REPORT DOCUMENTATION

Public reporting burden for this collect ion of information is estimated to average 1 hour
and completing and reviewing the collection
gathering and maintaining the data needed,
collection of information, including suggestions for reducing this burden, to Washington
Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and

1. AGENCY USE ONLY (Leave b/artk)

PAGE

Form

Approved

OMB

No. 0704-0188

per response, including the time for reviewing instructions, searching existing data sources,
of information. Send comments regarding this burden estimate or any other aspect of this
Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson
Budget, Paperwork Reduction Project (0704-0188), Washington, DC 20503.

2. REPORT DATE

3. REPORT TYPE AND DATES COVERED

January 1996

Handbook May 1991 - July 1995

4. TITLE AND SUBTITLE

5. FUNDING NUMBERS

Handbook for Transformation of Datums, Projections, Grids

and Common Coordinate Systems

QE5113UD01

6. AUTHOR(S)
See
Preface

7. PERFORMING

ORGANIZATION

NAME(S) AND ADDRESS(ES)

8. PERFORMING ORGANIZATION
REPORT NUMBER

U.S. Army Topographic Engineer Center

7701
Telegraph Road
Alexandria, VA 22315-3864

9.SPONSORING/MONITORINGAGENCY

11. SUPPLEMENTARY

TEC-SR-7

NAME(S) AND ADDRESS(ES)

19. SPONSORING/Monitoring
AGENCY REPORT NUMBER

NOTES

Beneficial comments or information that may be of use in improving this document

should be addressed to the Standards Division of the Digital Concepts & Analysis
Center at the above address.
12a. DISTRIBUTION/AVAILABILITY

STATEMENT

12b. DISTRIBUTION

Approved for public release; distribution is unlimited.

13. ABSTRACT

CODE

UNLIMITED

(Maximum200words)

This document provides Army organizations and agencies with general guidance on
selecting the appropriate methods for shifting between local geodetic datums and the
World Geodetic System (WGS), and for converting Cartesian and map projection
coordinates to and from geodetic coordinates. This guidance is provided to aid the
Army community in selecting datum shift algorithms, in developing, selecting, and
maintaining software using these algorithms, and in implementing this software to
support operational units. Equations are furnished for map projections and datums
commonly used within the Army, and references are provided for other, less commonly
encountered, map projections and datums.

14. SUBJECT TERMS

Cartography
Seodesy
7.

15. NUMBEROFPAGES

Datums
Coordinates

SECURITYCLASSIFICATION
OFREPORT

Unclassified
.
.
..
.-.--.---.
NtiN (540-01-2f30-5500

18.

Projections

SECURITYCLASSIFICATION
OFTHISPAGE

Unclassified

170
16. PRICE CODE
19.

sEcURITYCLASSIFICATIC)N
OFABSTRACT

Unclassified

20. LIMITATION OFABSTRACT

Unlimited
Standard
Form
Prescribed
298 102

298

(Rev.
2-89)

by ANSI Std. 239-18

TABLE OF CONTENTS
PAGE
LIST OF FIGURES
LIST OF TABLES
EXECUTIVE SUMMARY
PREFACE
1.
1.1

1.2
1.3
2.
2.1
2 .1.1
2 .1.2
2.2
2.3

ix
xi
xiii

SCOPE
Scope
Applicability
Application guidance

REFERENCED DOCUMENTS
Government documents
Specifications, standards and handbooks
Other government documents, drawings and
publications
Non-government publications
Order of precedence

2
2
2
2

1
1

3.
DEFINITIONS AND UNITS
3.1
Acronyms
3.2
Terms
3.2.1
Convergence of the meridian (Y)
3 .2.2
Coordinate
3 .2.3
Datum
3 .2.3.1
Horizontal datum
3 .2.3.2
Vertical datum
3 .2.4
Earth-fixed
3 .2.5
Elevation (orthometric height, H)
3 .2.6
Ellipsoid
3 .2.7
Equator
3.2.8
Equipotential surface
3.2.9
Geocentric Cartesian coordinates
3.2.10
Geodetic coordinates (geodetic position)
3.2.11
Geodetic height (ellipsoidal height, h)
Geodetic latitude ($)
3.2.12
Geodetic longitude (~)
3.2.13
3.2.14
Geographic coordinates
3.2.15
Geoid
Geoid separation (N)
3.2.16
3.2.17
Grid reference system
3.2.18
Map projection
3.2.19
Map scale
3.2.20
Meridian
3.2.21
Military Grid Reference System (MGRS)
3.2.22
Orthometric height
Parallel
3.2.23
3.2.24
Prime (initial) meridian
Reference ellipsoid
3.2.25
3.2.26
Scale factor (projection)
3.3
Units
iii

4
4
5
5
5
5
5
5
6
6
6
6
6
6
6
6
6
6
7
7
7
7
7.
7
7
7
7
7
8
8
8
8
8
8

TABLE OF CONTENTS
---

PAGE

3.4
3.5
3 .5.1
3.5.2

Sign conventions
Unit conversion factors
Degrees and radians
Specifying the unit of angular measure

8
8
9
9

FUNDAMENTAL CONCEPTS
Introduction
Reference surfaces
Reference ellipsoid
Geoid
Relationships among topography, the geoid,
and the reference ellipsoid
4.3
Earth-fixed coordinate systems
4.3.1
Cartesian coordinates (X, y, z)
Cartesian coordinate system/reference ellipsoid
4.3.2
relationship
First
eccentricity and flattening
4.3 .2.1
4.3 .2.2
Ellipsoid parameters
4.3.3
Geodetic coordinates (~, ~, h)
Latitude and longitude limits
4.3 .3.1
4.3.4
Coordinate conversion
4.3 .4.1
Geodetic to Cartesian coordinate conversion
4.3 .4.2
Cartesian to geodetic coordinate conversion
4.3 .4.2.1 Finding ~
4.3 .4.2.2 Finding @
4.3 .4.2.3 Calculating h
Representation of Geodetic Coordinates
4.4
Height Relationships
4.5
4.5.1
Elevations
Geoid separation
4.5.2
The relationships among H, h, and N
4.5.3
4.5.3.1
Notational use of H and h

10
10
11
11
11
11

5.
5.1
5.2
5.2.1
5.2.2
5.3
5.3.1
5.3.2
5.3.3

19
19
19
19
19
20
20
20

4.
4.1
4.2
4.2.1
4.2.2
4.2.3

5.4
5.5

GEODETIC SYSTEMS AND DATUMS

Introduction
Geodetic (horizontal) datums
Background
Geodetic datums and WGS 84
Vertical datums and elevations
Background
Vertical datums and mean sea level (MSL)
Relationship between local, vertical and horizontal
datums
World Geodetic System 1984 (WGS 84)
WGS 84 referenced elevations

DATUM SHIFTS
6.
Introduction
6.1
6.1.1
Shifting between two local datums
Shifting from NAD 27 to NAD 83
6.1.1.1
Seven-parameter geometric transformation
6.2
Transformation to WGS 84 Cartesian coordinates
6.2.1
iv

13
13
13
15
15
15
15
15
16
16
16
16
17
18
18
18
18
18
18

20
20
21
22
22
22
22
23
23

TABLE OF CONTENTS
6.2.2
6.2.3
6.2.4

Transformation to LGS Cartesian coordinates

Parameter Values
Accuracy of seven-parameter transformation

6.3

Three-parameter (AX, AY, AZ) geometric

transformation
Transformation to WGS 84 Cartesian coordinates
Three-step method: Transformation to
WGS 84 geodetic coordinates
Transformation to local geodetic coordinates
Three-step method: Transformation to
local geodetic coordinates
Molodensky shifts
AX, AY, AZ shift values
Accuracy of ~r AY, AZ shift parameters
Local datum distortion
WGS 72 to WGS 84 transformation
Selecting a conversion method
Direct WGS 72 to WGS 84 transformation
Conversion equations
Estimated errors (6@,~~,6~)
Two-step WGS 72 to WGS 84 transformation
Approximating geodetic heights for datum
transformation
Multiple Regression Equations
Vertical datum shifts
Satellite-derived WGS 72 elevations based
on measured WGS 72 geodetic heights
Survey-derived elevations
Unlisted datums

6.3.1
6.3.1.1
6.3.2
6.3.2.1
6.3.3
6.3.4
6.3.5
6.3.6
6.4
6.4.1
6.4.2
6.4.2.1
6.4.2.2
6.4.3
6.5
6.6
6.7
6.7.1
6.7.2
6.8

7.
MAP PROJECTIONS
7.1
Introduction
7.1.1
The map projection process
7.1.2
Properties of projections
7.1.3
Mapping equations
7.1.4
Conformal projection
7.1.5
Scale factor
7.1.6
Map scale
7.1.7
Convergence of the meridian
7.1.8
Information sources
7.2
Mercator projection
7.2.1
Meridians and parallels
7.2.2
Mercator mapping equations
7.2.2.1
Finding (X,y)
7.2.2.2
Point scale factor and convergence
of the meridian
7.2.2.3
Finding (~,~)
7.2.3
Accuracy
7.2.4
Area of coverage
7.3
Transverse Mercator (TM) projection
v

PAGE
24
24
24
24
25
25
26
26
26
28
28
29
29
30
30
30
32
32
32
33
33
33
33
33
34
34
34
34
34
34
34
35
35
35
35
35
36
36
37
37
38
38
38

TABLE OF CONTENTS

7 .3.1
7 .3.2
7 .3.2.1
7 .3.2.2
7 .3.2.2.1
7.3 .2.2.2
7 .3.2.3
7 .3.2.4
7 .3.2.5
7.3 .2.6
7 .3.3
7 .3.4
7.4
7.4.1
7 .4.1.1
7.4.1.2
7.4.2
7 .4.3
7.4.3.1
7 .4.3.2
7 .4.3.3
7 .4.4
7.4.5
7.5
7 .5.1
7 .5.1.1
7 .5.1.2
7.5.2
7 .5.2.1
7.5.2.2
7 .5.3
7 .5.4
7 .5.5
7.6
7 .6.1
7.6.1.1
7 .6.1.2
7 .6.1.3
7 .6.1.4
7 .6.1.5
7.6.2
7 .6.3
7.7
7.7.1
7 .7.2

EKE
38
Meridians and parallels
39
Transverse Mercator mapping equations
39
Finding (X,Y)
40
Finding (O,?L)
40
Finding the footpoint latitude (~1)
41
Finding (~,~) from @l
42
Point scale factor as a function of
x and @l
42
Point scale factor as a function of
($, k)
42
Convergence of the meridian in terms of
x and ~1
42
Convergence of the meridian in terms of
($, k)
Accuracy
43
Area of coverage
43
Universal Transverse Mercator (uTM) projection
43
UTM zones
44
Finding the UTM zone and central meridian
45
Non-standard width UTM zones
45
Reference source
45
UTM equations
45
Finding UTM coordinates (Xu~~,YU~~)
45
Computing geodetic coordinates
46
Discontinuity
46
Accuracy
46
Area of coverage
.
46
Lambert Conformal Conic projection
47
Lambert Conformal Conic with two standard parallels
47
Finding (X,y)
48
Finding (@,k) with two standard parallels
49
Lambert Conformal Conic with one standard parallel
49
Finding (x/y)
50
parallel
Finding ($,~) with one standard
50
Stalefactor and convergence
50
Accuracy
50
Area of coverage
51
Polar Stereographic projection
51
Polar Stereographic mapping equations
51
Finding (x,Y)
51
Finding (~,~)
53
Alternate method for finding @
54
Finding the point scale factor
54
Finding the convergence of the meridian
54
Accuracy
54
Area of coverage
54
projection
Universal Polar Stereographic (UPS)
55
mapping
equations
universal Polar Stereographic (UPS)
55
Finding ($,k)
vi

TABLE OF CONTENTS

2AGE
7.7.3
7.7.4
7.7.5
7.8
7.8.1
7.8.2
7.9
7.10
7.11

Finding point scale factor and convergence

of the meridian
Accuracy
Area of coverage
The U.S. Military Grid Reference System (MGRS)
MGRS coordinates in the UTM area
MGRS coordinates in the UPS area
The World Geographic Reference System (GEOREF)
Non-standard grids
Modification to map projections

SELECTIONS OF TECHNIQUES, SOFTWARE DEVELOPMENT ,

AND TESTING
The general conversion process
8.1
8.1.1
Application of the general conversion process
Procedural examples
8.1.2
Example 8.1 Horizontal geodetic coordinate-based
8.1.2.1
datum shifts
8.

8.1.2.2
8.1.2.3
8.2
8.2.1
8.2.2
8.2.3
8.2.3.1
8.2 .3.2
8.3
8.3.1
8.3.2
8.3 .2.1
8.3.3
8.3.4
8.3.5
8.4
8.4.1

8.4.2

8.4.3
8.4.4

Example 8.2

WGS 84 Cartesian coordinate to local

latitude, longitude, and elevation

57
57
57
57
57
58
63
63
63
66
66
70
70
70
71

Example 8.3

Local latitude, longitude, and elevation

to WGS 84 latitude, longitude, and
elevation
Choosing a datum shift method
Parameter availability
Accuracy
Implementation issues
preferred shift methods
Software testing
Error Analysis
System model
Error estimates
Error types
Reformulation
Error propagation
Example
Numerical examples
Example 8.4 Convert WGS 84 geodetic coordinates
to Universal Transverse Mercator
coordinates in the NAD 27 reference
system using a three-step datum shift
Example 8.5

Example 8.6
Example 8.7

72
73
74
74
75
76
76
77
77
78
79
79
80
81
85

85

Convert WGS 84 geodetic coordinates

to geodetic coordinates in the NAD 27
reference system using the standard
Molodensky method

88

Convert WGS 84 geodetic coordinates

to Mercator projection coordinates

90

Convert Mercator projection

vii

TABLE OF CONTENTS
PAGE

8.4.5

Example 8.8

coordinates in the WGS 84 reference

system to WGS 84 geodetic latitude
and longitude

90

Convert Universal Transverse

Mercator coordinates in the NAD 27
reference system to NAD 27 geodetic
latitude and longitude

92

8.4.6

Example 8.9

8.4.7

Example 8.10 Convert Lambert Conformal projection

(with two standard parallels) coordinates
in the WGS 84 reference system to WGS 84
96
geodetic latitude and longitude

8.4.8

Example 8.11 Convert geodetic latitude and

longitude on the International Ellipsoid
to Universal Polar Stereographic
coordinates

98

Example 8.12 Convert Universal Polar Stereographic

coordinates to geodetic latitude and
longitude

99

8.4.9

Convert WGS 84 latitude and

longitude to Lambert Conformal coordinates
94
using two standard parallels

101

List of test points

8.5
9.
9.1
9.2

NOTES
Intended use
Subject term (key word) listing

PAGE

APPENDICES
A.
B.
c.
D.
E.
F.

103
103
103

Reference Ellipsoid Parameters

Ax, AY, AZ Datum Shift Parameters
Datum List
WGS 84 Geoid Separation Computation
Seven-parameter Geometric Datum Shifts
old Hawaiian Datum with International Ellipsoid

viii

105
107
129
145
149
151

LIST OF FIGURES
PAGE
4.1

The relationship between the reference

ellipsoid, the geoid, and the physical
surface of the earth.

12

4.2

The geometric relationship between

Cartesian and geodetic coordinates.

14

6.1

Relationship between coordinate axis in a

sevenparameter geometric transformation.

23

6.2

Relationship between coordinate axis in a

three-parameter geometric transformation.

25

7.1

Meridians and parallels in the Mercator projection.

36

7.2

Meridians and parallels in the Transverse Mercator

Projection. & is the central meridian.

39

7.3

Meridians and parallels (dashed) and

a Universal Transverse Mercator Grid.

44

7.4

Meridians and parallels in the Lambert

Conformal Conic Projection.

47

7.5

Meridians and parallels in the Universal

Polar Stereographic Projection.

51

7.6

Meridians and parallels imposed on a

Universal Polar Stereographic grid.

55

7.7

UTM area grid for MGRS

59

7.8

UTM area grid for MGRS

60

7.9

North UPS area grid for MGRS

61

7.10

SouthUPS area grid for MGRS

62

8.1

A map projection to geodetic datum

conversion. Height information for H
and h is not included in the conversion.

66

8.2

Coordinate conversions within a datum.

67

8.3

Datum shifts.

68

8.4

The general conversion process.

69

8.5

System Model

84
ix

LIST OF FIGURES
PAGE
D.1

Coordinate System Associated with Geoid Separation Bi146

Linear Interpolation Method

LIST OF TABLES

E&E
3.1

Conversion Factors

6.1

Selection of WGS 72 to WGS 84 Transformation Method

30

7.1

Non-Standard Zone Limits

45

8.1

Accuracy Guidance for Equations

79

8.2

Probability Level Conversion Factors

81

A.1

Reference Ellipsoids

106

B.1

Ellipsoid Center Shift Transformation

Parameters

108

C.1

Countries and their Associated Datums

130

D.1

WGS 84 Geoid Separations

147

E.1

Seven Parameter Local Geodetic System to WGS 84

Datum Transformation

150

xi

Thisdocumentprovides
Army organizations
andagencies
withgeneral
guidance
on selecting
the
appropriate
methodsforshifting
betweenlocal
geodetic
datumsandtheWorldGeodetic
System
(WGS),andforconverting
Cartesian
andmap projection
coordinates
toandfromgeodetic
coordinates.
Thisguidance
isprovided
toaidtheArmy communityinselecting
datumshifi
algorithms,
indeveloping,
selecting,
andmaintaining
sofiware
usingthese
algorithms,
andin
implementing
this
sofiware
tosupport
operational
units.
Equations
arefurnished
formap
projections
anddatumscommonlyusedwithin
theArmy,andreferences
areprovided
forother,
less
commonlyencountered,
map projections
anddatums.
Beneficial
comments(corrections,
recommendations,
additions,
deletions)
orinformation
that
may beofuseinimproving
this
documentshould
beaddressed
totheStandards
Division
ofthe
Digital
Concepts
& Analysis
Center
(DCAC) at the Topographic Engineering Center (TEC).
Mail: U.S. Army Topographic Engineering
Road, Alexandria, VA 22315-3864.
Fax: ATTN:

Geospatial

Center, ATTN: CETEC-.GD-A

Information Division,(703)428-6834

Phone: (703) 428+761

...

Xlll

, 7701 Telegraph

PREFACE
This report was funded under the U.S. Army Topographic
Topographic Data Standards Program.

Engineering Centers (TEC) Digital

This report was co-authored at TEC by the following individuals:

Frederick Gloeckler
Richard Joy
Justin Simpson
Daniel Specht
The report is based on a draft of the Military Handbook, Datums, Projections, Grids and
Common Coordinate Systems, Transformation of (MIL-HDBK-600008),
and was revised during
the period July 1994 through July 1995. This Technical Report uses material from MIL-HDBK600008 co-authored by:
James Ackeret
Fred Esch
Chris Gard
Frederick Gloeckler
Daniel Oimoen
Juan Perez
MAJ Harry Rossander
Justin Simpson
Joe Watts
Tom Witte
This work was performed under the previous supervision of Juan Perez, and Richard A
Herrmann and the current supervision of John W. Hale, Chief, Standards Division and Regis J.
Orsinger, Director, Digital Concepts and Analysis Center.
Mr. Walter E. Boge was Director and COL Richard G. Johnson was Commander
Director of TEC at the time of publication of this report.

xv

and Deputy

1.

SCOPE

ScoDe. This Technical Report provides methods and

parameter values to shift positions between approximately 113
geodetic datums and provides methods for converting between
geodetic coordinates, Cartesian coordinates and map projection
coordinates. Guidance is provided on selecting methods
appropriate to the application, and on developing and testing
software to implement these conversions.
1.1

1.2 ADD licabilitv. The methods provided in this Technical

Report are used for a wide range of Army mapping, charting, and
positioning applications. Equations are furnished for the map
projections cormnonlyused within the Army. References are
provided for other map projections that occasionally may be
encountered. Special high accuracy applications such as
engineering, construction, and real estate boundary surveys are
outside the scope of this Technical Report. Appendix B provides
datum shift parameters for approximately 113 datums that are
currently available from the Defense Mapping Agency (DMA).
However, many more local datums exist. For guidance on datum
shift applications or for datum shift parameters not covered by
this Technical Report, contact TEC [address given in the preface] .
1.3 A~~lication auidknce. Transformation methods and
parameters should be appropriate to the application. A system
analyst should do a thorough evaluation of the system requiraents
before developing detailed specifications. This evaluation should
consider the intended use, data t~es, formats, and reqyired
accuracy of the output mapping, charting, and geodesy (MC&G) data,
as well as the source and accuracy of the input data. Also, any
hardware constraints must be considered. The methods presented in
this Technical Report can then be examined for how well they can
meet or be adapted to meet the requirements. Where more than one
method will satisfy a basic requirement, give priority to Armypreferred methods for standardization purposes. Additional
application guidance is furnished in Section 8.

2.

REFERENCED DOCUMENTS

2.1.1
ticms :~.
following specifications, standards, and handbooks
this document to the extent specified herein.

The
form a part of

SPECIFICATIONS. None
STANDARDS.
MIL-STD-2401

Department of Defense World Geodetic

System(WGS) (1994)

STANAG 2211
(Fifth Edition)

Geodetic Datums, ellipsoids, grids and

grid references (1991)

HANDBOOKS.

None

7nas ,
otheZ_Gc?wenlment documents ,
2.1.2
1ica~
. The following other Government documents, drawings, and
publications form a part of this document to the extent specified
herein. Unless otherwise specified, the issues are those cited in the
The date in parentheses indicates for each document the
solicitation.
edition that was used in preparation of this Technical Report.
Document ID/Agency

ZCi_tle

DMA

Department of Defense Glossary of

Mapping, Charting, and Geodetic Terms,
Fourth Edition. (1981)

DMA TM 8358.1

Datums, Ellipsoids, Grids, and

Grid Reference Systems, (1990)

DMA TM 8358.2

The Universal Grids: Universal

Transverse Mercator (UTM) and
Universal Polar Stereographic (UPS) .
(1989)

DMA Instruction 8000.1

with change 1, 4 Aug 92

Geodetic and Geophysical Sign

Geological Survey
Professional Paper 1395

Map Projections - A Working Manual.

(Snyder, 1987)

NOAA Manual NOS NGS 5

State Plane Coordinate

(Stem, 1989)

NSWC/DL TR-3624

Map Projection Equations. (Pearson,

1977)

Us.

Department of
Commerce, Coast and
Geodetic Survey,
Special Publication
No. 251

Conformal Projections in Geodesy and

Cartography. (Thomas, 1979)

Engineer Technical
Letter No. 1110-1-147

Engineering and Design Conversion to

North American Datum of 1983 (1990)

Federal Register
v. 55, no. 155, Friday,
August 10, 1990
Docket No. 900655-0165

Notice To Adopt Standard Method for

Mathematical Horizontal Datum
Transformation (NOAA, 1990)

Conventions and Fundamental

(1988)

Constants.

System of 1983.

Copies of; DMA Instruction 8000.1 and GS Professional Paper 1395 are
available from USGS, Branch of Distribution, Box 25286, Denver CO
80225. Copies of DoD Glossary of Mapping, Charting, and Geodetic
Terms; DMA TM 8358.1;
and DMA TM 8358.2 are available from the
Director, DMA Combat Support Center, ATTN: CCOR, 6001 MacArthur
Boulevard, Bethesda MD 20816-5001.
Copies of the NOAA Manual NOS NGS
5 and the Coast and Geodetic Survey Special Publication No. 251 are
available from National Geodetic Information Branch, N/CG174, National
Geodetic Survey, 1315 East-West Highway, Silver Spring MD 20910.
Copies of NSWC/DL TR 3624 (Accession Document Number = AD A037 381 on
paper, or GIDEP #E1512353 on cartridge) are available on from Defense
Technical Information Center, Bldg. S, Cameron Station, Alexandria, VA
22304-6145.
Copies of ETL 1110-1-147 are available through the
Department of the Army, U.S. Army Corps of Engineers, Washington, D.C.
203140-1000.
Copies of STANAG 2211 and MIL-STD 2401 are available
from the Defense Printing Service, ATTN: DODSSP, 700 Robbins Ave, Bld
4D Philadelphia PA 19111.

