal is related to the size of and density of the
reflecting obstacles. The simplest model to estimate rainfall from reflectivity is a Z-R
relationship, and the most commonly-used of these is:
Z = a Rb (8)
in which Z = reflectivity factor; R = the rainfall intensity; and a and b = empirical

coefficients. Thus, as a product of the weather radar, rainfall for cells of a grid that is
centered about a radar unit can be estimated. This estimate is the MAP for that cell
and does not necessarily suggest the rain depth at any particular point in the cell.
The NWS, Department of Defense, and Department of Transportation (Federal

Aviation Administration) cooperatively operate the WSR-88D network. They collect
and disseminate the weather radar data to federal government users. The NEXRAD
Information Dissemination Service (NIDS) was established to provide access to the
weather radar data for users outside of the federal government. Each WSR-88D unit
that is designated to support the NIDS program has four ports to which selected
vendors may connect. The NIDS vendors, in turn, disseminate the data to their
clients using their own facilities, charging the clients for the products provided and for
any value added. For example, one NIDS vendor in 1998 was distributing a 1-km x
1-km mosaic of data. This mosaic is a combined image of reflectivity data from
several radar units with overlapping or contiguous scans. Combining images in this
manner increases the chance of identifying and eliminating anomalies. It also
provides a better view of storms over large basins.
Figure 11 illustrates the advantages of acquiring weather radar data. Figure 11(a)
shows the watershed from Figure 10, but with an HRAP-like grid system
23
superimposed. Data from a radar unit will provide an estimate of rainfall in each cell
of the grid. Commonly these radar-rainfall estimates are presented in graphical
format, as illustrated in Figure 11(b), with color codes for various intensity ranges.
(This is similar to the images seen on television weather reports.)
(a) (b)
Figure 11. Weather radar provides rainfall "observations" on a grid.
With estimates of rainfall in grid cells, a "big picture" of the rainfall field over a
watershed is presented. With this, better estimates of the MAP at any time are
possible due to knowledge of the extent of the storm cells, the areas of more intense
rainfall, and the areas of no rainfall. By using successive sweeps of the radar, a time
series of average rainfall depths for cells that represent each watershed can be
developed.
Computations with Radar-Measured Precipitation

From the time-series of average rainfall depths, the required MAP series can be
computed, now accounting explicitly for the spatial variability of rainfall. The MAP
computations are relatively simple: MAP for each time step is the average of the
rainfall in the set of cells that represents the watershed.
The program includes algorithms for MAP computation from radar data that are
stored in either HRAP format or in HEC's standard hydrologic grid (SHG). (The latter
is described in the March 1996 issue of HEC's Advances in Hydrologic Engineering.)
Software for reformatting radar data provided by the NWS into the format required by
the program is available from HEC.
The radar-estimated precipitation should be compared or corrected to correlate with

field observations. Radar measures only the movement of water in the atmosphere,
not the volume of water falling on the watershed. Only options shown in Table 8 can
measure this. Ideally, the average rainfall would combine radar and raingage
networks; in the US, the NWS Stage 3 reports do so.
Hypothetical Storms
Standards-Based Design Concepts
Standards-based criteria are commonly used for planning and designing new water-
control facilities, preparing for and responding to floods, and regulating floodplain
activities (WEF/ASCE, 1992). With the standards-based criteria, a threshold or
standard is set for an acceptable level of risk to the public, and actions are taken to
24
satisfy this standard. For example, levees in parts of the western US have been
designed to provide protection from flooding should a selected large historical event
re-occur.
Standards-based criteria commonly limit risk by constraining the long-term average

time between exceedances of the capacity of drainage facilities. For example, the
criteria might limit development in a floodplain so that the annual probability is no
more than 0.01 that water rises above the first floor of structures. This limit is known
as the annual exceedance probability (AEP). To meet the standard, the specified
AEP discharge and stage must be estimated. In many cases, additional information
about the volume and time of runoff may be required. For example, runoff volume
must be estimated to provide information for sizing a detention pond for flood
protection.
When sufficient streamflow data are available for the stream of interest, design
discharges for specified AEP can be estimated using statistical-analysis methods. In
the US, guidelines for conducting such statistical analyses were proposed by the
Interagency Advisory Committee on Water Data and published in Bulletin 17B (1982).
The Bulletin 17B procedure uses recorded annual maximum discharge to calibrate a
log-Pearson type III statistical model, and uses this calibrated statistical model to
predict the flows with selected AEP. Designs based upon non-exceedance of this
flow will meet the standards.
The statistical-analysis procedure of Bulletin 17B is of limited use for estimating

discharge in many cases, because:
Few streams are gaged, and those that are, usually do not have a record
long enough for the statistical model to be fitted accurately.
Land-use changes alter the response of a watershed to rainfall, so

hypothetical-flood discharges determined with data for undeveloped or
natural conditions do not reflect discharges expected with developed
conditions.
The statistical-analysis procedure does not provide information about

runoff volume and timing.
Consequently, in many cases an alternative analysis procedure is required. A

common alternative analysis procedure relies upon use of rainfall of specified AEP
(also known as a design or hypothetical storm), coupled with a mathematical model
of the processes by which rainfall is transformed to runoff. The notion is that if
median or average values of all model parameters are used, the AEP of the
discharge computed from the hypothetical storm should equal the AEP of the
precipitation (Pilgrim and Cordery, 1975).
Three alternative standards-based storms are included:
1. A balanced frequency-based storm.
2. The standard project storm (SPS).
3. A user-defined storm depth and temporal distribution.
Frequency-Based Hypothetical Storm

The objective of the frequency-based hypothetical storm that is included in the
program is to define an event for which the precipitation depths for various durations
25
within the storm have a consistent exceedance probability. Use the following steps to
develop the storm:
1. Specify the total point-precipitation depths for the selected exceedance

probability for durations from 5 minutes through the desired total duration
of the hypothetical storm (but no longer than 10 days). Depths for
durations less than the time interval selected for runoff modeling are not
necessary. For example, if the analysis requires a 24-hour storm, and
the runoff from a 0.01-AEP event is sought, the user must specify the
0.01-AEP depths for durations from 5 minutes to 24 hours.
In the US, depths for various durations for a specified exceedance

probability may be obtained by consulting locally-developed depth-
duration-frequency functions, NOAA Atlas 2 for the western US (Miller, et
al., 1973) or NWS TP-40 (Herschfield, 1961) and TP-49 (Miller, 1964) for
the eastern US. If the depths are found from isopluvial maps in one of
these sources, the values should be plotted and smoothed by prior to
input to ensure that the storm hyetograph is reasonably shaped.
2. The program applies an area correction factor to the specified depths.

Precipitation estimates from depth-duration-frequency studies, such as
those presented in NOAA Atlas 2 or TP 40, commonly are point
estimates. However, intense rainfall is unlikely to be distributed
uniformly over a large watershed. For a specified frequency and
duration, the average rainfall depth over an area is less than the depth at
a point. To account for this, the U.S. Weather Bureau (1958) derived,
from averages of annual series of point and areal values for several
dense, recording-raingage networks, factors by which point depths are to
be reduced to yield areal-average depths. The factors, expressed as a
percentage of point depth, are a function of area and duration, as shown
in Figure 12.
In accordance with the recommendation of the World Meteorological

Organization (1994), point values should be used without reduction for
areas less than 9.6 sq. mi. Furthermore, in accordance with the
recommendation of HEC (USACE, 1982), no adjustment should be made
for durations less than 30 minutes. A short duration is appropriate for a
watershed with a short time of concentration. A short time of
concentration, in turn, is indicative of a relatively small watershed, which,
in turn, requires no adjustment.
26
100
95 24-hour
90
85 6-hour
% of point 80 3-hour
rainfall
for given area 75
70
1-hour
65
60
55
30-minutes
50
0 50 100 150 200 250 300 350 400
Area (sq mi)
Figure 12. Point depth reduction factors.
3. The program interpolates to find depths for durations that are integer
multiples of the time interval selected for runoff modeling. Linear
interpolation is used, with logarithmically transformed values of depth
and duration specified in Step 1.
4. Find successive differences in the cumulative depths from Step 3, thus

computing a set of incremental precipitation depths, each of duration
equal to the selected computation interval.
5. Use the alternating block method (Chow, Maidment, Mays, 1988) to

develop a hyetograph from the incremental precipitation values (blocks).
This method positions the block of maximum incremental depth at the
middle of the required duration. The remaining blocks are arranged then
in descending order, alternately before and after the central block. ZZ is
an example of this temporal distribution; this shows the rainfall depths for
a 24-hour hypothetical storm, with a 1-hour computation interval.
27
1.4
1.2
0.8
Depth (in)
0.6
0.4
0.2
Time (hr)
Figure 13. Example of distribution of frequency-based hypothetical storm.
The Standard Project Storm

The standard project storm (SPS) is
a relationship of precipitation versus time that is intended to be reasonably

characteristic of large storms that have or could occur in the locality of concern.
It is developed by studying the major storm events in the region, excluding the
most extreme. For areas east of 105 longitude the results of SPS studies are
published in EM 1110-2-1411 as generalized regional relationships for depth,
duration and area of precipitation. For areas west of 105 longitude, special
studies are made to develop the appropriate SPS estimates. The standard
project flood (SPF) [runoff from the SPS] is used as one convenient way to
compare levels of protection between projects, calibrate watershed models, and
provide a deterministic check of statistical flood frequency estimates. (USACE,
1989)
The SPS model included in the program is the SPS applicable to basins east of 105
longitude (east of the Rocky Mountains); it is limited to areas 10 to 1,000 square
miles. The SPS is rarely used now because of the emergence of risk-based design
techniques, the inconsistency of the method between different geographic regions,
the lack of a standard SPS west of 105 longitude, and no attached probability of
occurrence. The 0.002 annual exceedance probability event has all but replaced the
SPS for design and description purposes. However, to use the SPS model, an index
precipitation, the area over which the storm occurs, and a temporal distribution are
required. The index precipitation for an area can be estimated using the map in EM
1110-2-1411. EM 1110-2-1411 proposes using a shape factor (transposition
coefficient) to adapt the ideal SPS if the watershed is not of "ideal" shape, if the storm
is not centered over the watershed, or if the storm area is larger than the watershed
area. The shape factor can be determined using procedures specified in EM 1110-2-
1411. The temporal distribution can be the standard EM 1110-2-1411 distribution
(USACE, 1952) or the Southwestern Division PMP distribution. The latter is the
distribution of 100-yr precipitation at St. Louis, MO, as proposed by the NWS
(Fredrick, et al., 1977). A more detailed description of the SPS can be found in the
HEC Training Document No. 15 (USACE, 1982).
28
Once the SPS precipitation depth is specified, the program calculates a total storm
depth distributed over a 96-hour duration using:
4
Total depth = ( R24 HR (i ) SPFE ) (9)
i =1
where SPFE = standard-project-flood index-precipitation depth in inches; and R24HR(i)

th
= percent of the index precipitation occurring during the i 24-hour period. R24HR(i) is
given by:
3.5 if i = 1
15.5
if i = 2
R24 HR (i ) = ( 10 )
182.15 14.3537 * LOG e (TRSDA + 80) if i = 3
6.0 if i = 4
where TRSDA = storm area, in square miles.
Each 24-hour period is divided into four 6-hour periods. The ratio of the 24-hour
precipitation occurring during each 6-hour period is calculated as:
R6 HR (4) 0.033 if i = 1

0.055 ( SPFE 6.0) if i = 2
0.51

R6 HR ( i ) = 13.42 ( 11 )
(SPFE + 11.0 )0.93 if i = 3

0.5 (1.0 R6 HR (3) R6 HR (2) ) + 0.0165 if i = 4
th
where R6HR(i) = ratio of 24-hour precipitation occurring during the i 6-hour period.
th
The program computes the precipitation for each time interval in the j 6-hour
th
interval of the i 24-hour period (except the peak 6-hour period) with:
t
PRCP = 0.01 * R24 HR (i ) * R6 HR ( j ) * SPFE * ( 12 )
6
where t = computation time interval, in hours.
The peak 6-hour precipitation of each day is distributed according to the percentages
in Table 9. When using a computation time interval less than one hour, the peak 1-
hour precipitation is distributed according to the percentages in Table 10. (The
selected time interval must divide evenly into one hour.) When the time interval is
larger than shown in Table 9 or Table 10, the percentage for the peak time interval is
the sum of the highest percentages. For example, for a 2-hour time interval, the
values are (14 + 12)%, (38 + 15)%, and (11 + 10)%. The interval with the largest
percentage is preceded by the second largest and followed by the third largest. The
second largest percentage is preceded by the fourth largest, the third largest
percentage is followed by the fifth largest, and so on.
29
Table 9. Distribution of maximum 6-hour SPS in percent of 6-hour amount.

Duration (hr) EM 1110-2-1411 Criteria Southwestern Division
(Standard) Criteria (SWD)
1 10 4
2 12 8
3 15 19
4 38 50
5 14 11
6 11 8
Table 10. Distribution of maximum 1-hour precipitation in the SPS.

Time (min) Percent of Maximum 1- Accumulated Precent of
hr Precipitation in Each Precipitation
Time Interval
5 3 3
10 4 7
15 5 12
20 6 18
25 9 27
30 17 44
35 25 69
40 11 80
45 8 88
50 5 93
55 4 97
60 3 100
User-Defined Hypothetical-Storm Distribution

The User-Specified Hyetograph option allows the user to define the depth and
temporal distribution of a hypothetical storm. The hypothetical rainfall values entered
are interpreted as if it were rainfall at a gage.
For example, for drainage planning in the US, Soil Conservation Service (SCS), now
known as the Natural Resources Conservation Service (NRCS), hypothetical storms
are commonly used. These storms were developed by the SCS as averages of
rainfall patterns; they are represented in a dimensionless form in TR-55 (USDA,
1986). The choice of one of the storm types shown depends upon the location of the
watershed. For example, near Davis, CA, the appropriate storm is an SCS Type I
storm.
30
Storm Selection
The following important questions will help guide the selection of a proper
hypothetical storm:
What AEP event should be used when planning to use a risk-based

event? If the goal is to define a regulatory floodplain, such as the so-
called 100-yr floodplain, select a single hypothetical storm with the
specified AEP, compute the runoff from that storm, and assign to the
flow, volume, or stage the same AEP as that assigned to the storm.
On the other hand, if the goal is to define a discharge-frequency function,

the solution is to define hypothetical storms with AEP ranging from small,
frequent events (say 0.50 AEP) to large, infrequent events (such as the
0.002-AEP event.) With these, compute the runoff and assign to the
runoff peaks, volumes or states the same AEP as the hypothetical storm.
Chapter 3 of EM 1110-2-1415 (USACE, 1993) and Chapter 17 of EM
1110-2-1417 (USACE, 1994) provide more information about this
procedure.
What duration should the event be? The included hypothetical storm
options permit defining events that last from a few minutes to several
days. The selected storm must be sufficiently long so that the entire
watershed is contributing to runoff at the concentration point. Thus, the
duration must exceed the time of concentration of the watershed; some
argue that it should be 3 or 4 times the time of concentration (Placer
County, 1990).
The National Weather Service (Fredrick et al., 1977) reports that
in the contiguous US, the most frequent duration of runoff-

producing rainfall is about 12 hrat the end of any 6-hr period within
a storm, the probability of occurrence of additional runoff-producing
rain is slightly greater than 0.5at the end of the first 6 hr, the
probability that the storm is not over is approximately 0.75. It does
not drop below 0.5 until the duration has exceeded 24 hr.
Using observed data, Levy and McCuen (1999) showed that 24 hr is a

good hypothetical-storm length for watersheds in Maryland from 2 to 50
square miles. This leads to the conclusion that a 24-hr hypothetical
storm is a reasonable choice if the storm duration exceeds the time of
concentration of the watershed. Indeed, much drainage system planning
in the US relies on use of a 24-hr event, and the SCS events are limited
to storms of 24-hr duration. However, considering the likelihood of
longer or shorter storms, this length should be used with care.
Should a frequency-based hypothetical storm temporal distribution, the

SPS distribution, or another distribution be used? The answer to this
depends upon the information needs of the study. The SPS may be
chosen to provide hydrological estimates for design of a major flood-
control structure. On the other hand, a different distribution, such as the
triangular temporal distribution, may be selected if flows for establishing
frequency functions for determining optimal detention storage are
necessary.
31
Risk-Based Design Concepts

The program includes features for specifying and computing runoff from a variety of
standards-based storms, including frequency-based hypothetical storms. However,
this does not form the basis for the Corps' flood-damage reduction projects. Instead,
as outlined in EM 1110-2-1419 and EM 1110-2-1619, these projects are designed to
provide protection from a range of events, with project features selected to maximize
contribution to national economic development (NED), consistent with environmental
and policy constraints. In this context, the frequency-based hypothetical storm
capability may be used to estimate without-project and with-project flow or stage
frequency functions, with which expected annual damage reduction may be
computed.
Evaporation and Transpiration

Chapter 3 describes how, in common application, detailed accounting of evaporation
and transpiration are omitted, as these are insignificant during a flood. In the case of
shorter storms, such as the SPS, it may be appropriate to omit this accounting.
However, with the soil-moisture accounting (SMA) model, which is described in detail
in Chapter 5, it is possible to analyze watershed response to longer precipitation
recordsrecords that include both periods of rainfall and periods without rainfall.
During periods without rainfall, the watershed moisture state continues to change, as
water moves and is stored. Evaporation and transpiration are critical components of
this movement.
Evaporation, as modeled in the program, includes vaporization of water directly from

the soil and vegetative surface, and transpiration through plant leaves. This volume
of evaporation and transpiration combined is estimated as an average volume. The
evaporation and transpiration are combined and collectively referred to as
evapotranspiration (ET) in the SMA model and in the meteorological input to the
program. In this input, monthly-varying ET values are specified, along with an ET
coefficient. The potential ET rate for all time periods within the month is computed as
the product of the monthly value and the coefficient.
Chapter 5 describes in detail how specified ET rates are used in the soil-moisture
accounting model.
References
Chow, V.T., Maidment, D.R., and Mays, L.W. (1988). Applied hydrology. McGraw-
Hill, New York, NY.
Curtis, D.C., and Burnash, R.J.C. (1996). "Inadvertent rain gauge inconsistencies
and their effects on hydrologic analysis." California-Nevada ALERT users group
conference, Ventura, CA.
Ely, P.B., and Peters, J.C. (1984). "Probable maximum flood estimation eastern
United States." Water Resources Bulletin, American Water Resources Association,
20(3).
Fredrick, R.H., Myers, V.A., and Auciello, E.P. (1977). Five- to 60-minute precipitation
frequency for the eastern and central United States, Hydro-35. National Weather
Service, Silver Spring, MD.
32
Herschfield, D.M. (1961). Rainfall frequency atlas of the United States for durations
from 30 minutes to 24 hours and return periods from 1 to 100 years, TP 40. Weather
Bureau, US Dept. of Commerce, Washington, DC.
Interagency Advisory Committee on Water Data (1982). Guidelines for determining

flood flow frequency, Bulletin 17B. USGS, Reston, VA.
Levy, B., McCuen, R. (1999). "Assessment of storm duration for hydrologic design."
ASCE Journal of Hydrologic Engineering, 4(3) 209-213.
Linsley, R.K., Kohler, M.A., and Paulhus, J.L.H. (1982). Hydrology for engineers.
McGraw-Hill, New York, NY.
Miller, J.F. (1964). Two- to ten-day precipitation for return periods of 2 to 100 years in
the contiguous United States, TP 49. Weather Bureau, US Dept. of Commerce,
Washington, DC.
Miller, J.F., Frederick, R.H., and Tracey, R.J. (1973). Precipitation-frequency atlas of
the western United States, NOAA Atlas 2. National Weather Service, Silver Spring,
MD.
Pilgrim, D.H., and Cordery, I. (1975). "Rainfall temporal patterns for design floods."
Journal of the Hydraulics Division, ASCE, 101(HY1), 81-85.
Placer County Flood Control And Water Conservation District (1990). Stormwater
management manual. Auburn, CA.
US Department of Agriculture, USDA (1986). Urban hydrology for small watersheds,

Technical Release 55. Soil Conservation Service, Washington, DC.
USACE (1952). Standard project flood determinations, EM 1110-2-1411. Office of

Chief of Engineers, Washington, DC.
USACE (1982). Hydrologic analysis of ungaged watersheds using HEC-1, Training

Document 15. Hydrologic Engineering Center, Davis, CA.
USACE (1989). Digest of water resources policies and authorities, EP 1165-2-1.

Washington, DC.
USACE (1993). Hydrologic frequency analysis, EM 1110-2-1415. Office of Chief of

Engineers, Washington, DC.
USACE (1994). Flood-runoff analysis, EM 1110-2-1417. Office of Chief of Engineers,

Washington, DC.
USACE (1995). Hydrologic engineering requirements for flood damage reduction

studies, EM 1110-2-1419. Office of Chief of Engineers, Washington, DC.
USACE (1996). Risk-based analysis for flood damage-reduction studies, EM 1110-2-

1619. Office of Chief of Engineers, Washington, DC.
U.S. Weather Bureau (1958). "Rainfall intensity-frequency regime." USWB Technical

Paper 29, Washington, DC.
Water Environment Federation/American Society of Civil Engineers (1992). Design

and construction of urban stormwater management system. New York, NY.
33
World Meteorological Organization (1994). Guide to hydrological practices: Vol. II

Analysis, forecasting and other applications, WMO-No. 168. Geneva, Switzerland.
34
Chapter 5 Computing Runoff Volumes
CHAPTER 5
Computing Runoff Volumes

As illustrated by Figure 2, HEC-HMS computes runoff volume by computing the
volume of water that is intercepted, infiltrated, stored, evaporated, or transpired and
subtracting it from the precipitation. Interception and surface storage are intended to
represent the surface storage of water by trees or grass, local depressions in the
ground surface, cracks and crevices in parking lots or roofs, or a surface area where
water is not free to move as overland flow. Infiltration represents the movement of
water to areas beneath the land surface. Interception, infiltration, storage,
evaporation, and transpiration collectively are referred to in the program and
documentation as losses. This chapter describes the loss models and how to use
them to compute runoff volumes.
Basic Concepts
The program considers that all land and water in a watershed can be categorized as
either:
Directly-connected impervious surface.
Pervious surface.
Directly-connected impervious surface in a watershed is that portion of the watershed

for which all contributing precipitation runs off, with no infiltration, evaporation, or
other volume losses. Precipitation on the pervious surfaces is subject to losses. The
following alternative models are included to account for the cumulative losses:
The initial and constant-rate loss model.
The deficit and constant-rate model.
The SCS curve number (CN) loss model.
The Green and Ampt loss model.
With each model, precipitation loss is found for each computation time interval, and is
subtracted from the MAP depth for that interval. The remaining depth is referred to
as precipitation excess. This depth is considered uniformly distributed over a
watershed area, so it represents a volume of runoff.
Chapter 6 describes the two options for direct runoff hydrograph computations: the
unit hydrograph (UH) model and the kinematic-wave model. With a UH model, the
excess on pervious portions of the watershed is added to the precipitation on directly-
connected impervious area, and the sum is used in runoff computations. With the
kinematic-wave model, directly connected impervious areas may be modeled
separately from pervious areas if two overland flow planes are defined.
35
Initial and Constant Loss Model

Basic Concepts and Equations
The underlying concept of the initial and constant-rate loss model is that the
maximum potential rate of precipitation loss, fc, is constant throughout an event.
Thus, if pt is the MAP depth during a time interval t to t+t, the excess, pet, during the
interval is given by:
p fc if pt > f c
pet = t ( 13 )
0 otherwise
An initial loss, Ia, is added to the model to represent interception and depression
storage. Interception storage is a consequence of absorption of precipitation by
surface cover, including plants in the watershed. Depression storage is a
consequence of depressions in the watershed topography; water is stored in these
and eventually infiltrates or evaporates. This loss occurs prior to the onset of runoff.
Until the accumulated precipitation on the pervious area exceeds the initial loss
volume, no runoff occurs. Thus, the excess is given by:
0 if pi < I a

pet = pt f c if pi > I a and pt > f c ( 14 )
0 if pi > I a and pt < f c

Estimating Initial Loss and Constant Rate

The initial and constant-rate model, in fact, includes one parameter (the constant
rate) and one initial condition (the initial loss). Respectively, these represent physical
properties of the watershed soils and land use and the antecedent condition.
If the watershed is in a saturated condition, Ia will approach zero. If the watershed is

dry, then Ia will increase to represent the maximum precipitation depth that can fall on
the watershed with no runoff; this will depend on the watershed terrain, land use, soil
types, and soil treatment. Table 6-1 of EM 1110-2-1417 suggests that this ranges
from 10-20% of the total rainfall for forested areas to 0.1-0.2 inches for urban areas.
The constant loss rate can be viewed as the ultimate infiltration capacity of the soils.
The SCS (1986) classified soils on the basis of this infiltration capacity, and Skaggs
and Khaleel (1982) have published estimates of infiltration rates for those soils, as
shown in Table 11. These may be used in the absence of better information.
Because the model parameter is not a measured parameter, it and the initial
condition are best determined by calibration. Chapter 9 of this manual describes the
program's calibration capability.
36
Table 11. SCS soil groups and infiltration (loss) rates (SCS, 1986; Skaggs and
Khaleel, 1982)
Soil Group Description Range of Loss
Rates (in/hr)
A Deep sand, deep loess, aggregated silts 0.30-0.45
B Shallow loess, sandy loam 0.15-0.30
C Clay loams, shallow sandy loam, soils 0.05-0.15
low in organic content, and soils usually
high in clay
D Soils that swell significantly when wet, 0.00-0.05
heavy plastic clays, and certain saline
soils
Deficit and Constant Loss Model

