International Journal of

Environmental Research
and Public Health

Article
Spatial Layout Planning of Urban Emergency Shelter Based on
Sustainable Disaster Reduction
Wenlong Zhu 1, * , Houlong Xing 1, * and Wenlu Kang 2

1 Department of Urban and rural planning, School of Architecture and Design, China University of Mining and
technology, Xuzhou 221116, China
2 Department of Architecture, School of Architecture and Design, China University of Mining and technology,
Xuzhou 221116, China
* Correspondence: zhuwl@cumt.edu.cn (Z.W.); ts20190011a31@cumt.edu.cn (X.H.);
Tel.: +86-1381-3298-665 (Z.W.); +86-1954-9160-508 (X.H.)

Abstract: An urban emergency shelter provides a place of temporary life and shelter for victims after
a disaster. As a very important public service facility, its spatial layout is greatly related to the security
of lives and the property of the urban residents. Upholding the concept of sustainable disaster
reduction, this study took the central urban area of Xuzhou as an example. Based on the analysis of
ArcGIS software, this study analyzed and planned the spatial layout of emergency shelters in Xuzhou
and visualized the service area ratio, service population ratio, service capacity ratio, and service
overlap rate of each administrative district. Finally, 73 fixed emergency shelters were determined,
among which eight were classified as central shelters. At the same time, through consulting the
relevant data, it was found that similar problems such as potential safety hazard, blind areas, service
overlapping, and mismatch of shelter layout and actual needs also exist in other cities. Finally, in
light of the existing problems, relevant suggestions are provided for the adjustment and optimization
of the layout of emergency shelters.

Keywords: emergency shelter; central urban area of Xuzhou; ArcGIS; layout optimization

Citation: Zhu, W.; Xing, H.; Kang, W.

Spatial Layout Planning of Urban
1. Introduction
Emergency Shelter Based on Emergency shelters are important facilities for displacing victims in response to earth-
Sustainable Disaster Reduction. Int. J. quakes, fires, floods, epidemics, and other public emergencies, and are the guarantee of
Environ. Res. Public Health 2023, 20, urban emergency response and disaster prevention [1]. To maximize their shelter and relief
2127. https://doi.org/10.3390/ functions after a disaster, the scientific planning, layout, and maintenance of emergency
ijerph20032127 shelters are some of the key issues that need to be urgently studied in the field of disaster
Academic Editor: Wei Wu
prevention and mitigation in China [2].
Due to accelerating urbanization, and the highly-dense urban population and build-
Received: 19 December 2022 ings, all kinds of natural and man-made disasters have become more severe. More and
Revised: 6 January 2023 more experts and scholars recognize the great significance of reasonable planning and the
Accepted: 6 January 2023
construction of emergency shelters for the safety and sustainable development of cities. Yu
Published: 24 January 2023
Lixin proposed building additional shelters and other measures based on the analysis of
the layout and traffic accessibility of existing emergency shelters [3]; Chen Mingjie and Lv
Fei evaluated the network in terms of stability, vulnerability, and balance of the network
Copyright: © 2023 by the authors.
structure in the ancient city of Suzhou and proposed optimization measures [4]; Yu Sihan
Licensee MDPI, Basel, Switzerland.
matched the data of urban functional areas and population spatial distribution by way of
This article is an open access article cellular signaling data to project the changes in the spatial and temporal distribution of
distributed under the terms and the population; on this basis, the urban emergency shelters for earthquakes should be re-
conditions of the Creative Commons planned in accordance with the principles of accommodation, accessibility, and proximity
Attribution (CC BY) license (https:// to address the unreasonable layout of emergency shelters [5].
creativecommons.org/licenses/by/ Two indices are often used in the evaluation of the spatial distribution rationality of
4.0/). urban parks, namely, the service area ratio and service population ratio. The accuracy of