2.2

Non-Government

publications.

Heiskanen, weikko A., and Helmut Moritz; Physical Geodesy;

Institute of Physical Geodesy, Technical University, Graz,
Austria; 1967.
Krakiwsky, Edward J.; Conformal Map Projections
Department of Surveying Engineering, University
Fredrickson, N.B., Canada; September, 1973.

in Geodesy;
of New Brunswick,

Rapp, Richard H.; Geometric Geodesy - Part I; Department of

Geodetic Science and Surveying, The Ohio State University,

Columbus, Ohio; 1984.

Rapp, Richard H.; Geometric Geodesy - Part 11; Department of
Geodetic Science and Surveying, The Ohio State University,
Columbus, Ohio; 1987.
2.3 Order of precedence. MIL-STD-2401 shall take
precedence for all matters concerning WGS 84 and datum shift
parameters. Nothing in this document, however, supersedes
applicable laws and regulations unless a specific exemption has
been obtained.

3.

DEFINITIONS AND UNITS

3.1 Acronvms. The acronyms used in this Technical Report

are defined as follows:

e.
f.
9*
h.
i.
j.
k.
1.
m.
n.
o.
P*
q.
r.

BIH
CONUS
CTP
DMA
DoD
DTED
ECEF
GPS
LGS
MGRS
MC&G
MRE
MSL
NAD 27
NAD 83
NGS
NGVD 29
NOAA

s.
t.

NOS
NSWC /DL

u.
v.
w.
x.
Y.

TM
UPS
UTM
WGS 72
WGS 84

::
::

3.2

Bureau International de lHeure

continental United States
Conventional Terrestrial Pole
Defense Mapping Agency
Department of Defense
Digital Terrain Elevation Data
earthcentered, earthfixed
Global Positioning System
local geodetic system
Military Grid Reference System
mapping, charting and geodesy
multiple regression equation
mean sea level
North American Datum 1927
North American Datum 1983
National Geodetic Survey
National Geodetic Vertical Datum of 1929
National Oceanic and Atmospheric
Administration
National Ocean Survey
Naval Surface Weapons Center/Dahlgren
Laboratory
Transverse Mercator
Universal Polar Stereographic
Universal Transverse Mercator
World Geodetic System 1972
World Geodetic System 1984

Terms.

3.2.1 Convergence of the meridian (Y). The angle between

the projection of the meridian at a point on a map and the
projection North axis (typically represented by the y axis of the
projection) . The convergence of the meridian is positive in the
clockwise direction.
3.2.2 Coordinate. Linear or angular quantities that
designate the position that a point occupies in a given reference
frame or system. Also used as a general term to designate the
particular kind of reference frame or system, such as Cartesian
coordinates or spherical coordinates.
3.2.3 Datum. Any numerical or geometrical quantity or set
of such quantities specifying the reference coordinate system used
for geodetic control in the calculation of coordinates of points
on the earth. Datums may be either global or local in extent. A
local datum defines a coordinate system that is used only over a

region of limited extent. A global datum specifies the center of

the reference ellipsoid to be located at the earths center of
mass and defines a coordinate system used for the entire earth.
3.2.3.1 Horizontal datum. A horizontal datum specifies the
coordinate system in which latitude and longitude of points are
located. The latitude and longitude of an initial point, the
azimuth of a line from that point, and the semimajor axis and
flattening of the ellipsoid that approximates
the surface of the
earth in the region of interest define a horizontal datum.

3.2.3.2 Vertical datum. A vertical datum is the surface to

which elevations are referred. A local vertical datum is a
continuous surface, usually mean sea level, at which elevations
are assumed to be zero throughout the area of interest.
3.2.4

Earth-fixed.

Stationary with respect to the earth.

3.2.5 Elevation. Vertical distance measured along the local

plumb line from a vertical datum, usually mean sea level or the
geoid, to a point on the earth. Often used synonymously with
orthometric height.
3.2.6 ElliDsoid. The surface generated by an ellipse
rotating about one of its axes. Also called ellipsoid of
revolution.
3.2.7 Eauator. The line of zero geodetic latitude; the
great circle described by the semi-major axis of the reference
ellipsoid as it is rotated about the semi-minor axis.
3.2.8 Eauipotential surface. A surface with the same
potential, usually gravitational potential, at every point; a
level surface.
3.2.9 Geocentric Cartesian coordinates. Cartesian
coordinates (X, Y, Z) that define the position of a point with
respect to the center of mass of the earth.
3.2.10 Geodetic coordinates (aeodetic ~osition) . The
quantities of latitude, longitude, and geodetic height (~, i ,h)
that define the position of a point on the surface of the earth
Also , imprecisely,
with respect to the reference ellipsoid.
called geographic coordinates (See also geographic coordinates) .
3.2.11 Geodetic heiaht (ellipsoidal heiaht, h) . The height
above the reference ellipsoid, measured along the ellipsoidal
normal through the point in question. The geodetic height is
positive if the point is outside the ellipsoid.

3.2.12 Geodetic latitude (~). The angle between the plane

of the Equator and the normal to the ellipsoid through the
computation point. Geodetic latitude is positive north of the
equator and negative south of the Equator
(See figure 4.2).
3.2.13 Geodetic lonaitude (L ). The angle between the plane
of a meridian and the plane of the prime meridian. A longitude
can be measured from the angle formed between the local and prime
meridians at the pole of rotation of the reference ellipsoid, or
by the arc along the Equator intercepted by these meridians. See
figure 4.2.
3.2.14 Geo
~. r hic
The quantities of latitude
and longitude that define the position of a point on the surface
of the earth. See also geodetic coordinates.
3.2.15 Geoid. The equipotential surface of the earths
gravity field approximated by undisturbed mean sea level of the
oceans. The direction of gravity passing through a given point on
the geoid is perpendicular -to this equipotential surface.
3.2.16 Geoid se~aration (N). The distance between the geoid
and the mathematical reference ellipsoid as measured along the
ellipsoidal normal. This distance is positive outside, or
negative inside, the reference ellipsoid. Also called geoidal
height; undulation of the geoid. See figure 4.1.
3.2.17 Grid reference svstem. A plane-rectangular
coordinate system usually based on, and mathematically adjusted
to, a map projection in order that geodetic positions (latitudes
and longitudes) may be readily transformed into plane coordinates
and the computations relating to them may be made by the ordinary
methods of plane surveying (see also 3.2.21) .
3.2.18
points on a
coordinates
established
constructed

Man nroiection. A function relating coordinates of

curved surface (usually an ellipsoid or sphere) to
of points on a plane. A map projection may be
by analytical computation or, less commonly, may be
geometrically.

3.2.19 Ma~ scale. The ratio between a distance on a map

and
the corresponding actual distance on the earths surface.
3.2.20 Meridian. A north-south reference line, particularly
a great circle through the geographical poles of the earth, from
which longitudes and azimuths are determined; or the intersection
of a plane forming a great circle that contains both geographic
poles of the earth, and the ellipsoid.
3.2.21 Militarv Grid Reference Svstem (MGRS). MGRS is a
system developed by DMA for usage with UTM and UPS projections.

3.2.22

Orthometric height.

See elevation.

3.2.23 Parallel. A line on the earth, or a representation

thereof, that represents the same latitude at every point.
3.2.24 Prime (initial) meridian. A meridian from which the
longitudes of all other meridians are reckoned. This fieridian, of
longitude 0, was traditionally chosen to pass through the
Greenwich observatory in Greenwich, England. For new refined
coordinate systems, the location of the prime meridian is defined
by the International earth Rotation Service, Paris, France.
3.2.25 Reference ellinsoid. ~ ellipsoid whose dimensions
closely approach the dimensions of the geoid; the exact dimensions
are determined by various considerations of the section of the
earths surface concerned. Usually a hi-axial ellipsoid of
revolution.
roiection) . A multiplier for reducing
3.2.26 Scale factor (K)
a distance in a map projection to the actual.distance on the
chosen reference ellipsoid.
3.3 Units. In this Technical Report all distances are
expressed in meters. Angles are expressed in degrees, radians or
the combination of degrees, minutes, and seconds.
3.4 Sign conventions. The sign conventions used in this
Technical Report are defined as follows:
a.

Elevation. Positive outside or above the vert~cal datum,

negative inside or below the datum.

b.

Geodetic heiaht. Positive outside the reference

ellipsoid, negative inside the reference ellipsoid.

c.

Geode tic latitude. Positive in the northern hemisphere,

negative in the southern hemisphere.

d.

Geodetic lonaitude.

Zero at the prime meridian, positive

00 to 180, eastward from the prime meridian, and negative

0 to 180, westward from the prime meridian.

e.

Geoid separation. Positive outside the reference

ellipsoid, negative inside the reference ellipsoid.

f.

Cartes ian Z.

positive in the northern hemisphere.

The standard unit of distance

3.5 Unit conversion factors.
measurement is the meter. However, MC&G products in many parts of
the world were produced using other units of measurement. The
user must ascertain the units used in non-standard products and
appropriately scale them for use in the equations given in this

technical report. Some commonly used conversion factors are

listed in Table 3.1.
TABLE 3.1
1 meter
1 International foot
1 U.S. Survey foot
1 Int1 Nautical mile
1 Int1 Statute mile
1 degree
1 degree
1 minute
6400 U.S . roils

sion Factors.
coriver

=
=
=
=
=
=
=
=
=
=

3.28083333333 U.S. Survey feet

3.28083989501 International feet
0.3048 meter (Exact)
1200/3937 meter (Exact)
0.30480060960 meter
1852 meters (Exact)
6076.10333333 U.S. Survey feet
6076.11548556 International feet
1609.344 meters (Exact)
5280 International feet (Exact)

=
=
=
=

60 minutes
3600 seconds
60 seconds
360 degrees

3.5 .1 Dearees and radians. A circle contains 360 degrees or

2~ radians. If an angle contains d degrees or, equivalently, r
radians, then d and r are related by
r=

zd
180

(3.1)

3.5.2 S~ecifvina the unit of anaular measure. In this

Technical Report the unit of angular measure will be specified if
it matters. A general rule of thumb is that angles must be in
radians in an equation that was derived using calculus and
trigonometry. Radians are the usual units in computer evaluation
of trigonometric functions.

4.

FUNDAMENTAL CONCEPTS

Introduction. h understanding of geodesy helps one

4.1
understand how to represent the locations of points on or about
the surface of the earth. Terms printed in boldface are defined
in Section 3.2.
a.

The location of a point is described by a set of

coordinates. Some types of coordinates, such as
Cartesian (see Section 4.3.1), uniquely identify
three-dimensional locations of points in space. Their
use is computationally efficient in many applications.
However, users often have difficulty visualizing the
relationship of Cartesian coordinates to the surface of
the earth. Geodetic coordinates (see Section
4.3.3) make visualization easier by providing
horizontal coordinates on the surface of a figure
approximating the shape of the earth and a vertical
(height) coordinate above or below the reference
surface.

b.

Geodetic coordinates are defined by a geodetic

reference system. A geodetic reference system
includes, but is not limited to, a reference surface
(usually an ellipsoid) and a set of parameters
positioning and orienting the reference surface with
respect to the earth. Over the years, many geodetic
reference systems have been developed for various
localities. These systems are often called local datums
and are not rigorously tied to each other.

c.

When applications require that two or more points be

located in a common reference system, transformation
between two coordinate systems may be needed. For
example: both weapon and target must be located in the
same reference system if the indirect fire computations
are to be valid. As the ranges of weapons and target
acquisition systems increased and positional data were
derived from multiple sources, commanders often were
confronted with the problem of using positions on more
than one reference system. With the advent of
satellite geodesy one can establish global geodetic
reference systems and take measurements that
allow local datums to be related to the global
system. The standard global reference system for
DoD is the World Geodetic System 1984 (WGS 84). The
process of relating one reference system to another is
called a datum shift, or datum transformation.

10

d.

Coordinates generally are represented alphanumerically (e.g. latitude 38 51 22.45N and

longitude 77 26 34.56W). Position data for an area
often are represented graphically, as on a map. The
planar representation on a map provides a distorted
view of features on the curved earths surface.
Geodesists have developed various mathematical methods
to minimize certain types of map distortions, depending
on the intended use of the map.
The process of
relating coordinates of points on a curved surface
(usually an ellipsoid or sphere) to coordinates of
points on a plane is called map projection.
The

preferred map projections for DoD are the Universal

Transverse Mercator (UTM), used over most of the
landmass, and Universal Polar Stereographic (UPS), used
in the polar regions.
4.2
Reference Surfaces. Reference surfaces are used as the
bases for geodetic coordinate systems and elevations.
4.2.1

Reference elli~soid. A reference surface called the

reference
ellipsoid
(mathematically described in Section 4.3.2)
is used for defining geodetic coordinates (Section 4.3.3) because
it is regularly shaped, is mathematically tractable, and can
approximate the surface of the earth. Since the earth is
flattened at the poles, an oblate ellipsoid fits the shape of the
earth better than does a sphere. llnellipsoid of revolution is
the reference surface most cormnonlyused in geodesy and is used
throughout this Technical Report. A geodetic coordinate has a
simply defined relationship with the reference ellipsoid and any
other geodetic coordinate because of the regularity and relative
simplicity of the reference ellipsoids shape. The parameters for
the commonly used reference ellipsoids are listed in Appendix A.
A reference ellipsoid is often chosen to fit the geoid.
4.2.2 Geoid. The geoid is a closed surface of constant
gravity potential approximated by mean sea level (MSL) and the
theoretical extension of MSL through land areas. For purposes of
this technical report, the geoid is considered a reference surface
from which elevations are measured. A reference ellipsoid may fit
the geoid; for example, the maximum difference between the WGS 84
geoid model and the WGS 84 ellipsoid found in Appendix D, Table
D.1 is 102 meters.
4.2.3
Relationshi~s
amon~ to~oara~hv, the aeoid, and the
reference elli~soid. The relationships among topography, the
geoid, and the reference ellipsoid are illustrated in figure 4.1.

11

tin

-1-1 -

0gx

z
+

-I-4

12

4.3
Earth-fixed coordinate
Sv stems.
There are many
coordinate systems used to represent positions on or near the
The coordinate systems used in this Technical Report are
earth.
A coordinate system is earth-fixed if the axes
all earth-fixed.

are stationary with respect to the rotating earth. Two primary,

earthfixed coordinate systems, Cartesian and geodetic, are
discussed in Sections 4.3.1 and 4.3.3, respectively. Conversions
from one to the other are presented in Section 4.3.4.
4.3.1
Cartesian coordinates (x, Y, z). The Cartesian
coordinate system used in this Technical Report is a right-hand,
rectangular, three-dimensional (X, Y, Z) coordinate system with
an origin at (0,0,0). When the origin is also at the mass center
of the earth, the coordinate system is said to be geocentric.
Cartesian coordinate Svstem/reference elli~ soid
relationshi~. The relationship of the Cartesian coordinate system
to the reference ellipsoid is shown in figure 4.2. The Z-axis,
the axis of rotation (semi-minor axis) of the ellipsoid, is nearly
parallel to the axis of rotation of the earth. The Z-coordinate
is positive toward the North pole. The XY plane lies in the
equatorial plane, the plane swept out by the semi-major axis as
the ellipse is rotated. The X-axis lies along the intersection of
the plane containing the pkime (initial) meridian and the
equatorial plane. The X-coordinate is positive toward the
intersection of the prime meridian and equator. The reference
ellipsoid satisfies the equation
4.3.2

(4.1)
Here, a and b, respectively, denote the semi-major and semi-minor
axes of the ellipsoid. This surface can be generated by rotating
an ellipse defined by
(4.2)

about the Z-axis. Ideally, for global applications, the Z-axis

coincides with the mean axis of rotation of the earth and the
origin is the center of mass of the earth. Often, in practice, an
ellipsoidal surface is fitted to a local area so that the origin
and axes vary slightly from ideal coincidence.

13

P
(X,y,z)

($,k,h)

Reference
h

Figure 4.2

(0,0,0)

The geometric relationship

and geodetic coordinates.

14

between

Cartesian

4.3.2.1 First eccentricitv and flattening. Two important

constants are the (first) eccentricity & and flattening f that are
related to a and b by
(4.3)

Frequently a and f are given, rather than a

defining parameters of a reference ellipsoid

and b, as the

4.3.2.2 Elli~soid ~arameter~. Appendix A contains a list of

reference ellipsoids and their parameters. These can be divided
into two categories: those, such as the Bessel Ellipsoid, that are
used to approximate the shape of the earth for a local area and
those, such as the WGS 84 Ellipsoid, that are used to yield a
global approximation to the shape of the earth.
4.3.3 Geodet ic coordinates ($, ~, h). Geodetic coordinates
consist of geodetic latitude ($), geodetic longitude (~), and
geodetic height (h). Their relationship to the reference
ellipsoid is shown in figure 4.2. For an ellipsoid that satisfies
equation 4.1, the normal SP intersects the ellipsoid at Q. The
angle between the normal SP and the equatorial (X-Y) plane is
called the geodetic
latitude
(~) of point P.
The meridian plane
containing point P is defined as the halfplane containing the Zaxis and point P. The angle between the prime meridian (X-Z)
plane and the meridian containing point P is the geodetic
Geodetic longitude is not defined when
longitude
(~) of point P.
P lies on the Zaxis. The distance from Q to P is called the
geodetic

height

4.3.3.1

(h).

Latitude and lonaitude limits.

Latitude limits:
-90 to +900
90 at the South Pole
0 at the Equator
+90 at the North Pole
Longi tude limits .
.
00 at the prime meridian
18 OW (-180) to 180E (+180)
coo rdinate conversion.

Two basic earthfixed

coordinate systems, geocentric Cartesian coordinates and geodetic
coordinates, have been introduced in Sections 4.3.1 and 4.3.3.
The following section explains how coordinates in one system can
4.3.4

15

be transformed to the other. In Section 6, transformations

between coordinates on two different reference ellipsoids are
discussed. Map projections are treated in Section 7 and the
general conversion process is summarized in Section 8.1.
4.3.4.1 Geodetic to Cartesian coordinate conversion. If
geodetic coordinates (~, ~, h) are known, then the Cartesian
coordinates are given by
x=

(RN + h) COS $ COS k

Y=

(RN + h) cos ~sin

b2

z=

~R~+h

(4.4)

sin~

where RN, the radius of curvature in the prime vertical, is given

by
(4.5)

41 - &2 sin2 ~
See Figure 4.2. Equation sets 4.4 and 4.5 are exact.
h from elevation (H) is discussed on section 6.5.
4.3.4.2

Calculating

Cartesian to aeodetic coordinate conversion.

4.3.4.2.1 Findina k. In the reverse case, when (X, Y, z)

are known but (~ ,~, h) are unknown,
for

X * O,all Y,
X=o,
X=o,
X=o,

the lonaitude (~) is aiven bv

~ = arctan X
x
Y>0,2L=
90E
90W
Y<O,?L=

Y=

O, k

(+900)
(-900r
is undefined

(4.6)

270 E)

Equation set 4.6 is exact. Take care to ensure that the

evaluation of the arctan function yields a value of ~ lying in the
desired quadrant.
4.3.4.2.2 Findina d). When (X, y) is not (O, O), the general
equations for ~ and h are more involved. The method presented for
finding @ is an iterative procedure of Bowring. Following Rapp
(1984, pp. 123-124), one finds an initial approximation to a
variable ~ given by

16

(4.7)

Once

Po

is obtained, substitute it for ~ in the equation

tan

0=

clP

Z + &12 b sins
a &

4.

(4.8)

Here &2 and &12 are given by

=

2f - f2

(4.9)

&2

&2

This approximation to @ is substituted into

tan~=(l-f)tan(ll

(4.10)

to give an updated approximation to ~. This procedure, of using

the latest approximation of ~ (equation 4.10) to produce an
updated approximation of $ (equation 4.8), and then using this new
value of @ to update ~, can be continued until the updated value
of @ is sufficiently close to the previous value. For many
purposes, one iteration is sufficient when using the Bowring
method, so that $ can be found directly from equation 4.8, using
the original value of PO. Rapp (1984, p. 124) states that for
terrestrial applications, one such iteration yields values
accurate to within 0.1 millimeter. In the polar case (X = Or
Y= 0, z), the latitude ($) is defined to be +90 (900N) for
positive Z values and -900 (900S) for negative Z values.
Calculating h. once $ has been calculated, in
non-polar areas h can be found from
4.3.4.2.3

h=

l/x2+Y2_R
Cos q)

(4.11)

In polar regions it is preferable to use

h=

sin @

R~+&QRN

where RN is defined in equation 4.5.

17

(4.12)

For many applications h is not required and need not be

calculated. ~ exception to this statement is the computation of
height or elevation for the Global Positioning System (GPS).

Note:

4.4 Remesentation of Geod etic Coordinates Latitude is

given before longitude. Latitude and longitude usually are
numerically represented as: decimal degrees (e.g. 44.44 N);
integer degrees and decimal minutes
(e.g.44 44.44S) ; or integer
degrees, integer minutes and decimal seconds (e.g. 44 44 44.44
E) . For clarity, numeric fields representing
degrees, minutes and
seconds should be ended by the appropriate
synibol, respectively
IIo , , and . Numeric fields with a value of zero should be
indicated by a 00, and not left blank.
The hemisphere of both
latitude and longitude must be shown, preferably by placing an N
or S at the end of numeric latitude and an E or W after
numeric longitude.
4.5

Height

Relationships.

4.5.1
Elevations.
Throughout this Technical Report, the
term elevation (H) is used to denote the distance of a point

above the geoid or vertical reference surface as measured along

the plumb line. A plumb lime follows the direction of gravity and
is perpendicular to all equipotenti.al surfaces of the Earths
gravity field that intersect it.
points lying outside the geoid
For the purposes of
are defined as having positive elevation.
this Technci.al Report, elevation
and orthometric height are

considered equivalent. In practice, the reference surface for

measuring elevations may not exactly coincide with the geoid.
4.5.2 Geoid se~ara tion. Geoid separation (N) is the
distance from a reference ellipsoid to the geoid, measured along
the ellipsoidal normal. This is equivalent to the geodetic height
(see Section 4.3.3) of a point on the geoid. Geoid separation is
positive when the geoid lies outside the ellipsoid.
4.5.3 The relationship amona H, h, and N. Geodetic height
(h), geoid separation (N), and elevation (H) are depicted in
figure

4.1 and are related

by equation

h =H+N

4.13.

(4*13)

4.5.3.1 Notational us e of H and h. This Technci.al Report

uses the classical notation of H for elevation and h for geodetic
height.

18

5.

GEODETIC SYSTEMS AND DATUMS

Introduction. A geodetic system serves as the

5.1
framework for determining coordinates of points with respect to
the earth. Modern global geodetic systems, such as the World
Geodetic System 1984 (WGS 84), (see Section 5.4), have been
established using techniques of satellite geodesy. ~ically,
they are defined by a geocentric Cartesian coordinate system with
the Z-axis along the mean rotation axis of the earth and the Xaxis through a longitude reference.
a.