The program also includes a quasi-continuous variation on the initial and constant
model of precipitation losses; this is known as the deficit and constant loss model.
This model is different from the initial and constant loss model in that the initial loss
can "recover" after a prolonged period of no rainfall. [This model is similar to the loss
model included in computer program HEC-IFH (HEC, 1992).]
To use this model, the initial loss and constant rate plus the recovery rate must be
specified. The moisture deficit is tracked continuously, computed as the initial
abstraction volume less precipitation volume plus recovery volume during
precipitation-free periods. The recovery rate could be estimated as the sum of the
evaporation rate and percolation rate, or some fraction thereof.
SCS Curve Number Loss Model

The Soil Conservation Service (SCS) Curve Number (CN) model estimates
precipitation excess as a function of cumulative precipitation, soil cover, land use,
and antecedent moisture, using the following equation:
(P I a ) 2
Pe = ( 15 )
P Ia + S
where Pe = accumulated precipitation excess at time t; P = accumulated rainfall depth

at time t; Ia = the initial abstraction (initial loss); and S = potential maximum retention,
a measure of the ability of a watershed to abstract and retain storm precipitation.
Until the accumulated rainfall exceeds the initial abstraction, the precipitation excess,
and hence the runoff, will be zero.
From analysis of results from many small experimental watersheds, the SCS
developed an empirical relationship of Ia and S:
I a = 0.2 S ( 16 )
Therefore, the cumulative excess at time t is:
37
( P 0.2 S ) 2
Pe = ( 17 )
P + 0.8 S
Incremental excess for a time interval is computed as the difference between the
accumulated excess at the end of and beginning of the period.
The maximum retention, S, and watershed characteristics are related through an

intermediate parameter, the curve number (commonly abbreviated CN) as:
1000 10 CN
(foot pound system)
S= CN
25400 254 CN ( 18 )
(SI)
CN
CN values range from 100 (for water bodies) to approximately 30 for permeable soils
with high infiltration rates.
Publications from the Soil Conservation Service (1971, 1986) provide further
background and details on use of the CN model.
Estimating CN
The CN for a watershed can be estimated as a function of land use, soil type, and
antecedent watershed moisture, using tables published by the SCS. For
convenience, Appendix A of this document includes CN tables developed by the SCS
and published in Technical Report 55 (commonly referred to as TR-55). With these
tables and knowledge of the soil type and land use, the single-valued CN can be
found. For example, for a watershed that consists of a tomato field on sandy loam
near Davis, CA, the CN shown in Table 2-2b of the TR-55 tables is 78. (This is the
entry for straight row crop, good hydrologic condition, B hydrologic soil group.) This
CN is entered directly in the appropriate input form.
For a watershed that consists of several soil types and land uses, a composite CN is
calculated as:
CN composite =
A CN i i
( 19 )
A i
in which CNcomposite = the composite CN used for runoff volume computations; i = an

index of watersheds subdivisions of uniform land use and soil type; CNi = the CN for
subdivision i; and Ai = the drainage area of subdivision i.
Users of the SCS model as implemented in the program should note that the tables
in Appendix A include composite CN for urban districts, residential districts, and
newly graded areas. That is, the CN shown are composite values for directly-
connected impervious area and open space. If CN for these land uses are selected,
no further accounting of directly-connected impervious area is required.
Gridded SCS Curve Number Loss Model

Alternatively, the grid-based CN modeling option can be used. With this option, the
subdivisions in Equation 19 are grid cells. The description of each cell in the
database includes: the location of the cell, the travel distance from the watershed
outlet, the cell size, and the cell CN. The program computes precipitation excess for
38
each cell independently, using Equation 17, and routes the excess to the watershed
outlet, using the ModClark method.
Green and Ampt Loss Model

The Green and Ampt infiltration model included in the program is a conceptual model
of infiltration of precipitation in a watershed. According to EM 1110-2-1417
the transport of infiltrated rainfall through the soil profile and the infiltration
capacity of the soil is governed by Richards equation[which is] derived by
combining an unsaturated flow form of Darcy's law with the requirements of mass
conservation.
EM 1110-2-1417 describes in detail how the Green and Ampt model combines and
solves these equations. In summary, the model computes the precipitation loss on
the pervious area in a time interval as:
1 + ( i ) S f
ft = K ( 20 )
Ft
in which ft = loss during period t; K = saturated hydraulic conductivity; ( - i ) =

volume moisture deficit; Sf = wetting front suction; and Ft = cumulative loss at time t.
The precipitation excess on the pervious area is the difference in the MAP during the
period and the loss computed with Equation 20.
As implemented, the Green and Ampt model also includes an initial abstraction. This
initial condition represents surface ponding not otherwise included in the model.
Estimating Model Parameters

The Green and Ampt model in HEC-HMS requires specification of the parameters:
Initial loss. This is a function of the watershed moisture at the beginning

of the precipitation. It may be estimated in the same manner as the
initial abstraction for other loss models.
Hydraulic conductivity. Table 12 (which is derived from Table 6-2 of

EM 1110-2-1417) provides estimates of this parameter as a function of
texture class, which may be found from a soil survey. For additional
details regarding the derivation of information in this table, see Rawls, et
al. (1982).
Wetting front suction. This can be estimated as a function of pore size

distribution, which can, in turn, be correlated with texture class. Table 12
provides estimates of this.
Volume moisture deficit. This is ( - i ) in Equation 20, the soil porosity

less the initial water content. Rawls and Brakensiek (1982) and Rawls,
et al. (1982) have correlated the porosity with soil texture class; Table 12
shows this relationship. The initial water content must be between zero
and . For example, if the soil is saturated, i = ; for a completely dry
soil, i = 0. EM 1110-2-1417 suggests that the initial water content may
be related to an antecedent precipitation index.
39
Table 12. Texture class estimates (Rawls, et al., 1982)

Texture Class Porosity, Hydraulic Wetting front
3
(cm /cm) conductivity, suction (cm)
, saturated
(cm/hr)
Sand 0.437 21.00 10.6
Loamy sand 0.437 6.11 14.2
Sandy loam 0.453 2.59 22.2
Loam 0.463 1.32 31.5
Silt loam 0.501 0.68 40.4
Sandy clay loam 0.398 0.43 44.9
Clay loam 0.464 0.23 44.6
Silty clay loam 0.471 0.15 58.1
Sandy clay 0.430 0.12 63.6
Silty clay 0.479 0.09 64.7
Clay 0.475 0.06 71.4
Soil Moisture Accounting Loss Model

Models described thus far in this chapter are event models. They simulate behavior
of a hydrologic system during a precipitation event, and to do so, they require
specification of all conditions at the start of the event. The alternative is a continuous
modela model that simulates both wet and dry weather behavior. The soil moisture
accounting model (SMA) does this.

The SMA model is patterned after Leavesley's Precipitation-Runoff Modeling System
(1983) and is described in detail in Bennett (1998). The model simulates the
movement of water through and storage of water on vegetation, on the soil surface,
in the soil profile, and in groundwater layers. Given precipitation and potential
evapotranspiration (ET), the model computes basin surface runoff, groundwater flow,
losses due to ET, and deep percolation over the entire basin.
40
Precipitation Evapotranspiration
Canopy
interception
Surface
depression
Surface runoff
Infiltration
Tension Upper
zone zone Soil profile
storage storage storage
Percolation
Groundwater flow
Groundwater
layer 1 storage
Percolation
Groundwater flow
Groundwater
layer 2 storage
Deep
percolation
Figure 14. Conceptual schematic of the continuous soil moisture accounting

algorithm (Bennett, 1998)
Storage Component
The SMA model represents the watershed with a series of storage layers, as
illustrated by Figure 14. Rates of inflow to, outflow from, and capacities of the layers
control the volume of water lost or added to each of these storage components.
Current storage contents are calculated during the simulation and vary continuously
both during and between storms. The different storage layers in the SMA model are:
Canopy-interception storage. Canopy interception represents

precipitation that is captured on trees, shrubs, and grasses, and does not
reach the soil surface. Precipitation is the only inflow into this layer.
When precipitation occurs, it first fills canopy storage. Only after this
storage is filled does precipitation become available for filling other
storage volumes. Water in canopy interception storage is held until it is
removed by evaporation.
Surface-interception storage. Surface depression storage is the

volume of water held in shallow surface depressions. Inflows to this
storage come from precipitation not captured by canopy interception and
in excess of the infiltration rate. Outflows from this storage can be due to
infiltration and to ET. Any contents in surface depression storage at the
41
beginning of the time step are available for infiltration. If the water
available for infiltration exceeds the infiltration rate, surface interception
storage is filled. Once the volume of surface interception is exceeded,
this excess water contributes to surface runoff.
Soil-profile storage. The soil profile storage represents water stored in

the top layer of the soil. Inflow is infiltration from the surface. Outflows
include percolation to a groundwater layer and ET. The soil profile zone
is divided into two regions, the upper zone and the tension zone. The
upper zone is defined as the portion of the soil profile that will lose water
to ET and/or percolation. The tension zone is defined as the area that
will lose water to ET only. The upper zone represents water held in the
pores of the soil. The tension zone represents water attached to soil
particles. ET occurs from the upper zone first and tension zone last.
Furthermore, ET is reduced below the potential rate occurring from the
tension zone, as shown in Figure 15. This represents the natural
increasing resistance in removing water attached to soil particles. ET
can also be limited to the volume available in the upper zone during
specified winter months, depicting the end of transpiration by annual
plants.
1.0
0.9
0.8
0.7
0.6
Ratio of actual
ET to 0.5
potential ET 0.4
0.3
0.2
0.1
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Ratio of current storage to tension zone capacity
Figure 15. ET as a function of tension zone storage (Bennett, 1998)
Groundwater storage. Groundwater layers in the SMA represent

horizontal interflow processes. The SMA model can include either one
or two such layers. Water percolates into groundwater storage from the
soil profile. The percolation rate is a function of a user-specified
maximum percolation rate and the current storage in the layers between
which the water flows. Losses from a groundwater storage layer are due
to groundwater flow or to percolation from one layer to another.
Percolation from the soil profile enters the first layer. Stored water can
then percolate from layer 1 to groundwater layer 2 or from groundwater
layer 2 to deep percolation. In the latter case, this water is considered
lost from the system; aquifer flow is not modeled in the SMA.
42
Flow Component
The SMA model computes flow into, out of, and between the storage volumes. This
flow can take the form of:
Precipitation. Precipitation is an input to the system of storages.

Precipitation first contributes to the canopy interception storage. If the
canopy storage fills, the excess amount is then available for infiltration.
Infiltration. Infiltration is water that enters the soil profile from the ground
surface. Water available for infiltration during a time step comes from
precipitation that passes through canopy interception, plus water already
in surface storage.
The volume of infiltration during a time interval is a function of the volume

of water available for infiltration, the state (fraction of capacity) of the soil
profile, and the maximum infiltration rate specified by the model user.
For each interval in the analysis, the SMA model computes the potential
infiltration volume, PotSoilInfl, as:
CurSoilStore
PotSoilInfil = MaxSoilInfil - MaxSoilInfil ( 21 )
MaxSoilStore
where MaxSoilInfl = the maximum infiltration rate; CurSoilStore = the
volume in the soil storage at the beginning of the time step; and
MaxSoilStore = the maximum volume of the soil storage. The actual
infiltration rate, ActInfil, is the minimum of PotSoilInfil and the volume of
water available for infiltration. If the water available for infiltration
exceeds this calculated infiltration rate, the excess then contributes to
surface interception storage.
0.6
MaxSoilnfil
0.5
0.4
Potential infiltration
rate (in/hr) 0.3
0.2
0.1
MaxSoilStore
0
0 0.5 1 1.5 2
Soil profile storage volume (in)
Figure 16. Potential infiltration rate versus beginning of time step soil profile
storage.
Figure 16 illustrates the relationship of these, using an example with

MaxSoilInfil = 0.5 in/hr and MaxSoilStore = 1.5 in. As illustrated, when
the soil profile storage is empty, potential infiltration equals the maximum
infiltration rate, and when the soil profile is full, potential infiltration is
zero.
43
Percolation. Percolation is the movement of water downward from the

soil profile, through the groundwater layers, and into a deep aquifer.
In the SMA model, the rate of percolation between the soil-profile storage
and a groundwater layer or between two groundwater layers depends on
the volume in the source and receiving layers. The rate is greatest when
the source layer is nearly full and the receiving layer is nearly empty.
Conversely, when the receiving layer is nearly full and the source layer is
nearly empty, the percolation rate is less. In the SMA model, the
percolation rate from the soil profile into groundwater layer 1 is computed
as:
CurSoilStore CurGwStore
PotSoilPerc = MaxSoilPerc 1 ( 22 )
MaxSoilStore MaxGwStore
where PotSoilPerc = the potential soil percolation rate; MaxSoilPerc = a

user-specified maximum percolation rate; CurSoilStore = the calculated
soil storage at the beginning of the time step; MaxSoilStore = a user-
specified maximum storage for the soil profile; CurGwStore = the
calculated groundwater storage for the upper groundwater layer at the
beginning of the time step; and MaxGwStore = a user-specified
maximum groundwater storage for groundwater layer 1.
The potential percolation rate computed with Equation 22 is multiplied by

the time step to compute a potential percolation volume. The available
water for percolation is equal the initial soil storage plus infiltration. The
minimum of the potential volume and the available volume percolates to
groundwater layer 1.
A similar equation is used to compute PotGwPerc, the potential

percolation from groundwater layer 1 to layer 2:
CurGwStore CurGwStore
PotGwPerc = MaxPercGw 1 ( 23 )
MaxGwStore MaxGwStore
where MaxPercGw = a user-specified maximum percolation rate;

CurGwStore = the calculated groundwater storage for the groundwater
layer 2; and MaxGwStore = a user-specified maximum groundwater
storage for layer 2. The actual volume of percolation is computed as
described above.
For percolation directly from the soil profile to the deep aquifer in the
absence of groundwater layers, for percolation from layer 1 when layer 2
is not used, or percolation from layer 2, the rate depends only on the
storage volume in the source layer. In those cases, percolation rates are
computed as
CurSoilStore
PotSoilPerc = MaxSoilPerc ( 24 )
MaxSoilStore
and
CurGwStore
PotGwPerc = MaxPercGw ( 25 )
MaxGwStore
44
respectively, and actual percolation volumes are computed as described

above.
Surface runoff and groundwater flow. Surface runoff is the water that
exceeds the infiltration rate and overflows the surface storage. This
volume of water is direct runoff; the resulting runoff hydrograph is
computed with one of the models described in Chapter 6.
Groundwater flow is the sum of the volumes of groundwater flow from

each groundwater layer at the end of the time interval. The rate of flow is
computed as:
ActSoilPerc+CurGwi Store PotGwi Perc 12 GwFlowt TimeStep ( 26 )

GwFlowt + 1 =
RoutGwi Store + 12 TimeStep
where GwFlowt and GwFlowt+1 = groundwater flow rate at beginning of

the time interval t and t+1, respectively; ActSoilPerc = actual percolation
from the soil profile to the groundwater layer; PotGwiPerc = potential
percolation from groundwater layer i; RoutGwiStore = groundwater flow
routing coefficient from groundwater storage i; TimeStep = the simulation
time step; and other terms are as defined previously. The volume of
groundwater flow that the watershed releases, GwVolume, is the integral
of the rate over the model time interval. This is computed as
1
GwVolume= (GwFlowt +1 + GwFlowt ) TimeStep ( 27 )
2
This volume may be treated as inflow to a linear reservoir model to
simulate baseflow, as described in Chapter 7.
Evapotranspiration (ET). ET is the loss of water from the canopy

interception, surface depression, and soil profile storages. In the SMA
model, potential ET demand currently is computed from monthly pan
evaporation depths, multiplied by monthly-varying pan correction
coefficients, and scaled to the time interval.
The potential ET volume is satisfied first from canopy interception, then

from surface interception, and finally from the soil profile. Within the soil
profile, potential ET is first fulfilled from the upper zone, then the tension
zone. If potential ET is not completely satisfied from one storage in a
time interval, the unsatisfied potential ET volume is filled from the next
available storage.
When ET is from interception storage, surface storage, or the upper zone

of the soil profile, actual ET is equivalent to potential ET. When potential
ET is drawn from the tension zone, the actual ET is a percentage of the
potential, computed as:
ActEvapSoil = PotEvapSoil f (CurSoilStore, MaxTenStore ) ( 28 )
where ActEvapSoil = the calculated ET from soil storage; PotEvapSoil =

the calculated maximum potential ET; and MaxTenStore = the user
specified maximum storage in the tension zone of soil storage. The
function, f(), in Equation 28 is defined as follows:
45
o As long as the current storage in the soil profile exceeds the

maximum tension zone storage (CurSoilStore/MaxTenStore > 1),
water is removed from the upper zone at a one-to-one rate, the same
as losses from canopy and surface interception.
o Once the volume of water in the soil profile zone reaches the tension
zone, f() is determined similar to percolation. This represents the
decreasing rate of ET loss from the soil profile as the amount of
water in storage (and therefore the capillary force) decreases, as
illustrated in Figure 15.
Order of Model Computations

Flow into and out of storage layers is computed for each time step in the SMA model.
(Appendix B describes how the time step is selected.) The order of computations in
each time step depends upon occurrence of precipitation or ET, as follows:
If precipitation occurs during the interval, ET is not modeled.

Precipitation contributes first to canopy-interception storage.
Precipitation in excess of canopy-interception storage, combined with
water already in surface storage, is available for infiltration. If the volume
available is greater than the available soil storage, or if the calculated
potential infiltration rate is not sufficient to deplete this volume in the
determined time step, the excess goes to surface-depression storage.
When surface-depression storage is full, any excess is surface runoff.
Infiltrated water enters soil storage, with the tension zone filling first.
Water in the soil profile, but not in the tension zone, percolates to the first
groundwater layer. Groundwater flow is routed from the groundwater
layer 1, and then any remaining water may percolate to the groundwater
layer 2. Percolation from layer 2 is to a deep aquifer and is lost to the
model.
If no precipitation occurs, ET is modeled. Potential ET is satisfied first

from canopy storage, then from surface storage. Finally, if the potential
ET is still not satisfied from surface sources, water is removed from the
upper-soil profile storage. The model then continues as described above
for the precipitation periods.
Estimating Model Parameters

SMA model parameters must be determined by calibration with observed data. In
this iterative process, candidate parameter values are proposed, the model is
exercised with these parameters and precipitation and evapotranspiration inputs.
The resulting computed hydrograph is compared with an observed hydrograph for the
same period. If the match is not satisfactory, the parameters are adjusted, and the
search continues. Bennett (1998) and EM 1110-2-1417 offer guidance for this
calibration. The automatic calibration algorithm described in Chapter 9 may be used
to aid this search.
Applicability and Limitations of the Runoff-Volume Models

Selecting a loss model and estimating the model parameters are critical steps in
developing program input. Not all loss models can be used with all transforms. For
instance, the gridded loss methods can only be used with the ModClark transform.
Table 13 lists some positive and negative aspects of the alternatives. However,
46
these are only guidelines and should be supplemented by knowledge of, and
experience with, the models and the watershed. League and Freeze (1985) point out
that
In many ways, hydrologic modeling is more an art than a science, and it is likely
to remain so. Predictive hydrologic modeling is normally carried out on a given
catchment using a specific model under the supervision of an individual
hydrologist. The usefulness of the results depends in large measure on the
talents and experience of the hydrologist and understanding of the
mathematical nuances of the particular model and the hydrologic nuances of the
particular catchment. It is unlikely that the results of an objective analysis of
modeling methodscan ever be substituted for the subjective talents of an
experienced modeler.
47
Table 13. Positive and negative aspects of loss models.

Model Positive Negative
Initial and Constant "Mature" model that has been Difficult to apply to ungaged
used successfully in hundreds of areas due to lack of direct
studies throughout the US. physical relationship of
Easy to set up and use. parameters and watershed
properties.
Model is parsimonious; it
includes only a few parameters Model may be too simple to
necessary to explain the predict losses within event, even
variation of runoff volume (see if it does predict total losses
EM 1110-2-1417). well.
Deficit and Constant Similar to above Similar to above

Can be used for long-term
simulations (for example, for
period-of-record analyses.)
SCS CN Simple, predictable, and stable Predicted values not in
method accordance with classical
Relies on only one parameter, unsaturated flow theory.
which varies as a function of soil Infiltration rate will approach
group, land use and treatment, zero during a storm of long
surface condition, and duration, rather than constant
antecedent moisture condition. rate as expected.
Features readily grasped and Developed with data from small
reasonable well-documented agricultural watersheds in
environmental inputs. midwestern US, so applicability
Well established method, widely elsewhere is uncertain.
accepted for use in US and Default initial abstraction (0.2S)
abroad. does not depend upon storm
(From Ponce and Hawkins, characteristics or timing. Thus,
1996) if used with design storm,
abstraction will be same with
0.50-AEP storm and 0.01-AEP
storm.
Rainfall intensity not considered.
(Same loss for 25 mm rainfall in
1 hour or 1 day.)
Green and Ampt Parameters can be estimated for Not widely used, so less mature,
ungaged watersheds from not as much experience in
information about soils professional community.
Less parsimonious than simple
empirical models.
References
Bennett, T.H. (1998). Development and application of a continuous soil moisture
accounting algorithm for the Hydrologic Engineering Center Hydrologic Modeling
System (HEC-HMS). MS thesis, Dept. of Civil and Environmental Engineering,
University of California, Davis.
Leavesley, G. H., Lichty, R.W., Troutman, B.M., and Saindon, L.G. (1983).
Precipitation-runoff modeling system user's manual, Water-Resources Investigations
83-4238. United States Department of the Interior, Geological Survey, Denver, CO.
48
Loague, K.M., and Freeze, R.A. (1985). "A comparison of rainfall-runoff modeling
techniques on small upland catchments." Water Resources Research, AGU, 21(2),
229-248.
McFadden, Dudley E. (1994). Soil moisture accounting in continuous simulation

watershed models. Project report, Dept. of Civil and Environmental Engineering,
University of California, Davis.
Ponce, V.M., and Hawkins, R.H. (1996). "Runoff curve number: Has it reached
maturity?" Journal of Hydrologic Engineering, ASCE, 1(1), 11-19.
Rawls, W.J., and Brakensiek, D.L. (1982). "Estimating soil water retention from soil
properties." Journal of the Irrigation and Drainage Division, ASCE, 108(IR2), 166-
171.
Rawls, W.J., Brakensiek, D.L., and Saxton, K.E. (1982). "Estimation of soil water
properties." Transactions American Society of Agricultural Engineers, St. Joseph, MI,
25(5), 1316-2320.
Skaggs, R.W., and Khaleel, R. (1982). Infiltration, Hydrologic modeling of small

watersheds. American Society of Agricultural Engineers, St. Joseph, MI.
Soil Conservation Service (1971). National engineering handbook, Section 4:

Hydrology. USDA, Springfield, VA.
Soil Conservation Service (1986). Urban hydrology for small watersheds, Technical
Release 55. USDA, Springfield, VA.
USACE (1992). HEC-IFH user's manual. Hydrologic Engineering Center, Davis, CA.

Washington, DC.
49
Chapter 6 Modeling Direct Runoff
CHAPTER 6
Modeling Direct Runoff

This chapter describes the models that simulate the process of direct runoff of excess
precipitation on a watershed. This process refers to the "transformation" of
precipitation excess into point runoff. The program provides two options for these
transform methods:
Empirical models (also referred to as system theoretic models).