Int. J. Environ. Res. Public Health 2023, 20, 2127. https://doi.org/10.3390/ijerph20032127 https://www.mdpi.com/journal/ijerph
 the preciseness of the study as well as the service efficiency and economy of the
layout, another two indices are supplemented, namely, the service capacity ratio an
vice overlap rate. The four indices constitute the corresponding evaluation indices
spatial distribution of urban emergency shelters.
Int. J. Environ. Res. Public Health 2023, 20, 2127 2 of 16
At present, studies on the spatial layout of emergency shelters in China have m
focused on the location selection of accessibility, and quite little has been done on su
able two
these disaster
indicesreduction, which
has been proven is a disaster
in previous studiesreduction
[6]. Meanwhile,model thattoisincrease
in order integrated in
overall
the environment
preciseness andasinvolves
of the study theservice
well as the wholeefficiency
processand andeconomy
multiple oftypes of disasters
the spatial
the Xuzhou
layout, anothercentral urban
two indices area as the research
are supplemented, namely, theobject,
servicethis study
capacity used
ratio and the ArcGIS10
service
overlap rate. The four indices constitute the corresponding evaluation
work analysis module to optimize the spatial layout of emergency shelters in accor indices of the spatial
distribution of urban emergency shelters.
with the four indices, namely, the service area ratio, service population ratio, serv
At present, studies on the spatial layout of emergency shelters in China have mainly fo-
cused onratio,
pacity and service
the location selection overlap rate.and
of accessibility, This study
quite little provides
has been donenewon ideas and metho
sustainable
disaster reduction, which is a disaster reduction model that is integrated into the overall offers
research on the sustainable development of urban emergency shelters, which
ence for theand
environment optimization
involves the of the process
whole spatialand pattern of types
multiple urbanofemergency
disasters. Withshelters
the in X
Xuzhou central urban area as the research object, this study
and a scientific basis and suggestions for government decision‐makers. used the ArcGIS10.8 network
analysis module to optimize the spatial layout of emergency shelters in accordance with
the four indices, namely, the service area ratio, service population ratio, service capacity
2. Research
ratio, Area
and service andrate.
overlap Data Source
This study provides new ideas and methods for research on
the
2.1.sustainable
Overview development
of the Researchof urban
Area emergency shelters, which offers reference for the
optimization of the spatial pattern of urban emergency shelters in Xuzhou and a scientific
Xuzhou
basis and is listed
suggestions for as a national
government key city for flood prevention and earthquake rel
decision-makers.
Many disasters such as karst ground collapse, coal‐mined ground collapse, and q
2. Research Area
landslides haveand Data Source
occurred in Xuzhou, causing serious consequences including the
2.1. Overview of the Research Area
mation and subsidence of railroad roadbeds, and the subsidence and collapse of bui
Xuzhou
[8]. The is listed as a national
aforementioned key cityhazards
geological for floodhave
prevention
become andconstraints
earthquake relief [7].
to Xuzhou con
Many disasters such as karst ground collapse, coal-mined ground collapse, and quarry
tion and the sustainable development of the national economy, not only affecting th
landslides have occurred in Xuzhou, causing serious consequences including the deforma-
tainable development
tion and subsidence of the
of railroad area, and
roadbeds, but the
also seriously
subsidence andthreatening the lives
collapse of buildings [8].and pr
safety
The of urban and
aforementioned rural residents.
geological hazards have become constraints to Xuzhou construction
and the sustainable development
The research area is mainly of the national economy,
the planning not only
scope affecting
of the the sustainable
central urban area of Xu
development of the area, but also seriously threatening
as determined by the revised “The Overall Urban Planning of Xuzhou the lives and property safety (2007–202
of
urban and rural residents.
2017,Thecovering the administrative jurisdictions of Gulou District, Yunlong D
research area is mainly the planning scope of the central urban area of Xuzhou, as
Quanshanby
determined District, the “The
the revised former JiuliUrban
Overall District, and the
Planning urban (2007–2020)”
of Xuzhou area of Tongshan
in 2017, Distric
a total area
covering of about 573.19
the administrative km (Figure
2
jurisdictions 1). As
of Gulou shown
District, in “The
Yunlong Bulletin
District, of the Seven
Quanshan
District, the former Jiuli District, and the urban area of
tional Census of Xuzhou”, the permanent residents within the research Tongshan District, with a total area t
area of about 573.19 km2 (Figure 1). As shown in “The Bulletin of the Seventh National
1,798,800 at 00:00 on 1 November 2020. Currently, 22 emergency shelter and evac
Census of Xuzhou”, the permanent residents within the research area totaled 1,798,800 at
sites (parks,
00:00 on greenbelt,
1 November 2020.squares,
Currently,etc.) have been
22 emergency built,and
shelter with a total area
evacuation of 5,200,000
sites (parks,
meters, accounting
greenbelt, squares, etc.)forhave0.91%
beenofbuilt,
the with
research area(Figure
a total 2). The
area of 5,200,000 effective
square meters,shelter
about 250,000
accounting square
for 0.91% meters,
of the researchandarea(Figure
its capacity 2). is
Theabout 110,000
effective people.
shelter area is Among
about them
250,000 square meters, and its capacity is about 110,000 people. Among
are in Quanshan District, eight in Gulou District, four in Yunlong District, and one them, nine are
in Quanshan District, eight in Gulou District, four in Yunlong District, and one in the
Economic Development District.
Economic Development District.