DoD World Geodetic Systems have an associated reference

ellipsoid and a geoid model.
The geoid model is sometimes
used as a reference surface for elevations. World
Geodetic Systems provide a consistent framework for
determining geodetic positions.

b.

Local geodetic systems are established by a variety of

techniques. Typically, they consist of a geodetic datum
(see Section 5.2) for horizontal control and a vertical
datum (see Section 5.3) for elevations. Any given local
geodetic system covers only a small fraction of the
earths surface. Local geodetic datums and vertical
datums generally are
independent of each other.
.

5.2
Geodetic (horizontal) datums. A geodetic datum has a
earthfixed reference ellipsoid that may have been fit, in some
manner, to the surface of the earth in the area of interest.
Geodetic datums have parameters that define the size and shape of
the ellipsoid.
5.2.1 Background. Historically, local geodetic datums have
an origin, on the surface of the earth, that relates the geodetic
coordinate system to the ellipsoid. For these local datums, the
ellipsoid semi-minor axis generally does not coincide with the
earths mean rotation axis. However, the North American Datum
1983 (NAD 83) was developed using the same satellite geodesy
techniques used for global geodetic systems. This modern datum
has an origin at the center of the reference ellipsoid, and the
semi-minor axis coincides with the mean rotation axis of the
earth. Without reference to the word vertical, the word datum, as
used in this Technical Report, means geodetic (horizontal) datum.
5.2.2
Geodetic datums and WGS 84. Over the years, hundreds
of geodetic datums have been developed for various locations.
Currently, 113 local geodetic datums have been tied to WGS 84.
Coordinate transformations from local datums to WGS 84 are
discussed in Section 6.

19

5.3 Vertical datums and elevations. By definition, a

vertical datum is a surface of zero elevation. Elevations are
measured (positive upward) from the vertical datum. Ideally, a
vertical datum would closely approximate the geoid.
~.~.l Background. As a practical matter, it is impossible
to access the geoid surface directly for use as a vertical datum.
Historically, tide gage measurements were averaged over many years
to establish the local mean sea level (MSL) references for
The National Geodetic Vertical Datum of 1929
vertical datum.
(NGVD 29), which is based on tide gage measurements and precise
leveling surveys, is estimated to be within a few meters of the
geoid. There is greater uncertainty in the relationships between
the geoid and other local vertical datums.
5.3.2 Vertical datums and mean sea level (MSL). The exact
relationships between different vertical datums are unknown.
Therefore, when shifting datums, all elevations are considered
referenced to MSL regardless of the vertical datum used to
establish them. For Army applications, two exceptions to this
rule
a.

Conversion of elevations derived from WGS 72 geodetic

height measurements to WGS 84 coordinates (see Section
6.7.1).

b.

Civil works, engineering and construction applications

are outside the-scope of this Technical Report.

5.3.3 Relationshi~ between local vertical and horizontal

datums. Generally, vertical datums and horizontal datunlsare
independently defined. For example, NGVD 29 is the major vertical
datum, while the North American Datum 1927 (NAD 27) and NAD 83 are
the major horizontal datums used in North America.
World Geodetic Svstem 1984 (WGS 84). A world geodetic
system provides the basic reference frame, geometric figure and
gravimetric model for the earth, and provides the means for
relating positions on various local geodetic systems to an earth
centered, earth-fixed (ECEF) coordinate sYstem

5.4

standard DoD geodetic system. WGS

series of DoDdeveloped world
is a replacement for WGS 72, the
ECEF world geodetic system.

a.

WGS 84 is the current

84 is the latest in a
geodetic systems. It
previous DoD standard

b.

The origin of the WGS 84 coordinate system is the center

of mass of the earth. The WGS 84 Z-axis is parallel to
the direction of the Conventional Terrestrial Pole (CTP),
as defined by the Bureau International de lHeure (BIH).
The WGS 84 prime meridian is parallel to the BIH Zero
meridian. The X-axis is the intersection of the plane of
the WGS 84 reference meridian and the CTP equatorial

20

plane. The epoch is 1984.0. The Y-axis completes a

right-handed, earthfixed Cartesian coordinate system.
c*

The WGS 84 reference ellipsoid is a geocentric ellipsoid

of revolution. The WGS 84 coordinate system origin and
axes also serve as the geometric center and the X, Y, and
Z axes of the WGS 84 ellipsoid. WGS 84 ellipsoid
parameters are given in Appendix A.

d.

The WGS 84 geoid model can be used to convert between

elevations and geodetic heights (see Section 5.5) .

e.

Datum shift techniques and parameters needed to relate

positions on many local datums to WGS 84 were also
developed as part of WGS 84 (see Section 6 and Appendices
B and E).

f.

Only those portions of WGS 84 related to the

this Technical Report were described above.
comprehensive view of WGS 84 can be obtained
8350.2 (1991), 8350.2-A (1987), and 8350.2-B

purpose of
A
from DMA TRs
(1987).

--

WGS 84 referenced elevations. Sometimes elevations

must be generated from geodetic heights, without reference to a
local vertical datum. For example, satellite positioning systems,
like the Global Positioning System (GPS), establish coordinates
geometrically and cannot directly measure elevations. ~
elevation can be obtained from a WGS 84 geodetic height using:
3.5

H=

h WGS84

NWGS84

(5.1)

where NW~S84 is the value of WGS 84 geoid separation for the WGS 84
geodetic position of the point. Appendix D contains a method for
interpolating NW~S 84 from a table of grided values. Table D.1 is a
l-iSt Of NWGS84 on a 10 x 10 grid. It is useful for moderate
accuracy mapping and charting applications. TEC will furnish
qualified users with either digital 30 x 30 or 1 x 1 NWGS 84
tables or spherical harmonic expansion tables for survey and high
accuracy applications. Requests should be sent to the address in
the preface.

21

6.

DATUM SHIFTS

Several methods are available to shift

6.1 Introduction.
coordinates from one geodetic datum to another. To develop these
shifts, coordinates of one or more physical locations must be
for the local datum to WGS 84
known on both datums. ~ically,
shifts presented in this Technical Report, the WGS 84 coordinates
of local datum survey control points were determined using Doppler
satellite observations. TWO classes of datum shift methods are
geometric transformation and polynomial fitting techniques. Only
geometric transformati.ons are discussed here.
a.

The generalized geometric transformation model assumes

the origins of the two coordinate systems are offset from
each other; the axes are not parallel; and there is a
This sevenscale difference
between the two datums.

Data from
at least three wellspaced positions are needed to derive
a seven-parameter geometric transformation.

parameter model is discussed in Section 6.2.

b.

The number of parameters in a geometric transformation

model can be reduced. The commonly used threeparameter model neglects rotations between coordinate
systems axes and scale differences between datums. Only
the origins are offset. The threeparameter model is
discussed in Section 6.3. Data from one or more
positions are required to derive a three-parameter
geometric transformation.

c.

The WGS 72 to WGS 84 transformation, given in Section

6.4, is a special five-parameter geometric
transformation. When WGS 84 superseded WGS 72 as the
DoDpreferred reference system, this transformation was
created specifically to transform WGS 72 coordinates to
WGS 84 coordinates.

d.

Heights used for datum shifting are treated in section

6.5: Shifting of vertical datums is discussed in Section
6.7.

The datum
6.1.1 Shiftina between two local datum.
shift methods given in this Technical Report are for shifting from
a local datum to WGS 84 and from WGS 84 to a local datum. When
shifts are needed between two local datums, shift the coordinates
on the first local datum to WGS 84, then shift the WGS 84
coordinates to the second local datum.
6.1.1.1 Shiftina from NAD 27 to NAD 83.
The software package NADCON has been recommended as the standard
method for transformations between the North American Datum of
1927 (NAD27) and the North American Datum of 1983 (NAD 83) by The

22

Federal Geodetic Control Committee (Federal Register.) Corpscon,

a package incorporating NADCON, is mandated for the Army Corps of
Engineers (ETL 1110-1-147.) Because NADCON is a survey tool it is
beyond the scope of this Technical Report. Questions can be
addressed to the Survey Division of TEC at the address in the
preface, (703)355-2798, or to the National Geodetic Survey at
(301)713-3178.

6.2 Sevennarameter aeometric transformation.

The sevenparameter transformation, the most general transformation between

local and global Cartesian coordinates, is considered in
DMA TR 8350.2-A (1987). The underlying assumption for this seven
parameter model is that the local and WGS 84 Cartesian axes differ
by seven parameters: three rotational parameters (0, &, ~), a
scale change AS, and three origin shift parameters (AX, AY, AZ) .
See Figure 6.1.
I-7

&
x LGS

AY ~

XwGS 84
FIGURE 6.1.

Relationship between coordinate axes in a

seven-parameter geometric transformation.

6.2.1 Transformation to WGS 84 Cartesian coordinates.

the seven-parameter model, the local and WGS 84 Cartesian
coordinates are related by
x WGS 84 = x~GS + AX + @YLGS ~Z~GS + ASXLGS
Y WGS 84 = YLGS + AY - @XLGs + &ZLGs + ASYLGS
- &yLGs + ASzLGs
zWGS 84 = ZLGS + AZ + llfXLGs

23

For

(6.1)

The subscript LGS designates Local Geodetic System coordinates and

the subscript WGS 84 designates World Geodetic System 1984
coordinates.
6.2.2 Transformation to LGS Cartesian coordinates. The
equations for converting WGS 84 geocentric Cartesian coordinates
to local Cartesian coordinates are

x LGS

XWGS

84

Y LGS

YWGS

84

- AY

84

- AZ - ~XWGS 84 + ~ywGs 84 -

z LGS =Z

WGS

AX - OIYWGSg~ +

~zWGS

84

ASXWGS84

(6.2)

+ (.I)XWGS
84 - &ZwG~ 84 - ASYwG~ 84
SZWGS

84

6.2.3 Parameter values. The parameters AX, AY, AZ, ~, E,

~, and AS, for the European 1950 and ordinance Survey of Great
Britain 1936 datums are tabulated in Appendix E. Use of sevenparameter datum shifts for some applications in Europe and Great
Britain is proscribed by STANAG 2211. Since the values of the
rotational parameters @, &, and ware given in the table in arc
seconds, these values must be converted to radians before using
equation sets 6.1 and 6.2.
6.2.4 ~ccuracv of seven-~arameter transformation. The RMS
differences between WGS 84 surveyed and transformed coordinates
using the sevenparameter datum transformations are given in
Appendix E. For some datums, the accuracies for the sevenparameter transformation are not an improvement over those for the
three-parameter transformation. Equation sets 6.1 and 6.2 provide
good approximations to a rigorous orthogonal transformation when
@, &, ~, and AS are small.

6.3 Three-o arameter(AX, AY, AZ) aeometric transformation.

Frequently, it is not practical to determine the parameters for a
seven-parameter geometric datum shift or implement such a shift
method. Often, only an origin shift, three-parameter model is
used as shown in figure 6.2. The three-parameter datum
transformation satisfies the requirements of most mapping and
charting applications. The three-parameter model can be applied
directly to Cartesian coordinates (see Sections 6.3.1 and 6.3.2)
or used to compute shifts in geodetic coordinates via the
Molodensky equations (see Section 6.3.3)
a.

The datum shift parameters (Ax, AY, Az) are the

coordinates of the origin of the local reference
ellipsoid in the WGS 84 Cartesian coordinate system.

b.

Datum shift parameters for 113 local geodetic systems are

listed in Appendix B, Table B.1. The area covered by the
datum shift parameters, the local datum ellipsoid, and

24

the estimates of the errors in the datum shift parameters

also are given. Derivation of Ax, AY, and AZ datum shift
parameters is discussed in DMA TR 8350.2-A (1987).

Y LGS

x LGS
.

x WGS 84
FIGURE

6.2.

Relationship between coordinate axes in a

three-parameter geometric transformation.

6.3.1
Transformation to WGS 84 Cartesian coordinates. In
certain cases, AX, AY, and AZ datum shift parameters can be
applied directly to convert local geodetic system coordinates to
WGS 84 Cartesian coordinates as follows:
x WGS 84 = XLGS + AX
Y WGS 84 = YLGS + AY
zWGS 84 = ZLGS + AZ

(6.3)

Equation set 6.3 is a subset of equation set 6.1 with 63, ~, &, and
AS set to zero.
6.3.1.1 Three -ste~ method: Transformation to WGS 84 oeodetic
coordinates. Local geodetic coordinates can be shifted to WGS 84
coordinates in three steps:
a.
b.

Convert local geodetic coordinates to local geocentric

Cartesian coordinates using equation set 4.4. Ensure
that local ellipsoid parameters are used.
Shift the local geocentric Cartesian coordinates to
WGS 84 geocentric Cartesian coordinates using equation
set 6.3.

25

c.

Convert WGS 84 geocentric Cartesian coordinates to WGS

84 geodetic coordinates using the methods discussed in
Section 4.3.4.2. Ensure that WGS 1984 ellipsoid
parameters are used.

6.3.2 Transformation to local geodetic coordinates. To

convert WGS 84 Cartesian coordinates to the local geodetic system,
apply Ax, AY, and AZ datum shifts as follows

(6.4)

x LGS = XWGS ~~ - Ax
Y LGS - Y~Gs 84 - AY
zLGS =Z WGS 84 - AZ
Equation

set 6.4 is a subset

of equation

set 6.2 with

0,

~,

&, and

AS set to zero.
6.3.2.1

to local aeodetic
Three-ste~ method : Transformation
To
shift
WGS
84
geodetic
coordinates
to local
coordinates.
geodetic coordinates
a.

b.

c.

Convert WGS 84 geodetic coordinates to WGS 84 geocentric

Cartesian coordinates using equation set 4.4. Insure
that WGS 84 ellipsoid parameters are used.
Shift the WGS 84 geocentric Cartesian coordinates
to
local geocentric Cartesian coordinates using equation
set 6.4.
Convert the local geocentric Cartesian coordinates
to
local geodetic coordinates using a method of section
4.3.4.2.
Insure that local ellipsoid parameters
are
used.

Molodenskv
6.3.3
shifts.
The standard Molodensky method is
an approximation
to the threestep transformation
methods of
To use the Molodensw
methods to
Sections 6.3.1.1 and 6.3.2.1.

transform local geodetic coordinates to WGS 84 geodetic

coordinates, calculate A@, Ah, and Ah shifts using the standard
Molodensky formulas Rapp, 1987.

[-

sin $COS
&2 sin
+

LAx
Cos
w

+ sin$cos$(zN

[-sin ~Ax

- sin $sin LAY

0 Aa
+ &2 M sin2

+ cos ~AY]

[(N + Mcos

i- COS @AZ

$1

26

(1-f)Af]/[M + h]

(6.5)

Ah = cos @cos
-w

kAx + cos @sin LAY + sin @AZ

a(l - f)
Aa+
sin2 @ (Af)
w

in which
w2=l

&2

a(l -

M.

sin2 @

&2)

=2f-f2
&2
=

1-

The units of A@ and AL are radians.

set 4.5.

&2

Note that N = RN of equation

Note: Most applications do not require shifting geodetic heights,

so Ah does not need to be computed or applied. Geodetic heights,
which are used to compute datum shift parameters, are a special
construct that may not be well related to the local geodetic
system. Do not use Ah to shift local geodetic heights. If this
is a requirement, seek competent geodetic council.
a.

Apply the A$, AL, and Ah shifts as follows:

$WGS 84

LGS + OLGS

L GS

L(X + LLGS

84

(6.6)

hWGS 84 = h LGS + LGS

where 4hGsf LLGSI
AhLGs are computed from equation set
6.5 USing $LGS, ~LGS, hLGS, semimajor axis and flattening
for the local geodetic system. AX, AY, and AZ are found
in Appendix B. Aa and Af are calculated as follows;

27

Aa = aWGS 84 aLGS
Af = fwG~84 - fL~~
See section 6.5 for use of hL~S.

Note:
b.

The Molodensky
formulas can also be used
WGS 84 to the local geodetic system:

@LGS

o WGS 84 +

~WGS
h4GS

h LGS

hW~s84 + AhwGs84

from

(6.7)

84

?LLGS

GS 84 +

to convert

84

computed using $WGS 84,

h GS 84 h WGS 841 WGS 84 semi-major axis and flattening, and
the signs of Ax, AY, and Az, as given in Appendix B, are
reversed. Aa and Af are calculated as follows

where

$WGS

84f

&GS

84f

and

Aa = aLGS - aWGS
Af = fLGS -f WGS

hWGS

84

are

84
84

Note that ellipsoid parameters a and f are found in Table Al,

Appendix A.
6.3.4 AX, AY* and AZ shift values. Appendix B (Table B.1)
contains the AX, AY, and AZ values for the three-parameter methods
of converting local geodetic systems to WGS 84. The accuracies,
in terms of error estimates for AX, AY, and AZ, are given in
Appendix B. A further discussion of the accuracy of these
parameter shifts will be given in Section 6.3.5.
Accuracv of Ax, AY, and AZ shift ~arameters. Error
6.3.5
estimates for AX, AY, and AZ are tabulated in Appendix B. These
estimates include the errors associated with the Doppler stations
as well as the residual differences between transformed local
system coordinates and the reference WGS 84 coordinates used to
develop the datum shifts. (A Doppler station refers to a position
where the WGS 84 coordinates were obtained
processing
of TRANSIT satellite data.)

a.

by observing

and

For 38 datums, datum shifts were developed using a single

Doppler station. For these datums, no measure of
internal consistency of the datum shifts is calculable,
and DMA has assigned an accuracy value of 25 meters in
each coordinate. These error statistics do not reflect

28

any errors in the coordinates used to compute the datum

shift.
Datum shifts for eight datums were calculated from
sources other than Doppler stations for WGS 84 geodetic
coordinates. No error statistics are provided for these
datums.
Geodetic coordinate difference error estimates can be
approximated from rectangular coordinate errors using the
following spherical relationships:

(6.8)

Note that these relationships neglect any correlations

that may exist between the error estimates for Ax, 4Y
and Az.
6.3.6
~.
L
For some local datums, such
as the Australian Geodetic datums, one set of AX, AY, and AZ datum
shift parameters provides a consistent measure of the datum shifts
throughout the datum. For other datums, 4x, 4Y, and AZ shift
parameters for the whole local datum, called mean-value datum
shift parameters, poorly fit one or more regions within the datum.
Regional variation in the datum shift parameters may result from
using different, poorly connected surveys to establish the
geodetic control on the local datum; from rotation of the local
datum; and from scale differences. An extreme example of regional
datum shift variation is the 4Y component for the ARC 1950 datum.
The mean-value 4Y is -90 meters. However, 4Y ranges from -5 to 108 meters for different countries covered by the ARC 1950 datum.
Appendix B (Table B.1) contains regional 4x, 4Y, and AZ datum
shift parameters for 17 local datums. For these datums, one should
base the selection of mean or regional 4x, 4Y, and AZ shift
values on the accuracy required for the application. C)nlymean
value threeparameter shifts are available for the remaining
datums listed in Table B.1.

6.4
~GS 72 to WGS 84 transformation. WGS 84 has replaced
WGS 72 as the accepted geodetic reference system for most DoD
applications. The WGS 84 system, developed through a more
extensive set of satellite-derived and surface data than was
available at the start of WGS 72, is an improved geometric and
gravitational model of the earth. Whenever possible, WGS 72
coordinates should be converted to the WGS 84 reference system.

29

6.4.1 Selectina a conversion method. Before selecting a WGS

72 to WGS 84 conversion method, find the source of the WGS 72
coordinates. Select the conversion method using the guidance in
Table 6.1.
TABLE 6.1.

Selecting a WGS 72 to WGS 84 Transformation Method.

Source of WGS 72 Coordinates

Transformation Method

1.

Doppler

Direct

2.

Local

datum to WGS 72 transformation.

Local datum coordinates known.

Local datum to WGS 84

transformation
(6.2, 6.3, 6.6)

3.

Local datum to WGS 72 transformation.

Local datum coordinates unknown.

Two-Step (6.4.3)
Shift method known

4.

Local

5.

Unknown.

satellite

station.

(6.4.2)

Direct (6.4.2)
datum to WGS 72 transformation
using localized datum shift parameters
derived for nearby WGS 72 Doppler
station. Local datum coordinates unknown.
Direct (6.4.2)
Accuracy unknown

Direct WGS 72 to WGS 84 transformation. The direct

6.4.2
WGS 72 to WGS 84 transformation reflects fundamental changes
between the WGS 72 and WGS 84 systems. Changes include a shift in
coordinate system origin, a shift in longitudinal reference
(initial meridian), a change in system scale, and changes in
ellipsoidal parameters. For WGS 72 coordinates not originating in
the WGS 72 system (i.e., non-Dopplerderived coordinates) , take
care when employing the direct WGS 72 to WGS 84 transformation, as
any inaccuracies and uncertainties inherent in the WGS 72
coordinates will be directly transferred into the derived WGS 84
coordinates. These inaccuracies and uncertainties are generally a
result of local geodetic system to WGS 72 coordinate
transformations using mean rather than localized datum shifts.
Thus , every effort should be made to find the source of all WGS 72
coordinates before doing a direct WGS 72 to WGS 84 transformation.
6.4.2.1 Conversion eauations. It is important to note that
the relationship between the WGS 72 coordinates and the WGS 84
coordinates cannot be described by a three-parameter model (see
Section 6.3).

30

a. The equations for the direct WGS 72 to WGS 84 geodetic

coordinate transformation are

oWGS

84

WGS72 + A@

(6.9)

hWGS84 = hwGS72 + Ah
where
4.5 cos@WGS72 Af sin 2@W~Sy2
~
~
+
Q
WGS72

A@=

Ak = 0.554

(arc seconds)

(arc seconds)

Ah = 4.5 sin@WGS72

aWGS 72 Af sin2 @WGS72 - Aa + A~

(meters)

Af = 0.3121057 X 10-7
aWGS72

6378135

(meters)

Aa = 2.0

(meters~

Ar = 1.4

(meters)

Q =

~80

;3600

b. The equations for the direct WGS 84 to WGS 72 geodetic

coordinate transformation are:
= $WGS 84 + A~
hiGS 72 = ~WGS84 + ~
+ WGS 72

hWGS72

(6.10)

hwGSS4 + Ah

where

A~ =

Af Sin 2$WGS84

4.5 Cos~wGS 84

aWGS84 Q

A?L= -0.554

(arc seconds)

(arc seconds)

Ah = -4.5 sin$ - aWGS84 Af sin2@WGS84

Af = 0.3121057 X 10-7

31

Aa - Ar

(meters)

aWGS

84

6378137

(meters)

Aa = 2.o

(meters)

Ar = 1.4

(meters)

Q =

~80

:3600

based on the WGS 84 geoid model will be different from

elevations based on the WGS 72 geoid model due to differences in
the geoid models (see Section 6.7.1 for conversion).

Elevation

The estimated errors

6.4.2.2 Estimated errors (a@@&L
in a direct WGS 72 to WGS 84 transformation are
OA$ =~~=3m
CA

(6.11)

= 3mto4m

These error estimates do not include errors in the original WGS 72

coordinates. The accuracy estimate for ok reflects discrepancies
between WGS 72 and WGS 84 geoid models.
6.4.3
TWO-SteD WGS 72 to WGS 84 transformation.
If WGS
72 coordinates are known to have been derived from local geodetic
system coordinates using WGS 72 Ax, AY, and AZ datum shift
parameters,
the preferred method for transforming
them to the WGS
84 system is to do a two-step conversion.
First, do a WGS 72 to
local transformation
using the inverse of the transformation
originally used to shift local datum coordinates
to WGS 72. This
step will eliminate the inaccuracies
introduced using local to WGS
72 datum shifts.
Next, do a local geodetic system to WGS 84
conversion to produce coordinates within the WGS 84 system (see
Sections 6.2, 6.3, or 6.6).
Seppelin
(1974) provides mean value
local datum to WGS 72 datum shift parameters.
A?mroximatina
6.5
aeo detic heights for datum
Transformation.
When doing the Molodens~
approximation
(Section
6.3.3), ellipsoidal height (h) is often unavailable.
MSL height
(H) can be substituted
for h without introducing
errors in
horizontal position significant
to mapping and charting
applications.
The inpact of this substitution
is so small that
zero can be used in place of h, if H is unavailable.
NOTE : The generation
and usage of local datum ellipsoidal
heights are outside the scope of this Technical Report.
For such
applications,
consult TEC at the address in the foreward.