These are the traditional unit hydrograph (UH) models. The system
theoretic models attempt to establish a causal linkage between runoff
and excess precipitation without detailed consideration of the internal
processes. The equations and the parameters of the model have limited
physical significance. Instead, they are selected through optimization of
some goodness-of-fit criterion.
A conceptual model. The conceptual model included in the program is

a kinematic-wave model of overland flow. It represents, to the extent
possible, all physical mechanisms that govern the movement of the
excess precipitation over the watershed land surface and in small
collector channels in the watershed.
Basic Concepts of the Unit Hydrograph Model

The unit hydrograph is a well-known, commonly-used empirical model of the
relationship of direct runoff to excess precipitation. As originally proposed by
Sherman in 1932, it is "the basin outflow resulting from one unit of direct runoff
generated uniformly over the drainage area at a uniform rainfall rate during a
specified period of rainfall duration." The underlying concept of the UH is that the
runoff process is linear, so the runoff from greater or less than one unit is simply a
multiple of the unit runoff hydrograph.
To compute the direct runoff hydrograph with a UH, the program uses a discrete
representation of excess precipitation, in which a "pulse" of excess precipitation is
known for each time interval. It then solves the discrete convolution equation for a
linear system:
n M
Qn = P U
m =1
m n m +1 ( 29 )
where Qn = storm hydrograph ordinate at time nt; Pm = rainfall excess depth in time
interval mt to (m+1)t; M = total number of discrete rainfall pulses; and Un-m+1 = UH
ordinate at time (n-m+1)t. Qn and Pm are expressed as flow rate and depth
respectively, and Un-m+1 has dimensions of flow rate per unit depth. Use of this
equation requires the implicit assumptions:
1. The excess precipitation is distributed uniformly spatially and is of

constant intensity throughout a time interval t.
50
2. The ordinates of a direct-runoff hydrograph corresponding to excess

precipitation of a given duration are directly proportional to the volume of
excess. Thus, twice the excess produces a doubling of runoff
hydrograph ordinates and half the excess produces a halving. This is
the so-called assumption of linearity.
3. The direct runoff hydrograph resulting from a given increment of excess

is independent of the time of occurrence of the excess and of the
antecedent precipitation. This is the assumption of time-invariance.
4. Precipitation excesses of equal duration are assumed to produce

hydrographs with equivalent time bases regardless of the intensity of the
precipitation.
User-Specified Unit Hydrograph

A UH may be specified directly by entering all ordinates of the UH. That is, values of
Un-m+1 in Equation 29 may be specified directly and used for runoff computation.
Estimating the Model Parameters

Because it is a system theoretic model, the UH for a watershed is properly derived
from observed rainfall and runoff, using deconvolutionthe inverse of solution of the
convolution equation. To estimate a UH using this procedure:
1. Collect data for an appropriate observed storm runoff hydrograph and

the causal precipitation. This storm selected should result in
approximately one unit of excess, should be uniformly distributed over
the watershed, should be uniform in intensity throughout its entire
duration, and should be of duration sufficient to ensure that the entire
watershed is responding. This duration, T, is the duration of the UH that
will be found.
2. Estimate losses and subtract these from the precipitation. Estimate

baseflow and separate this from the runoff.
3. Calculate the total volume of direct runoff and convert this to equivalent
uniform depth over the watershed area.
4. Divide the direct runoff ordinates by the equivalent uniform depth. The
result is the UH.
Chow, Maidment, and Mays (1988) present matrix algebra, linear regression, and
linear programming alternatives to this approach.
With any of these approaches, the UH derived is appropriate only for analysis of
other storms of duration T. To apply the UH to storms of different duration, the UH
for these other durations must be derived. If the other durations are integral multiples
of T, the new UH can be computed by lagging the original UH, summing the results,
and dividing the ordinates to yield a hydrograph with volume equal one unit.
Otherwise, the S-hydrograph method can be used. This is described in detail in texts
by Chow, Maidment, and Mays (1988), Linsley, Kohler, and Paulhus (1982), Bedient
and Huber (1992), and others.
51
Application of the User-Specified UH

In practice, direct runoff computation with a specified-UH is uncommon. The data
necessary to derive the UH in the manner described herein are seldom available, so
the UH ordinates are not easily found. Worse yet, streamflow data are not available
for many watersheds of interest, so the procedure cannot be used at all. Even when
the data are available, they are available for complex storms, with significant
variations of precipitation depths within the storm. Thus, the UH-determination
procedures described are difficult to apply. Finally, to provide information for many
water resources development activities, a UH for alternative watershed land use or
channel conditions is often neededdata necessary to derive a UH for these future
conditions are never available.
Parametric and Synthetic Unit Hydrographs

What is a Parametric UH?
The alternative to specifying the entire set of UH ordinates is to use a parametric UH.
A parametric UH defines all pertinent UH properties with one or more equations,
each of which has one or more parameters. When the parameters are specified, the
equations can be solved, yielding the UH ordinates.
For example, to approximate the UH with a triangle shape, all the ordinates can be
described by specifying:
Magnitude of the UH peak.
Time of the UH peak.
The volume of the UH is knownit is one unit depth multiplied by the watershed
drainage area. This knowledge allows us, in turn, to determine the time base of the
UH. With the peak, time of peak, and time base, all the ordinates on the rising limb
and falling limb of the UH can be computed through simple linear interpolation. Other
parametric UH are more complex, but the concept is the same.
What is a Synthetic UH?

A synthetic UH relates the parameters of a parametric UH model to watershed
characteristics. By using the relationships, it is possible to develop a UH for
watersheds or conditions other than the watershed and conditions originally used as
the source of data to derive the UH. For example, a synthetic UH model may relate
the UH peak of the simple triangular UH to the drainage area of the watershed. With
the relationship, an estimate of the UH peak for any watershed can be made given
an estimate of the drainage area. If the time of UH peak and total time base of the
UH is estimated in a similar manner, the UH can be defined "synthetically" for any
watershed. That is, the UH can be defined in the absence of the precipitation and
runoff data necessary to derive the UH.
Chow, Maidment, and Mays (1988) suggest that synthetic UH fall into three
categories:
1. Those that relate UH characteristics (such as UH peak and peak time) to

watershed characteristics. The Snyder UH is such a synthetic UH.
2. Those that are based upon a dimensionless UH. The SCS UH is such a
synthetic UH.
52
3. Those that are based upon a quasi-conceptual accounting for watershed

storage. The Clark UH and the ModClark model do so.
All of these synthetic UH models are included in the program.
Snyder Unit Hydrograph Model

In 1938, Snyder published a description of a parametric UH that he had developed
for analysis of ungaged watersheds in the Appalachian Highlands in the US. More
importantly, he provided relationships for estimating the UH parameters from
watershed characteristics. The program includes an implementation of the Snyder
UH.
For his work, Snyder selected the lag, peak flow, and total time base as the critical
characteristics of a UH. He defined a standard UH as one whose rainfall duration, tr,
is related to the basin lag, tp, by:
t p = 5.5t r ( 30 )
(Here lag is the difference in the time of the UH peak and the time associated with
the centroid of the excess rainfall hyetograph, as illustrated in Figure 17.) Thus, if the
duration is specified, the lag (and hence the time of UH peak) of Snyder's standard
UH can be found. If the duration of the desired UH for the watershed of interest is
significantly different from that specified by Equation 30, the following relationship
can be used to define the relationship of UH peak time and UH duration:
tr tR
t pR = t p ( 31 )
4
in which tR = duration of desired UH; and tpR = lag of desired UH.
For the standard case, Snyder discovered that UH lag and peak per unit of excess
precipitation per unit area of the watershed were related by:
Up Cp
=C ( 32 )
A tp
where Up = peak of standard UH; A = watershed drainage area; Cp = UH peaking

coefficient; and C = conversion constant (2.75 for SI or 640 for foot-pound system).
53
tr
Discharge per
unit excess tp
precipitation Up
depth
Time
Figure 17. Snyder's unit hydrograph.
For other durations, the UH peak, QpR, is defined as:
U pR Cp
=C ( 33 )
A t pR
Snyder's UH model requires specifying the standard lag, tp, and the coefficient, Cp.
The program sets tpR of Equation 31 equal the specified time interval, and solves
Equation 31 to find the lag of the required UH. Finally, Equation 33 is solved to find
the UH peak. Snyder proposed a relationship with which the total time base of the
UH may be defined. I nstead of this relationship, the program uses the computed UH
peak and time of peak to find an equivalent UH with Clarks model (see the next
section). From that, it determines the time base and all ordinates other than the UH
peak.

Snyder collected rainfall and runoff data from gaged watersheds, derived the UH as
described earlier, parameterized these UH, and related the parameters to
measurable watershed characteristics. For the UH lag, he proposed:
t p = CC t ( LLc ) 0.3 ( 34 )
where Ct = basin coefficient; L = length of the main stream from the outlet to the
divide; Lc = length along the main stream from the outlet to a point nearest the
watershed centroid; and C = a conversion constant (0.75 for SI and 1.00 for foot-
pound system).
The parameter Ct of Equation 34 and Cp of Equation 32 are best found via

calibration, as they are not physically-based parameters. Bedient and Huber (1992)
report that Ct typically ranges from 1.8 to 2.2, although it has been found to vary from
0.4 in mountainous areas to 8.0 along the Gulf of Mexico. They report also that Cp
ranges from 0.4 to 0.8, where larger values of Cp are associated with smaller values
of Ct.
Alternative forms of the parameter predictive equations have been proposed. For
example, the Los Angeles District, USACE (1944) has proposed to estimate tp as:
54
LLc
t p = CC t ( )N ( 35 )
S
where S = overall slope of longest watercourse from point of concentration to the

boundary of drainage basin; and N = an exponent, commonly taken as 0.33.
Others have proposed estimating tp as a function of tC , the watershed time of

concentration (Cudworth, 1989; USACE, 1987). Time of concentration is the time of
flow from the most hydraulically remote point in the watershed to the watershed
outlet, and may be estimated with simple models of the hydraulic processes, as
described here in the section on the SCS UH model. Various studies estimate tp as
50-75% of tC.
SCS Unit Hydrograph Model

The Soil Conservation Service (SCS) proposed a parametric UH model; this model is
included in the program. The model is based upon averages of UH derived from
gaged rainfall and runoff for a large number of small agricultural watersheds
throughout the US. SCS Technical Report 55 (1986) and the National Engineering
Handbook (1971) describe the UH in detail.

At the heart of the SCS UH model is a dimensionless, single-peaked UH. This
dimensionless UH, which is shown in ZZ, expresses the UH discharge, Ut, as a ratio
to the UH peak discharge, Up, for any time t, a fraction of Tp, the time to UH peak.
Research by the SCS suggests that the UH peak and time of UH peak are related by:
A
UP =C ( 36 )
TP
in which A = watershed area; and C = conversion constant (2.08 in SI and 484 in

foot-pound system). The time of peak (also known as the time of rise) is related to
the duration of the unit of excess precipitation as:
t
Tp = + t lag ( 37 )
2
in which t = the excess precipitation duration (which is also the computational

interval in the run); and tlag = the basin lag, defined as the time difference between the
center of mass of rainfall excess and the peak of the UH. [Note that for adequate
definition of the ordinates on the rising limb of the SCS UH, a computational interval,
t , that is less than 29% of tlag must be used (USACE, 1998).]
When the lag time is specified, the program solves Equation 37 to find the time of UH
peak, and Equation 36 to find the UH peak. With Up and Tp known, the UH can be
found from the dimensionless form, which is built into the program, by multiplication.

The SCS UH lag can be estimated via calibration, using procedures described in
Chapter 9, for gaged headwater subwatersheds.
55
For ungaged watersheds, the SCS suggests that the UH lag time may be related to
time of concentration, tc, as:
t lag = 0.6 t c ( 38 )
Time of concentration is a quasi-physically based parameter that can be estimated

as:
t c = t sheet + t shallow + t channel ( 39 )
where tsheet = sum of travel time in sheet flow segments over the watershed land
surface; tshallow = sum of travel time in shallow flow segments, down streets, in gutters,
or in shallow rills and rivulets; and tchannel = sum of travel time in channel segments.
Identify open channels where cross section information is available. Obtain cross
sections from field surveys, maps, or aerial photographs. For these channels,
estimate velocity by Mannings equation:
CR 2 / 3 S 1 / 2
V = ( 40 )
n
where V = average velocity; R = the hydraulic radius (defined as the ratio of channel
cross-section area to wetted perimeter); S = slope of the energy grade line (often
approximated as channel bed slope); and C = conversion constant (1.00 for SI and
1.49 for foot-pound system.) Values of n, which is commonly known as Manning's
roughness coefficient, can be estimated from textbook tables, such as that in
Chaudhry (1993). Once velocity is thus estimated, channel travel time is computed
as:
L
t channel = ( 41 )
V
where L = channel length.
Sheet flow is flow over the watershed land surface, before water reaches a channel.
Distances are shorton the order of 10-100 meters (30-300 feet). The SCS
suggests that sheet-flow travel time can be estimated as:
0.007( NL) 0.8

t sheet = ( 42 )
( P2 ) 0.5 S 0.4
in which N = an overland-flow roughness coefficient; L = flow length; P2 = 2-year, 24-

hour rainfall depth, in inches; and S = slope of hydraulic grade line, which may be
approximated by the land slope. (This estimate is based upon an approximate
solution of the kinematic wave equations, which are described later in this chapter.)
Table 14 shows values of N for various surfaces.
Sheet flow usually turns to shallow concentrated flow after 100 meters. The average
velocity for shallow concentrated flow can be estimated as:
16.1345 S for unpaved surface

V = ( 43 )
20.3282 S for paved surface
56
From this, the travel time can be estimated with Equation 41.
Table 14. Overland-flow roughness coefficients for sheet-flow modeling (USACE,

1998)
Surface Description N
Smooth surfaces (concrete, asphalt, gravel, or bare soil) 0.011
Fallow (no residue) 0.05
Cultivated soils:
Residue cover 20% 0.06
Residue cover > 20% 0.17
Grass:
Short grass prairie 0.15
Dense grasses, including species such as weeping love grass, 0.24
bluegrass, buffalo grass, blue grass, and native grass mixtures
Bermudagrass 0.41
Range 0.13
1
Woods
Light underbrush 0.40
Dense underbrush 0.80
Notes:
1
When selecting N, consider cover to a height of about 0.1 ft. This is the only part of the
plant cover that will obstruct sheet flow.
Clark Unit Hydrograph Model

Clark's model derives a watershed UH by explicitly representing two critical
processes in the transformation of excess precipitation to runoff:
Translation or movement of the excess from its origin throughout the

drainage to the watershed outlet.
Attenuation or reduction of the magnitude of the discharge as the

excess is stored throughout the watershed.

Short-term storage of water throughout a watershedin the soil, on the surface, and
in the channelsplays an important role in the transformation of precipitation excess
to runoff. The linear reservoir model is a common representation of the effects of this
storage. That model begins with the continuity equation:
dS
= I t Ot ( 44 )
dt
in which dS/dt = time rate of change of water in storage at time t; It = average inflow
to storage at time t; and Ot = outflow from storage at time t.
With the linear reservoir model, storage at time t is related to outflow as:
S t = ROt ( 45 )
57
where R = a constant linear reservoir parameter. Combining and solving the

equations using a simple finite difference approximation yields:
Ot = C A I t + C B Ot 1 ( 46 )
where CA, CB = routing coefficients. The coefficients are calculated from:
t
CA = ( 47 )
R + 0.5t
CB = 1 C A ( 48 )
The average outflow during period t is:
Ot 1 + Ot
Ot = ( 49 )
2
With Clark's model, the linear reservoir represents the aggregated impacts of all
watershed storage. Thus, conceptually, the reservoir may be considered to be
located at the watershed outlet.
In addition to this lumped model of storage, the Clark model accounts for the time
required for water to move to the watershed outlet. It does that with a linear channel
model (Dooge, 1959), in which water is "routed" from remote points to the linear
reservoir at the outlet with delay (translation), but without attenuation. This delay is
represented implicitly with a so-called time-area histogram. That specifies the
watershed area contributing to flow at the outlet as a function of time. If the area is
multiplied by unit depth and divided by t, the computation time step, the result is
inflow, It, to the linear reservoir.
Solving Equation 46 and Equation 49 recursively, with the inflow thus defined, yields
values of Ot . However, if the inflow ordinates in Equation 46 are runoff from a unit of
excess, these reservoir outflow ordinates are, in fact, Ut, the UH.
[Note that as the solution of the equations is recursive, outflow will theoretically
continue for an infinite duration. The program continues computation of the UH
ordinates until the volume of the outflow exceeds 0.995 inches or mm. The UH
ordinates are then adjusted using a depth-weighted consideration to produce a UH
with a volume exactly equal to one unit of depth.]

Application of the Clark model requires:
Properties of the time-area histogram.
The storage coefficient, R.
As noted, the linear routing model properties are defined implicitly by a time-area
histogram. Studies at HEC have shown that, even though a watershed-specific
relationship can be developed, a smooth function fitted to a typical time-area
relationship represents the temporal distribution adequately for UH derivation for
most watersheds. That typical time-area relationship, which is built into the program,
is:
58
t
1.5
t
1.414 for t c
At t
= c 2
1.5 ( 50 )
A t tc
1 1.4141 t for t 2
c
where At = cumulative watershed area contributing at time t; A = total watershed

area; and tc = time of concentration of watershed. Application of this implementation
only requires the parameter tc, the time of concentration. This can be estimated via
calibration, as described in Chapter 9, or it can be estimated using the procedures
described earlier in the SCS UH section of this chapter.
The basin storage coefficient, R, is a index of the temporary storage of precipitation

excess in the watershed as it drains to the outlet point. It, too, can be estimated via
calibration if gaged precipitation and streamflow data are available. Though R has
units of time, there is only a qualitative meaning for it in the physical sense. Clark
(1945) indicated that R can be computed as the flow at the inflection point on the
falling limb of the hydrograph divided by the time derivative of flow.
ModClark Model
In Chapter 2, models are categorized as lumped-parameter models or distributed-
parameter models. A distributed parameter model is one in which spatial variability
of characteristics and processes are considered explicitly. The modified Clark
(ModClark) model is such a model (Kull and Feldman, 1998; Peters and Easton,
1996). This model accounts explicitly for variations in travel time to the watershed
outlet from all regions of a watershed.

As with the Clark UH model, runoff computations with the ModClark model explicitly
account for translation and storage. Storage is accounted for with the same linear
reservoir model incorporated in the Clark model. Translation is accounted for with a
grid-based travel-time model.
With the ModClark method, a grid is superimposed on the watershed. For each cell
of the grid representation of the watershed, the distance to the watershed outlet is
specified. Translation time to the outlet is computed as:
d cell
t cell = t c ( 51 )
d max
where tcell = time of travel for a cell, tc = time of concentration for the watershed, dcell =
travel distance from a cell to the outlet, and dmax = travel distance for the cell that is
most distant from the outlet.
The area of each cell is specified, and from this, the volume of inflow to the linear
reservoir for each time interval, t, is computed as the product of area and
precipitation excess. The excess is the difference in MAP on the cell and losses in
the cell. The inflows thus computed are routed through a linear reservoir, yielding an
outflow hydrograph for each cell. The program combines these cell outflow
hydrographs to determine the basin direct runoff hydrograph.
59
Setting Up and Using the ModClark Model

To use the ModClark model, a gridded representation of the watershed is defined.
Information about this representation is stored in a grid-parameter file; Figure 18
shows the contents of such a file. The file may be based upon an HRAP grid or
HEC's standard hydrologic grid, and it can be generated by any means. A
geographic information system (GIS) will permit automated preparation of the file;
guidance (GridParm; USACE, 1996) and software tools (HEC-GeoHMS; USACE,
1999) for this task are available from HEC.
Parameter Order: Xcoord YCoord TravelLength Area

End:
Subbasin: 85
Grid Cell: 633 359 88.38 3.76
Grid Cell: 634 359 84.51 0.18
Grid Cell: 633 358 85.55 16.13
Grid Cell: 632 358 82.55 12.76
Grid Cell: 625 348 13.75 12.07
Grid Cell: 626 348 17.12 0.09
Grid Cell: 622 347 21.19 3.26
Grid Cell: 623 347 15.56 9.96
End:
Subbasin: 86
Grid Cell: 637 361 59.13 6.79
Grid Cell: 638 361 59.04 6.95
Grid Cell: 636 361 56.68 1.17
Grid Cell: 636 360 55.08 16.38
Grid Cell: 636 347 67.96 2.45
Grid Cell: 637 347 71.72 7.41
Grid Cell: 638 347 72.57 8.78
Grid Cell: 639 347 73.32 0.04
End:
Figure 18. Contents of a grid-cell file.
Kinematic Wave Model

As an alternative to the empirical UH models, HEC-HMS includes a conceptual
model of watershed response. This model represents a watershed as an open
channel (a very wide, open channel), with inflow to the channel equal to the excess
precipitation. Then it solves the equations that simulate unsteady shallow water flow
in an open channel to compute the watershed runoff hydrograph. This model is
referred to as the kinematic-wave model. Details of the kinematic-wave model
implemented in the program are presented in HEC's Training document No. 10
(USACE, 1979).

Figure 19(a) shows a simple watershed for which runoff is to be computed for design,
planning, or regulating. For kinematic wave routing, the watershed and its channels
are conceptualized as shown in Figure 19(b). This represents the watershed as two
plane surfaces over which water runs until it reaches the channel. The water then
flows down the channel to the outlet. At a cross section, the system would resemble
an open book, with the water running parallel to the text on the page (down the
shaded planes) and then into the channel that follows the books center binding.
60
The kinematic wave overland flow model represents behavior of overland flow on the
plane surfaces. The model may also be used to simulate behavior of flow in the
watershed channels.
(a) (b)
Figure 19. Simple watershed with kinematic-wave model representation.
Overland-flow model. At the heart of the overland model are the fundamental
equations of open channel flow: the momentum equation and the continuity equation.
Flow over the plane surfaces is primarily one-dimensional flow. In one dimension,
the momentum equation is:
y V V 1 V
S f = S0 ( 52 )
x g x g t
where Sf = energy gradient (also known as the friction slope); S0 = bottom slope; V =
velocity; y = hydraulic depth; x = distance along the flow path, t = time; g =
acceleration due to gravity; (y/x) = pressure gradient; (V/g)(V/x) = convective
acceleration; and (1/g)(V/t) = local acceleration. [This equation, these terms, and
the basic concepts are described in detail in Chow (1959), Chaudhry (1993), and
many other texts.]
The energy gradient can be estimated with Manning's equation (Equation 40), which
can be written as:
CR 2 / 3 S 1f / 2
Q= A ( 53 )
N
where Q = flow, R = hydraulic radius, A = cross-sectional area, and N = a resistance
factor that depends on the cover of the planes (note that this is not Manning's n). For
shallow flow, bottom slope and the energy gradient are approximately equal and
acceleration effects are negligible, so the momentum equation simplifies to:
61
S f = So ( 54 )
Equation 53 can be simplified to:
Q = A m ( 55 )
where and m are parameters related to flow geometry and surface roughness.
The second critical equation, the one-dimensional representation of the continuity

equation, is:
V y y
A + VB + B =q ( 56 )
x x t
where B = water surface width; q = lateral inflow per unit length of channel; A(V/x)
= prism storage; VB(y/x) =wedge storage; and B(y/t) =rate of rise. [Again, the
equation, the terms, and the basic concepts are described in detail in Chow (1959),
Chaudhry (1993), and other texts.] The lateral inflow represents the precipitation
excess, computed as the difference in MAP and precipitation losses.
With simplification appropriate for shallow flow over a plane, the continuity equation
reduces to:
A Q
+ =q ( 57 )
t x
Combining Equations 56 and 57 yields
A A
+ mA(m1 ) =q ( 58 )
t x
This equation is a kinematic-wave approximation of the equations of motion. The
program represents the overland flow element as a wide rectangular channel of unit
width; =1.486S /N and m=5/3. N is not Manning's n, but rather an overland flow
1/2
roughness factor (Table 14).
Channel-flow model. For certain classes of channel flow, conditions are such that
the momentum equation can be simplified to the form shown as Equation 54. (These
cases are defined in Chapter 8.) In those cases, the kinematic-wave approximation
of Equation 58 is an appropriate model of channel flow. In the case of channel flow,
the inflow in Equation 58 may be the runoff from watershed planes or the inflow from
upstream channels.
Figure 20 shows values for and m for various channel shapes used in the program.
(The availability of a circular channel shape here does not imply that HEC-HMS can
be used for analysis of pressure flow in a pipe system; it cannot. Note also that the
circular channel shape only approximates the storage characteristics of a pipe or
culvert. Because flow depths greater than the diameter of the circular channel shape
can be computed with the kinematic-wave model, the user must verify that the results
are appropriate.)
62
Solution of equations. The kinematic-wave approximation is solved in the same

manner for either overland or channel flow:
The partial differential equation is approximated with a finite-difference

scheme.
Initial and boundary conditions are assigned.
The resulting algebraic equations are solved to find unknown hydrograph

ordinates.
The overland-flow plane initial condition sets A, the area in Equation 58, equal to
zero, with no inflow at the upstream boundary of the plane. The initial and boundary
conditions for the kinematic wave channel model are based on the upstream
hydrograph. Boundary conditions, either precipitation excess or lateral inflows, are
constant within a time step and uniformly distributed along the element.
Circular Section 1 1
0.804 2 6
= S D
n
D m = 5/4
Triangular Section 1
1
0.94 2 Z 3
= S
=
1 + Z
2
n
1 1 m = 4/3
Z Z
Square Section 1
0.72 2
= S
= n
m = 4/4
Rectangular Section 1 2
1.49 2 3
= S W
=
n
m = 5/3
Trapezoidal Section 2
1.49
1 5
1 3
=
Q= S 2 A3
2
1 1
n W + 2Y 1 + Z
Z Z
Figure 20. Kinematic wave parameters for various channel shapes (USACE,
1998)
63
In Equation 58, A is the only dependent variable, as and m are constants, so

solution requires only finding values of A at different times and locations. To do so,
the finite difference scheme approximates A/t as A/t, a difference in area in
successive times, and it approximates A/x as A/x, a difference in area at
adjacent locations, using a scheme proposed by Leclerc and Schaake (1973). The
resulting algebraic equation is:
m 1
Ai j Ai j 1 A j 1 + Ai j11 Ai j 1 Ai j11 q ij + q ij 1
+ m i = ( 59 )
t 2 x 2
Equation 59 is the so-called standard form of the finite-difference approximation. The

indices of the approximation refer to positions on a space-time grid, as shown in
Figure 21. That grid provides a convenient way to visualize the manner in which the
solution scheme solves for unknown values of A at various locations and times. The
index i indicates the current location at which A is to be found along the length, L, of
the channel or overland flow plane. The index j indicates the current time step of the
solution scheme. Indices i-1, and j-1 indicate, respectively, positions and times
removed a value x and t from the current location and time in the solution scheme.
Time
t(j+1)
j j
A i-1 Ai
t(j)
t j-1 j-1
A i-1 Ai
t(j-1)
x(i-1) x(i) x(i+1)