32

eauations (MRE).
for datum shifting is not recormnended.
6.6

Multi~le

recession

The use of MREs

Vertica 1 datum shifts. Because exact relationships

6.7
between vertical datums are unknown, shifting elevations between
different vertical datums should not be attempted, except for WGS
72 elevations computed from measured WGS 72 geodetic heights (see
Section 6.7.1.) otherwise, the following are assumed to be the
same:
::

Elevations on different local vertical datums.

Elevations computed from measured WGS 84 geodetic heights
and WGS 84 geoid separations.

It is estimated that errors resulting from this assumption will be

one to two meters. Should the accuracy be insufficient, contact
TEC at the address listed in the foreword for assistance.
~ elli riv
ions ba ed on measured
6.7.1
WGS 72 aeod etic he iahts. The WGS 84 geoid model i.s a closer
to the geoid than the older WGS 72 geoid model.
approximation
El~vations obtained from WGS 72 geodetic height measurements and
geoid separations may be improved by converting to WGS 84 using
H = HW~~VQ +

NW~~TQ + Ah

Nw~s 84

(6.12)

where HW~~72 is the elevation comPuted using the WGS 72 geoid,

N WGS72 is the WGS 72 geoid separation, and Ah is computed using
equation set (6.9).
6.7.2 Survev-derived elevat ions. Elevations obtained by
conventional leveling surveys, for points with WGS 72 geodetic
horizontal coordinates, should not be COnVerted to WGS 84
elevations. Simplified procedures to convert WGS 72-derived
elevations to WGS 84 are not recommended.
This Technical Report provides datum
6.8 Unlisted datws.
shift parameters between 113 local datums and WGS 84. Regional
shift parameters are also given for portions of many of these
datums. Contact TEC at the address given in the foreword for any
datums and their respective shift parameters that are not listed
in this Technical Report.

33

7.

MAP PROJECTIONS

7.1 Introduction. Map projections are used to represent the

earths features or points on a map or in a plane coordinate
system. A map projection is a systematic representation of a Dart
of the earths surface on a plane.
7.1.1 The man ~rojection ~rocess.
process can be done in three steps:

The map projection

(1) The positions of points on the surface of the earth ($,~,

h) are reduced to positions on the reference ellipsoid ($,
L,h=O).
(2) Next, a map projection is chosen and used to transform
the features geodetic coordinates ($, ~, h = O) into
planar map projection coordinates (x, y).

(3) Finally,

the map pro-jection coordinate= of the features

may be plotted, on a map, or stored for future use.

CJ-1
J.~.2

Properties

of ~roiections.

Once

the earths

surface

is projected on a plane, one can observe such properties as

distance, angle, direction, shape, and size. None of the
projections described in this Technical Report faithfully
preserves all these characteristics. Some type of distortion is
always introduced by a map projection. However, a given map
projection can preserve some of these properties, and if a
sufficiently small part of the earths surface is displayed, then
other properties can be represented
approximately.
Map
projections
presented in this Technical Report have areas of
coverage beyond which the pro-jection should not be used due to
increased distortion.

7.1.3 Ma~ning equations. The map projections discussed in

this Technical Report have mapping equations that are defined on a
reference ellipsoid and are functions of geodetic coordinates
(@, k). Geodetic height, h, does not enter into the projection
equations since h = O on a reference ellipsoid. The equations
that express projection coordinates, x and y, in terms of (~, ~)
are sometimes called mapping equations.
7.1.4

Conformal nroiection.
Each projection presented in
this Technical Report has the important property of being
conformal.
A conformal projection is angle preserving.
If two
lines on the ellipsoid meet at an angle of e, then in a conformal
projection
the image of these lines on the map meet at the same
angle, 0.

7.1.5 Scale factor. Two quantities associated with map

projections, which are often confused, are the scale factor of the
projection and the map scale. The scale factor of the pro-jection

34

is used in precise direction and distance calculations using a map

projections coordinates. The scale factor of a projection is the
ratio of arc length along a differentially small line in the plane
of the projection to arc length on the ellipsoid. This number
depends on both the location of the point and on the direction of
the line along which arc length is being measured. However, for
conformal projections, the scale factor of the projection is
independent of the direction of the line and depends only upon the
location of the point. Since all projections treated in this
Technical Report are conformal, the scale of the projection will
be referred to as the point scale factor (k).
7.1.6 MaQ scale. The map scale is the approximate
number
for converting map distances to terrestrial distances and is
Technically,
the map scale is the
usually published on the map.
ratio of a distance on the map to the corresponding
distance on
Although the map scale changes over the area of a
the ellipsoid.
map sheet, it is customary to assign a single value to denote the
This value i.s the true map scale along at least one
maps scale.
line on the map.
Usually, the map scale is given by a ratio l:n,
For
so that the larger the map scale, the smaller the value of n.
military maps of scale greater than 1:100,000 (n < 100,000), the
map scale changes only slightly over a map sheet, and the map
scale approxi-mates the ratio of distance on the map to distance on
the ellipsoid.
of the
Convergence of the meridian. The convergence
meridian
(y), often called grid convergence, is used to convert
between direction in the map projection (grid azimuth) and
geodetic azimuth. The convergence of the meridian at a point on a
map is the angle between the projection of the meridian at that
point and the projection North axis (typically represented by the
y axis of the projection). The convergence of the meridian is
positive in the clockwise direction. Convergence of the meridian
should not be confused with declination, which is the difference
between geodetic and magnetic north at the point.

7.1.7

7.1.8 ~.
Information on map projections
can be found in Krakiwsky (1973); NSWC/DL TR-3624 (1977); U.S.
Department of Commerce, Coast and Geodetic Survey, Special
publication No. 251 (1979); and Geological Survey Professional
Paper 1395 (1987).
7.2 Mercator ~roiection. The Mercator projection is a
nonformalprojection for which the point scale factor is one along
the equator. The equator lies on the line Y = O. This projection
is not defined at the poles.
7.2.1 Meridians and narallel~o Meridians and parallels
provide the framework for the Mercator projection. Meridians are
projected as parallel straight lines that satisfy the equation x =
constant. Evenly spaced meridians on the ellipsoid project to

35

evenly spaced straight lines on the projection. Parallels are

projected as parallel straight lines perpendicular to meridians
and satisfy the equation y = constant. Evenly spaced parallels on
the ellipsoid project to unevenly spaced parallels on the
projection. The spacing between projected parallels increases
with distance from the equator. See figure 7.1.
60

50~~

40
30
20
10
0
40

20

FIGURE

7.1.

7.2.2

Mercator

60

80

100

Meridians
(longitude) and parallels
in the Mercator projection.

120

(latitude)

rna?mina eaua tions.

7.2.2.1
Findina (x,v). The following discussion of the
Mercator mapping equations can be found in Geological Survey
First, choose a central meridian
Professional
Paper 1395 (1987).
(Ao) that represents
the zero point on the x axis.
Given a point
point
(~, k), in radians, on the ellipsoid, the corresponding

(x, y) in the plane is

x = a(h - Lo)

Y =aln
or

L&s2ul!l-=ino

() [(
a

Y=

Vnl-

sin@

l+&sin~

36

)1

(7.2)

where

a = semi-major axis of the ellipsoid

& = eccentricity
L o = longitude of the central meridian

$, kandkO

are expressed in radians.

7.2.2.2 Point scale factor and convergence of the meridian.

The equations for the point scale factor (k) and the convergence
of the meridian (~) are
k

a
Ncos$

(7.3)

PO
where
Findina ((b,
,k). Next, consider the inverse problem
of finding (@, h) when (x, y) is known. The longitude can be
found from
7.2.2.3

L a + Lo
where both ~ and ~. are expressed in radians. The value of $ can
be found iteratively. Let an intermediate parameter t be defined

t =e

(~)
(7.5)

After an initial value of

en

- 2 arctan t

iterate the following equation for

f-

.
@.

g
(1 - & sin @~) z

@n+l = z - 2 arctan t
[[ (1 + & sin on)1]

37

(7.6)

(7.7)

Substitute the initial value (@O) into equation 7.7 to find the
next candidate ~l. Similarly, substitute ~1 into equation 7.7 to
obtain the updated candidate, @2. Continue this process until the
difference between successive values of ~~ is sufficiently small.
Then set @ = ~~. All angles are measured in radians.
7.2.3

A~.

The Mercator equations for X, Y, ~~ Y, and ~

The iterative equation for $ can be updated until any
are exact.
desired accuracy is obtained.

Area of coveraa~. In the Mercator projection, as the

latitude (~) approaches the poles, the y coordinate approaches
infinity. Area and length distortion increases with distance from
the equator. For example, the point scale factor is approximately
2 at 60 latitude and 5.7 at 80 latitude.
7.2.4

7.3 Transverse Mercator (TM) ~roiection. The Transverse

Mercator projection is a conformal projection for which the point
scale factor equals one along the central meridian. The line
Y= O is the projection of the equator, and the line x = O is the
projection of the central meridian. The Universal Transverse
Mercator (UTM) projection (see Section 7.4) is a modification of
the Transverse Mercator projection.
7.3.1 Meridians and ~arallels. Both the central meridian
and the equator are represented as straight lines. No other
meridian or parallel is projected onto a straight line. Since the
point scale factor is one along the central meridian, this
projection is most useful near the central meridian. Scale
(See Figure 7.2.)
distortion increases away from this meridian.

38

)LO
FIGURE 7.2.

7.3.2

Meridians and parallels in the Transverse

Mercator Projection. LO is the central
meridian.

Transverse Mercator ma~~ina eauations.

7.3.2.1 Findina (x,v). The presentation here follows U.S.

Department of Commerce, Coast and Geodetic Survey, Special
Publication No. 251 (1979). A point with geodetic coordinates (0,
~) in radians has Transverse Mercator coordinates (x,Y) that are
given by
x=

120
N

S$+2
N
+

As

COS3

~
(1

tz

+ lp)

(7.8)

5 - 18t2 + t4 + 14~2 58t2~z)

A2

sin $COS $
A4

24
N

fi5COS5@

Y=

COS()

sin @cos3 $(5 - tz

+ gqz

+ 4q4)

A6
Tzo

sin @coss O(61 - 58tz + t4 + 270~z - 330tz1lz)

39

where
a

(7.9)

N=
Ah - &2 sin2 @
tan @
t=
~= &cos ~
&2
C12 _
\.J.

VI

e2 = a2 - b2
a2
a[AO$ - A2 sin 2$ + Ad sin 44)
AG sin 6$ + A8 sin 8$]
175 ~
3
5
=-6_
AO = 1-: -&2_ ~4_
163848

;(ez + 54
~(&4
256

- %8)

+ $6

+ 36

- +)

~;;2 (&6 - $8)

315
&8
131072
in radians
X-k.
central meridian in radians
At the poles, where @ = + or - 7V2, the Transverse Mercator
coordinates are x = Oandy=
s+, a quadrant arc on the ellipsoid.
7.3.2.2

,.L).
Findina (d)

To find (~,
7.3.2.2.1 Findina the foot~oint latitude (OIL
?L), given x and Y, first compute the footpoint latitude (@l) using
the following iterative method (See Krakiwsky 1973). The equation
to be iterated to yield the footpoint latitude (~1) is

(7.10)

where
s @ =

a[Ao - 2A2cos2@ + 4A4COS4$ - 6A6COS64 + 8A8cos8@

and S$, Ao1 A2, %,

(7.11)

A6, and A8 are defined in equation set 7.9.

first candidate for @l be ~ = Y/a.

When this value of @
the
updated
candidate for @l is
is substituted into equation 7.10,
I
o . Similarly, the substitution of ~ for ~ in equation 7.10

Let the

40

gives the next updated value of ~. When successive values of ~

are sufficiently close, then the footpoint latitude is found by
setting @l = $ . A non-iterative method for finding @l can be
found

in NOAA Report

NOS NGS 5 (1989).

7.3. 2.2.2 Findina (@,L) from @l.

k) can be found from the approximations

once ~1 is determined,

+ )1

($,

B~x6

720 (NI

(7.12)

where
(7.13)

BG = 61 + 90t: + 46q; + 45t;

&2
&12

(1 - &2)
ezcosz $1

= tan @l
a

If@l=&f

then ~= A 90 (a pole) and ~ is undefined.

7.3.2.3 Point scale factor as a function of x and (bI. As a

function of x and ~1, the point scale factor is given by
(7.14)

(7.15)
+

and tl , ~1, and NI are given by equation set 7.13.

,,~).
7.3 .2.4 Point sca le factor as a function of ((b
factor
is
given
by
the
point
scale
of (q, k) f

As

function

A2

=1+

COS2 d)
2

Ad COS4 d)

D2

24

D4

Nx@_!t!

(7.16)

720

in which
D2 = 1+qs
D4 =54tQ + 14~Q + 13~4 - z8t2~2 - 48t2~4

(7.17)

t, and ~ are given in equation set 7.9.

and

7 .3.2.5 Convergence of the meridian in terms of x and d)l.

The convergence of the meridian, in terms of x and ~1, is
9
E2XQ

1 -
3 ()
NI

315

(7.18)
-1

where
E2

l+t;

E4

2 + 5t;

EG

17 + 77t: + lo5t: +

and tl,

~ 1,

(7.19)
+ 2Tf

3t:

and N1 are given by equation set 7-13.

7 .3.2.6 Convergence of the meridian in terms of (@, k).

convergence of the meridian, in terms of (@r k), is

42

The

Y= A sin

4(

#x@41?
2
3

.#&@k+F6~,
15

315

(7.20)

in which

1+

31p+ 2q4
2- ts + 15q2 + 35q4 - 15t2q2 - 5ot2q4
17 - 26ts + 2ta

(7.21)

and A, t, and ~ are given in equation set 7.9.

Accuracv. The Transverse Mercator equations for x,
Yf 41 L qp ~~1k, and Y are approximations. Within 40 of the
central meridian, the equations for X, y, $, and ~ have an error
of less than 1 centimeter.
(See Geological Survey Professional
Paper 1395, 1987.)
7.3.3

7.3.4
Area of coverage. The equations given here for the
Transverse Mercator projection can be used within 4 of the central
meridian. If these equations are used with UTM coordinates, then
consult Section 7.4.1 for guidance on the area of coverage for UTM
coordinates.
.
7.4
Universal Transverse Mere ator (UTM) ~roiection.
The
UTM projection is a family of projections that differ from the
Transverse Mercator projection in these ways (see Figure 7.3):
a.
b.
c.

d.

The longitudes of the central meridians (in degrees) are

3 +6n, n=
O, ...59.
The point scale factor along the central meridian is
0.9996.
The Y value, called the northing, has an origin of O
meters in the northern hemisphere at the equator. The y
value, called the southing, has an origin of 10,000,000
meters at the equator and decreases toward the pole in
the southern hemisphere.
The x value, called the casting, has an origin of 500,000
meters at the central meridian.

7.4.1 UTM zones.

For the UTM grid system, the ellipsoid is

divided into 60 longitudinal zones of 6 degrees each. Zone number
one lies between 1800 E and 1860 E. The zone numbers increase
consecutively in the eastward direction. It is intended that a
UTM projection should be used only in one of the 6-degree zones,
plus the overlap area. In each zone, the UTM projection extends
for a 40 km overlap into the two adjacent zones. Moreover, UTM
coordinates are used only between 800 S and 84o N plus the overlap
region. The UTM projection extends to 84o 30 N and 80 30 S to
overlap with the Universal Polar Stereographic (UPS) projections.

43

I
I
-

I
I
I
- -

- -

9 -

Meridians and parallels (dashed) and a

Universal Transverse Mercator Grid.

FIGURE 7.3.

7.4.1.1 Findina the UTM zone and central meridian. To

find UTM coordinates from geodetic coordinates, first compute the
UTM zone. A standard zone can be computed directly from geodetic
longitude. If the geodetic longitude (k) is expressed in positive
radians, then the zone (Z) is
180~
(7.22)
forO<i<~a
z= Greatest Integer < 31 + 67c
)

z=

Greatest

once the zone (z) is known, the central meridian is given by

equation set 7.23. If a longitude lies on a UTM zone boundary,
then the user must choose the desired UTM zone. If a map is being
used, then the UTM zone appears on the map. The central meridian
of zone (z) is

)L =

(6z - 183)*

for (z > 31)

L = (6z + 177)+

for (z S 30)

with LO expressed in radians.

44

(7.23)

7.4.1.2 Non-standard width UTM zones. Non-standard

width UTM zones are used in North Atlantic regions and Norway as
delineated in Table 7.1 (see DMA TM 8358.1 (1990). These nonstandard zones may also be applied in the 30 UPS overlap region.
TABLE 7.1. Non-standard zone limits.
Latitude
Lower
Upper

Zone
31
32
31
32
33
34
35
36
37

56o N
56
72
72
72
72
72
72
72

Lonqitude
West
East

64o N
64
84
84
84
84
84
84
84

00

3
0
9
21
33

3 E
12
9
not used
21
not used
33
not used
42

Central Meridian
3 E
9
3
15
27
39

7.4.2 Reference source. The standard references for the UTM

projection are DMA TM 8358.1 (1990) and DMA TM 8358.2 (1989).
7.4.3

UTM eauations.

7.4.3.1 Findina UTM coordinates xuTM=UTM~ - Having found

the central meridian, next compute Transverse Mercator coordinates
(x, y) , using equation sets 7.8 and 7.9. Then, the UTM coordinates
can be found from:
UTM = 0.9996
YUTM = 0.9996
YUTM = 0.9996
k UTM = 0.9996

XT~ + 500,000

YTM
YT~ + 10,000,000

(7.24)
(Northern hemisphere)
(Southern hemisphere)

kTM

YUTM = YTM
7.4.3.2 Com~utina aeo detic coordinates. To compute geodetic
coordinates from UTM coordinates, a zone, and a hemisphere, first
solve efiation set 7.25 fOr (XTM, YTM)l equation set 7.23 for Lo,
and then compute ($,k) from equation set 7.12.
1

XTM =

- 500,000)
0.9996 (xulIM
1
YTM = 0.9996 (YUTM)
1
YTM = 0.9996
(YuTM- 10,000,000)
k TM =

1
0.9996

k UTM

YTM = YUTM

45

(7.25)
(Northern hemisphere)
(Southern hemisphere)

Discontinuity. Notice that UTM coordinates are

discontinuous in Y at the equator.
7.4.3.3

Accuracv. The accuracy of the UTM equations is based

on the accuracy of the Transverse Mercator equations. Within the
area of coverage, the equations for x, YJ @f and ~ have an error
(See Geological Survey Professional
of less than one centimeter.
Paper 1395, 1987.)
7.4.4

Area of covera~e. The area of coverage for UTM

7.4.5
coordinates is defined by zone limits, latitude limits, and
overlap.
Zone limits:

60 zones, extending 3 to each side

of the central meridian. See Table 7.1
for exceptions to this general rule.

Latitude limits:

North: 84o
South: 80

Zone overlap:

40 km on either side of the zone

limits.

Polar overlap:

30 toward the poles

North: 84o 30
South: 80 30

Lambert Conformal Conic nroiection. The Lambert

conformal conic projection is a conformal projection in which the
projected parallels are arcs of concentric circles centered at the
pole. The projected meridians are radii of concentric circles
that meet at the pole. Moreover, there are one or two parallels,
called standard parallels, along which the point scale factor is
one. (See Figure 7.4.)
7.5

46

Pole

90%

90E

00

45W
FIGURE 7.4.

45E

Meridians and parallels in the Lambert Conformal

Conic Projection.
(Not drawn to scale.)

7.5.1 Larnbert conformal Conic with two standard Qarallels.

This presentation of the mapping equations follows Krakiwsky
The first case to be considered is the case of two
(1973)
standard parallels, @l and oz. The point scale factor is one
along both standard parallels. Choose a fixed central meridian,
k o

7.5.1.1 Findina (x,v) . Begin by fixing a latitude (@O)

that is below the area of interest. For a given (~, ~) the
mapping equations are

x=

r sin LA

Y =ro-

(7.26)

r cos LA

where

A=L-ko
r=

Ke

-I@
(7.27)
-Lqo

ro=Ke
N1 COS 01
N2 COS @2
K=
-Lql =
-Lq2
Le
Le
e = base of natural logarithms (lne=
+1 = first standard parallel
2 = second standard parallel

N=

a
41 - &2 sin2 $

47


r
q

1(

=lntan~+2

-4

1 - E sin
l+ &sin@

&2

ln(Nl cos @l) - ln(Nz cos @2)

cb - ql
qj = q evaluated at $j , j = 0,1,2
Nj = N evaluated at $j , j = 1,2
L =

.lA) with two standard ~arallels. For the

Findina (0
reverse case, when (x, y) is given along with @o, @l, @2, and LO,
first compute
L, rO, K, Nl, N 2J qO, ql, and q2 from equation set
7.27.
Then compute h and q from
7 .5.1.2

8
+0

?L

(7.28)

L
in which

r=

rO- y
Cos e

0 = arctan

(7.29)
x
rO- y

Note: When evaluating the arc tangent, the user must take
care to insure that the resulting angle lies in the desired
quadrant.
once q is found, then ~ can be determined using the Newton
Raphson method, which is explained next. Define functions f and
f, and an initial approximation ~ by

48

(7.30)

f($)

1 -

=
(1-

oi
Then

&2

= 2 arctan

successive

&2

CTin2 (p)cos

(eq) - 7T/2

iterations

for @ are found

from

the

equation

(7.31)

In particular,
substitute $ into this equation to obtain $ , the
first approximation
to ~.
Then set $ = $ II
substitute
~ into
the equation to obtain $ , the second appro~imation
to (j).
Continue this procedure until the differences between successive
candidates
for $ are sufficiently
small.
7.5.2
Lambert Conformal Conic with one standard ~arallel.
In the second case, the point scale factor is one along one
standard parallel, @ ~. Fix a central meridian, Lo.
7.5.2.1

Findinq

(x,v).

The mapping

x=
y=rO-

equations

are

r sin LA
r cos LA

(7.32)

where

A=L-kO
L = sin ~.

(7.33)

$
K = No cot $0 eqn 0

-Lq
r=Ke
-LqO
rO=Ke
e = base of natural
a
N=
1

logarithms

q=ln
~

= q evaluated

at ~.

NO = N evaluated

at $0

49

( in e = 1 ),

To
Findina
($1x) with one standard Darallel.
suppose that X, y, $., and ~o
transform in the other direction,
First, compute L, K, NO, ~, r, and r. from equation
are given.
Then, as in the previous case, (~, ~) can be found from
set 7.33.
7.5.2.2

(7.34)

L
in which
ro- y
r=

(7.35)

Cos e

8 = arctan

v
*
rO y

Again, ($ can be obtained from q by using

(see equation sets 7.30 and 7.31).

7.5.3 Scale factor and convergence.

convergence of the meridian are given by
k=

Y=

rL

the Newton-Raphson

The scale

factor

method

and

(7.36)

N COS $
L?L

the values of r, L, and N

parallels,
For
one
standard parallel, the
are given m equation set 7.27.
values of r, L, and N are found in equation set 7 .33.
In

the

case

of two standard

For the Lambert conformal conic projection

7.5.4
Accuracv.
with either one or two standard parallels, the equations for X, y,
The equation for $ can be iterated
?L, q, k, and Y are exact.
until

the

desi-red

accuracy

is

obtained.

The
family
of Lambert
conformal
Area
of covera~e.
7.5.5
conic
projections
and
their
limiting
cases is used worldwide.
the point
scale
factor
increases
of this
family,
a given member
a point moves away from the standard parallel(s) .

50

For
as

7.6
Polar StereoaraDhic
Droiection.
The Polar
Stereographic
projection is a limiting case of the Lambert
conformal conic projection when the one standard parallel
In this conformal projection meridians are
approaches a pole.
straight lines, parallels are concentric circles, and the point
scale factor is one at the pole (see Figure 7.5) .

180

180

135%

li

90E

45E
Figure

7.5.

7.6.1

90%

90E

450W

45E

Meridians and parallels in the Universal Polar

Left : South Zone,
Stereographic
projection.
Right : North Zone
(Not drawn to scale.)

Polar

StereoqraQhic

maDRin~

equ ations

There

is

considerable
diversity in the treatment that various authors give
to the polar Stereographic projection.
Some authors interchange
The
the x and y axis; others change the sign convention.
discussion here is largely taken from NSWC/DL TR-3624 (1977) since
it most closely conforms to DMA TM 8358.2 (1989) .

7.6.1.1
Stereographic

Findin~

(x,v)

coordinates

Given

(~,

~),

the

Polar

are

(7.37)

x = r sin ~
(Northern
(Southern

Y=-rcos~
Y=rcos~
in

hemisphere)
hemisphere)

which

(w-)(

r = K tan

2a2
b
a2

-&
l+&

()
1

- b2

&2

a2

51

& sin

- &

sin

1~1
1$1

(7.38)

1$1 denotes

the absolute value of $.

($, ~) is expressed in radians.

Nevertheless,
these equations can
At the poles, h is not defined.
At
$ = ~ ?L/2, pick an
be used at the poles since r is zero there.
Then, equation sets 7.37 and 7.38 yield
arbitra~
value for ~.
Polar Stereographic
coordinates
of (O, O) .
(0,k) . The discussion of the conversion
7.6.1.2
Findina
from (x, Y) to (~, ~) follows Krakiws@
(1973).
If
otherwise,
(x, y) = (o, o), then $ = ~ 900 and ~ is undefined.
the longitude is given by
k=

arctan

(Northern

hemisphere)

(7.39)

?L=

arctan

(Southern hemisphere)

If (x, y) # (O, O), then x, Y or both are nonzero.

these non-zero values to find q from the equations
-q

IYI
K IcOS il

e=

-q

e=

K sin h

K=

when

Ixl > Iy

when

Iyl > lx

2az 1 - & ~
bl+&

()

Use

(7.40)

(7.41)

As in the Lambert conformal conic projection, @ can be found from

Define functions f and f , and
q using the Newton-Raphson
method.
an initial approximation
@o by

f($) =

f($) =

o0

-q+~ln

1
1-

[(*)(=:$)]
&2

(1- &2 sin2$)cos

= 2 arctan eq - ~

52

(7.42)

Then,

successive

iterations

for $ are found

from the equation

(7.43)
In

particular,

first

approximation

obtain
until

substitute

the

@2,
the

between
is

7.6.1.3
approximate
Professional

this

to

Alternate

sin2~

value

of

positive
method

+ B

$.

for

sin4~

sin6~

the

of

the
~

by

d.

for

to

procedure
@

are

at

the

This

southern
-~.
Alternatively,

Geological

equation

this

computers.

In

findina
in

into

infinity

many

@l, the

obtain

candidates

$.

value

to

Continue

approaches
for

equation
for $ is given
Paper
1395
(1987).

4=X+A

$1

successive

though

this

equation

substitute

satisfactory

a positive

replace

hemisphere,

Then

Even

method

yields

$.

into

approximation

small.

this

procedure

second

differences

sufficiently
poles,

to

$0

sin8~

an

Survey

(7.44)

where

x=:-z

(in radians)

x#o
tan

K=

IK

Cos

2az
b

y#o

I-&;

()
I+&

1
5
24+36+*8

29

B=~4+~&6

+811

81

4279 *
D = 161280E
_

&*

11520

c=~6+D*

~2

13

a2 -

bz

az

53

(7.45)

this procedure yields

in the southern hemisphere.

Since

7.6.1.4
factor

Findina
is given by

k=l
k=

a positive

the Doint

scale

value

of $, replace

factor.

The point

@ by -$

scale

at the poles
r
N COS ~

(7.46)

elsewhere

where

K an(:-

2a21

+)

l+&sinl$l$

(7.47)

-&~

()

bl+e

7.6.1.5
convergence

Findina the convergence

of the meridian is
y=k

(Northern

=
Y~

(Southern

of the meridian.

The

(7.48)

hemisphere)
hemisphere)

projection,
7.6.2
Accuracy.
For the Polar Stereographic
The equation for $
the equations
for x, y, ~, and q are exact.
can be iterated until the desired accuracy is obtained.
Area of coveraae.
There is no general agreement
7.6.3
the area of coverage for the Polar Stereographic
projection.
Section 7.7.5 gives the area of coverage for UPS coordinates.

on

The
7.7 Universal Polar StereoaraDhic
(UPS) Droiection.
Universal polar Stereographic
(UPS) projection is the standard
military grid used in polar regions (see Figure 7.6) . The main
reference for the UPS grid is DMA TM 8358.2 (1989) . The UPS grid
is a family
Stereographic

of

two projections
that
differ
projection
in these
ways:

from

the

Polar

a. Both the x and y values

(see Figure
7.6),
called casting
and northing, respectively,
have an origin of 2, 000,000
meters.
b. The scale factor at the origin is 0.994.

54

180

90W

90E

00

Figure

7.6.

Meridians and parallels imposed on a Universal

Polar Stereographic
grid (north zone) .

The limits of the system ake north of 84ON and south of 80 S. In

order to provide a 30minute overlap with the UTM grid, the UPS
grid contains an overlap that extends to 83o 30N and 79o 30s.
Although the UPS grid originally was defined to use the
International
Ellipsoid, in practice several reference ellipsoids
have been used.
The ellipsoid appropriate
to the selected datum
should be used.

7.7.1

em

ations.
given
by

x UPS
and

The

(x,

the

Universal
For

Polar

known

stereoara~hic

geodetic

coordinates,

(UPS)
maQ~inQ
the UPS
casting

2,000,000 + 0.994r sin 1

UPS northing

(7.49)

i s given by

Yups

2,000,000 - 0.994r

cos

(Northern

hemisphere)

YUPS

2,000,000

cos

(southern

hemisphere)

parameter

is

is

0.994r

computed

using

equation

set

(7.50)

7.38.

7.7.2 Findina (@rk) . If a hemisphere and UPS coordinates

Y) are known, convert the UPS values to Polar Stereographic

coordinates
using
equation
set 7.51,
geodetic
coordinates
as described
in

55

and then
Sections

transform
7.6.1.2

to and 7.6.1.3.

xp~

(7.51)

- 2,000,000)

0.;94 x
1

Yps = 0.;94 s - 2,000,000)

at the poles where x = 2,000,000 and Y = 2,000,000
particular,
otherwise longitude and
then, $ = ~ 90 and h is undefined.
latitude can be found from
In

L = arctaIl.
L = +

arctan

(Xus

(Yus

%Js

(Yus

2,000,000)
2,000,000)
2,000,000)
2,000,000)

(Northern hemisphere)
(Southern

(7.52)

hemisphere)

which

in

(7.53)

x=
(xup~- 2,000,000)
tan

=
0.994K

sin

Xps

(yup~- 2,000,000)
0.994K COS k
Yps #

tan~=-

2,000,000

(Northern
zfooo~ooo

(yps- 2,000,000)
0.994K COS k
yups # 2,000,000

tan:=+

()

a2

_ b2

za2

K=

&2

(southern

hefi,isphere)

hemisphere)

l+&

az
The values
In

the

of A, 13, C, and D are found using

southern

hemisphere,

replace

56

the

equation

computed

value

set 7.45.
of

with

7.7.3

meridian.
are

Finding Doint scale factor and convergence of the

The point scale factor and convergence of the meridian

0.994 at the poles

where

0.994

are

given

-r
A
Cos

(7.54)

elsewhere

(p

(Northern
(Southern

hemisphere)
hemisphere)

in equation

set 7.47.

7.7.4
Accuracv.
The accuracy of the UPS equations is based
upon the accuracy of their polar Stereographic
counterparts.
For
the Polar Stereographic
projection,
the equations for x, y, ~, and
The
equation
for
q are exact.
can be iterated until the desired
accuracy is obtained.

7.7.5 Area of coveraae.

The area of coverage
coordinates is defined by zone limits and overlap.
North

zone:

The north

polar

area

84o - 900

south

zone:

The

polar

area

80 - 900

UTM overlap:

south

for UPS

30 overlap,
North:
83 30
South : 790 30

7.8 The U.S. Militarv Grid Reference Svstem (MGRS) . The

MGRS is an alpha-numeric
system, based upon the UTM and UPS map
projections,
for identifying positions.
An MGRS coordinate
consists
of a zone
designation,
alphabetic
100,OOO-meter
grid
square
designator,
and numeric coordinates within the 100,000
meter
grid
square.
MGRS coordinates
are defined for the primary
UTM
and UPS areas,
but not for the overlap areas.
7.8.1 MGRS coordinates in the UTM area.
In the UTM area, MGRS
coordinates are based on the ellipsoid, geodetic latitude,
UTM
zone, easting and northing as follows:
a. The first two characters of the MGRS coordinate
two digits of the numeric UTM zone (see sections 7.4.1.1
7.4.1.2. ) Leading zeros must be included.

are the
and

b. The third character of the MGRS coordinate is a letter

representing
a band of geodetic latitude.
Beginning at 80 South
and proceeding northward,
20 bands are lettered C through X,

57

omitting
I and O.
is 12
84N) , which

The bands
high.

are

all

8 high

except

band

(72oN

to

The fourth and fifth characters of the MGRS coordinate

are a pair of letters representing
the 100,000meter
grid square.
The letter pair can be selected from figure 7.7 and 7.8 (from DMA
First, reduce the UTM northing by multiples of
TM 8358.1.)
2,000,000 until the resulting value is in the range of O to
Second, reduce the UTM northing and casting
1,999,999 meters.
Third, locate the
values to the nearest 100,000 meters.
corresponding
100,000 meter northing and casting grid lines for
The grid square identifier
the UTM zone number on the figure.
will be immediately above and to the right of the intersection
of
the casting and northing grid lines on the figure.
c.

The alphabetic method of designating

100,000Note:
older products
meter grid squares has changed over time.
may have different grid square designations
than those
Also, some software was programmed
to
described above.
compute MGRS coordinates based on geographic position and
The boundaries between
a map of preferred ellipsoids.
Therefore,
preferred ellipsoids have changed with time.
MGRS coordinates
computed with such software may not agree
MGRS computation
software
with current map products.
If MGRS
should always require input of the ellipsoid.
coordinates
are to be used in conduction with a map
product they should be checked against the map product
to verify compatibility.
d. The remainder of the MGRS coordinate consists of the
numeric casting and northing values within the 100,000meter
grid
The left half of the digits is the casting grid value,
square.
which is read to the right of the 100,000meter
casting grid line
The right half of the digits is the
established
in step c. above.
northing grid value, which is read northward of the 100,000meter
Both the casting
northing grid line established
in step c. above.
and northing grid values are within the range of O to 100,000
meters (i.e., zone and grid square designators,
only) to 1 meter
followed by 10 digits, five for
(i.e., zone and grid designators
casting grid value and five for northing grid value) . Both
casting and northing value must have the same resolution
and must
include leading zeros.

7.8.2

MGRS coordinates in the UPS area.

MGRS
coordinates
are based
on the ellipsoid,
longitude,
casting,
and northing
as follows:

a.
A in the
B in the
Y in the
Z in the
and 7.10.

In the UPS

geodetic

area,

latitude

The first character of the MGRS coordinate

is
Southern and Western Hemispheres,
Southern and Eastern Hemispheres,
Northern and Western Hemispheres,
and
Northern and Eastern Hemispheres
as shown in figures

58

and

7.9

000(308

Ill

000001

000009
000400s
000OOP
ooooo&
UI00000
z
W 00000j3
000001
!

--

000009
000400s
00000$
Oooooc
u Oooooz
000008

(II

000OOL
,

000009
00000s

I --

Ooooot
t

000OOE
(IJOl)ol)oz

u 000008
000001
I

----

000009
00000s
Ooooot

00000s
UJOfjoooz

+
Ill

()()0()(38
000001
000009
00000s
Ooooot
Oooooc

u (JOlyol)z

Oo(lofjg

IJ.1

000OOL
000009
000(00$
Ooooob
000OOE
w 000IJJZ

E
0

E
0

0
0

0
0
0

0-

0
0
0

In

59

E
0

UJ 000 ()()8

000 00/

000009
00000<
000009
000Oof
UJO(J()Ooz

UI 000008

000Oof
000009
I

l-l!l[

000 00s
000 00;

,,

00000[

~ 000 00?

u 000 OIJ8
000001
000009
000 00s
Ooooot
00000:
UJOl)oooz

UJ

000008
000 00f
000009
000 00s
000 Oot

-m

mm

000OOE

UJ 000 Ooz

w IJOO1)08
000001
000 O(I9
000 00!
000 Oob
000 00[
~ 000 Ooz

w 000008
000 00f

000005
00000!
000Oor
00000:
u rl~,flo~:

60

dd

ldLQ

0
0
0
00

1-

FUJ
LI

0..

-/1

+17

.C(j[

.96

006

Sd

0
0
0
00

62

The second and third characters of the MGRS coordinate

b.
are letters representing
a 100,000 meter grid square.
The letter
pair can be selected from figure 7.9 or 7.10 (from DMA TM 8358.1.)
Reduce the UPS easting and northing values to the nearest 100,000
Locate the corresponding
100,000 meter northing and
meters.
easting grid lines for the UTM zone number on the figure.
The
grid square identifier will be immediately above and to the right
of the intersection
of the easting and northing grid lines on the
figure.

coord;~ate

The method for calculating the remainder

is the same as in paragraph 7.8.1 d.

of the MGRS

7.9
The World GeoaraDhic Reference Svstem (GEOREF) . The
GEOREF is an alpha-numeric
system for reporting positions for air
defense and strategic air operations.
It is based upon geodetic
coordinates . GEOREF is described in DMA TM 8358.1 (l$3gO).

7.10 Non-standard arids.

The specifications
and diagrams of
nonstandard
grids, used as the prima~
or secondary grid on DMA
produced maps, are contained in DMA TM 8358.1 (1990) .
7.11 Modificat ions to maD Droiections . Providing equations
for transforming
between geodetic coordinates and eve~
local map
projection
is beyond the scope of this Technical Report.
However,
many local map projections
are minor modifications
of the map
projections
given in this Technical Report.
Typically, these
modifications
include use of false Eastings (E) and Northings
(N)
at the projection
origin, a non-unity scale factor (ko) along the
principal
line
of the projection,
and units
other
than
meters.
Sometimes
the local
projection
origin
is offset
from
the origin
used
in this
Technical
Report.
chapter
four
of DMA TM 8352.1
(1991)
gives
parameters
for several
local
projections.
If
given

in

longitude
following
1.

2.

3.

the

local

this
(~)

projection

Technical
to

is

Report

coordinates

based

on

convert
in

the

one

of

geodetic

local

the

projections

latitude

projection

by

(@)
using

and
the

projection;

Make sure ($, 1) are on the ellipsoid for the local

projection
and determine the ellipsoids parameters,
and either f or b.

Convert (~, k)to map projection coordinates (XP, YP)

using the appropriate map projection equations given in
this Technical Report. Note that the units for XP and YP
are in meters.
If
the

the

local

origin,

projection
convert

them

63

has

false

coordinates,

to

units

of

meters

E,
if

N,

at

necessary.

4.

The map projection coordinates (xP, YP) and scale factor

(kP) can be converted to local projection coordinates
using equations 7.55 if the origin of the local
projection is the same as the origin of that projection
Those origins are
as used in this technical report.
central meridian (&) and equator for Transverse
Mercator and Mercator projections, and standard
parallel(s) and equator for the Lambert Conformal
projection.
x

koxp +

Y = koyp +

(7.55)

= kokp

7 = Y.

where
false X-value (in meters) at the origin

E
N

false Y-value (in meters) at the origin

ko = scale factor along the appropriate line.

( Along the equator for the Mercator projection,
along the standard parallel(s) for the Lambert
Conformal Conic projection, and along the central
meridian for the Transverse Mercator projection. )
5.

Convert (X, Y) to the units that are locally used. It

is the users responsibility to determine if this
procedure is valid. This can be done by comparing
computed mapping coordinates to test points where the
relationship between (~, k) and (x, y) is known,
independently.

The above procedure is intended to be used only when the

origin of the local map projection coincides with the origin used
in this Technical RePort. Some Transverse Mercator projections
have origins of ($., ~) instead of (O, ~) on the equator. For
some of these projections, agreement with test points can be
obtained by changing the computation of Y in equation set 7.55 to

64

Y =

koym

+ N -

kospo

(7.56)

where
ko = scale factor on the central meridian, L
YTM = Transverse Mercator Y value computed from equation
set 7.8
N = false Northing in meters of the origin
S$O = value computed using equation set 7.9
(arc length in meters on the ellipsoid from the
equator to the latitude
$0.)

65

8.

SELECTION OF TECHNIQUES, SOFTWARE DEVELOPMENT, AND TESTING

8.1 The aeneral conversion Qrocess. The general coordinate

conversion process can be divided into three basic parts:
1.

Conversion between map projection coordinates and

geodetic coordinates (see Figure 8.1) .

2.

Coordinate conversion between geodetic coordinates and

Cartesian coordinates within a datum (see Figure 8.2) .

3.

Coordinate transformation between datums (datum shifts)

(see Figure 8.3).

The general conversion process which includes all methods

presented in this Technical Report, is described
graphically in Figure 8.4.
a.

Figure 8.1 represents the map projection coordinate (x,Y)

to the geodetic coordinate (@,k) conversion described in
Section 7. As noted in Section 7.1, map projection
coordinates do not include the height information
necessary for finding the geodetic height (H) or
elevation (h). For conversions from map projection to
geodetic coordinates, the user may have to provide this
information. Elevations can often be inferred from
contour lines on the map or they can be obtained from
DTED or direct measurement. Geodetic height may be
measured by using GPS or by calculating from an elevation
and geoid separation, or for some applications, may be
assumed to be O (zero).
A
X,y

Figure 8.1.

Map projection to geodetic coordinate

conversion. Height information for H and h is
not included in the conversion.

where
X,Y
$
k
A

=
=
=
=

map projection coordinates

geodetic latitude
geodetic longitude
map projection coordinates to/from latitude,
longitude conversion

66

b.

Figure 8.2 represents coordinate conversions within a

datum as described in Section 4.3.4 and 6.5.

Y,

H
FIGURE 8.2 .

Coordinate conversions within a datum,

where
x, Y, z
$, L h
H
B
c

Cartesian coordinates
geodetic coordinates (latitude, longitude,
geodetic height)
elevation
geodetic height to/from elevation conversion
geodetic coordinate to/from Cartesian coordinate
conversion

Note that a conversion between Cartesian coordinates (X,Y,Z)

and local datum coordinates (@,~,H) is not a direct conversion
and requires going through the geodetic coordinates ($,k,h) .
c.

d.

Figure 8.3 represents the datum shift portion of the

general conversion process as described in Section 6.
Two attributes of figure 8.3 are important to note.
1.

Direct local datum to local datum shifts are not

performed.

2.

Elevations do not change when changing datums.

Exceptions to the second statement are explained
in Section 6.7.

In general, a direct local datum to local datum shift

should not be done. Instead, the local datum to local
datum shift should be done in two steps.
1. An (old) local datum to WGS 84 shift.
2. A WGS 84 to (new) local datum shift.

67

WGS 84

L-

Figure 8.3. Datum shifts.

Note elevation values (H) do not generally change when
changing datums,
where
x, Y, z
(/), k, h
H
c

BB
cc
e.

Cartesian coordinates
geodetic coordinates (latitude, longitude,
geodetic height)
elevation
geodetic coordinate to/from Cartesian coordinate
conversion
Cartesian coordinate-based datum shifts
geodetic coordinate-based datum shifts
elevations are unchanged

The aeneral conversion process, which includes all

meth~ds presented in this Technical Report, is described
graphically in figure 8.4. Figure 8.4 is the combination
of figures 8.1 through 8.3,

68

WGS 84
x, Y, z

k,

h
B
H

Figure 8.4.

The general conversion process,

where
X,y=
Y,z=

k,h=
H
A
B
c
AA
BB
cc
f.

map projection coordinates

Cartesian coordinates
geodetic coordinates (latitude, longitude,
geodetic height)
elevation
map projection coordinates to/from latitude,
longitude conversion
geodetic height to/from elevation conversion
geodetic coordinate to/from Cartesian coordinate
conversion
Cartesian coordinate-based datum shifts
geodetic coordinate-based datum shifts
elevations are unchanged

In figures 8.1 through 8.4, coordinate conversions are

denoted as a single letter (A, B, C). They are described
in Section 7, Section 6.5 and Section 4.3.4,
respectively. Datum shifts are denoted as double letters
(AA, BB, CC) and are described in section 6.

69

c3- The Cartesian coordinate-based datum shifts (AA) include:

Seven-parameter geometric transformation
Ax,AY,Az direct application

(6.2)
(6.3.1 & 6.3.2)

Both of these datum shift methods are geometric

transformations.
h.

The geodetic coordinate based datum shifts (BB) include:

(6.3.3)
Molodensky shifts
(6.3.1.1 & 6.3.2.1)
Ax, AY, AZ three-step method
Direct WGS 72 to WGS 84 transformation (6.4.2)

i.

For the purposes of this Technical Report, the elevation

to elevation shift (CC) is the equality function (=);
that is, the elevation for the same location on different
datums is considered to be the same (H = H). MI
exception to this rule is the WGS 72derived elevations
to WGS 84 elevations shift. Section 6.7 describes the
elevation to elevation shift process.

8.1.1 A~plication of the aeneral Conversion Drocess. The

general conversion process described above can be used to solve
coordinate conversion/datum shift problems in five steps:
a.

Locate beginning point and ending point of the problem on

figure 8.4.

b.

Determine valid paths between beginning and end points.

c.

Evaluate alternatives and select a path.

d.

Locate the needed equations and parameters in this

Technical Report.

e.

Implement the selected path.

8.1.2

Procedural examQles.

8.1.2.1 Exam~le 8.1. Horizontal geodetic coordinate-based

datum shifts (($,~,h)LG~ to ($,1)~~~*~) (Local datum 9eodetic
heights (hLGs ) typically are not available. It is not recommended
they be transformed to WGS 84) .
This Technical Report has discussed two methods of doing
datum shifts. These methods are the three-step method, and
Molodensky shifts. Either the three-parameter AX, AY, AZ shift or
the seven-parameter shift may be used in the three-step method.
shift of figures 8.3 and 8.4.
This example corresponds to a IIBBII

70

a.

b.

c.

A~~roach 1 Three-ste~ method usina a sevenDarameter

shift.
1.

Compute (X,Y, Z)LGS from (@,A,h)LGSusing equation sets

4.4 and 4.5, and the reference ellipsoid parameters
from Table Al.