Distance
Figure 21. Finite difference method space-time grid.
With the solution scheme proposed, the only unknown value in Equation 59 is the
current value at a given location, Ai j . All other values of A are known from either a
solution of the equation at a previous location and time, or from an initial or boundary
condition. The program solves for the unknown as:
m 1
t A + Ai 1
[A ]
j 1 j 1
Ai = q a t + Ai
j j 1
m i
j 1
Ai j11 ( 60 )
x
i
2
The flow is computed as:
Qi j = Ai j[ ]m
( 61 )
This standard form of the finite difference equation is applied when the following
stability factor, R, is less than 1.00 (see Alley and Smith, 1987):
64
q t + A j 1 m A j 1 m ; q > 0
R= ( 62 )
q a x
i 1 i 1
a

a
or
m 1 t
R = mAi j11 qa ; qa = 0 ( 63 )
x
If R is greater than 1.00, then the following finite difference approximation is used:
Qi j Qi j1 Ai j1 Ai j11
+ = qa ( 64 )
x t
j
where Qi is the only unknown. This is referred to as the conservation form. Solving
for the unknown yields:
Qi j = Qi j1 + qx
x j
t
[
Ai 1 Ai j11 ] ( 65 )
When Qi j is found, the area is computed as
1
Q j m
Ai j = i ( 66 )

Accuracy and stability. HEC-HMS uses a finite difference scheme that ensures
accuracy and stability. Accuracy refers to the ability of the solution procedure to
reproduce the terms of the differential equation without introducing minor errors that
affect the solution. For example, if the solution approximates A/x as A/x, and a
very large x is selected, then the solution will not be accurate. Using a large x
introduces significant errors in the approximation of the partial derivative. Stability
refers to the ability of the solution scheme to control errors, particularly numerical
errors that lead to a worthless solution. For example, if by selecting a very small x,
an instability may be introduced. With small x, many computations are required to
simulate a long channel reach or overland flow plan. Each computation on a digital
computer inherently is subject to some round-off error. The round-off error
accumulates with the recursive solution scheme used by the program, so in the end,
the accumulated error may be so great that a solution is not found.
An accurate solution can be found with a stable algorithm when x/t c , where c =
average kinematic-wave speed over a distance increment x. But the kinematic-
wave speed is a function of flow depth, so it varies with time and location. The
program must select x and t to account for this. To do so, it initially selects x =
ctm where c = estimated maximum wave speed, depending on the lateral and
upstream inflows; and tm = time step equal to the minimum of:
1. One-third the plane or reach length divided by the wave speed.
2. One-sixth the upstream hydrograph rise time for a channel.
3. The specified computation interval.
65
Finally, x is chosen as: the minimum of this computed x and the reach, or plane
length divided by the number of distance steps (segments) specified in the input form
for the kinematic-wave models. The minimum default value is two segments.
When x is set, the finite difference scheme varies t when solving Equation 61 or
Equation 66 to maintain the desired relationship between x, t and c. However, the
program reports results at the specified constant time interval.
Setting Up and Using the Kinematic Wave Model

To estimate runoff with the kinematic-wave model, the watershed is described as a
set of elements that include:
Overland flow planes. Up to two planes that contribute runoff to

channels within the watershed can be described. The combined flow
from the planes is the total inflow to the watershed channels. Column 1
of Table 15 shows information that must be provided about each plane.
Subcollector channels. These are small feeder pipes or channels, with

principle dimension generally less that 18 inches, that convey water from
street surfaces, rooftops, lawns, and so on. They might service a portion
of a city block or housing tract, with area of 10 acres. Flow is assumed
to enter the channel uniformly along its length. The average contributing
area for each subcollector channel must be specified. Column 2 of Table
15 shows information that must be provided about the subcollector
channels.
Collector channels. These are channels, with principle dimension

generally 18-24 inches, which collect flows from subcollector channels
and convey it to the main channel. Collector channels might service an
entire city block or a housing tract, with flow entering laterally along the
length of the channel. As with the subcollectors, the average
contributing area for each collector channel is required. Column 2 of
Table 15 shows information that must be provided about the collector
channels.
The main channel. This channel conveys flow from upstream

subwatersheds and flows that enter from the collector channels or
overland flow planes. Column 3 of Table 15 shows information that must
be provided about the main channel.
The choice of elements to describe any watershed depends upon the configuration of
the drainage system. The minimum configuration is one overland flow plane and the
main channel, while the most complex would include two planes, subcollectors,
collectors, and the main channel.
The planes and channels are described by representative slopes, lengths, shapes,
and contributing areas. Publications from HEC (USACE, 1979; USACE, 1998)
provide guidance on how to choose values and give examples.
The roughness coefficients for both overland flow planes and channels commonly are
estimated as a function of surface cover, using, for example, Table 14, for overland
flow planes and the tables in Chow (1959) and other texts for channel n values.
66
Table 15. Information needs for kinematic wave modeling.

Overland Flow Planes Collectors and Main Channel
Subcollectors
Typical length Area drained by channel Channel length
Representative slope Representative channel Description of channel
Overland-flow roughness length shape
coefficient Description of channel Principle dimensions of
Area represented by plane shape channel cross section
Loss model parameters (see Principle dimensions of Channel slope

Chapter 5) representative channel Representative Mannings
cross section roughness coefficient
Representative channel Identification of upstream
slope inflow hydrograph (if any)
Representative
Mannings roughness
coefficient
Applicability and Limitations of Direct Runoff Models

Choice of a direct runoff model from amongst the included options depends upon:
Availability of information for calibration or parameter estimation.

Use of the parametric UH models requires specifying model parameters.
Use one of the empirical parameter predictors, such as Equation 35, to
compute parameters. However, the optimal source of these parameters
is calibration, as described in Chapter 9. If the necessary data for such
calibration in an urban watershed is not available, then the kinematic-
wave model may be the best choice, as the parameters and information
required to use that model are related to measurable and observable
watershed properties.
Appropriateness of the assumptions inherent in the model. Each of

the models is based upon one or more basic assumptions; if these are
violated, then avoid the use of the model. For example, the SCS UH
model assumes that the watershed UH is a single-peaked hydrograph. If
all available information indicates that the shape of the watershed and
the configuration of the drainage network causes multiple peaks for even
simple storms, then the SCS UH should not be used.
Likewise, the kinematic wave model is not universally applicable: Ponce

(1991) for example, argues that because of numerical properties of the
solution algorithms, the method "is intended primarily for small
2
watersheds [those less than 1 sq mi (2.5 km )], particularly in the cases
in which it is possible to resolve the physical detail without compromising
the deterministic nature of the model." Thus, for a larger watershed, one
of the UH models is perhaps a better choice.
User preference and experience. A combination of experience and

preference should guide the choice of models. As noted in Chapter 5,
experience is a critical factor in the success of a modeling effort.
However, be careful in using a particular model with a given parameter
just because that seems to be the standard of practice. For example, do
not automatically assume that tlag = 0.6 tc for the SCS UH method.
67
Instead, make best use of available data to confirm this parameter

estimate.
References
Alley, W.M. and Smith, P.E. (1987). Distributed routing rainfall-runoff model, Open file
report 82-344. U.S. Geological Survey, Reston, VA.
Bedient, P.B., and Huber, W.C. (1992). Hydrology and floodplain analysis. Addison-
Wesley, New York, NY.
Chaudhry, H.C. (1993). Open-channel hydraulics. Prentice Hall, NJ.
Chow, V.T. (1959). Open channel flow. McGraw-Hill, New York, NY.
Hill, New York, NY.
Clark, C.O. (1945). "Storage and the unit hydrograph." Transactions, ASCE, 110,
1419-1446.
Cudworth, A.G. (1989). Flood hydrology manual. US Department of the Interior,

Bureau of Reclamation, Washington, DC.
Dooge, J.C.I. (1959). "A general theory of the unit hydrograph." Journal of
Geophysical Research, 64(2), 241-256.
Kull, D., and Feldman, A. (1998). "Evolution of Clark's unit graph method to spatially
distributed runoff." Journal of Hydrologic Engineering, ASCE, 3(1), 9-19.
Leclerc, G. and Schaake, J.C. (1973). Methodology for assessing the potential
impact of urban development on urban runoff and the relative efficiency of runoff
control alternatives, Ralph M. Parsons Lab. Report 167. Massachusetts Institute of
Technology, Cambridge, MA.
Peters, J. and Easton, D. (1996). "Runoff simulation using radar rainfall data." Water
Resources Bulletin, AWRA, 32(4), 753-760.
Ponce, V.M. (1991). "The kinematic wave controversy." Journal of hydraulic

engineering, ASCE,. 117(4), 511-525.
Soil Conservation Service (1971). National engineering handbook, Section 4:

Hydrology. USDA, Springfield, VA.
Soil Conservation Service (1986). Urban hydrology for small watersheds, Technical
Report 55. USDA, Springfield, VA.
USACE (1944). Hydrology, San Gabriel River and the Rio Hondo above Whittier
Narrows flood control basin. US Army Engineer District, Los Angeles, CA.
USACE (1979). Introduction and application of kinematic wave routing techniques

using HEC-1, Training Document 10. Hydrologic Engineering Center, Davis, CA
USACE (1987). Derivation of a rainfall-runoff model to compute n-year floods for

Orange County watersheds. US Army Engineer District, Los Angeles, CA.
68

Washington, DC.
USACE (1996). GridParm: Procedures for deriving grid cell parameters for the
ModClark rainfall-runoff model. Hydrologic Engineering Center, Davis, CA.
USACE (1998). HEC-1 flood hydrograph package user's manual. Hydrologic

USACE (1999). GeoHEC-HMS user's manual (Draft). Hydrologic Engineering Center,

Davis, CA.
69
Chapter 7 Modeling Baseflow
CHAPTER 7
Modeling Baseflow
Two distinguishable components of a streamflow hydrograph are (1) direct, quick
runoff of precipitation, and (2) baseflow. Baseflow is the sustained or "fair-weather"
runoff of prior precipitation that was stored temporarily in the watershed, plus the
delayed subsurface runoff from the current storm. Some conceptual models of
watershed processes account explicitly for this storage and for the subsurface
movement. However this accounting is not necessary to provide the information for
many water resources studies.
The program includes three alternative models of baseflow:
Constant, monthly-varying value.
Exponential recession model.
Linear-reservoir volume accounting model.
Basic Concepts and Implementation

Constant , Monthly-Varying Baseflow
This is the simplest baseflow model included in the program. It represents baseflow
as a constant flow; this may vary monthly. This user-specified flow is added to the
direct runoff computed from rainfall for each time step of the simulation.
Exponential Recession Model

The program includes a exponential recession model to represent watershed
baseflow (Chow, Maidment, and Mays, 1988). The recession model has been used
often to explain the drainage from natural storage in a watershed (Linsley et al,
1982). It defines the relationship of Qt, the baseflow at any time t, to an initial value
as:
Qt = Q 0 k t ( 67 )
where Q0 = initial baseflow (at time zero); and k = an exponential decay constant.
The baseflow thus computed is illustrated in Figure 22. The shaded region represents
baseflow in this figure; the contribution decays exponentially from the starting flow.
Total flow is the sum of the baseflow and the direct surface runoff.
As implemented in the program, k is defined as the ratio of the baseflow at time t to

the baseflow one day earlier. The starting baseflow value, Q0, is an initial condition
3
of the model. It may be specified as a flow rate (m /s or cfs), or it may be specified
3 2
as a flow per unit area (m /s/km or cfs/sq mi).
70
Discharge
Direct surface
runoff
Baseflow
Time
Figure 22. Initial baseflow recession.
The recession baseflow model is applied both at the start of simulation of a storm
event, and later in the event as the delayed subsurface flow reaches the watershed
channels, as illustrated in Figure 23. Here, after the peak of the direct runoff, a user-
specified threshold flow defines the time at which the recession model of Equation 67
defines the total flow. That threshold may be specified as a flow rate or as a ratio to
the computed peak flow. For example, if the threshold is specified as a ratio-to-peak
3 3
of 0.10, and the computed peak is 1000 m /s, then the threshold flow is 100 m /s.
Subsequent total flows are computed with Equation 67, with Q0 = the specified
threshold value.
Discharge
Total flow
Threshold
Flow defined
Initial by recession
baseflow
recession
Baseflow
Time
Figure 23. Baseflow model illustration.
At the threshold flow, baseflow is defined by the initial baseflow recession.

Thereafter, baseflow is not computed directly, but is defined as the recession flow
less the direct-surface-runoff. When the direct-surface runoff eventually reaches zero
(all rainfall has run off the watershed), the total flow and baseflow are identical.
After the threshold flow occurs, the streamflow hydrograph ordinates are defined by
the recession model alone, unless the direct runoff plus initial baseflow recession
contribution exceeds the threshold. This may be the case if subsequent precipitation
causes a second rise in the hydrograph, as illustrated in Figure 24. In that case,
ordinates on the second rising limb are computed by adding direct runoff to the initial
recession, as illustrated.
71
Discharge
Threshold
Initial flow
Initial flow Initial flow
recession recession
Time
Figure 24. Recession with multiple runoff peaks.
Linear Reservoir Model

The linear-reservoir baseflow model is used in conjunction with the continuous soil-
moisture accounting (SMA) model that is described in Chapter 5. This baseflow
model simulates the storage and movement of subsurface flow as storage and
movement of water through reservoirs. The reservoirs are linear: the outflow at each
time step of the simulation is a linear function of the average storage during the time
step. Mathematically, this is identical to the manner in which Clark's UH model
represents watershed runoff, as described in Chapter 6.
The outflow from groundwater layer 1 of the SMA is inflow to one linear reservoir, and
the outflow from groundwater layer 2 of the SMA is inflow to another. The outflow
from the two linear reservoirs is combined to compute the total baseflow for the
watershed.
Estimating Baseflow Model Parameters

Constant , Monthly-Varying Baseflow
The parameters of this model are the monthly baseflows. These are best estimated
empirically, with measurements of channel flow when storm runoff is not occurring.
In the absence of such records, field inspection may help establish the average flow.
For large watersheds with contribution from groundwater flow and for watersheds
with year-round precipitation, the contribution may be significant and should not be
ignored. On the other hand, for most urban channels and for smaller streams in the
western and southwestern US, the baseflow contribution may be negligible.
Exponential Recession Model

The parameters of this model include the initial flow, the recession ratio, and the
threshold flow. As noted, the initial flow is an initial condition. For analysis of
hypothetical storm runoff, initial flow should be selected as a likely average flow that
would occur at the start of the storm runoff. For frequent events, the initial flow might
be the average annual flow in the channel. Field inspection may help establish this.
As with the constant, monthly-varying baseflow, for most urban channels and for
smaller streams in the western and southwestern US, this may well be zero, as the
baseflow contribution is negligible.
The recession constant, k, depends upon the source of baseflow. If k = 1.00, the
baseflow contribution will be constant, with all Qt = Q0. Otherwise to model the
72
exponential decay typical of natural undeveloped watersheds, k must be less than

1.00. Table 16 shows typical values proposed by Pilgrim and Cordery (1992) for
2
basins ranging in size from 300 to 16,000 km (120 to 6500 square miles) in the US,
eastern Australia, and several other regions. Large watersheds may have k values at
the upper end of the range, while smaller watersheds will have values at the lower
end.
Table 16. Typical recession constant values.

Flow Component Recession Constant, Daily
Groundwater 0.95
Interflow 0.8-0.9
Surface runoff 0.3-0.8
The recession constant can be estimated if gaged flow data are available. Flows
prior to the start of direct runoff can be plotted, and an average of ratios of ordinates
spaced one day apart can be computed. This is simplified if a logarithmic axis is
used for the flows, as the recession model will plot as a straight line.
The threshold value can be estimated also from examination of a graph of observed
flows versus time. The flow at which the recession limb is approximated well by a
straight line defines the threshold value.
Linear Reservoir Model

The linear reservoir model is used with soil-moisture accounting model. It is best
calibrated using procedures consistent with those used to calibrate that model.
References
Hill, New York, NY.
Pilgrim, D.H, and Cordery, I. (1992). "Flood runoff." D.R. Maidment, ed., Handbook of
hydrology, McGraw-Hill, New York, NY.
73
Chapter 8 Modeling Channel Flow
CHAPTER 8
Modeling Channel Flow

This section describes the models of channel flow that are included in the program;
these are also known as routing models. The routing models available include:
Lag
Muskingum
Modified Puls, also known as storage routing
Kinematic-wave
Muskingum Cunge
Each of these models computes a downstream hydrograph, given an upstream

hydrograph as a boundary condition. Each does so by solving the continuity and
momentum equations. This chapter presents a brief review of the fundamental
equations, simplifications, and solutions to alternative models.
The routing models that are included are appropriate for many, but not all, flood
runoff studies. The latter part of this chapter describes how to pick the proper model.
Open-Channel Flow Equations and Solution Techniques

Basic Equations of Open-Channel Flow
At the heart of the routing models included in the program are the fundamental
equations of open channel flow: the momentum equation and the continuity equation.
Together the two equations are known as the St. Venant equations or the dynamic
wave equations.
The momentum equation accounts for forces that act on a body of water in an open
channel. In simple terms, it equates the sum of gravitational force, pressure force,
and friction force to the product of fluid mass and acceleration. In one dimension, the
equation is written as:
y V V 1 V
S f = S0 ( 68 )
x g x g t
where Sf = energy gradient (also known as the friction slope); S0 = bottom slope; V =
velocity; y = hydraulic depth; x = distance along the flow path; t = time; g =
acceleration due to gravity; y/x = pressure gradient; (V/g)(V/x) = convective
acceleration; and (1/g)(V/t) = local acceleration.
The continuity equation accounts for the volume of water in a reach of an open
channel, including that flowing into the reach, that flowing out of the reach, and that
stored in the reach. In one-dimension, the equation is:
74
V y y
A + VB + B =q ( 69 )
x x t
where B = water surface width; and q = lateral inflow per unit length of channel. Each
of the terms in this equation describes inflow to, outflow from, or storage in a reach of
channel, a lake or pond, or a reservoir. Henderson (1966) described the terms as
A(V/x) = prism storage; VB(y/x) = wedge storage; and B(y/t) = rate of rise.
The momentum and continuity equations are derived from basic principles,
assuming:
Velocity is constant, and the water surface is horizontal across any

channel section.
All flow is gradually varied, with hydrostatic pressure prevailing at all

points in the flow. Thus vertical accelerations can be neglected.
No lateral, secondary circulation occurs.
Channel boundaries are fixed; erosion and deposition do not alter the
shape of a channel cross section.
Water is of uniform density, and resistance to flow can be described by empirical

formulas, such as Manning's and Chezy's equation.
Approximations
Although the solution of the full equations is appropriate for all one-dimensional
channel-flow problems, and necessary for many, approximations of the full equations
are adequate for typical flood routing needs. These approximations typically combine
the continuity equation (Equation 69) with a simplified momentum equation that
includes only relevant and significant terms.
Henderson (1966) illustrates this with an example for a steep alluvial stream with an
inflow hydrograph in which the flow increased from 10,000 cfs to 150,000 cfs and
decreased again to 10,000 cfs within 24 hours. Table 17 shows the terms of the
momentum equation and the approximate magnitudes that he found. The force
associated with the stream bed slope is the most important. If the other terms are
omitted from the momentum equation, any error in solution is likely to be insignificant.
Thus, for this case, the following simplification of the momentum equation may be
used:
S f = S0 ( 70 )
If this simplified momentum equation is combined with the continuity equation, the
result is the kinematic wave approximation, which is described in Chapter 6.
75
Table 17. Relative magnitude of momentum equation terms for steep channel,
rapidly-rising hydrograph (from Henderson, 1966)
Flow Component Recession Constant, Daily
So (bottom slope) 26
y 0.5
(pressure gradient)
x
V V 0.12 0.25
(convective acceleration)
g X
1 V 0.05
(local acceleration)
g t
Other common approximations of the momentum equation include:
Diffusion wave approximation. This approximation is the basis of the

Muskingum-Cunge routing model that is described subsequently in this
chapter.
y
S f = S0 ( 71 )
x
Quasi-steady dynamic-wave approximation. This approximation is

often used for water-surface profile computations along a channel reach,
given a steady flow. It is incorporated in computer programs HEC-2
(USACE, 1990) and HEC-RAS (USACE, 1998).
y V V
S f = S0 ( 72 )
x g x
Solution Methods
In HEC-HMS, the various approximations of the continuity and momentum equations
are solved using the finite difference method. In this method, finite difference
equations are formulated from the original partial differential equations. For example,
V/t from the momentum equation is approximated as V/t , a difference in velocity
in successive time steps t, and V/x is approximated as V/x, a difference in
velocity at successive locations spaced at x. Substituting these approximations into
the partial differential equations yields a set of algebraic equations. Depending upon
the manner in which the differences are computed, the algebraic equations may be
solved with either an explicit or an implicit scheme. With an explicit scheme, the
unknown values are found recursively for a constant time, moving from one location
along the channel to another. The results of one computation are necessary for the
next. With an implicit scheme, all the unknown values for a given time are found
simultaneously.
Parameters, Initial Conditions, and Boundary Conditions

The basic information requirements for all routing models are:
A description of the channel. All routing models that are included in the
program require a description of the channel. In some of the models,
76
this description is implicit in parameters of the model. In others, the

description is provided in more common terms: channel width, bed slope,
cross-section shape, or the equivalent.
Energy-loss model parameters. All routing models incorporate some

type of energy-loss model. The physically-based routing models, such
as the kinematic-wave model and the Muskingum-Cunge model use
Manning's equation and Mannings roughness coefficients (n values).
Other models represent the energy loss empirically.
Initial conditions. All routing models require initial conditions: the flow
(or stage) at the downstream cross section of a channel prior to the first
time period. For example, the initial downstream flow could be estimated
as the baseflow within the channel at the start of the simulation, as the
initial inflow, or as downstream flow likely to occur during a hypothetical
event.
Boundary conditions. The boundary conditions for routing models are

the upstream inflow, lateral inflow, and tributary inflow hydrographs.
These may be observed historical events, or they may be computed with
the precipitation-runoff models included in the program.
Modified Puls Model

The Modified Puls routing method, also known as storage routing or level-pool
routing, is based upon a finite difference approximation of the continuity equation,
coupled with an empirical representation of the momentum equation (Chow, 1964;
Henderson, 1966).
For the Modified Puls model, the continuity equation is written as
Q A
+ =0 ( 73 )
x t
This simplification assumes that the lateral inflow is insignificant, and it allows width
to change with respect to location. Rearranging this equation and incorporating a
finite-difference approximation for the partial derivatives yields:
S t
I t Ot = ( 74 )
t
where It = average upstream flow (inflow to reach) during a period t; Ot = average

downstream flow (outflow from reach) during the same period; and St = change in
storage in the reach during the period. Using a simple backward differencing scheme
and rearranging the result to isolate the unknown values yields:
S t Ot I t 1 + I t S t 1 Ot 1
+ = + ( 75 )
t 2 2 t 2
in which It-1 and It = inflow hydrograph ordinates at times t-1 and t, respectively; Ot-1
and Ot = outflow hydrograph ordinates at times t-1 and t, respectively; and St-1 and St
= storage in reach at times t-1 and t, respectively. At time t, all terms on the right-
77
hand side of this equation are known, and terms on the left-hand side are to be
found. Thus, the equation has two unknowns at time t: St and Ot.
A functional relationship between storage and outflow is required to solve Equation