2.

Compute (X,Y,Z)WGS 84 from (XJY,Z)LGS using the sevenparameter shift equation set 6.1 and shift parameters
from Table El.

3.

Compute (@jk)wGS84 from (X,Y,Z)WGS 84using ewation

sets 4.6 through 4.10.

Armroach 2 - Three-ste~ method usin a Ax,

AY,

AZ

shifts

1.

Compute (X,y, Z)LGS from (@,~,h)LGSusing equation sets

4.4 and 4.5. Reference ellipsoid parameters are
found in Table Al.

2.

Compute (X,Y,Z)WGS 84 from (X,Y,Z)LGS using equation

set 6.3, and shift parameters from Table B.1

3.

Compute (@,~)WGS84 from (X,Y, Z)WGS 84and equation

sets 4.6 through 4 .10.

AR~roach 3 - Usina Molodenskv shift

1.

Compute @and
AL from (@,k,hLGs) using the
Molodensky equation set 6.5,shift parameters from
Table B.1, and the local datums reference ellipsoid
parameters given in Table Al.

2.

Compute ($,~)WGS84 from ($,~)LGSusing e~ation

set 6.6.

WGS 84 Cartesian coordinate to local

8.1.2.2 Exam~le 8.~
longitude,
elevation.
latitude,
Several paths for converting (X,Y,Z)WGS 84 to ($LGS,kLGs,H ) can
be traced in figure 8.4. The steps of three approaches are
provided below.
a.

AR~roach 1 - Usina a seven-~arameter datum shift

1.

Compute (X,Y, Z)LGS from (X,Y, Z)WGS84using the sevenparameter shift equation set 6.2 and parameters from
Table El.

71

b.

c.

2.

Compute (~,k)LGSfrom (X,Y, Z)LGS using equation sets

4.6 through 4.10 with the parameters of the local
ellipsoid.

3.

Compute hwGsQd from (x,y,z)wGs84using equations 4.7

through 4.11 (Use equation 4.12 in polar regions) .

4.

Compute H from hwGs84 and $JWGS84using e~ation

resulting in @ LGS,k LGS! H.
2 - Usin~ a direct AX,

Amroach

AY,

5.1,

AZ datum shift

1.

Compute (X,Y,Z)LGS from (X,y, Z)WGS 84 using the AX,

AY, AZ shift equation set 6.4 and shift parameters
from Table B.1.

2.

Compute (@,~)LGSfrom (X,Y,Z)LGS using equation sets

4.6 through 4.11 with the parameters of the local
ellipsoid.

3.

Compute hwGs84 from (X,Y,Z)WGS 84using equations 4.7

through 4.11. (Use equation 4.12 in polar regions).

4.

COmpUte H frOm hwGs84 and NW~S 84using e~ation

resulting in $ LGS,L LGSf H.

5.11

A~~roach 3 - Usina a Molodenslrv datum shift

1.

COmpUte (@h,h)WGS 84 from (X,YJZ)WGS 84using e~ation

sets 4.6 through 4.11. (In polar regions, equation
set 4.11 should be replaced by equation set 4.12) .

2.

COmpUte ($,k)LGsfrOm ($,L)wGs84 using the

Molodensky equation set 6.5, equation set 6.7 and
shift parameters from Table B.1.

3.

COmpUte H from hwGs84 and NWGS 84using equation 5.1/

resulting in QLGS, ~ LGS1 H.

Local latitude, longitude, elevation

8.1.2.3 ExamDle 8.~
($LGS,~L~S, H) to WGS 84 latitude, longitude, elevation
(@WGS84, &S B4,H)

Example 8.3 is a extension of example 8.1. Example 8.3 shows

the recommended methods for determining geodetic height used in
each of the geometric datum shift methods. The steps of three
alternate approaches are provided as follows.

72

a.

Amroach
1.

Compute (X,Y,Z)~~~ from (@,~,H)LGS using equation sets

4.4 and 4.5.

2.

Compute (XJylZ)~@ 84 from (XJY,Z)~~S using the seven

parameter shift equation set 6.1 and shift parameters
from Table El.

3.

Compute (@,~)WGS84 from (X,YJZ)WGS84using equation

sets 4.6 through 4.10.

4.

b.

Take H from the input coordinates resulting in

L GS 841 H.

Amroach

2 - Usina a Ax,

AY,

AZ datum shift

1.

Compute (X,Y, Z)LGS from (@,~,H)LGs using equation sets

4.4 and 4.5.

2.

COmpUte (X,yJZ)WGS 84 from (&Y,Z)LGs using AX, Ay,

AZ shift equation set 6.3, and shift parameters from
Table B.1.
.

3.

Compute (@,h)WIS 84 from (X,ylz} WGS84using equation

sets 4.6 through 4.10.

4.

c.

1 - Usina a seven~arameter datum shift

Take H from the input coordinate resulting n $WGS84>

L GS 841 H.

A~~roach 3 - Usin~ a Molodenslw datum shift

1.

2.

Compute ($,~)wGS84 from ($, h, H)LGSusin9 Molodens&

equation set 6.5, equation set 6.6, and shift
parameters from Table B.1.
Take H from the input coordinate resulting in $w~s84,

kWGS 841 H.
8.2 Choosina a datum shift method. Several factors should
be considered in selecting a datum shift method. These factors
include: shift parameter availability, timing requirements,
accuracy requirements, software availability, hardware
limitations, standardization considerations and potential future
requirements. If future requirements are considered as part of
the system development, future product improvements may be
minimized. These requirements may include the need to provide
data to other systems and to meet higher accuracy requirements.
The discussion below applies to a wide range of Army applications.

73

Special, high accuracy applications are outside the scope of this

Technical Report (see Section 1-2) .
The threeparameter shift is recommended for mapping and charting
and most other military application. For Western Europe (EUR-M),
and Ordnance Survey of Great Britain 1936 (OGB-M), STANAG 2211
specifies the sevenparameter shifts. The use of MREs is not
recorrunended,although it should be noted that STANAG 2211
specifies the use of MREs for geodetic and geophysical
applications where data is available in geodetic coordinates.
Parameter sets for AX, AY, AZ shifts are found in Appendix B, and
sevenparameter geometric shifts are contained in Appendix E. Be
sure to use the same method throughout a project, identifying the
method chosen in the project documentation so that future users of
the data can replicate or backtrace your results.
8.2.1 Parameter availability. For many datums, only mean
value threeparameter AX, AY, and AZ shifts are given in this
Technical Report. Based on system implementation tradeoffs, the
system developer can: apply the shifts directly to Cartesian
coordinates (see Sections 6.3.1 and 6.3.2); use the threestep
method (see Section 6.3.1.1 and 6.3.2.1); or apply shifts to
geodetic coordinates using the Molodensky approximations (see
Section 6.3.3). For other datums, some co~ination of seven
parameter geometric shifts, and mean and regional AX, AY, and AZ
shifts are available. Select a parameter set from those that meet
or come closest to meeting the accuracy requirements of the output
product. Base the final selection on standardization requirements
and system implementation tradeoffs.
8.2.2 Accuracv. Errors in the shift parameters are the
dominant error source in datum shifting. Errors introduced by any
of the datum shift algorithms given in this Technical Report are
small compared to the errors in the shift parameters and can
generally be neglected. However, take care that accuracy is not
degraded by the software and hardware implementation of these
algorithms.
a.

For the large datums,the regional AX, AY, AZ shifts and

sevenparameter shifts are generally considered the most
accurate, and mean value AX, AY, AZ shifts have the
poorest accuracy. The estimated accuracies of the shift
parameters are given in Appendices B and E for each
listed datum. The errors in geodetic coordinates may be
approximated from the AX, AY, AZ errors listed in
Appendix B, using the equation set 6.8.

b.

The error estimates and general guidance above must be

used judiciously. For a given datum, all types of shift

74

parameters were derived from essentially the same set of

control data, and errors were estimated by using a
comparison to the same data. For small datums, with
poorly distributed control, the better fitting shift type
may not be more accurate. The seven-parameter model
cannot accurately represent datum rotations and scale
changes unless the control is geometrically well
distributed across the datum.
c.

Regional Ax, AY, AZ shifts were developed when there were

sufficient data to identify regional variations over the
datum as a whole. Regional AX, AY, AZ shifts are more
accurate than mean-value AX, AY, AZ shifts.

8.2.3 Implementation issues. System implementation issues

may influence the selection of transformation method(s) .
a.

Computational speed and efficiency are important for some

applications, such as realtime navigation. The shift
type and method might be selected from those that meet
accuracy requirements to minimize computational time.
Alternately, shifts might be computed in a background
process and applied as fixed values for a period of time
or over a given area. The system designer must analyze
the impact of such design strategies on system accuracy
and performance.

b.

For some applications, system computational and data

storage capability are considerations in selecting and
implementing transformations. The system designer must
determine that word precision, range, and accuracy of
mathematical functions preserve the required accuracy.

c.

Code complexity, modularity, and software maintenance are

important system design concerns. While the datum shift
algorithms themselves are relatively simple, selection
logic often is complex.

d.

Both future requirements and present needs should be

considered. Factors for consideration include: timing
requirements, current accuracy requirements, potential
future accuracy requirements (such as requirements to
provide data for future systems), effects of improved
databases, and the effects of improvements in other
inputs . In other words, the savings in future planned
product improvements should be evaluated against todays
costs.

e.

It is possible that a software developer may choose to

use a transformation technique that is more accurate than
is required. In such a case, the results are not
necessarily superior to those arising from a less

75

accurate, but acceptable, technique. The data are simply

degraded less by the more accurate method; the
transformed data can be no better than the input data.
For example, techniques that achieve geodetic accuracy
must be used when geodetic accuracy is needed. Such
techniques can be used with a 1:250,000-scale product,
but the transformed data still have the accuracy of a
1:250,000 product.
8.2.3.1 Preferred shift methods. preferred shift methods
and shift parameters may change as additional information becomes
available. Shift parameters and identification information should
be treated as data apart from the algorithm code so that they may
be updated without modifying and recompiling the code. Some
examples of implementation tradeoffs are:
a.

If Cartesian coordinate are already available, AX, AY, AZ

shifts can be applied directly (see Sections 6.3.1 and
6.3.2) rather than through the Molodensky approximations
(see Section 6.3.3).

b.

If the program contains algorithms for converting between

Cartesian and geographic coordinates (see Section 4.3.4) ,
the three-step method (see Sections 6.3.1.1 and 6.3.2.1)
can be used instead of Molodens& approximations (see
Section 6.3.3) to apply AX, AY, Az shifts.

c.

Both the Molodens@ approximations (Section 6.3.3) and

Cartesian-to-geographic conversion equations (Section
4.3.4.2) become singular at the poles. Selection of
one technique or the other will reduce the logic needed
to guard against singularity. An alternate equation to
compute geodetic height from Cartesian coordinates in
polar regions is given as equation 4.12.

8.2.3.2 Software testinq. When developing software based on

methods discussed in this Technical Report, it is advisable to
check the software against test points. Test points can be found
in Section 8.5.
a.

Issues relating to software development and system

performance might force a software developer to use a
specialpurpose transformation technique that is not
mentioned in this Technical Report. The results of the
specialpurpose software should be compared to those of
the techniques in this Technical Report to insure that
the differences are acceptable.

b.

This
Another useful debugging aid is circular testing.
is done by using the output from a transformation as the
input for the inverse function; in other words trying to
verify the original numbers are returned within

76

acceptable accuracy. Since the code for a transformation

and its inverse might contain mistakes that cancel each
other, circular testing can not prove that code is
correct. At best, it can demonstrate that code is
faulty.
c.

The following is a list of points that might prove

troublesome to coordinate transformation software.
the code at these points, if it is appropriate:

Test

Poles
Polar regions
Discontinuity in ~ (transition between 0,360 degrees or
-180, 180 degrees)
origin expressed in (~, L)

(p=o
1=0
Both hemispheres in $
Both hemispheres in ~
x= o

(North and South)

(East and West)

Y=o
All four X-Y quadrants
Discontinuities in x and Y at zone boundaries, equator,
etc.
8.3 Error analvsis. Error analysis is an important part of
the system design process. Error analysis will estimate the
accuracy of the systems output products. This is a prediction
of how well the system will meet its accuracy requirements. The
system analyst can use error analysis to perform system
implementation studies. The major steps in error analysis are:
a.

Develop the system model.

b.

Determine the estimated errors for each source in the

model.

c.

Reformulate the estimated errors to a common description.

d.

Propagate the errors through the model.

Each of these topics is discussed below.

8.3.5.

An example is given in

8.3.1 Svstem model. The system model simply is a

description of the data used in the process and all the steps
(processes) used to process the data. The data include the raw
input data and parameters, like datum shift parameters, used in
the processing of the data. Both hardware-driven and softwaredriven processes should be included in the model. For example,
errors in digitizing and plotting are primarily hardware related;

77

coordinate conversion and datum shifting are primarily software

processes. For the datum shifting and coordinate conversion, the
system model will probably be a simple, serial process. That is,
the output of one process will be the input for the next, with no
feedback to prior processes. For trade-off studies, alternate
system models can be developed to evaluate accuracy versus
implemental ion.
8.3.2 Error estimates. Once the system error model has been
described, error estimates must be assigned to each process and
will.data source in the model. When this is done, the analyst -.
quickly determine which error sources can be neglected and which
must be propagated through the model.
a.

Estimates of the expected errors in the input data may be

provided with the data or in product specifications.
Also, the original source or intended use of the data may
provide an indication of its accuracy (e.g. digital
data derived from a cartographic source will have an
accuracy no better than the cartographic source) .

b.

Without proper guidance, it is often difficult to

estimate the error in the data. This is especially true
for cartographically derived elevations, because map
accuracy standards for elevation allow for error in both
the horizontal location and the vertical direction.
Therefore, the allowable elevation error is dependent on
the topography and may vary within a map sheet.

c.

In addition to providing datum shift parameters,

Appendices B and E give the estimated errors in the
parameters for AX, AY, AZ; and seven-parameter
(Note: Correlations between
shifts, respectively.
parameters, which may be significant, were not available.
Therefore, they are not given in the appendices. )

d.

The accuracy and behavior of the equations presented in

this Technical Report are well defined. Table 8.1
identifies the location of accuracy guidance for the
coordinate conversion, datum shift, and map projection
equations. Additional references describing equation
accuracy and behavior are noted in the text and listed in
section 2.

e.

The errors introduced by system implementation factors

such as computer word size, accuracy and precision of
mathematical functions, truncation of terms, etc. must be
quantified.

f.

The errors introduced by hardware processes must be

determined from equipment specifications, analysis,
and/or testing.

78

Table 8.1. Accuracv auidance for eauations

Accuracy auidance section(s)

Subi ect

Coordinate Conversion Eauations

Cartesian/geodetic conversion
Height conversion

4.3.4
4.4.3

Datum shift Eauations

Seven-parameter geometric transformation
Ax,AY,AZ shifts
Direct WGS 72 to WGS 84 transformation
Vertical datum shifts

6.2.4
6.3.1, 6.3.2
6.4.2.2
6.7, 6.7.1, 6.7.2

MaD ~roiection Eauations

Mercator projection
Transverse Mercator projection
Universal Transverse Mercator (UTM) grid
Lambert Conformal Conic projection
.
Polar Stereographic
Universal Polar Stereographic

7.2.3
7.3.3
7.4.4
7.5.4
7.6.2
7.7.4

Shift Parameter Accuracv values

AX,AY,AZ shifts
Seven-parameter geometric transformation
Direct WGS 72 to WGS 84 transformation

Appendix B
Appendix E
Section 6.4.2.2

8.3.2.1 Error tv~es. Depending on the source, error

estimates may be described in several ways. Error descriptions
may have a bias component and a random (noise) component. A set
of error estimates describing the accuracy in multiple dimensions,
for example, ~~, ~*~, ~A~, may have correlations between the
parameters. Generally, with MC&G products, the error descriptions
do not differentiate between biased, correlated, and independent
random errors; the assumption is made that the error estimates are
unbiased, independent and random. The resultant error analysis
will be less accurate than if the true statistical natures of the
error sources were known.
8.3.3 Reformulation. once the error sources have been
quantified, they must be converted to a common frame of reference.
The form used for the system output error estimate is often a
convenient frame of reference.

79

a.

The reference for the error estimates must be established

(i.e. the ellipsoid or the map sheet). It may be
necessa~ to scale from one reference to another.

b.

The error estimates should be converted to the same

units .

c.

The datum shift parameter error estimates, ON, ~A~, ~&,

given in Appendix B, can be reformulated tO ~A~, Gw, ~AH
using equation set 6.8.

d.

Error estimates typically are linear, circular, or

spherical statistical values with associated
probability levels. Error estimates should be converted
to the same type of statistic at the same probability
level. Two independent linear standard deviations
(68.27% probability), 6X and CY, can be converted to a
circular standard deviation (39.35% probability) , using
(ax + 0,)
2
if the values of 6X and OY are reasonably close.
Factors to convert between different levels of
probability are given in Table 8.2.
Oc =

(8.1)

8.3.4 Error ~ro~aaation. Assuming the error estimates are

statistically independent and have been converted to the same
units and statistical type, they may be propagated through a
simple, serial system model using the Root-Sum-Square (KSS)
technique.
.
1

RSS =

[(NI)Q + (N2)2 + .... + (NJ2]2

(8.2)

If error estimates have known biases or correlations, then a more

sophisticated error propagation technique should be used.

80

Table 8.2. l?robabilitvLevel Conversion Factors

Notation
Error

Standard Error
(One Sigma)

Form

Map Accuracy
Standard

Probable
Error

Linear

(68.27%)

PE (50%)

?!QS(90%)

Circular

ac

(39.35%)

CEP (50%)

CMAS (90%)

Spherical

OS

(19.9%) I

SEP (50%)

I**SAS.

I
S~herical

Accuracv

Near Certainty
Error

30 (99.73%)
3.5CJC(99.78%)

I **SAS (90%) I 4(SS(99.89%)

Standard

Linear Error (One dimensional)

To I

5og

90%

68.27% I

99.73%

From

50%
68.27
90
99.73

1.0000
0.6745
0.4101
0.2248

1.4826
1.0000
0.6080
0.3333

2.4387
1.6449
1.0000
0.5483

4.4475
3.0000
1.8239
1.0000

Circular Error (Two dimensional)

To

39.35%

50%

1.0000
0.8493
0.4660
0.2857

1.1774
1.0000
0.5486
0.3364

90%

99.78%

From
39.35%
50
90
99.78

2.1460
1.8227
1.0000
0.6131

3.5000
2.9726
1.6309
1.0000

Spherical Error (Three Dimensional)

I

To

19.9%

50~

1.000
0.650
0.400
0.250

1.538
1.000
0.615
0.385

90%

99.89%

From
19.9%
50
90
99.89
(Note:

2.500
1.625
1.000
0.625

4.000
2.600
1.600
1.000

Errors are assumed to be random and uncorrelated. )

8.3.5 Exam~le. The commander wants a printed product

showing the transportation routes near Seoul, Korea with WGS 84,
UTM coordinates, at a scale of 1:50,000. The data source is a
Class A, 1:50,000-scale topographic map on the Tokyo datum. The
product will be created by digitizing the transportation routes on
the map sheet, shifting the coordinates from the Tokyo datum to
WGS 84 using the three-step method, and plotting the shifted data.

81

a.

The system model is shown in Figure 8.5. In this model,

it is assumed that computerrelated issues will
not degrade coordinate conversion and datum shift
calculations. Also, it is assumed that the
representation of the digitized data will not change
through the rest of the process. Often this assumption
is not valid because digital data representation must be
changed to a predetermined format. Such data
modifications can introduce error.

b.

The error estimates for each of the elements in the model

are
Error

Element
Map accuracy (horizontal)
Digitizer
UTM to Geodetic conversion
Geodetic to XYZ conversion
Datum shift equations
Datum shift parameters: ON
CAY
OAz
XYZ to Geodetic conversion
Geodetic to UTM conversion
Plotter

0.02 CMAS(90%)
~o.oo5
<1 cm
0.0
0.0
5m
3m
3m
0.1 m
<1 cm
*0.OI

Only the map, digitizer, datum shift parameter, and plotter

err~rs are significant and will be considered in the rest of
the analysis.
c.

The error estimates must be reformulated to the same

units and statistical basis. Since the commander wants
to know the estimated error on the ground, we will
convert to circular error, i.n meters, on the ellipsoid,
at 50 percent probability (CEP).
1.
Map error must be scaled from the map to the
ellipsoid and changed from CMAS to CEP.

Map Error = 0.02 CMAS

= 25.4 mCMAS
= 13.9 mCEP

0.0254 m/in x 50,000

x 0.5486

2. It is assumed that the digitizer error is a near

cer tainty I circular error specification that must be
sca led from the map to the ellipsoid and converted to
CEP .
Digitizer error = 0.005 max x 0.0254 m/in x 50,000
= 6.4 m max x 0.3364
= 2.1 mCEP

82

The datum shift parameter error estimates, GM,

~&r must be converted to ~A@ and GM, then to CEP.
3.

~*~,

and ~&, $= 37.50

Using the above values for 6&, ~Ay,
and k = 127, equation set 6.8 yields ~AQ = 3.34 m
and CJ& = 4.38 m.
CEP = 1.1774 X ~C = 1.1774 x (~A@ + G~)/2

= 4.5 m

4. It is assumed that the plotter error is a near

certainty, circular error (99.78% probability)
specification that must be scaled from the map to the
ellipsoid and converted to CEP.
Plotter error = 0.01 max x 0.0254 m/in x 50,000
= 12.7 m max x 0.3364
= 4.3 m CEP
5.

The RSS error for the complete system model is

1

RSS = [(13.9)2 + (2.1)2 + (4.5)2 + (4.3)2]2

15.4 mCEP
d.

From the above analysis, it is evident that the error in

the source data is the prime contributor to error in the
output product for this application. For other types of
applications, such as surveying, this may not be the
case.

83

Fiaure 8.5.

Svstem Model

Digitize
t

Convert UTM to
Geodetic

Convert Geodetic
to XYZ

Datum
Shift

Shift To&o Datum

to WGS 84

Convert XYZ
to Geodetic

Convert Geodetic
to UTM

Plot
I

84

Datum Shift
Parameters
AX, AY, AZ

8.4 Numerical Exam~les.

8.4.1 Examnle 8.4. Convert WGS 84 geodetic coordinates
($,h,h)W~~84 to Universal Transverse Mercator coordinates
XUTMYUTM)in the NAD 27 reference system using a three-step AX,
AY, AZ datum shift.
This example converts WGS 84 geodetic coordinates to
Universal Transverse Mercator projection coordinates in two steps.
First, the WGS 84 geodetic coordinates are converted to NAD 27
geodetic coordinates using the three-step AX, AY, AZ datum shift.
Geodetic height is needed for this conversion. Next, the NAD 27
geodetic latitude and longitude are converted to UTM coordinates.
Geodetic height is not needed for this step.
Beginning WGS 84 coordinates:
$=

420561

52.16311N= 0.74g580g18883376 radians

~=

288 22 24.326E = 1.891473118 radians

.

h = 203.380 m

Step 1: Convert WGS 84 geodetic coordinates to NAD 27 geodetic

coordinates using the threestep AX,AY,AZ datum shif t.
1.1 Convert (@,i,h)wGS~q to (X,ylZ)WG~ ~Q using equation sets
4.4 and 4.5.
a

6378137

0.00335281066474
2f . fz = 0.00669437999013

&2

a(l - f) = 6356752.3142

RN
xWGS 84

= 6388070.57383009

1
=

(RN

h)

COS(/)COS~=

1473933.5413

Y WGS 84 =

(R N

h)

cos$sink=

-4437679.0666

85

z WGS 84

RN+h
~2

sin$=

4323399.2717

1.2 Convert (X,y,z)w~~Ba to (XIY, z)NAD27 usiw ewation

6.4 and AX, AY, AZ datum shift values from Table B.1 NAS-C.
x NAD

27= x WGS 84

Y NAD

27 = Y WGS

z NAD

27 =

- AX

84

z WGS 84 -

AY

1473933.5413 - (-8.0) = 1473941.5413 m

= -4437679.0666 - 160.0 = -4437839.0666 m

AZ = 4323399.2717 - 176 = 4323223.2717 m

1.3 Convert (X,Y,Z)NAD27 to ($, k)NAD

given in Section 4.3.4.2.

kNAD

27 =

set

27

following equations

= 5.033056560575

arctan
(2::)

Calculating $ is an iterative process.