75. Once that function is established, it is substituted into Equation 75, reducing the
equation to a nonlinear equation with a single unknown, Ot. This equation is solved
recursively by the program, using a trial-and-error procedure. [Note that at the first
time t, the outflow at time t-1 must be specified to permit recursive solution of the
equation; this outflow is the initial outflow condition for the storage routing model.]
Defining the Storage-Outflow Relationship

The storage-outflow relationship required for the Modified Puls routing model can be
determined with:
Water-surface profiles computed with a hydraulics model. Steady-

flow water surface profiles, computed for a range of discharges with
programs like HEC-2 (USACE, 1990), HEC-RAS (USACE, 1998), or a
similar model, define a relationship of storage to flow between two
channel cross sections.
Figure 25 illustrates this; it shows a set of water-surface profiles between

cross section A and cross section B of a channel. These profiles were
computed for a set of steady flows, Q1, Q2, Q3, and Q4.
For each profile, the volume of water in the reach, Si, can be computed,
using solid geometry principles. In the simplest case, if the profile is
approximately planar, the volume can be computed by multiplying the
average cross-section area bounded by the water surface by the reach
length. Otherwise, another numerical integration method can be used. If
each computed volume is associated with the steady flow with which the
profile is computed, the result is a set of points on the required storage-
outflow relationship.
This procedure can be used with existing or with proposed channel

configurations. For example, to evaluate the impact of a proposed
channel project, the channel cross sections can be modified, water
surface profiles recalculated, and a revised storage-outflow relationship
developed.
78
= Profile 4
= Profile 3
= Profile 2 S4
S3
Profile 1
=
Q4
S2
Q3
S1
Q2
Q1
A B
Figure 25. Steady-flow water-surface profiles and storage-outflow curve.
Historical observations of flow and stage. Observed water surface

profiles, obtained from high water marks, can be used to define the
required storage-outflow relationships, in much the same manner that
computed water-surface profiles are used. Each observed discharge-
elevation pair provides information for establishing a point of the
relationship.
Sufficient stage data over a range of floods is required to establish the

storage-outflow relationship in this manner. If only a limited set of
observations is available, these may best be used to calibrate a water-
surface profile-model for the channel reach of interest. Then that
calibrated model can be exercised to establish the storage-outflow
relationship as described above.
Calibration, using observed inflow and outflow hydrographs for the

reach of interest. Observed inflow and outflow hydrographs can be
used to compute channel storage by an inverse process of flood routing.
When both inflow and outflow are known, the change in storage can be
computed using Equation 74. Then, the storage-outflow function can be
developed empirically. Note that tributary inflow, if any, must also be
accounted for in this calculation.
Inflow and outflow hydrographs also can be used to find the storage-
outflow function by trial-and-error. In that case, a candidate function is
defined and used to route the inflow hydrograph. The outflow
hydrograph thus computed is compared with the observed hydrograph.
If the match is not adequate, the function is adjusted, and the process is
repeated. Chapter 9 provides more information regarding this process,
which is referred to as calibration.
Estimating Other Model Parameters

Chapter 6 of this manual describes how an accurate solution of the finite difference
form of the kinematic-wave model requires careful selection of x and t; this is also
79
true for solution of the storage-routing model equations. For the kinematic-wave
model, an accurate solution can be found with a stable algorithm when x/t c,
where c = average wave speed over a distance increment x. This rule applies also
with storage routing. As implemented in the program, x for the finite difference
approximation of Q/x is implicitly equal to the channel reach length, L, divided by
an integer number of steps. The goal is to select the number of steps so that the
travel time through the reach is approximately equal the time step t. This is given
approximately by:
L
steps = ( 76 )
ct
The number of steps affects the computed attenuation of the hydrograph. As the
number of routing steps increases, the amount of attenuation decreases. The
maximum attenuation corresponds to one step; this is used commonly for routing
though ponds, lakes, wide, flat floodplains, and channels in which the flow is heavily
controlled by downstream conditions. Strelkoff (1980) suggests that for locally-
controlled flow, typical of steeper channels:
S0
steps = 2 L ( 77 )
y0
where yo = normal depth associated with baseflow in the channel. EM 1110-2-1417

points out that this parameter, however, is best determined by calibration, using
observed inflow and outflow hydrographs.
Muskingum Model
The Muskingum routing model, like the modified Puls model, uses a simple finite
difference approximation of the continuity equation:
I t 1 + I t Ot 1 + Ot S t S t 1
= ( 78 )
2 2 t
Storage in the reach is modeled as the sum of prism storage and wedge storage. As
shown in Figure 26, prism storage is the volume defined by a steady-flow water
surface profile, while wedge storage is the additional volume under the profile of the
flood wave. During rising stages of the flood, wedge storage is positive and is added
to the prism storage. During the falling stages of a flood, the wedge storage is
negative and is subtracted from the prism storage.
80
Water Surface Profile Negative Wedge

storage
Wedge storage
Wedge storage
rage
Prism Sto
rage
Prism Sto
rage
Prism Sto
Figure 26. Wedge storage (from Linsley et al., 1982)
The volume of prism storage is the outflow rate, O, multiplied by the travel time
through the reach, K. The volume of wedge storage is a weighted difference
between inflow and outflow, multiplied by the travel time K. Thus, the Muskingum
model defines the storage as:
S t = KOt + KX (I t Ot ) = K [ XI t + (1 X )Ot ] ( 79 )
where K = travel time of the flood wave through routing reach; and X = dimensionless
weight (0 X 0.5).
The quantity X It + (1-X) Ot is a weighted discharge. If storage in the channel is

controlled by downstream conditions, such that storage and outflow are highly
correlated, then X = 0.0. In that case, Equation 82 resolves to S = KO; this is the
linear reservoir model that was described in Chapter 6. If X = 0.5, equal weight is
given to inflow and outflow, and the result is a uniformly progressive wave that does
not attenuate as it moves through the reach.
If Equation 78 is substituted into Equation 79 and the result is rearranged to isolate

the unknown values at time t, the result is:
t 2 KX t + 2 KX 2 K (1 X ) t
Ot = I t + I t 1 + Ot 1 ( 80 )
2 K (1 X ) + t 2 K (1 X ) + t 2 K (1 X ) + t
The program solves Equation 80 recursively to compute ordinates of the outflow

hydrograph, given the inflow hydrograph ordinates (It for all t), an initial condition
(Ot=0), and the parameters, K and X.

Constraints on the parameters. As noted, the feasible range for the parameter X is
(0, 0.5). However, these other constraints apply to selection of X and the parameter
K:
As with other routing models, an accurate solution requires selection of

appropriate time steps, distance steps, and parameters to ensure
81
accuracy and stability of the solution. With Muskingum routing, as with

modified Puls routing, the distance step, x, is defined indirectly by the
number of steps into which a reach is divided for routing. And as with
other models, x/t is selected to approximate c , where c = average
wave speed over a distance increment x. With the Muskingum model,
the wave speed is K/L, so the number of steps should be approximately
K/t.
The parameters K and X and the computational time step t also must
be selected to ensure that the Muskingum model, as represented by
Equations 8-15 and 8-16, is rational. That means that the parenthetical
terms must be non-negative; the values of K and X must be chosen so
that the combination falls within the shaded region shown in Figure 27.
t / K 1
0
0.0 0.5 1.0
X
Figure 27. Feasible region for Muskingum model parameters.
Calibrating the model using observed flows. If observed inflow and outflow
hydrographs are available, the Muskingum model parameter K can be estimated as
the interval between similar points on the inflow and outflow hydrographs. For
example, K can be estimated as the elapsed time between the centroid of areas of
the two hydrographs, as the time between the hydrograph peaks, or as the time
between midpoints of the rising limbs. Once K is estimated, X can be estimated by
trial and error.
Chapter 9 describes the calibration capability of the program; this may be used with
parameters of the Muskingum model. In that case, both K and X may be estimated
by trial-and-error.
Estimating the parameters for ungaged watersheds. If gaged flows required for
calibration are not available, K and X can be estimated from channel characteristics.
For example, EM 1110-2-1417 proposes estimating K as follows:
Estimate the flood wave velocity, Vw , using Seddons law, as:
1 dQ
Vw = ( 81 )
B dy
82
where B = top width of the water surface, and dQ/dy = slope of the
discharge rating curve at a representative channel cross section. As an
alternative, EM 1110-2-1417 suggests estimating the flood wave velocity
as 1.33-1.67 times the average velocity, which may be estimated with
Manning's equation and representative cross section geometric
information.
Estimate K as:
L
K= ( 82 )
Vw
Experience has shown that for channels with mild slopes and over-bank
flow, the parameter X will approach 0.0. For steeper streams, with well-
defined channels that do not have flows going out of bank, X will be
closer to 0.5. Most natural channels lie somewhere in between these
two limits, leaving room for engineering judgement. Cunge (1969)
estimated X as
1 Qo
X = 1 ( 83 )
2 BS o cx
where Qo = a reference flow from the inflow hydrograph; B = top width of
flow area; So = friction slope or bed slope; c = flood wave speed
(celerity); and x = the length of reach. The reference flow is an average
value for the hydrograph, midway between the base flow and the peak
flow (Ponce, 1983).
Lag Model
Basic Concept
This is the simplest of the included routing models. With it, the outflow hydrograph is
simply the inflow hydrograph, but with all ordinates translated (lagged in time) by a
specified duration. The flows are not attenuated, so the shape is not changed. This
model is widely used, especially in urban drainage channels (Pilgrim and Cordery,
1993).
Mathematically, the downstream ordinates are computed by:
It t < lag
Ot = ( 84 )
I t lag t lag
where Ot = outflow hydrograph ordinate at time t; It = inflow hydrograph ordinate at

time t; and lag = time by which the inflow ordinates are to be lagged.
Figure 28 illustrates the results of application of the lag model. In the figure, the
upstream (inflow) hydrograph is the boundary condition. The downstream
hydrograph is the computed outflow, with each ordinate equal to an earlier inflow
ordinate, but lagged in time.
83
lag
Discharge
Upstream Downstream
hydrograph hydrograph
(inflow) (outflow)
Time
Figure 28. Lag example.
The lag model is a special case of other models, as its results can be duplicated if
parameters of those other models are carefully chosen. For example, if X = 0.50 and
K = t in the Muskingum model, the computed outflow hydrograph will equal the
inflow hydrograph lagged by K.
Estimating the Lag

If observed flow hydrographs are available, the lag can be estimated from these as
the elapsed time between the time of the centroid of areas of the two hydrographs,
between the time of hydrograph peaks, or between the time of the midpoints of the
rising limbs.
Kinematic Wave Model

The kinematic-wave channel routing model is based upon a finite difference
approximation of the continuity equation and a simplification of the momentum
equation. This is described in detail in Chapter 6.
Setting Up the Model

Information required to used the kinematic-wave channel routing model is shown in
Table 18. This information, for the most part, can be gathered from maps, surveys,
and field inspection. Manning's n can be estimated using common procedures.
84
Table 18. Kinematic wave routing model information requirements.

Flow Component
Shape of the cross section: Is it trapezoidal, rectangular, or circular?
Principle dimension: bottom width of the channel, diameter of the conduit.
Side slope of trapezoidal shape.
Length of the reach.
Slope of the energy grade line.
Manning n, roughness coefficient for channel flow.
Muskingum-Cunge Model
Although popular and easy to use, the Muskingum model includes parameters that
are not physically based and thus are difficult to estimate. Further, the model is
based upon assumptions that often are violated in natural channels. An extension,
the Muskingum-Cunge model, overcomes these limitations.
The model is based upon solution of the following form of the continuity equation,
(with lateral inflow, qL, included):
A Q
+ = qL ( 85 )
t x
and the diffusion form of the momentum equation:
y
S f = So ( 86 )
x
Combining these and using a linear approximation yields the convective diffusion
equation (Miller and Cunge, 1975):
Q Q 2Q
+c = 2 + cq L ( 87 )
t x x
where c = wave celerity (speed); and = hydraulic diffusivity. The wave celerity and
the hydraulic diffusivity are expressed as follows:
dQ
c= ( 88 )
dA
and
Q
= ( 89 )
2 BS o
where B = top width of the water surface. A finite difference approximation of the
partial derivatives, combined with Equation 80, yields:
Ot = C1 I t 1 + C 2 I t + C 3 Ot 1 + C 4 (q L x) ( 90 )
85
The coefficients are:
t
+ 2X
C1 = K ( 91 )
t
+ 2(1 X )
K
t
2X
C2 = K ( 92 )
t
+ 2(1 X )
K
t
2(1 X )
C3 = K ( 93 )
t
+ 2(1 X )
K
t
2
C4 = K ( 94 )
t
+ 2(1 X )
K
The parameters K and X are (Cunge, 1969; Ponce, 1978):
x
K= ( 95 )
c
1 Q
X = 1 ( 96 )
2 BS o cx
But c, Q, and B change over time, so the coefficients C1, C2, C3, and C4 must also
change. The program recomputes them at each time and distance step, t and x,
using the algorithm proposed by Ponce (1986).
Again, the choice of these time and distance steps is critical. The steps are selected
to ensure accuracy and stability. The t is selected as the minimum of the following:
user time step from the control specifications; the travel time through the reach; or
th
1/20 the time to rise of the peak inflow with the steepest rising limb, rounded to the
nearest multiple or divisor of the user time step. Once t is chosen, x is computed
as:
x = ct ( 97 )
The value is constrained so that:
1 Q
x < ct + o ( 98 )
2 BS o c
Here Qo = reference flow, computed from the inflow hydrograph as:
86
Qo = Q B +
1
2
(
Q peak Q B ) ( 99 )
where QB = baseflow; and Qpeak = inflow peak.
Setting Up the Model and Estimating Parameters

The Muskingum-Cunge model included in the program can be used in either of two
configurations:
Standard configuration. In this configuration, a simple description of a

representative channel cross section is provided. Or, one of the
alternative shapes shown in Figure 20 is selected. The principle
dimensions of the section are specified, along with channel roughness,
energy slope, and length. The length and roughness can be estimated
from maps, aerial photographs, and field surveys. The energy slope can
be estimated as the channel bed slope, in the absence of better
information.
1 8
Left Right
overbank Main channel
overbank
n left overbank n main channel n right overbank

2
7
3
5
Figure 29. Format for describing channel geometry with 8 points.
8-point cross-section configuration. If one of the standard cross-

section shapes will not represent will the channel geometry, the
alternative is to use the so-called 8-point cross section configuration.
With this, a representative cross section is described for the routing
reach, using 8 pairs of x, y (distance, elevation) values. These values
are defined specifically as illustrated in Figure 29. Points labeled 3 and 6
represent the left and right banks of the channel at the representative
cross section. Points 4 and 5 are within the channel. Points 1 and 2
represents the left overbank, and points 7 and 8 represent the right
overbank.
The reach length, roughness coefficient(s), and energy grade also must be specified.
As with the standard configuration, the length and roughness can be estimated from
maps, aerial photographs, and field surveys, and the energy slope can be estimated
as the channel bed slope, in the absence of better information.
With either configuration of the Muskingum-Cunge model, if the channel properties

vary significantly along the routing reach, the reach may be subdivided and modeled
as a series of linked subreaches, with the properties of each defined separately.
87
Applicability and Limitations of Routing Models

Each routing model that is included in the program solves the momentum and
continuity equations. However, each omits or simplifies certain terms of those
equations to arrive at a solution. To select a routing model, one must consider the
routing method's assumptions and reject those models that fail to account for critical
characteristics of the flow hydrographs and the channels through which they are
routed. These include (but are not limited to) the following:
Backwater effects. Tidal fluctuations, significant tributary inflows, dams,

bridges, culverts, and channel constrictions can cause backwater effects.
A flood wave that is subjected to the influences of backwater will be
attenuated and delayed in time. The kinematic wave and Muskingum
models cannot account for the influences of backwater on the flood
wave, because these are based on uniform-flow assumptions. Only the
modified Puls model can simulate backwater effects, and it can do so in
only the case of time-invariant downstream conditions. To model this
with the modified Puls model, the effects of the backwater must be
determined and included when developing the storage-discharge
relationship.
Practically, none of the routing models that are included in the program
will simulate channel flow well if the downstream conditions have a
significant impact on upstream flows. The internal structure of the
program is such that computations move from upstream watersheds and
channels to those downstream. Thus downstream conditions are not yet
known when routing computations begin. Only a complete hydraulic
system model can accomplish this.
Floodplain storage. If flood flows exceed the channel carrying capacity,

water flows into overbank areas. Depending on the characteristics of the
overbanks, that overbank flow can be slowed greatly, and often ponding
will occur. This can be significant in terms of the translation and
attenuation of a flood wave.
To analyze the transition from main channel to overbank flows, the

model must account for varying conveyance between the main channel
and the overbank areas. For one-dimensional flow models, this is
normally accomplished by calculating the hydraulic properties of the
main channel and the overbank areas separately, then combining them
to formulate a composite set of hydraulic relationships. This cannot be
accomplished with the kinematic-wave and Muskingum models. The
Muskingum model parameters are assumed constant. However, as flow
spills from the channel, the velocity may change significantly, so K
should change. While the Muskingum model can be calibrated to match
the peak flow and timing of a specific flood magnitude, the parameters
cannot easily be used to model a range of floods that may remain in
bank or go out of bank. Similarly, the kinematic wave model assumes
constant celerity, an incorrect assumption if flows spill into overbank
areas.
In fact, flood flows through extremely flat and wide flood plains may not
be modeled adequately as one-dimensional flow. Velocity of the flow
across the floodplain may be just as large as that of flow down the
channel. If this occurs, a two-dimensional flow model will better simulate
88
the physical processes. EM 1110-2-1416 (1993) provides more

information on this complex subject.
Interaction of channel slope and hydrograph characteristics. As

channel slopes lessen, assumptions made to develop many of the
models included in the program will be violated: momentum-equation
terms that were omitted are more important if the channel slope is small.
For example, the simplification for the kinematic-wave model is

appropriate only if the channel slope exceeds 0.002. The Muskingum-
Cunge model can be used to route slow-rising flood waves through
reaches with flat slopes. However, it should not be used for rapidly-rising
hydrographs in the same channels, because it omits acceleration terms
of the momentum equation that are significant in that case. Ponce
(1978) established a numerical criterion to judge the likely applicability of
various routing models. He suggested that the error due to the use of
the kinematic wave model is less than 5 percent if:
TS o u o
171 ( 100 )
do
where T = hydrograph duration; uo = reference mean velocity, and do =

reference flow depth. (These reference values are average flow
conditions of the inflow hydrograph.) He suggested that the error with
the Muskingum-Cunge model is less than 5 percent if:
12
g
TS o 30 ( 101 )
do
where g = acceleration of gravity.
Configuration of flow networks. In a dendritic stream system, if the

tributary flows or the main channel flows do not cause significant
backwater at the confluence of the two streams, any of the hydraulic or
hydrologic routing methods can be applied. However, if significant
backwater does occur at confluences, then the models that can account
for backwater must be applied. For full networks, where the flow divides
and possibly changes direction during the event, none of the simplified
models that are included in the program should be used.
Occurrence of subcritical and supercritical flow. During a flood, flow

may shift between subcritical and supercritical regimes. If the
supercritical flow reaches are short, this shift will not have a noticeable
impact on the discharge hydrograph. However, if the supercritical-flow
reaches are long, these should be identified and treated as separate
routing reaches. If the shifts are frequent and unpredictable, then none
of the simplified models are appropriate.
Availability of data for calibration. In general, if observed data are not

available, the physically-based routing models will be easier to set up
and apply with some confidence. Parameters such as the Muskingum X
can be estimated, but the estimates can be verified only with observed
flows. Thus these empirical models should be avoided if the watershed
and channel are ungaged.
89
Table 19 summarizes the model selection criteria.
Table 19. Guidelines for selecting routing model.

If this is true then consider this model
No observed hydrograph data available for
Kinematic wave; Muskingum-Cunge
calibration
Significant backwater will influence discharge
Modified Puls
hydrograph
Modified Puls, Muskingum-Cunge
Flood wave will go out of bank, into floodplain
with 8-point cross section
TS o u o
Channel slope > 0.002 and 171 Any
do
Channel slopes from 0.002 to 0.0004 and
TS o u o Muskingum-Cunge; modified Puls;
171 Muskingum
do
12
g
d
Channel slope < 0.0004 and TS o 30 Muskingum-Cunge
o
12
Channel slope < 0.0004 and TS g

o < 30 None
do
Modeling Confluences as Junctions

Figure 30 illustrates a simple stream confluence, also known as a stream junction.
Here two channels intersect, flow is combined, and water travels downstream.
Inflow 2
Inflow 1
Outflow
Figure 30. Stream confluence.
Such a confluence can be modeled with the program. To do so, the program uses the
following simplification of the continuity equation, which is based upon an assumption
that no water is stored at the confluence:
I
r
t
r
Ot = 0 ( 102 )
90
r
in which It = the flow in channel r at time t; and Ot = outflow from the confluence in
period t. Rearranging yields:
Ot = I
r
t
r
( 103 )
That is, the downstream flow at time t equals the sum of the upstream flows. This
equation is solved repeatedly for all times t in the simulation duration.
Setting Up a Confluence Model

The confluence model included in the program requires the stream system
configuration be specified using the graphical user interface. No parameters are
required for the model.
Limitations of the Confluence Model

The confluence model is appropriate only if the fundamental assumption of no
storage at the confluence is valid. This may not be true if backwater conditions exist
at the confluence. In that case, the stream system can be represented well with an
unsteady open-channel network model, such as UNET (USACE, 1996).
Modeling Bifurcations as Diversions

Figure 31 illustrates a bifurcationa split in the flow in a channel. Such a bifurcation,
in which the water flows downstream in one of two channels, can be modeled with
the program.
Inflow
Main Secondary
channel outflow channel outflow
Figure 31. Stream bifurcation.
A bifurcation is modeled in the program with a simple one-dimensional approximation

of the continuity equation. In that case:
I t Ot Otsec ondary = 0
main
( 104 )
main
in which Ot = average flow passing downstream in the main channel during time
interval t; It = average channel flow just upstream of the bifurcation during the
secondary
interval; and Ot = average flow into the secondary channel during the interval.
The distinction between main and secondary channels is arbitrary.
91
Setting Up a Bifurcation Model

The diversion model included in the program requires the secondary channel flow be
specified as a function of the inflow upstream of the diversion. That is, Equation 104
must be represented as:
Otmain = I t f ( I t ) ( 105 )
in which f(It) = a functional relationship of main channel inflow and secondary channel
flow. The relationship can be developed with historical measurements, a physical
model constructed in a laboratory, or a mathematical model of the hydraulics of the
channel.
Limitations of the Bifurcation Model

The diversion model included in the program is applicable to stream systems in which
the necessary relationship between main channel inflow and secondary channel flow
can be developed. Often this is impossible, because the secondary channel flow will
not be a unique function of main channel inflow. Instead, it will depend upon
downstream conditions in one or both channels, and upon the temporal distribution of
the inflow hydrographs. In that case, an unsteady-flow network model, such as
UNET (USACE, 1997), must be used instead to represent properly the complex
hydraulic relationship.
References
Chow, V.T. (1964). Handbook of applied hydrology. McGraw-Hill , New York, NY.
Cunge, J. A. (1969). "On the subject of a flood propagation computation method

(Muskingum method)." Journal of Hydraulic Research, 7(2), 205-230.
Henderson, F.M. (1966). Open channel flow. MacMillan Publishing Co. Inc., New
York, NY.
Miller, W.A., and Cunge, J.A. (1975). "Simplified equations of unsteady flow." K.
Mahmood and V. Yevjevich, eds., Unsteady flow in open channels, Vol. I, Water
Resources Publications, Ft. Collins, CO.
Pilgrim, D.H, and Cordery, I. (1983). "Flood runoff." D.R. Maidment, ed., Handbook of
hydrology, McGraw-Hill, New York, NY.
Ponce, V. M. (1983). Development of physically based coefficients for the diffusion

method of flood routing, Final Report to the USDA. Soil Conservation Service,
Lanham, MD.
Ponce, V. M., and Yevjevich, V. (1978). "Muskingum-Cunge method with variable

parameters." Journal of the Hydraulics Division, ASCE, 104(HY12), 1663-1667.
Ponce, V.M. (1986). "Diffusion wave modeling of catchment dynamics." Journal of

the Hydraulics Division, ASCE, 112(8), 716-727.
Strelkoff, T. (1980). Comparative analysis of flood routing methods. Hydrologic

Engineering Center, Davis, CA
92
USACE (1986). Accuracy of computed water surface profiles, Research Document

26. Hydrologic Engineering Center, Davis, CA.
USACE (1990). HEC-2 water surface profiles users manual. Hydrologic Engineering
Center, Davis, CA.
USACE (1993). River hydraulics, EM 1110-2-1416. Office of Chief of Engineers,

Washington, DC.