1.3.1 Compute
an initial approximation to a variable ~,using

equation 4.7. Parameters a and b for the Clarke 1866 ellipsoid

are found in Table A.

P0 =

aZ

arctan

b I/~

= 0.74788820090716142 radians
.

1.3.2 Substitute PO for

approximation for +.

9=

arctan

in equation 4.8 to get an

Z + &2 b sins

~,

11~-

= 0.74958142853674489 radians

a E2 COS3 ~

EQ = 2f - fQ = 0:0067686579973
and
&12

~2
=

l-~2

0.0068147849459

For most applications, this first iteration is sufficient.

Step 2: Convert the NAD 27 geodetic latitude and longitude to
Universal Transverse Mercator projection coordinates.
2.1 Find the UTM zone z and the central meridian lo from
equations 7.22 and 7.23.

86

180 k _ 29
~n

greatest integer S

z=

19

z=

L o = (6z + 177) ~

= 5.078908123303 radians

2.2 Calculate Transverse Mercator projection coordinates

(xm,ym) using equations 7.8 and 7.9.
2.2.1 Calculate terms in equation set 7.9 using &2 and E2 as
calculated in Step 1.3.2.
1

AO=l-~2-~4-~e
A2=~ 3

3
+ 36

&4

77
_ --&8

0.2542555420215

10-2

0.2698010542695

10-5

(
A6 =

- ZGW8

A4=~

455

15
+ ---@6

1
+ ~4

&2

175 *
- 16384E = 0.998305681856

(6 -%

) =

0.3502448582027

10-8

315
~8
-0.5044416191579 X 10-11
131072
=

A8=-

[AO@ - A2 sin 2$ + A4 sin 4$ - A6 sin 6$ + g sin 8$1

so =

so =

4756711.680282
a

N=R~=
Ah

- &2

= 6388250.562973
sin2 @

Note that N = RN differs from RN calculated in SteP l.1~ aS SteP 2

is in the NAD 27 reference frame and Step 1.1 is in the WGS 84
reference frame.
~=
t=

&cos (-p=0.6042571576496

10-1

tan$ = tan(0.74958094822852639) = 0.9308149267480

A=k-+u=o

-0.04585156272818 radians

87

2.2.2 Calculate the Transverse Mercator projection

coordinates (x~~,ym).

xry~= N

COS$

As

COS3

(l

t2 +

q2)

A5COS5$

(5 -

120
N

18t2

+ t4

+ 14q2 -

58t2q2)

A2

YTM = S$+2
N A4
+

sin $cos

sin $cos3 $(5 - t2 + 9~2 + 4~4)

24

N A6
+ ~20 sin $COS5 $(61 - 58t2 + t4 + 270~2 - 330t2~2)
%M

-214408.9715411

YTM

4760061.989246

2.3
Convert the Transverse Mercator projection coordinates
(Xm,ym) to Universal Transverse Mercator projection coordinates
(XUTM,YUTM)using equation set 7.24 (Northern hemisphere).
XUTM = 0.9996 XTM + 5001000 = 0.9996
YUTM

0.9996

YTM

0.9996

(-214408.9715411)

+ 500,000

(4760061.989246)

The resulting Universal Transverse Mercator coordinates (XuTM,Yum)

in the NAD 27 reference system are:
XUTM = 285676.792
YUTM = 4758157.964
Convert WGS 84 geodetic coordi~~Examle 8.2.
($,~,h) to geodetic coordinates (@,~,h) in the NAD 27 re~~. ...
system using the standard Molodens@ method.
8.4.2

This example converts WGS 84 geodetic coordinates to NAD 27

coordinates using the standard Molodens~ method.
Begiming

WGS 84 coordinates:

$= 42 5652 .163 N = 0.749580918883376 radians

~ = 108 2224.326 W = 1.891473118 radians
h = 203.380

88

Step 1: Compute Aa, Af, wZ, M, N, &z, and &z using equation set
6.5. obtain aW~~ 84, A NAD 27 fWGS 84 and fNAD27 from Table Al.
Aa = aWGS84

-69.4

aCLARKE 1866 =

Af = fw~~ 84 -f

CLARKE 1866 =

3.726463863

10-5

Zf - fz = 0.006694379990

&2 =

&2
&12

1 -&2

w=

41 &2

N=

RN.;.

a(l -

0.006739496742

= 0.9984449805

~i~20

6388070.574
&2)

= 6365086.681

W3

Step 2: Compute A@ and AL, and Ah using equation set 6.5, and
Ax, AL and AZ datum shifts from Appendix B for datum code NASA.
In this axample, the signs of AX, AY, and AZ are reversed from
those of Appendix B due to the direction of the datum shift (WGS
84 to local coordinates) .
&2sin~cos@Aa
A(j)

[ (-sin$cos~Ax
sin~cos@(2N

sin@sin~Ay

+ COS$AZ +

+ &2Msin2$) (1 - f)Af]/[M

A~

0.131

Ak

[-sinkAx + coskAyI
=2.615
{(N + h)cos$l

Ah

cos$cosiAx

Ah

28.65 m

+ cos$sin~Ay

+ sin$Az - WAa

w
+

h]

a(l - f)
w

sinzf Af

Step 3: Compute the NAD 27 geodetic coordinates (~,k,h) from

equation set 6.7 that results in the followina valuesl:

89

$NAD
L

AD 27 =

h NAD

L) WGS

$ WGS 84 +

27 =

L GS 84 +

A@=
AL=

42o 5652.294N
108 2221.711W

HW~~84 + AH = 232.03 m

27 =

8.4.3 Exam~le 8.6. Convert WGS 84 geodetic coordinates

84 to Mercator projection coordinates (x,Y) .

(~,

This example converts WGS 84 geodetic coordinates to Mercator

projection coordinates within the WGS 84 reference system.
Geodetic height (h) is not used in this conversion. The Mercator
projection central meridian LO is 288o = 5.026548246 radians
Beginning WGS 84 coordinates are
4z0

5fjI

52.16311N

GS 84 = 288

22

24.326E = 5.033057346 radians

$WGS
L

84 =

().74958()9189

radians

Step 1: Use equation 7.1 and WGS 84 datum parameters a and f

from Table A.1 to calculate X.
x = a(k - ~0) = 6378137(5.033057346 - 5.026548246) = 41569.357
Step 2: Calculate & from f using equation 4.9, then use equation
7.2 to calculate y.
& = ~~

= 0.0818191908426

The resulting Mercator projection coordinates (X,Y) are

X = 41569.357
Y = 5274911.868
8.4.4 ExamDle 8.7_. Converst Mercator projection coordinates
(X,Y) in the WGS 84 reference system to WGS 84 geodetic latitude
and longitude ($,k).

90

This example converts WGS 84 Mercator coordinates to WGS 84

latitude and longitude.

Beginning Mercator coordinates

central meridian

(ho):

X = 41569.3572
Y = 5274911.8684
Lo = 288oE = 5.0265482457 radians
Step 1: Compute the longitude (~) using equation 7.4 where a is
the semi-major axis of the WGS 84 ellipsoid listed in Table Al.

Ao =5.0330657221 radians = 288o 2224.326E or

71 37 35.674W
Step 2:
t

~(-~)

Compute an intermediate value t from equation 7.5.

=

~(-

5274911.8684
) =
6378137

.43734619376064

Step 3: Compute an initial value of the geodetic latitude ($0)

.
from equation 7.6.

$0=

T - 2 arctan(t) = 0.746233651 radians

Step 4: Iterate equation 7.7 until the updated value of

sufficiently close to the previous value of $.
&2 = 2f - fz = 0.00669437999013
E = 0.0818191908426
Qn+l

$1

~-

2 arctan

r111

& sin ~.& sin on.

= 0.7495688569418636

The geodetic latitude will be given by the last value of

calculated.
Final ~ c 0.7495809188833704 radians = 42o 56 52.163 N
The resulting WGS 84 geodetic latitude and longitude ($,1)
$=

42o 561 52.163N

91

is

h=

ZGGO

221

24.326IEor 71 37 35.674W

8.4.5 ExamDle 8.8. Convert Universal Transverse Mercator

coordinates (XUTM~YUTM)in the NAD 27 reference system to NAD 27
geodetic latitude and longitude($, ~).
This example converts NAD 27 UTM coordinates to NAD 27
geodetic latitude and longitude. The UTM coordinates are
converted to Transverse Mercator projection coordinates. The
Transverse Mercator projection coordinates are then converted to
NAD 27 latitude and longitude.
Beginning coordinates are
XUTM = 285,677.332
YUTM= 4,758,154.856
Zone = 19
Step 1: Convert Universal Transverse Mercator coordinates
(XUTM,YUTM)to Transverse Mercator coordinates (XTM,yTM)using
equation set 7.25.
1

XTM

500,000)
0.9996 (XUTM-

YTM

0.9996

= -214,408.4314

1
YUTM

47600588796

Step 2: Convert Transverse Mercator coordinates (XTM,YTMLto

geodetic latitude and longitude($, h).
2.1 Calculate the UTM zone central meridian
equation 7.23.

()

L o = (6z + 177) ~

(~0) using

= 5.078908123303 radians

2.2 Compute the eccentricity (&) and second eccentricity (&)

of the Clarke 1866 ellipsoid, the reference ellipsoid for NAD 27,
using equation set 7.9.
&2

Zf

f2

oe6jJ68657997zcjl

10-2

&2
&12

=
p&2=

0.6814784945915

10-2

2.3 Calculate the subscripted-A terms of equation set 7.9,

using &2 calculated in Step 2.1. The subscripted-A terms are used
in the calculation of the footpoint latitude.

92

1
~&2_

AO=l-

:4841

A2

E2+~&4+

175
&6

15

~&c-4096

- ~G3B4

455

&8 =

0.998305681856

&8 = 0.2542555420215 X 10-2

)

Aa =

15
&4+
256 (

AC =

41
35
= 0.3502448582027 X 10-8
3072 (c - -)
32 8

AB =

315
- ~3~072 E8 = - 0.5044416191579 X 10-11

3
~&6-=

77 EB = O.2698O1O542695 X 10-5
)

2.4 Compute the footprint latitude, Q 1, using equation 7.10

and a from Table Al.
Make an initial estimate for $1:
@1 =

YTM

0.7463005398433

radians

2.5 Calculate So using equation set 7.9 and the estimate of

L

01.

So = a[AO@ - A2 sin 2($)+ A4 sin 4($ - A6 sin 6$ + A8 sin 8@]

SO = 0.4735829138159 X 107
2.6 Calculate S$ using equation set 7.11 and the estimate of
$, $1SO = a[AO - 2A2COS 24) + 4A4 COS 4@ - 6AG C.os 6@ + 8AB cos 8$1
SO = 0.6364798134276 X 107
2.7 Find an updated value of the footpoint latitude, o 1 using
equation 7.10.

Sq-y
$1 (updated) =

$1 (current) -

S ,~

= 0.7501073756426 radians

Iterate Steps 2.5 through 2.7 until successive values of

$ 1 are as close as is desired.
2.8

93

For this example, the final value of $1 is:

Final @l = 0.7501073020563 radians
2.9 Calculate values used in equation set 7.12 using
equation set 7.13.
tl = tan @l = tan (0.7501073020563)
a

NI =
jl

&2

- 6,388,261.949762
sin2 ~1

a (1 - &2)
~J1_E,
sin201J

m2

&12

COS2

o1 =

~ =6,365,044.225886

0.3647692467262 X 10-2

B~ = 1 + 2t12 + ~12 = 2.740138642336

Bd = 5 + 3t12 + ~12 - 4~14 - 9t12~12 = 7.579827062321
B5 = 5 + 28t12 + 24t14 + 6~12 + 8t12~12 = 47.45050110665
BG = 61 + 90t12 + 46~12 + 45t14 - 252t12~12 - 3~14 - 66t12~14 9otlq~lQ + 225t14~14 = 172.1890097418
2.10 Calculate the geodetic latitude and longitude ($,L)
using equation set 7.12.

$ = 0.74958094314009481 radians = 42 56 52.168

h = 5.03305669671

radians = 288 22 22.464

8.4.6 ExamDle 8.2. Convert WGS 84 latitude and longitude

(@,l) to Lambert Conformal coordinates (X,y) using two standard
parallels.
This example converts WGS 84 latitude and longitude to Lambert
conformal projection coordinates in the WGS 84 reference system.

94

Beginning WGS 84 coordinates are

56 52.163IN= 0.749580918883376 radians

@WGS 84 =

AzO

288 22 24.326E = 5.033065722110382 radians

GS 84 =

Standard parallels:
@1 =

42o

30J 00.000 = 0.7417649320975901 radians

+ 2=

430

00

00.00011 =

0.7504915783575617 radians

Origin:

40 =

42o

301 00.0001! = 0.7417649320975901 radians

~=

288o 001 00.000,,= 5.026548245743669 radians

Step 1:
& =

~~

Step 2:

Calculate eccentricity & from flattening f.

= 0.0818191908426
Calculate q at $, $1,$2 and $0 using equation set 7.27.

q($) = 0.8270301921136922
q(@l) = 0.8164293462094778
q(@2) = 0.8282703592274435
q($o) = 0.8164293462094778
Step 3: Calculate N at $1 and @2using
from Table A-1.

N(@i) =

1 - E2 sin2 @i

95

equation set 7.27 and a

N($l) = 6387903.46785173
N($2) = 6388090.05758619
Step 4:

Calculate L from equation set 7.27.

ln(Nlcos@l) - ln(Nzcos$z)
L=

= o. 6788029306559900

Q-ql

Step 5:
K=

Calculate K from equation set 7.27.

lcOs@~
~=~2076~69
L &-Lql = L &-Lqz

Step 6:

25203636
.

Calculate r and r. from e~ation

set 7.27.

r = K&-Lq(@ = 6888432.18075871
rO = K&-@@o) = 6938179.32109766
Step 7: Calculate the Lambert coordinates
set 7.26.
x = r sin L(Ly=ro-

(x,Y) using equation

LO) = 30474.8898082

r cos L(L - LO) = 49814.5521555

8.4.7 ExamDle 8.10. Convery Lambert Conformal projection

(with two standard parallels) coordinates (x/Y) in the WGS 84
reference system to WGS 84 geodetic latitude and longitude ($,X).
This example converts WGS 84 Lambert Conformal coordinates to
WGS 84 latitude and longitude.
Beginning Lambert Conformal coordinates, standard parallels
(@l,$2) and origin (~o,ko) are
= 30474.890
49814.552
Y=
0 1 = 4Z0 3010011 =
=
0 2 = 430 0000
0 = 4Z0 301(3011 =
k:=zBBo
001()()11=
x

0.7417649320976
0.7504915783576
0.7417649320976
5.0265482457437

96

radians
radians
radians
radians

Step 1: Compute &, L, rO,k,Nl,Nz,~,ql,

7.27. Obtain a from Table Al.
&=l/2f-f2

and qz using equation set

= 0.0818191908426

0.8164293462094778
0.8164293462094778
0.8282703592274435
a

4 1-

&2 sin2 $i

6387903.467851727
6388090.057586191
.

ln(Nlcos$l)-in (Nzcos@2)
cb -%
NICOS@l
Le-Lql
Step 2:
r.

N2COS$2
= 12076169.25203651
Le-Lq2

Compute r. using equation set 7.27.

Ke-J@(@Oj = 6938179.321097849

Step 3:

Compute r and e from equation set 7.29.

(3= arctan

r=

= 0.6788029306559713

x
rO-y

= 4.4240820859501107 X 1O-3 radians

rO-y

Step 4:

Cose

= 6888432.18091526

Compute geodetic longitude (h) using equation set 7.28.

L=g+ko=

5.0330657221511255

97

radians = 288o 22 24.326

Compute q using equation set 7.28.

Step 5:

in ($)
q.

= 0.8270301920802524

Find an initial approximation

Step 6:
latitude from equation set 7.30.
o

($) for the geodetic

n
I = 2 arctan (eq) - ~
= 0.746233650926621

Step 7:

Compute f(@)

and f (~) from emation

set 7.30.

f($) = -q+~ln
1

[(=)*(;:::;:!)
X 10-3
f(q))= -4.5493507336696604
1- ex
= 1.357001593456580
(1 - &2 sin2 $)cos@

f(q))=
Step

Find an updated approximation

latitude using equation 7.31.

8:

I
II=
o

_JQILL

=
f(q))

($) for the geodetic

0.7495861530575
.

Step 9: Iterate steps 7 and 8 until successive values of

sufficiently close after many iterations.
Final (($))
= 0.74958091885881052 radians = 42o 56

52.163

are

The resulting WGS 84 coordinates are:

@ = 42 56 52.163
~ = 288o 22 24.326I
Exam~le 8.11. Convert geodetic latitude and longitude
($,~) on the International Ellipsoid to Universal Polar
Stereographic coordinates (x,Y).
8.4.8

This example converts geodetic latitude and longitude to

projection coordinates. Coordinates are computed on the
International Ellipsoid.

Beginning coordinates are

$=
~=

8717 14.400 S = -1.523451361952464 radians

132 14 52.303 E = 2.308160619653466 radians

Step 1: Compute the eccentricity (&) of the International

Ellipsoid using equation set 7.38 and a and b from Table Al.
&2 = 2f - fz= 0.0067226700623
0.08199189022
E=
Step 2:

K=
Step 3:

Compute k using equation set 7.38.

2az
(1-E);
[
1
b
(1 + e)

12713920.17558161

Compute r using equation set 7.38.

r=
Step 4:

Compute UPS coordinates using equations 7.49 and 7.50.

x=

2,000,000 + 0.994 r sin ~ = 2222991.410

y.

2,000,000 - 0.994 r cos k = 1797464.051

8.4.9 Examnle 8.lJ. Convert Universal Polar Stereographic

coordinates (x,Y) to geodetic latitude and longitude (~,~).
This example converts UPS coordinates calculated on the
International Ellipsoid to geodetic latitude and longitude on the
International Ellipsoid.
Beginning UPS coordinates in the Southern hemisphere are

x=

2222991.410

Y=

1797464.051

Step 1: Convert Universal Polar Stereographic coordinates

(xupS,yuP~)to Polar Stereographic coordinates (XPS,YPS).

99

x Ps

TTDC!

2000000

224337.434205231

0.994

Y UPS

Y Ps

- 2000000

0.994

-203758.500100604

Compute the longitude (~) from equation 7.39.

Step 2:

7C + arctan

L=

x Ps
YPS

k n 2.308160619653466

radians = 132 1452.303 E

Compute k using equation 7.41, and a and b from Table

Step 3:
Al.

e
(1 _ &) ;

2a2
K=

= 12713920.17558161

Compute q using equation set 7.40.

Step 4:

Iwps

e-q
K

(1 + e)

!COS

Al

3.73652495299514

Step 5: Compute an initial estimate ($.) of the geodetic latitude

using equation set 7.42.

$o
Step 6:

= 2 arctan eq - ~ = 1.52313176628106

radians

L1

Compute f($o) and f($o) using equation set 7.42.

f($) =
fu.po)
f ($)

-cl+

1
~ in

sin
sin )(
~

1+

[[1-

-&

1
sin ~
1 + & sin Q

)1
c

-0.006730109061328
1-

&2

(1 - &2 sin2$) cos~

20.9875717033244

Step 7: Compute an updated geodetic latitude $1, using equation

set
7.43.
f ($n-J
= 1.523452437445710
0 n = 4 n1 - f ($J

100

Step 8: Iterate steps 6 and 7 until successive values of the

geodetic latitude are sufficiently close.
Final

($) = 1.523451361952464 radians = 87o 17 14.4S.

The resulting geodetic latitude and longitude ($,h) coordinates

are:
@ = 87o 17 14.4S
~ = 132 1452.303E
8.5 List of test Doints. The test points listed here are
taken from DMA TM 8358.2 (1989) and Snyder (1987).
Mercator
Central Meridian = 180
Ellipsoid
= Clarke 1866
@=35N
~= 75 w

(285 E)

x = 11,688,673.7
y.
4,139,145.6

m
m

k = 1.2194146
Universal Transverse Mercator
Central Meridian = 2850 E (75 W)
= Clarke 1866
Ellipsoid
$=

4030

k=

73 30 W

x.
y.

k =

(286 30 E)

627,106.5 m
4,484,124.4 m
0.9997989

Lambert Conformal Conic

Ellipsoid
= Clarke 1866
Standard parallels = 33 N
= 45 N
Origin:

Q = 230 N
k: = 96o W

(264o E)

101

@=35
N
k = 75 W (285 E)
x=

Y=

1,894, 410.9
1,564,649.5

m
m

k = 0.9970171
Universal Polar Stereographic
Ellipsoid = International

87o 17 14.400S
h = 132 14 52.303E

Y=

1,797,464.051

x = 2,222,991.410

m
m

102

9.

NOTES

9.1 Intended use. This Technical Report provides guidance

on coordinate conversion and datum shifting for a wide range of
Army mapping, charting, and positioning applications. Guidance is
provided to aid the Army and others in developing software for
applications of datum shifts and other transformations and in
implementing this software in support of operational units.
9.2

Subiect (kev word) listinq

Cartography
Charts
Coordinate Systems
Datums
Elevation
Geodetic Satellites
Geodesy
Geodetic Coordinates
Geographic Coordinates
Grids
Hydrographic Surveying
Latitude
Longitude
Map Projection
Mapping
Maps
MC&G
Mean Sea Level
Military Standards
Photogrammetry
Projections
Surveying
Topographic Maps
Topography
UTM

103

APPENDIX
REFERENCE
10.

ELLIPSOID

A
PARAMETERS

GENERAL

10.1 Sco~e. This appendix identifies reference ellipsoids with

associated geometric parameters.
20.

APPLICABLE DOCUMENTS.

The ellipsoids and associated parameters are taken from DMA

TR 8350.2 (Edition 2, Insert 1.) Note: Use semi-major axis a and
flattening f to calculate E and E . Do not use semi-minor axis b,
which is a derived geometric constant.
30.

GENERAL REQUIREMENTS.

30.1 A?mlication. The ellipsoid furnishes a simple, consistent,

and uniform reference system for geodetic coordinates.

105

u)

c1

H
o
m

-1+

0
0
0
w

.$Jmo

-ricrlti

APPENDIX

AX, AY, and AZ DATUM SHIFT PARAMETERS

10.

GENERAL

ScoQe. This appendix identifies the ellipsoid center shift

transformation parameters used to transform a local geodetic system
to WGS 84. An identifier, the reference ellipsoid, AX, AY, AZ
shifts, and the estimated errors (0) in the shifts are listed for
each local geodetic system.

10.1

20.

APPLICABLE DOCUMENTS.

The datum shift parameters and associated information for each

local geodetic system are taken from DMA TR 8350.2 (Second Edition,
Insert 1.)
30.