Washington, DC.
USACE (1995). HEC-DSS users guide and utility manuals. Hydrologic Engineering
Center, Davis, CA.
USACE (1997). UNET one-dimensional unsteady flow through a full network of open
channels users manual. Hydrologic Engineering Center, Davis, CA.
USACE (1998). HEC-RAS river analysis system users manual. Hydrologic

93
Chapter 9 Calibrating the Models
CHAPTER 9
Calibrating the Models
What is Calibration?
Each model that is included in the program has parameters. The value of each
parameter must be specified to use the model for estimating runoff or routing
hydrographs. Earlier chapters identified the parameters and described how they
could be estimated from various watershed and channel properties. For example,
the kinematic-wave direct runoff model described in Chapter 6 has a parameter N
that represents overland roughness; this parameter can be estimated from
knowledge of watershed land use.
However, as noted in Chapter 2, some of the models that are included have
parameters that cannot be estimated by observation or measurement of channel or
watershed characteristics. The parameter Cp in the Snyder UH model is an example;
this parameter has no direct physical meaning. Likewise, the parameter x in the
Muskingum routing model cannot be measured; it is simply a weight that indicates
the relative importance of upstream and downstream flow in computing the storage in
a channel reach. Equation 85 provides a method for estimating x from channel
properties, but this is only approximate and is appropriate for limited cases.
How then can the appropriate values for the parameters be selected? If rainfall and
streamflow observations are available, calibration is the answer. Calibration uses
observed hydrometeorological data in a systematic search for parameters that yield
the best fit of the computed results to the observed runoff. This search is often
referred to as optimization.
Summary of the Calibration Procedure

In HEC-HMS, the systematic search for the best (optimal) parameter values follows
the procedure illustrated in Figure 32. This procedure begins with data collection.
For rainfall-runoff models, the required data are rainfall and flow time series. For
routing models, observations of both inflow to and outflow from the routing reach are
required. Table 20 and Table 21 offer some tips for collecting these data.
The next step is to select initial estimates of the parameters. As with any search, the
better these initial estimates (the starting point of the search), the quicker the search
will yield a solution. Tips for parameter estimation found in previous chapters may be
useful here.
94
Collect
rainfall / Select
Start runoff data starting
estimate
Improve
estimate
Simulate
runoff
No
Finish Compare
computed
&
Yes OK?
observed
Figure 32. Schematic of calibration procedure.
Given these initial estimates of the parameters, the models included in the program
can be used with the observed boundary conditions (rainfall or upstream flow) to
compute the output, either the watershed runoff hydrograph or a channel outflow
hydrograph.
At this point, the program compares the computed hydrograph to the observed
hydrograph. For example, it computes the hydrograph represented with the dashed
line in Figure 33 and compares it to the observed hydrograph represented with the
solid line. The goal of this comparison is to judge how well the model "fits" the real
hydrologic system. Methods of comparison are described later in this chapter.
If the fit is not satisfactory, the program systematically adjusts the parameters and
reiterates. The algorithms for adjusting the parameters are described later in this
chapter.
When the fit is satisfactory, the program will report the optimal parameter values.
The presumption is that these parameter values then can be used for runoff or
routing computations that are the goal of the flood runoff analyses.
95
Table 20. Tips for collecting data for rainfall-runoff model calibration.
Rainfall and runoff observations must be from the same storm. The runoff time series
should represent all runoff due to the selected rainfall time series.
The rainfall data must provide adequate spatial coverage of the watershed, as these
data will be used with the methods described in Chapter 4 to compute MAP for the
storm.
The volume of the runoff hydrograph should approximately equal the volume of the
rainfall hyetograph. If the runoff volume is slightly less, water is being lost to infiltration,
as expected. But if the runoff volume is significantly less, this may indicate that flow is
stored in natural or engineered ponds, or that water is diverted out of the stream.
Similarly, if the runoff volume is slightly greater, baseflow is contributing to the total
flow, as expected. However, if the runoff volume is much greater, this may indicate that
flow is entering the system from other sources, or that the rainfall was not measured
accurately.
The duration of the rainfall should exceed the time of concentration of the watershed to
ensure that the entire watershed upstream of the concentration point is contributing to
the observed runoff.
The size of the storm selected for calibration should approximately equal the size of the
storm the calibrated model is intended to analyze. For example, if the goal is to predict
runoff from a 1%-chance 24-hour storm of depth 7 inches, data from a storm of duration
approximately 24 hours and depth approximately 7 inches should be used for
calibration.
Table 21. Tips for collecting data for routing model calibration.
The upstream and downstream hydrograph time series must represent flow for the
same period of time.
The volume of the upstream hydrograph should approximately equal the
volume of the downstream hydrograph, with minimum lateral inflow. The
lumped routing models in HEC-HMS assume that these volumes are equal.
The duration of the downstream hydrograph should be sufficiently long so that
the total volume represented equals the volume of the upstream hydrograph.
The size of the event selected for calibration should approximately equal the
size of the event the calibrated model is intended to analyze. For example, if
the study requires prediction of downstream flows for an event with depths of
20 feet in a channel, historical data for a event of similar depth should be used
for calibration.
96
18.0
16.0
14.0
12.0
Discharge (m 3/s)
Simulated time series
10.0 Observed time series
8.0
6.0
4.0
2.0
0.0
0 5 10 15 20 25 30 35
Time (hr)
Figure 33. How well does the computed hydrograph "fit"?
Goodness-of-Fit Indices
To compare a computed hydrograph to an observed hydrograph, the program
computes an index of the goodness-of-fit. Algorithms included in the program search
for the model parameters that yield the best value of an index, also known as
objective function. Only one of four objective functions included in the program can
be used, depending upon the needs of the analysis. The goal of all four calibration
schemes is to find reasonable parameters that yield the minimum value of the
objective function. The objective function choices (shown in Table 22) are:
Sum of absolute errors. This objective function compares each

ordinate of the computed hydrograph with the observed, weighting each
equally. The index of comparison, in this case, is the difference in the
ordinates. However, as differences may be positive or negative, a
simple sum would allow positive and negative differences to offset each
other. In hydrologic modeling, both positive and negative differences are
undesirable, as overestimates and underestimates as equally
undesirable. To reflect this, the function sums the absolute differences.
Thus, this function implicitly is a measure of fit of the magnitudes of the
peaks, volumes, and times of peak of the two hydrographs. If the value
of this function equals zero, the fit is perfect: all computed hydrograph
ordinates equal exactly the observed values. Of course, this is seldom
the case.
Sum of squared residuals. This is a commonly-used objective function

for model calibration. It too compares all ordinates, but uses the squared
3
differences as the measure of fit. Thus a difference of 10 m /sec
3
"scores" 100 times worse than a difference of 1 m /sec. Squaring the
differences also treats overestimates and underestimates as
undesirable. This function too is implicitly a measure of the comparison
of the magnitudes of the peaks, volumes, and times of peak of the two
hydrographs.
97
Table 22. Objective functions for calibration.

Criterion Equation
Sum of absolute errors NQ
(Stephenson, 1979) Z= q
i =1
O (i ) q S (i )
Sum of squared residuals NQ

(Diskin and Simon, 1977) Z= [q
i =1
O (i ) q S (i )]
2
Percent error in peak q S ( peak ) q O ( peak )

Z = 100
q O ( peak )
Peak-weighted root mean 1
1 NQ
2 q O (i ) + q O ( mean)
2

(q O (i ) q S (i ) )
square error objective
function (USACE, 1998) Z =
NQ i =1 2q O (mean)
Note:
1
Z = objective function; NQ = number of computed hydrograph ordinates; qO(t) = observed
flows; qS(t) = calculated flows, computed with a selected set of model parameters; qO(peak)
= observed peak; qO(mean) = mean of observed flows; and qS(peak) = calculated peak
Percent error in peak. This measures only the goodness-of-fit of the

computed-hydrograph peak to the observed peak. It quantifies the fit as
the absolute value of the difference, expressed as a percentage, thus
treating overestimates and underestimates as equally undesirable. It
does not reflect errors in volume or peak timing. This objective function
is a logical choice if the information needed for designing or planning is
limited to peak flow or peak stages. This might be the case for a
floodplain management study that seeks to limit development in areas
subject to inundation, with flow and stage uniquely related.
Peak-weighted root mean square error. This function is identical to the

calibration objective function included in computer program HEC-1
(USACE, 1998). It compares all ordinates, squaring differences, and it
weights the squared differences. The weight assigned to each ordinate
is proportional to the magnitude of the ordinate. Ordinates greater than
the mean of the observed hydrograph are assigned a weight greater than
1.00, and those smaller, a weight less than 1.00. The peak observed
ordinate is assigned the maximum weight. The sum of the weighted,
squared differences is divided by the number of computed hydrograph
ordinates; thus, yielding the mean squared error. Taking the square root
yields the root mean squared error. This function is an implicit measure
of comparison of the magnitudes of the peaks, volumes, and times of
peak of the two hydrographs.
In addition to the numerical measures of fit, the program also provides graphical
comparisons that permit visualization of the fit of the model to the observations of the
hydrologic system. A comparison of computed hydrographs can be displayed, much
like that shown in Figure 33. In addition, the program displays a scatter plot, as
shown in Figure 34. This is a plot of the calculated value for each time step against
the observed flow for the same step. Inspection of this plot can assist in identifying
model bias as a consequence of the parameters selected. The straight line on the
plot represents equality of calculated and observed flows: If plotted points fall on the
line, this indicates that the model with specified parameters has predicted exactly the
98
observed ordinate. Points plotted above the line represents ordinates that are over-
predicted by the model. Points below represent under-predictions. If all of the
plotted values fall above the equality line, the model is biased; it always over-
predicts. Similarly, if all points fall below the line, the model has consistently under-
predicted. If points fall in equal numbers above and below the line, this indicates that
the calibrated model is no more likely to over-predict than to under-predict.
The spread of points about the equality line also provides an indication of the fit of
the model. If the spread is great, the model does not match well with the
observations random errors in the prediction are large relative to the magnitude of
the flows. If the spread is small, the model and parameters fit better.
20.0
15.0
Simulated flow
10.0
5.0
0.0
0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00
Observed flow
Figure 34. Scatter plot of optimization results.
The program also computes and plots a time series of residualsdifferences

between computed and observed flows. Figure 35 is an example of this. This plot
indicates how prediction errors are distributed throughout the duration of the
simulation. Inspection of the plot may help focus attention on parameters that require
additional effort for estimation. For example, if the greatest residuals are grouped at
the start of a runoff event, the initial loss parameter may have been poorly chosen.
99
5.00
4.00
3.00
Residual (m3/s) 2.00
1.00
0.00
-1.00
-2.00
-3.00
-4.00
0 10 20 30 40 50 60 70 80 90 100
Time (hr)
Figure 35. Residual plot of optimization results.
Search Methods
As noted earlier, the goal of calibration is to identify reasonable parameters that yield
the best fit of computed to observed hydrograph, as measured by one of the
objective functions. This corresponds mathematically to searching for the
parameters that minimize the value of the objective function.
As shown in Figure 32, the search is a trial-and-error search. Trial parameters are
selected, the models are exercised, and the error is computed. If the error is
unacceptable, the program changes the trial parameters and reiterates. Decisions
about the changes rely on the univariate gradient search algorithm or the Nelder and
Mead simplex search algorithm.
Univariate-Gradient Algorithm
The univariate-gradient search algorithm makes successive corrections to the
k
parameter estimate. That is, if x represents the parameter estimate with objective
k k+1
function f(x ) at iteration k, the search defines a new estimate x at iteration k+1 as:
x k +1 = x k + x k ( 106 )
in which x = the correction to the parameter. The goal of the search is to select x
k k
so the estimates move toward the parameter that yields the minimum value of the
objective function. One correction does not, in general, reach the minimum value, so
this equation is applied recursively.
The gradient method, as used in the program, is based upon Newton's method.
Newton's method uses the following strategy to define x :
k
100
The objective function is approximated with the following Taylor series:
df ( x k ) ( x k +1 x k ) 2 d 2 f ( x k )
f ( x k +1 ) = f ( x k ) + ( x k +1 x k ) + ( 107 )
dx 2 dx 2
k+1
in which f(x ) = the objective function at iteration k; and df()/dx and
2 2
d f()/dx = the first and second derivatives of the objective function,
respectively.

k+1 k+1
Ideally, x should be selected so f(x ) is a minimum. That will be true
k+1
if the derivative of f(x ) is zero. To find this, the derivative of Equation
107 is found and set to zero, ignoring the higher order terms. That yields
df ( x k ) d 2 f (x k )
0= + ( x k +1 x k ) ( 108 )
dx dx 2
This equation is rearranged and combined with Equation 106, yielding
df ( x k )
x k = 2 dx k ( 109 )
d f (x )
dx 2
2 2
The program uses a numerical approximation of the derivatives df()/dx and d f()/dx
at each iteration k. These are computed as follows:

k k
Two alternative parameters in the neighborhood of x are defined as x 1 =
k k k
0.99x and x 2 = 0.98x , and the objective function value is computed for
each.
Differences are computed, yielding 1 = f(x 1) f(x ) and 2 = f(x 2)

k k k
k
f(x 1)
The derivative df()/dx is approximated as 1, and d f()/dx is

2 2
approximated as 2 - 1. Strictly speaking, when these approximations

are substituted in Equation 109, this yields the correction x in Newton's
k
method.
As implemented in the program, the correction is modified slightly to incorporate HEC

staff experience with calibrating the models included. Specifically, the correction is
computed as:
x k = 0.01 Cx k ( 110 )
in which C is as shown in Table 23.

k+1
In addition to this modification, the program tests each value x to determine if, in
k+1 k k+2
fact, f(x ) < f(x ). If not, a new trial value, x is defined as
x k + 2 = 0.7 x k + 0.3 x k +1 ( 111 )

k+2 k
If f(x ) > f(x ), the search ends, as no improvement is indicated.
101
Table 23. Coefficients for correction in the univariant gradient search.

2 - 1 1 C
1
>0 0.5
2
<0 >0 50
0 -33
If more than a single parameter is to be found via calibration, this procedure is

applied successively to each parameter, holding all others constant. For example, if
Snyders Cp and tp are sought, Cp, is adjusted while holding tp at the initial estimate.
Then, the algorithm will adjust tp, holding Cp at its new, adjusted value. This
successive adjustment is repeated four times. Then, the algorithm evaluates the last
adjustment for all parameters to identify the parameter for which the adjustment
yielded the greatest reduction in the objective function. That parameter is adjusted,
using the procedure defined here. This process continues until additional
adjustments will not decrease the objective function by at least 1%.
Nelder and Mead Algorithm

The Nelder and Mead algorithm searches for the optimal parameter value without
using derivatives of the objective function to guide the search. Instead this algorithm
relies on a simpler direct search. In this search, parameter estimates are selected
with a strategy that uses knowledge gained in prior iterations to identify good
estimates, to reject bad estimates, and to generate better estimates from the pattern
established by the good.
The Nelder and Mead search uses a simplexa set of alternative parameter values.
For a model with n parameters, the simplex has n+1 different sets of parameters.
For example, if the model has two parameters, a set of three estimates of each of the
two parameters is included in the simplex. Geometrically, the n model parameters
can be visualized as dimensions in space, the simplex as a polyhedron in the n-
dimensional space, and each set of parameters as one of the n+1 vertices of the
polyhedron. In the case of the two-parameter model, then, the simplex is a triangle in
two-dimensional space, as illustrated in Figure 36.
Parameter 2 Set 2
Set 1
Set 3
Parameter 1
Figure 36. Initial simplex for a 2-parameter model.
102
The Nelder and Mead algorithm evolves the simplex to find a vertex at which the
value of the objective function is a minimum. To do so, it uses the following
operations:
Comparison. The first step in the evolution is to find the vertex of the
simplex that yields the worst (greatest) value of the objective function
and the vertex that yields the best (least) value of the objective function.
In Figure 37, these are labeled W and B, respectively.
Reflection. The next step is to find the centroid of all vertices, excluding
vertex W; this centroid is labeled C in Figure 37. The algorithm then
defines a line from W, through the centroid, and reflects a distance WC
along the line to define a new vertex R, as illustrated Figure 37.
C R
W
xi (reflected)= xi (centroid)+ 1.0 [xi (centroid)- xi (worst)]

Figure 37. Reflection of a simplex.
Expansion. If the parameter set represented by vertex R is better than, or

as good as, the best vertex, the algorithm further expands the simplex in
the same direction, as illustrated in Figure 38. This defines an expanded
vertex, labeled E in the figure. If the expanded vertex is better than the
best, the worst vertex of the simplex is replaced with the expanded vertex.
If the expanded vertex is not better than the best, the worst vertex is
replaced with the reflected vertex.
C R E
xi (expanded) = xi + 2.0 [xi (reflected) - xi (centroid)]

Figure 38. Expansion of a simplex.
Contraction. If the reflected vertex is worse than the best vertex, but
better than some other vertex (excluding the worst), the simplex is
contracted by replacing the worst vertex with the reflected vertex. If the
reflected vertex is not better than any other, excluding the worst, the
simplex is contracted. This is illustrated in Figure 39. To do so, the
worst vertex is shifted along the line toward the centroid. If the objective
function for this contracted vertex is better, the worst vertex is replaced
with this vertex.
103
Con Cen
W
xi (contracted) = xi (centroid) - 0.5 [xi (centroid) - xi (worst)]

Figure 39. Contraction of a simplex.
Reduction. If the contracted vertex is not an improvement, the simplex is

reduced by moving all vertices toward the best vertex. This yields new
vertices R1 and R2, as shown in Figure 40.
R2
W R1
xi,j (reduced) = xi (best) + 0.5 [xi,j - xi (best)]

Figure 40. Reduction of a simplex.
The Nelder and Mead search terminates when either of the following criterion is
satisfied:
n (z zc )
2
< tolerance
j
( 112 )
j =1, j worst n 1
in which n = number of parameters; j = index of a vertex, c = index of

centroid vertex; and zj and zc = objective function values for vertices j
and c, respectively.
The number of iterations reaches 50 times the number of parameters.
The parameters represented by the best vertex when the search terminates are
reported as the optimal parameter values.
Constraints on the Search

The mathematical problem of finding the best parameters for a selected model (or
models) is what systems engineers refer to as a constrained optimization problem.
That is, the range of feasible, acceptable parameters (which systems engineers
would call the decision variables) is limited. For example, a Muskingum x parameter
that is less than 0.0 or greater than 0.5 is unacceptable, no matter how good the
resulting fit might be. Thus, searching outside that range is not necessary, and any
104
value found outside that range is not be accepted. These limits on x, and others
listed in Table 24, are incorporated in the search.
During the search with either the univariant gradient or Nelder and Mead algorithm,
the program checks at each iteration to ascertain that the trial values of the
parameters are within the feasible range. If they are not, the program increases the
trial value to the minimum or decreases it to the maximum before it continues.
In addition to these inviolable constraints, the program will also consider user-
specified soft constraints. These constraints define desired limits on the parameters.
For example, the default range of feasible values of constant loss rate is 0-300
mm/hr. However, for a watershed with dense clay soils, the rate is likely to be less
than 15 mm/hra much greater value would be suspect. A desired range, 0-15
mm/hr, could be specified as a soft constraint. Then if the search yields a candidate
parameter outside the soft constraint range, the objective function is multiplied by a
penalty factor. This penalty factor is defined as:
n
Penalty = 2 ( x
i =1
i ci + 1) ( 113 )
in which xi = estimate of parameter i; ci = maximum or minimum value for parameter

i; and n = number of parameters. This "persuades" the search algorithm to select
parameters that are nearer the soft-constraint range. For example, if the search for
uniform loss rate leads to a value of 300 mm/hr when a 15 mm/hr soft constraint was
specified, the objective function value would be multiplied by 2(300-15+1) = 572.
Even if the fit was otherwise quite good, this penalty will cause either of the search
algorithms to move away from this value and towards one that is nearer 15 mm/hr.
105
Table 24. Calibration parameter constraints.

Model Parameter Minimum Maximum
Initial and constant- Initial loss 0 mm 500 mm
rate loss
Constant loss rate 0 mm/hr 300 mm/hr
SCS loss Initial abstraction 0 mm 500 mm
Curve number 1 100
Green and Ampt loss Moisture deficit 0 1
Hydraulic conductivity 0 mm/mm 250 mm/mm
Wetting front suction 0 mm 1000 mm
Deficit and constant- Initial deficit 0 mm 500 mm
rate loss
Maximum deficit 0 mm 500 mm
Deficit recovery factor 0.1 5
Clark's UH Time of concentration 0.1 hr 500 hr
Storage coefficient 0 hr 150 hr
Snyder's UH Lag 0.1 hr 500 hr
Cp 0.1 1.0
Kinematic wave Lag 0.1 min 30000 min
Baseflow Mannings n 0 1
3 3
Initial baseflow 0 m /s 100000 m /s
Recession factor 0.000011 -
Muskingum routing K 0.1 hr 150 hr
X 0 0.5
Number of steps 1 100
Kinematic wave
N-value factor 0.01 10
routing
Lag routing Lag 0 min 30000 min
References
Diskin, M.H. and Simon, E. (1977). "A procedure for the selection of objective
functions for hydrologic simulation models." Journal of Hydrology, 34, 129-149.
Stephenson, D. (1979). "Direct optimization of Muskingum routing coefficients."

Journal of Hydrology, 41, 161-165.
US Army Corps of Engineers, USACE (1998). HEC-1 flood hydrograph package

users manual. Hydrologic Engineering Center, Davis, CA.
USACE (2000). HEC-HMS hydrologic modeling system users manual. Hydrologic

106
Chapter 10 Modeling Water-Control Facilities
CHAPTER 10
Modeling Water-Control Facilities

This chapter describes how the program can be used for modeling two types of
water-control facilities: diversions and detention ponds or reservoirs.
Diversion Modeling
Figure 41 is a sketch of a diversion. This diversion includes a bypass channel and a
control structure (a broad-crested side-channel weir). When the water-surface
elevation in the main channel exceeds the elevation of the weir crest, water flows
over the weir from the main channel into the by-pass channel. The discharge rate in
the diversion channel is controlled by the properties of the control structure. The
discharge rate in the main channel downstream of the control is reduced by the
volume that flows into the diversion channel.
Diversion channel
Diversion control
structure
Main channel
Figure 41. Illustration of a diversion structure.
A diversion is modeled in the same manner as a stream bifurcation by using a simple

one-dimensional approximation of the continuity equation. In that case:
Otmain = I t Otbypass ( 114 )

main
in which Ot = average flow passing downstream in the main channel during time
interval t ; It = average main channel flow just upstream of the diversion control
bypass
structure during the interval; and Ot = average flow into the by-pass channel
during the interval.
107
Setting Up a Diversion Model

The diversion model included in the program requires specifying the by-pass channel
flow as a function of the main channel flow upstream of the diversion. That is,
Equation 114 is represented as:
Otmain = I t f ( I t ) ( 115 )
in which f(It ) = the functional relationship of main channel flow and diversion channel
flow. The relationship can be developed with historical measurements, a physical
model constructed in a laboratory, or a mathematical model of the hydraulics of the
structure. For example, flow over the weir in Figure 41 can be computed with the
weir equation:
O = CLH 1.5 ( 116 )
in which O = flow rate over the weir; C = dimensional discharge coefficient that
depends upon the configuration of the weir; L = effective weir width; H = total energy
head on crest. This head is the difference in the weir crest elevation and the water-
surface elevation in the channel plus the velocity head, if appropriate. The channel
water-surface elevation can be computed with a model of open channel flow, such as
HEC-RAS (USACE, 1998a). For more accurate modeling, a two-dimensional flow
model can be used to develop the relationship.
Return Flow from a Diversion

The bypass channel may be designed to return flow to the main channel downstream
of the protected area, as illustrated in Figure 42. This is modeled in the program by
linking a diversion/bifurcation model with channel routing models for the main and
bypass channels and a confluence model at the downstream intersection of the
bypass and main channels, as shown in Figure 42. Chapter 8 provides more
information about modeling a confluence.
108
Diversion control
structure
Diversion channel
Property to be protected
Main channel
Confluence
Figure 42. Illustration of a diversion return flow.
Applicability and Limitations of the Diversion Model

The diversion model is applicable to water-control systems in which the necessary
relationship between main channel and bypass channel flow can be developed.
If a backwater condition can exist at the control structure (due to downstream

conditions such as the confluence of the diversion and the main channel), then an
unsteady-flow network model, such as UNET (USACE, 1997), must be used to
properly represent the complex hydraulic relationship.
Reservoir and Detention Modeling

A reservoir or detention pond mitigates adverse impacts of excess water by holding
that water and releasing it at a rate that will not cause damage downstream. This is
illustrated by the hydrographs shown in Figure 43. In this figure, the target flow
(release from detention pond) is 113 units. The inflow peak is as shown in the figure;
186 units. To reduce this peak to the target level, storage is provided. Thus the
volume of water represented by the shaded area is stored and then released
gradually. The total volume of the inflow hydrograph and the volume of the outflow
hydrograph (the dotted line) are the same, but the time distribution of the runoff is
altered by the storage facility.
Figure 44 is a sketch of a simple detention structure. The structure stores water

temporarily and releases it, either through the outlet pipe or over the emergency
spillway. The configuration of the outlet works and the embankment in this illustration
serves two purposes. It limits the release of water during a flood event, thus
protecting downstream property from high flow rates and stages, and it provides a
method of emptying the pond after the event so that the pond can store future runoff.
(Also, check that this change in timing of the peak does not adversely coincide with
flows from other parts of the basin.)
109
200
180 Inflow
160
140
120 Target flow

Discharge
100
80
60 Outflow
40
20
0
0 50 100 150 200 250
Time
Figure 43. Illustration of the impact of detention.
The reservoir outlet may consist of a single culvert, as shown in Figure 44. It may
also consist of separate conduits of various sizes or several inlets to a chamber or
manifold that leads to a single outlet pipe or conduit. The rate of release from the
reservoir through the outlet and over the spillway depends on the characteristics of
the outlet (in this case, a culvert), the geometric characteristics of the inlet, and the
characteristics of the spillway.
Top of embankment
Top of emergency
spillway / overflow
Orifice Outlet pipe (culvert)
Figure 44. Simple detention structure.

Outflow from an impoundment that has a horizontal water surface can be computed
with the so-called level-pool routing model (also known as Modified Puls routing
model). That model discretizes time, breaking the total analysis time into equal
intervals of duration t. It then solves recursively the following one-dimensional
approximation of the continuity equation:
S
I avg - O avg = ( 117 )
t
in which Iavg = average inflow during time interval; Oavg = average outflow during time
interval; S = storage change. With a finite difference approximation, this can be
written as:
110
I t + I t+1 - Ot + Ot+1 = S t+1 - S t ( 118 )

2 2 t
in which t = index of time interval; It and It+1 = the inflow values at the beginning and
th
end of the t time interval, respectively; Ot and Ot+1 = the corresponding outflow
values; and St and St+1 = corresponding storage values. This equation can be
rearranged as follows:
2 S t +1 2 St
( + O t +1 ) = ( I t + I t +1 ) + ( - Ot ) ( 119 )
t t
All terms on the right-hand side are known. The values of It and It+1 are the inflow
hydrograph ordinates, perhaps computed with models described earlier in the
th
manual. The values of Ot and St are known at the t time interval. At t = 0, these are
the initial conditions, and at each subsequent interval, they are known from
calculation in the previous interval. Thus, the quantity (2St+1 / t + Ot+1) can be
calculated with Equation 119. For an impoundment, storage and outflow are related,
and with this storage-outflow relationship, the corresponding values of Ot+1 and St+1
can be found. The computations can be repeated for successive intervals, yielding
values Ot+1 , Ot+2, ... Ot+n , the required outflow hydrograph ordinates.
Setting Up a Reservoir Model

To model detention with the program, the storage-outflow relationship for the existing
or proposed reservoir must be specified. The storage-outflow relationship (or
elevation-storage-outflow or elevation-area-outflow relationship) that is developed
and provided will depend on the characteristics of the pond or reservoir, the outlet,
and the spillway. Figure 45 illustrates how the relationship in a simple case might be
developed. HEC-RAS or other hydraulics software can develop storage-outflow
relationships for complex structures.
111
Elevation Elevation
Outlet flow Spillway flow
(b)
Elevation
Total flow
(c)
Elevation Elevation
Surface area Volume

(d) (e)
Volume
Total flow
(f)
Figure 45. Illustration of the procedure for defining storage-outflow a relationship.
Figure 45(a) is the pond outlet-rating function; this relates outflow to the water-
surface elevation in the pond. The relationship is determined with appropriate weir,
orifice, or pipe formulas, depending on the design of the outlet. In the case of the
configuration of Figure 44, the outflow is approximately equal to the inflow until the
capacity of the culvert is exceeded. Then water is stored and the outflow depends on
the head. When the outlet is fully submerged, the outflow can be computed with the
orifice equations:
O = KA 2 gH ( 120 )
in which O = flow rate; K = dimensional discharge coefficient that depends upon the
configuration of the opening to the culvert; A= the cross-sectional area of the culvert,
normal to the direction of flow; H = total energy head on outlet. This head is the
difference in the downstream water-surface elevation and the upstream (pond) water-
surface elevation.
112
Figure 45(b) is the spillway rating function. In the simplest case, this function can be
developed with the weir equation (Equation 116). For more complex spillways, refer
to EM 1110-2-1603 (1965), to publications of the Soil Conservation Service (1985),
and to publications of the Bureau of Reclamation (1977) for appropriate rating
procedures.
Figure 45(a) and (b) are combined to yield (c), which represents the total outflow
when the reservoir reaches a selected elevation.
Figure 45(d) is relationship of reservoir surface area to water-surface elevation; the

datum for the elevation here is arbitrary, but consistent throughout the figure. This
relationship can be derived from topographic maps or grading plans. Figure 45(e) is
developed from this with solid-geometry principles.
For an arbitrarily-selected elevation, the storage volume can be found in (e), the total
flow found in (c), and the two plotted to yield the desired relationship, as shown in (f).
With this relationship, Equation 116 can be solved recursively to find the outflow
hydrograph ordinates, given the inflow.
Applicability and Limitations of the Detention Model

The reservoir model that is included in the program is appropriate for simulating
performance of any configuration of outlets and pond. However, the model assumes
that outflow is inlet-controlled. That is, the outflow is a function of the upstream
water-surface elevation. If the configuration of the reservoir and outlet works is such
that the outflow is controlled by a backwater effect (perhaps due to a downstream
confluence), then the reservoir model should not be used. Instead, an unsteady-flow
network model, such as UNET (USACE, 1997) must be used to properly represent
the complex relationship of storage, pond outflow, and downstream conditions.
Further, if the reservoir is gated, and the gate operation is not uniquely a function of
storage, then a reservoir system simulation model, such as HEC-5 (USACE, 1998b),
should be used.
References
Bureau of Reclamation (1977). Design of small dams. U.S. Dept. of the Interior,
Washington, D.C.
Soil Conservation Service (1985). Earth dams and reservoirs, Technical Release 60.
USDA, Springfield, VA.
USACE (1965). Hydraulic design of spillways, EM 1110-2-1603. Office of Chief of

Engineers, Washington, DC.
USACE (1997). UNET one-dimensional unsteady flow through a full network of open
channel users manual. Hydrologic Engineering Center, Davis, CA.
USACE (1998a) HEC-RAS users manual. Hydrologic Engineering Center, Davis,

CA.
USACE (1998b). HEC-5 simulation of flood control and conservation systems user's
manual. Hydrologic Engineering Center, Davis, CA.
113
Appendix A CN Tables
APPENDIX A
CN Tables
The four pages in this section are reproduced from the SCS (now NRCS) report
Urban hydrology for small watersheds. This report is commonly known as TR-55.
The tables provide estimates of the curve number (CN) as a function of hydrologic
soil group (HSG), cover type, treatment, hydrologic condition, antecedent runoff
condition (ARC), and impervious area in the catchment.
TR-55 provides the following guidance for use of these tables:
Soils are classified into four HSGs (A, B, C, and D) according to their
minimum infiltration rate, which is obtained for bare soil after prolonged
wetting. Appendix A [of TR-55] defines the four groups and provides a
list of most of the soils in the United States and their group classification.
The soils in the area of interest may be identified from a soil survey
report, which can be obtained from local SCS offices or soil and water
conservation district offices.
There are a number of methods for determining cover type. The most
common are field reconnaissance, aerial photographs, and land use
maps.
Treatment is a cover type modifier (used only in Table 2-2b) to describe

the management of cultivated agricultural lands. It includes mechanical
practices, such as contouring and terracing, and management practices,
such as crop rotations and reduced or no tillage.
Hydrologic condition indicates the effects of cover type and treatment on

infiltration and runoff and is generally estimated from density of plant and
residue cover on sample areas. Good hydrologic condition indicates that
the soil usually has a low runoff potential for that specific hydrologic soil
group, cover type and treatment. Some factors to consider in estimating
the effect of cover on infiltration and runoff are: (a) canopy or density of
lawns, crops, or other vegetative areas; (b) amount of year-round cover;
(c) amount of grass or close-seeded legumes in rotations; (d) percent of
residue cover; and (e) degree of surface roughness.
The index of runoff potential before a storm event is the antecedent

runoff condition (ARC). The CN for the average ARC at a site is the
median value as taken from sample rainfall and runoff data. The curve
numbers in table 2-2 are for the average ARC, which is used primarily for
design applications.
The percentage of impervious area and the means of conveying runoff

from impervious areas to the drainage systems should be considered in
computing CN for urban areas. An impervious area is considered
connected if runoff from it flows directly into the drainage systems. It is
also considered connected if runoff from it occurs as shallow
concentrated shallow flow that runs over a pervious area and then into a
drainage system. Runoff from unconnected impervious areas is spread
over a pervious area as sheet flow.
114
SCS TR-55 Table 2-2a Runoff curve numbers for urban areas1
Cover description Curve numbers for hydrologic

soil group
Cover type and hydrologic condition Average percent A B C D

impervious area2
Fully developed urban areas

Open space (lawns, parks, golf courses, cemeteries,
etc.)3:
Poor condition (grass cover < 50%) . . . . . . . . . . . . . 68 79 86 89
Fair condition (grass cover 50% to 75%) . . . . . . . . . 49 69 79 84
Good condition (grass cover > 75%) . . . . . . . . . . . . 39 61 74 80
Impervious areas:
Paved parking lots, roofs, driveways, etc.
(excluding right-of-way) . . . . . . . . . . . . . . . . . . . . 98 98 98 98
Streets and roads:
Paved; curbs and storm sewers (excluding
right-of-way) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 98 98 98
Paved; open ditches (including right-of-way) . . . . 83 89 92 93
Gravel (including right-of-way) . . . . . . . . . . . . . . 76 85 89 91
.
Dirt (including right-of-way) . . . . . . . . . . . . . . . . . 72 82 87 89
Western desert urban areas:
Natural desert landscaping (pervious areas only)4 . . 63 77 85 88
Artificial desert landscaping (impervious weed
barrier, desert shrub with 1- to 2-inch sand
or gravel mulch and basin borders) . . . . . . . . . . . . 96 96 96 96
Urban districts:
Commercial and business . . . . . . . . . . . . . . . . . . . . . 85 89 92 94 95
Industrial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 81 88 91 93
Residential districts by average lot size
1/8 acre or less (town houses) . . . . . . . . . . . . . . . . . . 65 77 85 90 92
1/4 acre . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 61 75 83 87
1/3 acre . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 57 72 81 86
1/2 acre . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 54 70 80 85
1 acre . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 51 68 79 84
2 acre . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 46 65 77 82
Developing urban areas
Newly graded areas (pervious areas only,
no vegetation)5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 86 91 94
Idle lands (CN's are determined using cover types
similar to those in table 2-2c
1
Average runoff condition, and Ia = 0.2S.
2
The average percent impervious area shown was used to develop the composite CN's. Other assumptions are as follows: impervious areas
are directly connected to the drainage system, impervious areas have a CN of 98, and pervious areas are considered equivalent to open
space in good hydrologic condition. CN's for other combinations of conditions may be computed using figure 2-3 or 2-4.
3
CN's shown are equivalent to those of pasture. Composite CN's may be computed for other combinations of open space cover type.
4
Composite CN's for natural desert landscaping should be computed using figures 2-3 or 2-4 based on the impervious area percentage (CN
= 98) and the pervious area CN. The pervious area CN's are assumed equivalent to desert shrub in poor hydrologic condition.
5
Composite CN's to use for the design of temporary measures during grading and construction should be computed using figure 2-3 or 2-4,
based on the degree of development (imperviousness area percentage) and the CN's for the newly graded pervious areas.
115
SCS TR-55 Table 2-2b Runoff curve numbers for cultivated agricultural lands1

soil group
Cover type Treatment2 Hydrologic condition3 A B C D
Fallow Bare soil 77 86 91 94

Crop residue cover Poor 76 85 90 93
(CR)
Good 74 83 88 90
Row crops Straight row (SR) Poor 72 81 88 91

Good 67 78 85 89
SR + CR Poor 71 80 87 90
Good 64 75 82 85
Contoured (C) Poor 70 79 84 88
Good 65 75 82 86
C + CR Poor 69 78 83 87
Good 64 74 81 85
Contoured & terraced Poor 66 74 80 82
(C & T)
Good 62 71 78 81
C & T + CR Poor 65 73 79 81
Good 61 70 77 80
Small grain SR Poor 65 76 84 88

Good 63 75 83 87
SR + CR Poor 64 75 83 86
Good 60 72 80 84
C Poor 63 74 82 85
Good 61 73 81 84
C + CR Poor 62 73 81 84
Good 60 72 80 838
C&T Poor 61 72 79 82
Good 59 70 78 81
C & T + CR Poor 60 71 78 81
Good 58 69 77 80
Close-seeded SR Poor 66 77 85 89
or broadcast Good 58 72 81 85
legumes or C Poor 64 75 83 85
rotation Good 55 69 78 83
meadow C&T Poor 63 73 80 83
Good 51 67 76 80
1
2
Crop residue cover applies only if residue is on at least 5% of the surface throughout the year.
3
Hydrologic condition is based on combination of factors that affect infiltration and runoff, including (a) density and canopy of vegetative
areas, (b) amount of year-round cover, (c) amount of grass or close-seeded legumes in rotations, (d) percent of residue cover on the land
surface (good 20%), and (e) degree of surface roughness.
Good: Factors impair infiltration and tend to increase runoff.
Poor: Factors encourage average and better than average infiltration and tend to decrease runoff.
116
SCS TR-55 Table 2-2c Runoff curve numbers for other agricultural lands1

soil group
Cover type and hydrologic condition Hydrologic condition A B C D
Pasture, grassland, or range continuous Poor 68 79 86 89

forage for graving.2 Fair 49 69 79 84
Good 39 61 74 80
Meadow continuous grass, protected from 30 58 71 78

grazing and generally mowed for hay.
Brush brush-weed mixture with brush Poor 48 67 77 83

the major element.3 Fair 35 56 70 77
Good 304 48 65 73
Woods grass combination (orchard Poor 57 73 82 86

or tree farm).5 Fair 43 65 76 82
Good 32 58 72 79
Woods.6 Poor 45 66 77 83
Fair 36 60 73 79
Good 304 55 70 77
Farmsteads buildings, lanes, driveways, 59 74 82 86

and surrounding lots.
1
2
Poor: <50% ground cover or heavily grazed with no mulch.
Fair: 50 to 75% ground cover and not heavily grazed.
Good: >75% ground cover and lightly or only occasionally grazed.
3
Poor: <50% ground cover.
Fair: 50 to 75% ground cover.
Good: >75% ground cover.
4
Actual curve number is less than 30; use CN=30 for runoff computations.
5
CN's shown were computed for areas with 50% woods and 50% grass (pasture) cover. Other combinations of conditions may be
computed from the CN's for woods and pasture.
6 Poor: Forest litter, small trees, and brush are destroyed by heavy grazing or regular burning.
Fair: Woods are grazed but not burned, and some forest litter covers the soil.
Good: Woods are protected from grazing, and litter and brush adequately cover the soil.
117
SCS TR-55 Table 2-2d Runoff curve numbers for arid and semiarid rangelands1

soil group
Cover type Hydrologic A3 B C D

condition2
Herbaceous mixture of grass, weeds, and Poor 80 87 93

low-growing brush, with brush the Fair 71 81 89
minor element. Good 62 74 85
Oak-aspen mountain brush mixture of oak brush, Poor 66 74 79

aspen, mountain mahogany, bitter brush, maple, Fair 48 57 63
and other brush Good 30 41 48
Pinyon-juniper pinyon, juniper, or both; Poor 75 85 89

grass understory. Fair 58 73 80
Good 41 61 71
Sagebrush with grass understory. Poor 67 80 85

Fair 51 63 70
Good 35 47 55
Desert shrub major plants include saltbrush, Poor 63 77 85 88

greasewood, creosotebush, blackbrush, bursage, Fair 55 72 81 86
palo verde, mesquite, and cactus. Good 49 68 79 84
1
2
Poor: <30% ground cover (litter, grass, and brush overstory).
Fair: 30 to 70% ground cover.
Good: >70% ground cover.
3
Curve numbers for group A have been developed only for desert shrub.
118
Appendix B Soil Moisture Accounting Model Details
APPENDIX B
Soil Moisture Accounting Model Details

This appendix includes additional description of features of the soil moisture
accounting model.
Time Interval Selection

Models included in the program rely on the solution of differential equations to
estimate watershed runoff. To solve the equations, the models use a finite-difference
approximation, as described in Chapter 6. A discrete time interval (t) is selected for
the approximation, and for this time interval, the program commonly uses the value
defined by the user in the control specifications. So, for example, if the control
specification calls for a 10-minute time interval, the curve number loss model is
applied to compute infiltration for successive 10-minute intervals, and the unit
hydrograph equations are solved to compute runoff hydrograph ordinates at 10-
minute intervals. For these cases, the time interval is user-specified and is constant.
To ensure accuracy of solution of SMA model equations, the program determines

and uses internally a computational time interval. This interval may be the user-
specified interval, or it may be a fraction of that value. In either case, the program
reports hydrograph ordinates at the user-specified interval. The time interval is
selected as follows:
1. HEC-HMS finds a minimum time interval for each storage volume with
potential to outflow, using procedures shown in Table 25.
2. HEC-HMS selects the minimum interval from Step 1. If the user-

specified value is less, it is used instead.
3. If the time interval calculated in Step 2 is larger than one-quarter of the

time required to fill the combined available canopy, surface and soil
profile storage, the interval is reduced to that value.
4. If the interval from Step 3 is greater than the precipitation data interval,
the computational interval is set equal the precipitation interval.
5. If the interval from Step 4 is greater than 12 hours, the computational

interval is reduced to 12 hours. If the interval is less than 1 minute, the
interval is increased to 1 minute.
6. If the interval from Step 5 is greater than the remaining time in the user-
specified interval, the computational interval is set equal the remaining
time.
7. If the interval from Step 6 is less than the remaining time in the user-
specified interval, the computational interval is adjusted so it is an even
divisor of the remaining time.
8. If the remaining time less the interval found in Step 7 is less than one
minute, the computational interval is set equal to the time remaining in
the user-specified interval.
119
Appendix B Soil Moisture Accounting Model Details
The time required to fill or drain storages varies throughout the simulation period, so
the program varies the computational time interval throughout the simulation. To do
so, it repeats these steps for each user-specified interval. So, for example, during
periods in which water is moving rapidly into and out of the storages in the SMA, the
program may select and use ten 1-minute computational intervals to account for soil
moisture fluxes during a 10-minute user-specified interval. However, as the
movement slows, the program may select a longer computational intervalperhaps
using two 5-minute computational intervals during the 10-minute user-specified
interval.
Table 25. Minimum time step for storages.

Storage Minimum Time Step
Canopy interception storage 1 CurCanStore
TimeStep =
4 PotEvapTrans
Calculated only if evapotranspiration losses can occur
and when the current canopy interception storage at the
beginning of the time step exceeds the nominal storage
volume.
Surface interception storage 1 CurSurfStore
TimeStep =
4 PotSoilInfl + PotEvapTrans
Calculated when potential evapotranspiration or
infiltration losses > 0, and CurSurfStore > 0.
Soil profile storage 1 CurSoilStore
TimeStep =
4 PotSoilPerc + PotEvapTrans
Calculated when percolation or evapotranspiration can
occur from the soil profile, and CurSoilStore > 0.0001
inches.
Groundwater storage 1 CurGw1Store
TimeStep =
4 PotGw1Perc
Calculated when percolation (loss) can occur from a
groundwater layer, and the current volume in a
groundwater layer > 0
1
TimeStep = RoutGw1Store
16
Calculated when the groundwater storage volume divided
by the linear reservoir routing coefficient > 0
Precipitation intensity 1 MaxCanStore + MaxSurfStore + MaxSoilStore
TimeStep =
4 PrecipTimeStep
Calculated when PrecipTimeStep > 0

Note:
1 TimeStep = time step for storage; CurCanStore = current canopy interception storage;
CurSurfStore = current surface interception storage; CurSoilStore = current soil profile
storage; MaxCanStore = maximum canopy interception storage; MaxSurfStore = maximum
surface interception storage; MaxSoilStore = maximum soil profile storage; CurGw1Store =
current groundwater storage; PotEvapTrans = potential ET; PotSoilInf = potential infiltration;
PotSoilPerc = potential percolation from soil profile; PotGw1Perc = potential percolation
from groundwater layer; RoutGw1Store = coefficient for groundwater linear reservoir model;
PrecipTimeStep = time step for specification of precipitation data.
120
Appendix C Glossary
APPENDIX C
Glossary
This glossary is a collection of definitions from throughout the technical reference
manual plus definitions of other pertinent terms. Many of the definitions herein are
from the electronic glossary available from the USGS internet website at
http://water.usgs.gov/wsc/wsc_glo.htm and the USBR website at
http://www.usbr.gov/cdams/glossary.html.
Term Definitions
Annual Flood
The maximum peak discharge in a water year.

Annual Flood Series
A list of annual floods.

Antecedent Conditions
Watershed conditions prevailing prior to an event; normally used to

characterize basin wetness, e.g., soil moisture. Also referred to as initial
conditions.
Area-Capacity Curve
A graph showing the relation between the surface area of the water in a
reservoir and the corresponding volume.
Attenuation
The reduction in the peak of a hydrograph resulting in a more broad, flat

hydrograph.
Backwater
Water backed up or retarded in its course as compared with its normal or

natural condition of flow. In stream gaging, a rise in stage produced by a
temporary obstruction such as ice or weeds, or by the flooding of the stream
below. The difference between the observed stage and that indicated by the
stage-discharge relation, is reported as backwater.
Bank
The margins of a channel. Banks are called right or left as viewed facing in
the direction of the flow.
121
Appendix C Glossary
Bank Storage
The water absorbed into the banks of a stream channel, when the stages rise
above the water table in the bank formations, then returns to the channel as
effluent seepage when the stages fall below the water table.
Bankfull Stage
Maximum stage of a stream before it overflows its banks. Bankfull stage is a

hydraulic term, whereas flood stage implies damage. See also flood stage.
Base Discharge
In the US Geological Survey's annual reports on surface-water supply, the

discharge above which peak discharge data are published. The base
discharge at each station is selected so that an average of about three peaks
a year will be presented. See also partial-duration flood series.
Baseflow
The sustained or fair weather flow in a channel due to subsurface runoff. In

most streams, baseflow is composed largely of groundwater effluent. Also
known as base runoff.
Basic Hydrologic Data
Includes inventories of features of land and water that vary spatially (topographic and
geologic maps are examples), and records of processes that vary with both place
and time. Examples include records of precipitation, streamflow, ground-water, and
quality-of-water analyses.
Basic hydrologic information is a broader term that includes surveys of the water
resources of particular areas and a study of their physical and related economic
processes, interrelations and mechanisms.
Basic-Stage Flood Series
See partial duration flood series.

Bifurcation
The point where a stream channel splits into two distinct channels.
Boundary Condition
Known or hypothetical conditions at the boundary of a problem that govern its

solution. For example, when solving a routing problem for a given reach, an
upstream boundary condition is necessary to determine condition at the
downstream boundary.
Calibration
Derivation of a set of model parameter values that produces the best fit to
observed data.
122
Appendix C Glossary
Canopy Interception
Precipitation that falls on, and is stored in the leaf or trunk of vegetation. The
term can refer to either the process or a volume.
Channel
An naturally or artificially created open conduit that may convey water. See
also watercourse.
Channel Storage
The volume of water at a given time in the channel or over the flood plain of
the streams in a drainage basin or river reach. Channel storage can be large
during the progress of a flood event.
Computation Duration
The user-defined time window used in hydrologic modeling.

Computation Interval
The user-defined time step used by a hydrologic model for performing

mathematical computations. For example, if the computation interval is 15
minutes and the starting time is 1200, hydrograph ordinates will be computed
at 1200, 1215, 1230, 1245, and so on.
Concentration Time
See time of concentration.

Confluence
The point at which two streams converge.