GENERAL REQUIREMENTS

The AX, AY, and AZ shifts are lsed for datum

30.1 Amlication.
transformations using the three-step method discussed in section
6.3,
and shown by e~ation sets 6.3 and 6.4.

___

-.-
---

- - -- -

The following symbology is used in Table B.1.

*
**

WGS 84 minus local geodetic system.

Derived from non-Dobbler sources, accuracy unknown.

***

Derived from a single Doppler station, accuracy not verified.

See Appendix F to transform Old Hawaiian Datum coordinates on

the International Ellipsoid.

++

Also known as Hito XVIII, 1963.

+++

Use A = 6,377,483.865 meters for the Bessel 1941 Wllipsoid in

Namibia.

&

Derived from non-satellite sources, accuracy unknown.

107

o
m
CQ
l-l

d
d

(n
fj

r-i
F1

la

..
c1
H

m
m

m
I

Lf)omwmm?+m

CNl+oJm

Inmcnmulmww

oloForlbm
l-l
I

lllr+lll
1

J-1
I

c1

WI
g

x
bq

u
Id

II
o
o

a
co
l-l

<

-i
4.J

WI

L
4JCN
GO

HT-I

a
vi
o
u)
,fJl
l-i
d
w

Ax

ASTRODOS7114
St .

Helena

ASTRONOMIC
Marcus

Island

(FRIG)

1961

Tern

Island

1952

and

Tasmania

Island

GEODETIC

Australia

and

Tasmania

Island

cJy

550

25

Az

CJz

-494

25

124

25

-234

25

-25

25

114

25

-116

25

-333

25

-133

-48

148

-134

-48

149

25

-129

25

145

25

-127

20

-769

20

472

20

-73

20

213

20

296

20

1924

TRN

International
1924

**

Australia

AUSTRALIAN

International

ASQ

***

GEODETIC

-320

Ay

1924

ISLAND

AUSTRALIAN

International

Island***

STATION

ASTRO/TERN

SHE

CJx

1966

AUA

Australian
National

1984

AUG

Australian
National

.
AYBELLA

LIGHTHOUSE

PHA

Clarke

1880

IBE

International

.!9

Djibouti***

BELLEVUE
Efate

(IGN)
and Erromango

1924

Islands

BERMUDA
Bermuda

1957
Islands

BER

Clarke

1866

*
*

:
4J

C
d

v-i

El0
G
fd

$!
.

In

k
*
*

u
i?

0
r

:
02

L
tt
M
r
t+
H
m

C/i
0

H
u

c1
b
*

P
0
n)

H
M
w

M
P
P

%
(n
o

E
A

ILL

Wd

Wmrm

CO(A)O
cmolP

b
m

(4

C!o
4

m
)P
I

(n
.

i-l
.

F)

&

c1
Cu

M
lb

R
&
M

CQ
A
I

6
N

RRm(+n

*PM
*NW

x
M
z

P
r
H
:

H
u

i!P
I

z
lb

*
Ow

El

co
Its

P
*

i-

*
*

H
q
H

co
h

I
C2

mlb

ii)tumkn

(DIE+
(D I!iz

?!

Cn

a
m
E
H
H
q
H

m
w

.r
r

0
H
u

m
0
(2

ill
H
Cn

n
*(I-

fii

I&l

iu

ti

(-t
o

-!.l
4

a)
lb
I

(-)
0
%

Hl

A
c1
.
I

Q
ul
1-

(9
M

(n(c)

oh

(n
D
cl)

%
*

M
Cu (J
w

8
G

(-t

m
*

*
*
*
H

(3
(D

&

(D

o
H
u

d-

I(3

tD
M
(n

1:

1%

Is
co
o

c1
>

tQtQtQrQNrQ

R E

1tD
Cn

Cn

H1

Hl

(n

m
IP
I

rQrQh3fQ

W(mwoulo

F1
Id

*
*
*

c)

*
*
*
(d
d

to

Iii
10

*
P
1-

iD

I
H

u
m
z

H
H

0
H
u

ix)
m
m

E
r

RI

m
b
r

u
I

C/l

F1
Id

*
*
*

c)

*
*
*
(d
d

i3tio

(3

*
*
*

m
1P

t+
o

(n

II

c1

(0

h-l

r+-1
-(n (.
a

cl)

H
b
I

f-t

(2

!5
I

II

c1

APPENDIX C
DATUM LIST
10.

GENERAL

This appendix consists of a list of countries and

geographic areas and their associated datums.
10.1

20.

Scope.

APPLICABLE

DOCUMENTS.

This section is not applicable to this appendix.

30.

GENERAL

REQUIREMENTS

30.1 ADDlication. The intended use of this appendix is to establish a

table of countries or geographic areas for quick reference to their
respective datum.

129

TABLE C-1
Countries and Their Associated

COUNTRY/GEO~

Datums

HIC

Afghanistan

Herat North

Alaska

North American

1927, 1983

Alberta

North American

1927, 1983

Antigua

Antigua

Island Astro 1943

North American

Caxnpo Inchauspe,

Argentina

South American
Ascension

1927

Island

Ascension

1969

Island 1958

Australia

Australian Geodetic 1966, 1984

Austria

European 1950, 1979

Bahama Islands

Cape

Canaveral,

North

American

Bahrain Island

Ain el ABD 1970

Baltra, Galapagos

South American

1927

1969

Islands
Bangladesh

Indian
North American

Barbuda

1927

North American 1927

TABLE C-1
(cent d)
Countries and Their Associated
Datums

COUNTRY/GEC)GQH

IC

Belgium

European

Belize

North

Bolivia

Provisional

American

South
ARc

Brazil

Corrego
Columbia

Brufiei

Burkina

Caicos

1969

Alegre,

South

American

1969

North

American

1927, 1983

1948

Adindan

58

Point
Burundi

American

1950

Timbalai
Faso

1927

South

American

Botswana

British

1950

ARc 1950
Islands

Cameroon

North

American

1927

North

American

1927, 1983

North

American

1927

Adindan
Minna

Canada

Canal

Canary

Zone

Islands

Pico

de Las Nieves

1956,

TABLE
Countries

Cam

(centd)

C1

and Their

Associated

DATUM

TRY/G EOGW?J?HIC AREA

Caribbean

Caroline

Cayman

North
Islands

Brat

Island

American

L.C.

1951

5 Astro

1961

America

North

Chatham

Island

Chatham

Anna

Astro 1971

Island

Provisional

Island

1927, 1983

American

South
cocos

1927

Astro

Kusaie

Central

Chile

Datums

South

1956,

1969, PS Chile 1963

American

1 Astro

American

1965
.

Colombia

Bogota

Observatory,

South

American

Congo

Point

Noire

CONUS

North

American

Corvo

Island

Costa

Rica

Cuba

(Azores)

1969

1948

Observatorio

1927,

Meteorologic

North

American

1927

North

American

1927

European

1950

1983

1939

TABLE
Countries

CouNTRY/GEO

GRAPHIC

1950

073 Astro

Aybella

Dominican

Republic

East

Canada

East

Falkland

Island

Island

Eastern

United

Ecuador

States

Datums

QAxW

ISTS

Djibouti

Eftate

Associated

European
Garcia

Easter

and Their

(centd)

AREA

Denmark
Diego

C-1

1969

Lighthouse

North

American

1927,

1983

North

American

1927,

1983

Sapper

Hill

Easter

Island

North

1943

American

1967
1927,

1983

Provisional
South American
South American
1969
Island

Bellevue

(IGN)

Egypt

European
19S0
Old Egyptian

El Salvador

North

England

European
of Great

1950, Ordnance
Britain
1936

Belleme

(IGN)

Erromango

Island

1956

American

1927,

1983
Survey

TABLE
Countries

COUN TRY/GEO GWHIC

and Their

Associated

Datum

AREA

Mass awa

Eritrea

Espirito

Santo

Island

Santo

(DOS)

1965

Adindan

Ethiopia

Faial

(cent d)

C-1

Island

Federal

(Azores)

Republic

Germany

of

(before

Federal

States

Graciosa

Base

European

1950

1948

1990)

of

Kusaie

1951

Micronesia
European

Finland

Flores

Island

Florida

(Azores)

1950 , 1979

Observatorio

Cape

Meteorologic

Canaveral

North

American

France

European

Gabon

Mporaloko

Ghana

Leigon

Gibraltar

European

1950

1950

1927, 1983

1939

TABLE
Countries

COUNTRY/GEOG~C

Gizo

Grand

Island

Canyon

Greece

(
Greenland

and Their

(cent d)

Associated

Island (New
Georia Islands)

Graciosa

C-1

DATW

DOS

1968

Graciosa

North

Base

American

European
(Hayes

Datums

North

SW

1950

1950

American

1927

Peninsula)

Guadacanal

Island

Gux

1 Astro

Guam

Guam

1963

Guatemala

North

Guinea-Bissau

Bissau

Guinea

Dabola

Guyana

Provisional
South

Hawaii

Honduras

Old

American

1927,

South

American

1983

American

1969

Hawaiian

North

American

1983

North

American

1927,

1983

1956

TABLE
Countries

(cent d)

C-1

and Their

Associated

COUNTRY/GFIOG~HIC

Hong

Kong

Datums

P~TW

Hong

Kong

1963

Iceland

Hjorsey 1955

India

Indian

Iran

European

1950

Iraq

European

1950

Ireland

European

1950, Ireland

Isle of

Ordnance
1936

Survey

Israel

European

1950

Italy

European

1950

Iwo Jima

Astro

Beacon

Jamaica

North

American

Japan

Tokyo

Johnston

Island

Island

Jordan

European

1950

Kalimantan
Island
(Indonesia)

Gunung

Segara

of Great

Johnston

1965

1927

1961

Britain

TABLE
Countries

C-1

(centd)
Associated
Datums

and Their

COUNTRY/GEOGRWHIC

Kauai

Old Hawaiian
North

Kenya

American

1983

ArC 1960

Kerguelen

Island

Kerguelen Island

Korea

Tokyo

Kuwait

European

1950

European

1950

1949

Lebanon

Leeward

Islands

Fort

Thomas

Antigua

Island

Montserrat
Liberia

Liberia

Luxembourg

European

Madagascar

Tananarive

Madeira

Porto

Mahe

Malawi

Islands

Island

Mahe

Arc

Astro

Island

1950

1971

Observatory

1936

1943

Astro

1964

Santo

1950

1955

1958

TABLE
Countries

and

C-1

Their

(centd)

Associated

CWNTRY/GEOGRAPHIC

DATW

Malaysia

Timbalai
Republic

Maldives,

of

Adindan

Malta

European

Manitoba

North
Islands

Marshall

Islands

1948

Gan

Mali

Marcus

Datums

1950

American

Astronomic

Wake

1927,

Station

Eniwetok

1983

1952

1960
.

Mascarene

Masirah

Island

Island

Reunion

Nahrwan

(Oman)

Maui

Old Hawaiian
North

American

1983

Mexico

North

American

1927,

Micronesia

Kusaie

1951

Midway

Midway

Astro

Mindanao

Island

Island

Luzon

1961

1983

TABLE
Countries

C-1

and Their

(cent d)

Associated

Datums

Montserrat

Montserrat

Morocco

Merchich

Namibia

Schwarzeck

Nepal

Indian

Netherlands

European

Nevis

Fort

New Brunswick

North

American

1927,

1983

New Foundland

North

American

1927,

1983

New

Geodetic

Zealand

Island

1950,

Thomas

Datum

North

American

Niger

Point

58

Nigeria

Minna

Northwest

Norway

Ireland

Territories

Ireland

North

1979

1949

1927

1965

American

European

1958

1955

Nicaragua

Northern

Astro

1950,

1927,

1979

1983

TABLE
Countries

c~~TRy/GEoGR-=HIc

Nova

C-1

and Their

(centd)

Associated

T~

Scotia

North

Oahu

Datums

American

1927,

1983

Old Hawaiian
North

American

1983

American

1927,

Okinawa

Tokyo

Oman

Oman

Ontario

North

Paraguay

Chua Astro,

South

American

1969

Peru

Provisional

South

American

1956,

South
Philippines

Luzon

Phoenix

Canton

Pico

Islands

Island

Pitcairn

Porto

Island

Santo

Portugal

Puerto

(Azores)

Island

American

Astro

Base

Pitcairn

Astro

Santo

European

Rico

1966

Graciosa

Porto

Puerto

SW

1967

1936

1950

Rico

1969

1983

TABLE
Countries

C-1

and Their

(cent d)

Associated

Qatar

Qatar

National

Quebec

North

American

Salvage

Selvagem

San Salvador

Santa

Maria

Island

Islands

North

Datums

1927,

Grade

1983

1938

American

1927

Sao Braz

(Azores)

S~o Jorge

(Azores)

Graciosa

S~o Miguel

Island

S;o Braz

Sarawak

and Sabah

Base

Timbalai

1948

Sardinia

European

1950, Rome 1940

Saskatchewan

North

Saudi

Ain

American

El Abd

1927,

1970,

1983

European

1950,

Nahrwan
Scotland

European
Great

Senegal

Shetland

1950,
Britain

Ordnance

Survey

of

Survey

of

1936

Adindan
Islands

European
Great

1950,
Britain

Ordnance
1936

TABLE
Countries

(centd)

C-1

and Their

Associated

Datums

HIC

Sicily

European

Singapore

Kertau

Somalia

Afgooye

South

Africa

Cape

South

Chile

Provisional

South

Geogia

South

Greenland

Islands

1950

1948,

Sri Lanka

Kandawala

1950,

Dos

St. Kitts

Fort

Thomas

Sudan

Adindan
Island

Chilean

1968

Astro

Sumatra

Asia

Qornoq

European

Island

South

ISTS Astro

Spain

St. Helena

South

Djakarta

(Indonesia)

Surinam

Zanderij

Swaziland

Arc

1950

1979

71~4

1955

(Botavia)

1963

TABLE
Countries

C-1

and Their

(centd)

Associated

Datums

Co~TRy/~U213A-PSJULA-E.AE

Sweden

European

1950, 1979

Switzerland

European

1950, 1979

Syria

European

1950, 1979

Taiwan

Hu-Tzu-Shan

=C

Tanzania

Tasmania

Island

Terceira

Islands

1960

Australian

Graciosa

Geodetic

Base

1966,

1984

SW

(Azores)

Tern

Astro/Tern

Island

Tristan

and

Tobago

da Cunha

Naparima,

Tristan

Tunisia

Carthage

Turk

North

and Caicos
Islands

United

Uruguay

(Frig)

1961

Indian

Thailand

Trinidad

Island

Arab

Emirates

Nahrwan

Yacare

BWI;

South American

Astro

American

1927

1969

TABLE
Countries

(centd)

C-1

and Their

Associated

Datums

COUNTRY/GEW~C

Venezuela

Provisional
South American
1969
South American

Vietnam

Indian

Virgin

Puerto

Islands

Rico

Viti Levu Island

(Fiji Island)

Indian

Wales

Ordnance Survey of
Great Britain
1936

Wake

Western
West

Wake

Atoll

Island

European

Europe

Kertau

Malaysia

Astro

1952

19S0
1948

North

American

1927, 1983

Yukon

North

American

1927, 1983

Zaire

Arc

Western

United

States

1956,

1950
.

Zambia

Arc

1950

Zimbabwe

Arc

1950

APPENDIX D
WFS 84 Geoid Separation Computation
10.0

GENERAL .

10.1 Sco~e. This appendix provides the bilinear method for

interpolating geoid separation and WGS 84 Geoid Se~arations on
a
.
100 x 10 grid. These separations are expressed in meters.
20.

APPLICABLE

DOCUMENTS.

This section is not applicable to this appendix.

30.

GENERAL REQUIREMENTS.

30.1 The Bilinear Inter~olation Method of Calculating WGS 84

Geoid Se~aration (NWGS84~* The bilinear interpolation method uses
interpolation in two directions ($,x) to approximate the geoid
separation NWc~84, from four known WGS 84 geoid separations given
in a lookup table. The hi-linear interpolation method has the
form given in equation set D.1. Figure D.1 shows the relationship
between the positions of the known NW~~ *4values and the
interpolation point.

Np($,h

= aO + alX + aQY + a~XY

(D.1)

Geoid Separation (N) to be interpolated at point

P($, x) .

aO

NI

al

Nz -

a2

N4 -

a3

N1 + N3

x
=

-N2-N4

(L - L,)
(L2 - il)

145

@ = Geodetic latitude of Point P

k = Geodetic longitude of Point P
N 1, NQ, N~, Na = Known geoid heights at grid points used in the
interpolation process.

Figure F-1 Coordinate System Associated with

Geoid Separation Bi-Linear Interpolation Method
30.2 NWGS 84 Tables. NWGS 81 values are tabulated on a 10 x 10
grid in Table D.1. The values in Table D.1 are suitable for use
in moderate accuracy mapping and charting applications. Qualified
users may obtain digitallystored tables of NWGS84 values on
30 x 30 and 1 x 1 grids for survey and high accuracy mapping
and charting applications.

146

m 0

r-l

m Cn (
w m L
I

v
4

Ln
r)

=r
In

I
v
m

0
w

t-l

ml m m m c
I 4 N m L
I
I
I

0
m
t-i

Kr a) m w c
r--l t-i P--) KJ+ ~

0
w

l-l

I
1= CN

0
y

w <
m ~

0
m
!+

I
m
m
I

ml m Cn P
ml ml N (=
I
I
I

ml

l-l

N In w w F c
m r-i A A c P1 I I I
I

t-l N t-i u

ml

1-1-

PI

0
t-i

l-l

u c
I cc

o-

m a) 1= w
t+ m N +

c c
(-q

Oti
co -I-i

Q+ a 0 (-N T 0
m
N m m OJ

0:
l--

0 COm w u 0
0-) m m ml
m

t-i l-n u) w u 0
Kr Uf m O1
m

0
m

0 u) In 0 T 0
m
m =3 r-) m
I

m m 0 l-l ~
m m m ml

t-i w u) r- rm m N l-l

0
ml

Or-1
UI
@
-d

w w 0 w!0
CNml l-l l-i I m

a)

or-i

r-i
.
c1
lam
I-J
LQU2

E+-

II

QZ

al m w W=a 0
I-i m l--i !-l I m
I

II

gz

0
rl

0 00
Kr U)w
I

Oc
F a
I

m
I

0
v
m
0
m
m
0
N

m
0
r-i
m

0
m
NCn

o
u+

II

ml

o
m
ml

-r-i

(w

GO

00
-A

r-i

ml

fax

o
0
N

o
cm
r-i
r-i
.

o
m

(2

!--l

co

t-l

E-+

-@W
II

w)

gz

-f-l

b-l

00

a-d
tJ -

APPENDIX

Seven-Parameter Geometric Datum Shifts

10.

GENERAL

10.1 Sco~e. This appendix provides the data needed to perform

sevenparameter geometric datum shifts between two local geodetic
systems and WGS 84.
20.

APPLICABLE

DOCUMENTS.

This section is not applicable to this appendix.

30.

GENERAL REQUIREMENTS

30.1 A~~lication. Table E1 lists the coverage area, ellipsoid

applicable to the local geodetic system, ellipsoid center shifts
(AX, AY, AZ), axes rotations (&, ~, 0), and scale factor change
(As) for use in seven-parameter
geometric
datum shifts between
two local
geodetic
systems
and
WGS
84.
The
rootmean
square differences
between Doppler-derived
WGS 84 geodetic
coordinates and WGS 84 coordinates, computed from local system
coordinates using the seven-parameter model, are provided as an
indication
of
the
datum
shift
consistency.

149

TABLE E.1

Seven-Parameter Local Geodetic System to WGS 84

Datum Transformations

DATUM :

EUROPEAN 1950 (EUR-M)

COVERAGE AREA:
ELLIPSOID:

AX = -1o2

Western Europe

International

AY = -102 m

& = 0.413

AZ = -129 m
@ = 0.385

~ = 0.18411

RMS DIFFERENCES
A$=2m

: WGS

AS = 2.4664

10-6

84 surveyed, minus computed coordinates

AH=2m

Ak=3m

DATUM :

ORDNANCE

COVERAGE

AREA:

SURVEY

OF

GREAT

BRITAIN

1936

England, Isle of Man, Scotland, Shetland Islands,

and Wales

ELLIPSOID:

AX = 446 m
& = -0.945

Airy

AZ = 544 m

AY = -99 m
~ = -0.26111

m = -0.435

AS = -20.8927 X 10-6

RMS DIFFERENCES : WGS 84 surveyed minus computed coordinates

A@=2m

Ak=2m

AH=lm

150

APPENDIX F
Old Hawaiian Datum with International Ellipsoid
10.

GENERAL

scoDe. This appendix provides the information needed to

shift coordinates between the Clarke 1866 and International in the
Old Hawaiian datum.
10.1

20.

APPLICABLE DOCUMENTS
DMA SGG MEMORANDOM FOR RECORD
SUBJECT: Old Hawaiian Datum with
International Ellipsoid
13 June 1990
Army Map Service GEODETIC MEMO
No. 687, SUBJECT: Hawaiian Islands,
UTM Coordinates and International Spheroid
Geographic Positions. P.O.# 23832-002,
29 November 1950.

30.

GENERAL REQUIREMENTS

30.1 Background. Positions on the old Hawaiian datum may be on

either the Clarke 1866 ellipsoid or the International ellipsoid.
Users of Old Hawaiian datum positions must determine which
ellipsoid applies. Datum shift parameters from Old Hawaiian datum
to WGS 84 have been detemined only for the Clarke 1866 ellipsoid.
Old Hawaiian datum coordinates on the international ellipsoid must
be converted to the Clarke 1866 ellipsoid before being shifted to
WGS 84. The conversion between International and Clarke 1866
ellipsoids has been defined in terms of UTM coordinates. Geodetic
latitude and longitude must be converted to UTM before changing
ellipsoids . Height conversions should not be attempted as no
conversion method is available.
30.2 A?mlication. To convert Old Hawaiian datum UTM coordinates
on the International ellipsoid (EINT,NINT) to UTM coordinates on
the Clarke 1866 ellipsoid (EC66,NC66) use:
In UTM Zone 4

EC66= EINT- 3
lJC66 = NIN~ -

(1.1)

169

In UTM Zone 5

EC6G= E1~T+ 13
NC66= NINT - 169

(1.2)

151

To convert Old Hawaiian datum UTM coordinates on the Clarke

1866 ellipsoid to the International ellipsoid use:
In UTM Zone 4
EI~~ = E~66 + 3

(1.3)

N1~T = NC66 + 169

In UTM Zone 5
(1.4)

EIN~ = E~66 - 13
NINT = NC66 + 169

30.3 Exam~le. Convert geodetic survey coordinates on the old

Hawaiian datum, International ellipsoid, to WGS 84 geodetic
coordinates.
=

o INT

k INT
This

a.

21 19 37.431N

= 157

58

position

is

25.61OW
on the island of oahu.

Convert $INT,

UTM :

Zone = 4
EINT = 606,428
NINT = 2,358,722
b.

Convert EINT, NINT to the Clarke 1866 ellipsoid using

equation set 1.1, for zone 4.
E~66 = 606,425
NC66 = 2,358,553

c.

Convert ECGG, NC66 to geodetic coordinates on the

Clarke 1866 ellipsoid.
o C66 = 21 19 37.425N
L C66 = 157 58 25.631W

d.

AX, AY, AZ shift parameters for datum identifier code

OHA-D should be used for shifting geodetic survey
coordinates on oahu.

e.

Shift @c66, ~c66 to WGS 84 using, for example, the

three-step method.

152

19 26.0691JN
k WGS 84 = 157 58 15.766W
@WGS

84

zl

153