Continuous Model
A model that tracks the periods between precipitation events, as well as the
events themselves. Compare event-based model.
Correlation
The process of establishing a relation between a variable and one or more

related variables. Correlation is simple if there is only one independent
variable and multiple when there is more than one independent variable. For
gaging station records, the usual variables are the short-term gaging-station
record and one or more long-term gaging-station records.
Dendritic
Channel pattern of streams with tributaries that branch to form a tree-like pattern.
Depression Storage
The volume of water contained in natural depressions in the land surface,

such as puddles.
123
Appendix C Glossary
Detention Basin
Storage, such as a small unregulated reservoir, which delays the conveyance

of water downstream.
Diffusion
Dissipation of the energy associated with a flood wave; results in the

attenuation of the flood wave.
Direct Runoff
The runoff entering stream channels promptly after rainfall or snowmelt. Superposed
on base runoff, it forms the bulk of the hydrograph of a flood. The terms base runoff
and direct runoff are time classifications of runoff. The terms groundwater runoff and
surface runoff are classifications according to source. See also surface runoff
Discharge
The volume of water that passes through a given cross-section per unit time;
commonly measured in cubic feet per second (cfs) or cubic meters per second
3
(m /s). Also referred to as flow.
In its simplest concept discharge means outflow; therefore, the use of this term is not
restricted as to course or location, and it can be applied to describe the flow of water
from a pipe or from a drainage basin. If the discharge occurs in some course or
channel, it is correct to speak of the discharge of a canal or of a river. It is also
correct to speak of the discharge of a canal or stream into a lake, a stream, or an
ocean.
Discharge data in US Geological Survey reports on surface water represent the total
fluids measured. Thus, the terms discharge, streamflow, and runoff represent water
with sediment and dissolved solids. Of these terms, discharge is the most
comprehensive. The discharge of drainage basins is distinguished as follows:
Yield. Total water runout or crop; includes runoff plus underflow.
Runoff. That part of water yield that appears in streams.
Streamflow. The actual flow in streams, whether or not subject to

regulation, or underflow.
Each of these terms can be reported in total volumes or time rates. The
differentiation between runoff as a volume and streamflow as a rate is not accepted.
See also streamflow and runoff.
Discharge Rating Curve
See stage discharge relation.

Distribution Graph
A unit hydrograph of direct runoff modified to show the proportions of the

volume of runoff that occurs during successive equal units of time.
124
Appendix C Glossary
Diversion
The taking of water from a stream or other body of water into a canal, pipe, or
other conduit.
Drainage Area
The drainage area of a stream at a specified location is that area, measured

in a horizontal plane, which is enclosed by a drainage divide.
Drainage Divide
The rim of a drainage basin. See also watershed.

Duration Curve
See flow-duration curve for one type.

ET
See evapotranspiration.
Effective Precipitation
That part of the precipitation that produces runoff. Also, a weighted average of
current and antecedent precipitation that is "effective" in correlating with runoff.
Evaporation
The process by which water is changed from the liquid or the solid state into
the vapor state. In hydrology, evaporation is vaporization and sublimation
that takes place at a temperature below the boiling point. In a general sense,
evaporation is often used interchangeably with evapotranspiration or ET.
See also total evaporation.
Evaporation Demand
The maximum potential evaporation generally determined using an

evaporation pan. For example, if there is sufficient water in the combination
of canopy and surface storage, and in the soil profile, the actual evaporation
will equal the evaporation demand. A soil-water retention curve describes the
relationship between evaporation demand, and actual evaporation when the
demand is greater than available water. See also tension zone.
Evaporation Pan
An open tank used to contain water for measuring the amount of evaporation.
The US National Weather Service class A pan is 4 feet in diameter, 10 inches
deep, set up on a timber grillage so that the top rim is about 16 inches from
the ground. The water level in the pan during the course of observation is
maintained between 2 and 3 inches below the rim.
Evapotranspiration
Water withdrawn from a land area by evaporation from water surfaces and
moist soils and plant transpiration.
125
Appendix C Glossary
Event-Based Model
A model that simulates some hydrologic response to a precipitation event.

Compare continuous model.
Exceedance Probability
Hydrologically, the probability that an event selected at random will exceed a

specified magnitude.
Excess Precipitation
The precipitation in excess of infiltration capacity, evaporation, transpiration,

and other losses. Also referred to as effective precipitation.
Excess Rainfall
The volume of rainfall available for direct runoff. It is equal to the total rainfall
minus interception, depression storage, and absorption.
Falling Limb
The portion of a hydrograph where runoff is decreasing.

Field Capacity
The quantity of water which can be permanently retained in the soil in

opposition to the downward pull of gravity. Also known as field-moisture
capacity.
Field-Moisture Deficiency
The quantity of water, which would be required to restore the soil moisture to
field-moisture capacity.
Flood
An overflow or inundation that comes from a river or other body of water, and
causes or threatens damage. Any relatively high streamflow overtopping the
natural or artificial banks in any reach of a stream. A relatively high flow as
measured by either gage height or discharge quantity.
Flood Crest
See flood peak.

Flood Event
See flood wave.

Flood Peak
The highest value of the stage or discharge attained by a flood; thus, peak
stage or peak discharge. Flood crest has nearly the same meaning, but
since it connotes the top of the flood wave, it is properly used only in referring
to stagethus, crest stage, but not crest discharge.
126
Appendix C Glossary
Floodplain
A strip of relatively flat land bordering a stream, built of sediment carried by

the stream and dropped in the slack water beyond the influence of the
swiftest current. It is called a living flood plain if it is overflowed in times of
highwater; but a fossil flood plain if it is beyond the reach of the highest flood.
The lowland that borders a river, usually dry but subject to flooding. That
land outside of a stream channel described by the perimeter of the maximum
probable flood.
Flood Profile
A graph of elevation of the water surface of a river in flood, plotted as

ordinate, against distance, measured in the downstream direction, plotted as
abscissa. A flood profile may be drawn to show elevation at a given time,
crests during a particular flood, or to show stages of concordant flows.
Flood Routing
The process of progressively determining the timing and shape of a flood

wave at successive points along a river.
Flood Stage
The gage height of the lowest bank of the reach in which the gage is situated.
The term "lowest bank" is, however, not to be taken to mean an unusually low
place or break in the natural bank through which the water inundates an
unimportant and small area. The stage at which overflow of the natural
banks of a stream begins to cause damage in the reach in which the
elevation is measured. See also bankfull stage.
Flood Wave
A distinct rise in stage culminating in a crest and followed by recession to

lower stages.
Flood-Frequency Curve
A graph showing the number of times per year on the average, plotted as abscissa,
that floods of magnitude, indicated by the ordinate, are equaled or exceeded. Also, a
similar graph but with recurrence intervals of floods plotted as abscissa.
Floodway
A part of the floodplain otherwise leveed, reserved for emergency diversion of water
during floods. A part of the floodplain which, to facilitate the passage of floodwater, is
kept clear of encumbrances.
The channel of a river or stream and those parts of the floodplains adjoining the
channel, which are reasonably required to carry and discharge the floodwater or
floodflow of any river or stream.
Flow-Duration Curve
A cumulative frequency curve that shows the percentage of time that

specified discharges are equaled or exceeded.
127
Appendix C Glossary
Gaging Station
A particular site on a stream, canal, lake, or reservoir where systematic

observations of gage height or discharge are obtained. See also stream-
gaging station.
Ground Water
Water in the ground that is in the zone of saturation, from which wells,
springs, and groundwater runoff are supplied.
Groundwater Outflow
That part of the discharge from a drainage basin that occurs through the
ground water. The term "underflow" is often used to describe the
groundwater outflow that takes place in valley alluvium, instead of the surface
channel, and thus is not measured at a gaging station.
Groundwater Runoff
That part of the runoff that has passed into the ground, has become ground
water, and has been discharged into a stream channel as spring or seepage
water. See also base runoff and direct runoff.
Hydraulic Radius
The flow area divided by the wetted perimeter. The wetted perimeter does
not include the free surface.
Hydrograph
A graph showing stage, flow, velocity, or other property of water with respect
to time.
Hydrologic Budget
An accounting of the inflow to, outflow from, and storage in, a hydrologic unit,
such as a drainage basin, aquifer, soil zone, lake, reservoir, or irrigation
project.
Hydrologic Cycle
The continuous process of water movement between the oceans,

atmosphere, and land.
Hydrology
The study of water; generally focuses on the distribution of water and

interaction with the land surface and underlying soils and rocks.
Hyetograph
Rainfall intensity versus time; often represented by a bar graph.
128
Appendix C Glossary
Index Precipitation
An index that can be used to adjust for bias in regional precipitation, often
quantified as the expected annual precipitation.
Infiltration
The movement of water from the land surface into the soil.
Infiltration Capacity
The maximum rate at which the soil, when in a given condition, can absorb
falling rain or melting snow.
Infiltration Index
An average rate of infiltration, in inches per hour, equal to the average rate of
rainfall such that the volume of rain fall at greater rates equals the total direct
runoff.
Inflection Point
Generally refers the point on a hydrograph separating the falling limb from the
recession curve; any point on the hydrograph where the curve changes
concavity.
Initial Conditions
The conditions prevailing prior to an event. See also to antecedent

conditions.
Interception
The capture of precipitation above the ground surface, for example by

vegetation or buildings.
Isohyet
Lines of equal rainfall intensity.

Isohyetal Line
A line drawn on a map or chart joining points that receive the same amount of
precipitation.
Lag
Variously defined as time from beginning (or center of mass) of rainfall to

peak (or center of mass) of runoff.
Lag Time
The time from the center of mass of excess rainfall to the hydrograph peak.
Also referred to as basin lag.
129
Appendix C Glossary
Loss
The difference between the volume of rainfall and the volume of runoff.
Losses include water absorbed by infiltration, water stored in surface
depressions, and water intercepted by vegetation.
Mass Curve
A graph of the cumulative values of a hydrologic quantity (such as

precipitation or runoff), generally as ordinate, plotted against time or date as
abscissa. See also double-mass curve and residual-mass curve.
Maximum Probable Flood
See probable maximum flood.
Meander
The winding of a stream channel.

Model
A physical or mathematical representation of a process that can be used to

predict some aspect of the process.
Moisture
Water diffused in the atmosphere or the ground.

Objective Function
A mathematical expression that allows comparison between a calculated

result and a specified goal. In the program, the objective function correlates
calculated discharge with observed discharge. The value of the objective
function is the basis for calibrating model parameters.
Overland Flow
The flow of rainwater or snowmelt over the land surface toward stream
channels. After it enters a stream, it becomes runoff.
Parameter
A variable, in a general model, whose value is adjusted to make the model

specific to a given situation. A numerical measure of the properties of the
real-world system.
Parameter Estimation
The selection of a parameter value based on the results of analysis and/or

engineering judgement. Analysis techniques include calibration, regional
analysis, estimating equations, and physically based methods. See also
calibration.
130
Appendix C Glossary
Partial-Duration Flood Series
A list of all flood peaks that exceed a chosen base stage or discharge,
regardless of the number of peaks occurring in a year. Also called floods
above a base. See also basic-stage flood series.
Peak
The highest elevation reached by a flood wave. Also referred to as the crest.
Peak Flow
The point of the hydrograph that has the highest flow.

Peakedness
Describes the rate of rise and fall of a hydrograph.

Percolation
The movement, under hydrostatic pressure, of water through the interstices

of a rock or soil.
Precipitation
As used in hydrology, precipitation is the discharge of water, in liquid or solid

state, out of the atmosphere, generally upon a land or water surface. It is the
common process by which atmospheric water becomes surface or
subsurface water. The term precipitation is also commonly used to designate
the quantity of water that is precipitated. Precipitation includes rainfall, snow,
hail, and sleet, and is therefore a more general term than rainfall.
Probable Maximum Flood
The largest flood for which there is any reasonable expectancy in this climatic
era.
Probable Maximum Precipitation
The largest precipitation for which there is any reasonable expectancy in this
climatic era.
Rain
Liquid precipitation.
Rainfall
The quantity of water that falls as rain only. Not synonymous with
precipitation.
Rainfall Excess
See excess rainfall.
131
Appendix C Glossary
Rating Curve
The relationship between stage and discharge.

Reach
A segment of a stream channel.

Recession Curve
The portion of the hydrograph where runoff is predominantly produced from

basin storage (subsurface and small land depressions); it is separated from
the falling limb of the hydrograph by an inflection point.
Recurrence Interval
The average interval of time within which the given flood will be equaled or
exceeded once. When the recurrence interval is expressed in years, it is the
reciprocal of the annual exceedance probability.
Regulation
The artificial manipulation of the flow of a stream.

Reservoir
A pond, lake, or basin, either natural or artificial, for the storage, regulation,
and control of water.
Residual-Mass Curve
A graph of the cumulative departures from a given reference such as the

arithmetic average, generally as ordinate, plotted against time or date, as
abscissa. See also mass curve.
Retention Basin
Similar to detention basin but water in storage is permanently obstructed from

flowing downstream.
Return Period
See recurrence interval.
Rising Limb
Portion of the hydrograph where runoff is increasing.

Runoff
That part of the precipitation that appears in surface streams. It is the same
as streamflow unaffected by artificial diversions, storage, or other works of
man in or on the stream channels.
132
Appendix C Glossary
Saturation Zone
The portion of the soil profile where available water storage is completely
filled. The boundary between the vadose zone and the saturation zone is
called the water table. Note, that under certain periods of infiltration, the
uppermost layers of the soil profile can be saturated. See vadose zone.
SCS Curve Number
An empirically derived relationship between location, soil-type, land use,

antecedent moisture conditions and runoff. A SCS curve number is used in
many event-based models to establish the initial soil moisture condition, and
the infiltration characteristics.
Snow
A form of precipitation composed of ice crystals.

Soil Moisture Accounting
A modeling process that accounts for continuous fluxes to and from the soil
profile. Models can be event-based or continuous. When using a continuous
simulation, a soil moisture accounting method is used to account for changes
in soil moisture between precipitation events.
Soil Moisture
Water diffused in the soil, the upper part of the zone of aeration from which
water is discharged by the transpiration of plants or by soil evaporation. See
also field-moisture capacity and field-moisture deficiency.
Soil Profile
A description of the uppermost layers of the ground down to bedrock. In a

hydrologic context, the portion of the ground subject to infiltration,
evaporation and percolation fluxes.
Soil Water
See soil moisture.
Stage
The height of a water surface in relation to a datum.

Stage-Capacity Curve
A graph showing the relation between the surface elevation of the water in a
reservoir usually plotted as ordinate, against the volume below that elevation
plotted as abscissa.
Stage-Discharge Curve
A graph showing the relation between the water height, usually plotted as
ordinate, and the amount of water flowing in a channel, expressed as volume
per unit of time, plotted as abscissa. See also rating curve.
133
Appendix C Glossary
Stage-Discharge Relation
The relation expressed by the stage-discharge curve.

Stemflow
Rainfall or snowmelt led to the ground down the trunks or stems of plants.
Storage
Water artificially or naturally impounded in surface or underground reservoirs. The

term regulation refers to the action of this storage in modifying downstream
streamflow.
Also, water naturally detained in a drainage basin, such as ground water, channel
storage, and depression storage. The term drainage basin storage or simply basin
storage is sometimes used to refer collectively to the amount of water in natural
storage in a drainage basin.
Storm
A disturbance of the ordinary average conditions of the atmosphere which,

unless specifically qualified, may include any or all meteorological
disturbances, such as wind, rain, snow, hail, or thunder.
Stream
A general term for a body of flowing water. In hydrology the term is generally
applied to the water flowing in a natural channel as distinct from a canal.
More generally as in the term stream gaging, it is applied to the water flowing
in any channel, natural or artificial.
Stream Gaging
The process and art of measuring the depths, areas, velocities, and rates of
flow in natural or artificial channels.
Streamflow
The discharge that occurs in a natural channel. Although the term discharge
can be applied to the flow of a canal, the word streamflow uniquely describes
the discharge in a surface stream course. The term streamflow is more
general than runoff, as streamflow may be applied to discharge whether or
not it is affected by diversion or regulation.
Stream-Gaging Station
A gaging station where a record of discharge of a stream is obtained. Within

the US Geological Survey this term is used only for those gaging stations
where a continuous record of discharge is obtained.
Sublimation
The process of transformation directly between a solid and a gas.
134
Appendix C Glossary
Surface Runoff
That part of the runoff that travels over the soil surface to the nearest stream
channel. It is also defined as that part of the runoff of a drainage basin that
has not passed beneath the surface since precipitation. The term is misused
when applied in the sense of direct runoff. See also runoff, overland flow,
direct runoff, groundwater runoff, and surface water.
Surface Water
Water on the surface of the earth.

Tension Zone
In the context of the program, the portion of the soil profile that will lose water
only to evapotranspiration. This designation allows modeling water held in
the interstices of the soil. See also soil profile.
Time of Concentration
The travel time from the hydraulically furthermost point in a watershed to the
outlet. Also defined as the time from the end of rainfall excess to the
inflection point on the recession curve.
Time of Rise
The time from the start of rainfall excess to the peak of the hydrograph.
Time to Peak
The time from the center of mass of the rainfall excess to the peak of the
hydrograph. See also to lag time.
Total Evaporation
The sum of water lost from a given land area during any specific time by
transpiration from vegetation and building of plant tissue; by evaporation from
water surfaces, moist soil, and snow; and by interception. It has been
variously termed evaporation, evaporation from land areas,
evapotranspiration, total loss, water losses, and fly off.
Transpiration
The quantity of water absorbed and transpired and used directly in the
building of plant tissue, in a specified time. It does not include soil
evaporation. The process by which water vapor escapes from the living
plant, principally the leaves, and enters the atmosphere.
Underflow
The downstream flow of water through the permeable deposits that underlie a
stream and that are more or less limited by rocks of low permeability.
135
Appendix C Glossary
Unit Hydrograph
A direct runoff hydrograph produced by one unit of excess precipitation over

a specified duration. For example, a one-hour unit hydrograph is the direct
runoff from one unit of excess precipitation occurring uniformly over one hour.
Vadose Zone
The portion of the soil profile above the saturation zone.

Water Year
In US Geological Survey reports dealing with surface-water supply, the 12-

month period, October 1 through September 30. The water year is
designated by the calendar year in which it ends and which includes 9 of the
12 months. Thus, the year ended September 30, 1959, is called the 1959
water year.
Watercourse
An open conduit either naturally or artificially created which periodically or

continuously contains moving water, or which forms a connecting link
between two bodies of water. River, creek, run, branch, anabranch, and
tributary are some of the terms used to describe natural channels. Natural
channels may be single or braided. Canal and floodway are terms used to
describe artificial channels.
Watershed
An area characterized by all direct runoff being conveyed to the same outlet. Similar
terms include basin, drainage basin, catchment, and catch basin.
A part of the surface of the earth that is occupied by a drainage system, which
consists of a surface stream or a body of impounded surface water together with all
tributary surface streams and bodies of impounded surface water.
136
Index
INDEX
algorithm, 8 limitations, 109

annual exceedance probability (AEP), 25 return flow, 108
application, 8 setup, 108
attenuation, 57 evaporation, 32
backwater, 88 floodplain storage, 88
baseflow model geographic information system (GIS), 60
recession, 72 green and ampt loss rate model
threshold, 73 concepts, 39
bifurcation model parameters, 39
concepts, 91 HEC-1, 1
limitations, 92 HEC-2, 76, 78
setup, 92 HEC-DSS, 9
calibration HEC-HMS
constraints, 104 application, 14
definition, 94 graphical user interface (GUI), 15
objective function, 97 models included, 12
plots, 98 on-line help, 2
procedure, 94 overview, 1
search methods, 100 setup, 14
channel-flow models users manual, 2
approximations, 75 web-site address, 2
concepts, 74 HEC-IFH, 37
equation solution methods, 76 HEC-RAS, 76, 78
parameters, 76 hypothetical-storm
Clark UH SCS, 30
concepts, 57 impervious surface, 35
parameters, 58 infiltration, 35
storage coefficient, 58 information
time-area histogram, 58 flood-runoff, 4
confluence model initial and constant loss rate model
concepts, 90 concepts, 36
limitations, 91 parameters, 36
setup, 91 initial condition, 7
constant monthly baseflow model kinematic wave model
concepts, 70 channel flow, 62, 84
parameters, 72 concepts, 60
deficit and constant loss rate model, 37 overland flow, 61
design flood setup, 66
concepts, 24 solution of equations, 63
design storm lag model
alternating block temporal distribution, 27 concepts, 83
area correction factor, 26 parameters, 84
concepts, 24 linear reservoir baseflow model
depths, 26 concepts, 72
duration, 31 parameters, 73
frequency based, 25 loss, 35
selection, 31 mean areal precipitation (MAP)
standard project (SPS), 28 arithmetic mean, 18
user-specified, 30 inverse-distance squared method, 20
detention model. see reservoir model isohyetal, 18
diversion model temporal distribution, 19
concepts, 107 Thiessen polygon, 18
137
Appendix C Glossary
ModClark model runoff process

concepts, 59 components of, 10
setup, 60 representation of, 11
model SCS CN loss rate model
categories of mathematical, 5 composite CN, 38
constituents of, 6 concepts, 37
forms of, 4 gridded, 38
primer, 4 parameters, 38
modified Puls model SCS UH
concepts, 77 basin lag, 55
parameters, 79 concepts, 55
storage-outflow relationship, 78 parameters, 55
Muskingum model SMA loss rate model
parameters, 81 flow component, 43
Muskingum-Cunge model parameters, 46
concepts storage component, 41
parameters, 87 Snyder UH
networks, 89 basin lag, 53
precipitation concepts, 53
design, 24 parameters, 54
historical, 16 peaking coefficient, 53
measurement, 16 standard hydrologic grid (SHG), 24, 60
minimum number of raingages, 22 standard project storm (SPS)
radar, 21 concept, 28
runoff computation requirements, 17 index rainfall, 28
radar temporal distribution, 29
HRAP grid, 23 transposition coefficient, 28
NEXRAD Information Dissemination state variable, 7
Service, 23 subcritical and supercritical flow, 89
Stage 3, 24 time of concentration, 56
use with HEC-HMS, 24 translation, 57
WSR-88D, 23 transpiration, 32
Z-R relationship, 23 UNET, 91, 92, 109, 113
recession baseflow model unit hydrograph
parameters, 72 unit hydrograph (UH)
reservoir model assumption of linearity, 51
concepts, 110 assumption of time-invariance, 51
limitations, 113 convolution, 50
setup, 111 of different duration, 51
storage-outflow relationship, 111 parametric, 52
roughness coefficient synthetic, 52
channel, 56 user-specified, 51
overland, 56
138

TechRefManualMar2000 HEC HMS

Uploaded by

Copyright:

Available Formats

TechRefManualMar2000 HEC HMS

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

TechRefManualMar2000 HEC HMS

Uploaded by

Copyright:

Available Formats

US Army Corps

Hydrologic Modeling System

Technical Reference Manual

Approved for Public Release Distribution Unlimited CPD-74B

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION

Hydrologic Engineering Center, HEC CPD-74B

11. SUPPLEMENTARY NOTES

12A. DISTRIBUTION / AVAILABILITY STATEMENT 12B. DISTRIBUTION CODE

13. ABSTRACT (Maximum 200 words)

14. SUBJECT TERMS 15. NUMBER OF PAGES

Technical Reference Manual

US Army Corps of Engineers

Phone (530) 756-1104

BY USING OR COPYING THIS SOFTWARE, USER AGREES TO ABIDE BY THE COPYRIGHT

DEFICIT AND CONSTANT LOSS MODEL........................................................................................... 37

Basic Equations of Open-Channel Flow ................................................................................ 74

CN TABLES .................................................................................................................... 114

SOIL MOISTURE ACCOUNTING MODEL DETAILS ................................................. 119

What's in this Manual?

Precipitation-specification options which can describe an observed

Models of naturally occurring confluences and bifurcations.

Models of water-control measures, including diversions and storage

A continuous soil-moisture-accounting model used to simulate the long-

The program also includes:

An automatic calibration package that can estimate certain model

Other Program References

The Hydrologic Modeling System HEC-HMS Users Manual (USACE,

The HEC-HMS on-line help system is a component of the computer

Organization of this Manual

Because of the importance of model calibration, Chapter 9 describes the automated

Table 1. Summary of contents of HEC-HMS Technical Reference Manual.

USACE (2000). HEC-HMS hydrologic modeling system users manual. Hydrologic

Planning and designing new hydraulic-conveyance and water-control

Operating and/or evaluating existing hydraulic-conveyance and water-

Preparing for and responding to floods.

Regulating floodplain activities.

The program includes models in a third categorymathematical models. In this

Table 2. What is a mathematical model?

Table 3. Categorization of mathematical models (from Ford and Hamilton, 1996)

Solving the Constituents

Models and Computer Programs

Programs may be classified broadly as those developed for a specific set of

Ford, D.T., and Hamilton, D. (1996). "Computer models for water-excess

Meta Systems (1971). Systems analysis in water resources planning. Water

Woolhiser, D.A, and Brakensiek, D.L. (1982). "Hydrologic system synthesis."

stemflow & Land

infiltration overland flow

Representation of the Runoff Process

Figure 2. Typical representation of watershed runoff.

Synopsis of Included Models

Models that compute runoff volume.

Models of direct runoff, including overland flow and interflow.

Models of channel flow.

Table 4. Runoff-volume models.

Table 5. Direct-runoff models.

Table 6. Baseflow models.

Table 7. Routing models.

Program Setup and Application

1. Start a new project.

2. Create gage data.