Figure1.1.Planning
Figure scope.
Planning scope.
Int. J. Environ. Res. Public Health 2023, 20, x FOR PEER REVIEW
Int. J. Environ. Res. Public Health 2023, 20, 2127 3 of 16

Figure 2. Current emergency shelters.

Figure 2. Current emergency shelters.
2.2. Data Source
The research
2.2. Data Source data were divided into two main categories: geospatial data and demo-
graphic data. The geospatial data came from the Geographic Data Sharing Platform of the
The
Chinese research
Academy data were
of Sciences divided
including into two main
administrative categories:
division, geospatial
road networks data and d
and remote
graphic data. The
sensing images, geospatial
etc. The POI data data came
of parks, fromand
schools, thesports
Geographic Data
venues came Sharing
from Platform
the open
platforms of Gaode Map and Baidu Map, and were corrected by matching with
Chinese Academy of Sciences including administrative division, road networks a the actual
satellite maps. The demographic data were obtained from the Xuzhou Bureau of Statistics,
mote sensing images, etc. The POI data of parks, schools, and sports venues came
“Xuzhou Statistical Yearbook 2020”, and “the Bulletin of the Seventh National Population
the open
Census platforms of Gaode Map and Baidu Map, and were corrected by matching
of Xuzhou”.
the actual satellite maps. The demographic data were obtained from the Xuzhou B
3. Research
of Statistics,Design and Model
“Xuzhou Construction
Statistical Yearbook 2020”, and “the Bulletin of the Seventh Na
3.1. Research Design
Population Census of Xuzhou”.
By way of the GIS-network analysis, the spatial analysis module and network analysis
module of GIS were mainly used to analyze the spatial layout of the emergency shelters.
3. Research Design and Model Construction
Since urban emergency shelters have the usual characteristics of small scale, large
3.1. Research
number, Design
and wide distribution, this study did not explore the specific spatial distribution,
but only proposes planning guidance. The focus of this study is directed toward the spatial
By way of the GIS‐network analysis, the spatial analysis module and network
layout of the central and fixed emergency shelters (Table 1).
sis module of GIS were mainly used to analyze the spatial layout of the emergency
ters.
Table 1. Planning and construction requirements of various urban emergency shelters.
Since urban emergency shelters have the usual characteristics of small scale,
Per Capita Effective
Nature of number,
Floor Spaceand wide distribution,
Effective this study
Service didTime
Walking not explore
Widththe
of specific spatial distrib
Essential
Project 2) Radius
Shelters but only proposes planning
(hm Shelter Area guidance.
(km) The focus of this study is directed toward th
(min) Evacuation Facilities
(m2 ) Channels
Emergency
tial layout of the central and fixed emergency shelters (Table 1).
≥0.1
shelters
Temporary
The research ≥ 1
was designed0.5in three steps: ≤10 ≥4 —

Fixed
Middle or Step 1: To determine the “service demand points” and “alternative facility p
emergency 1–20 ≥2 2–3 ≤60 ≥7 Basic facilities
shelters
short term for emergency shelters. “Service demand point” is the gathering point of citizens fo
daily≥activities
20 such as residence, employment, and life, and is alsoGeneral the starting po
Fixed Middle or long Referable fixed Referable fixed
emergency generally, ≥ 2
travel during disaster prevention emergencyand avoidance; ≥ 15 facilities
emergency “alternative facility point” is the po
term Integrated
shelters
site for emergency shelters, and those sitesshelters
over 50 shelters
that meet the basic conditions facilities of disaste
ters can be used as alternative.
Step
The 2: To construct
research was designed theinlocation models of the emergency shelter, analyze and
three steps:
out the
Stepdistance costs the
1: To determine of the “service
“service demanddemand
points” point” and “alternative
and “alternative facility
facility points” for point
emergency shelters. “Service demand point” is the gathering
propose the priority of the site selection of emergency shelters. point of citizens for their daily
activities such as residence, employment, and life, and is also the starting point for travel
Step 3: To make planning suggestions to optimize the spatial layout based on t
during disaster prevention and avoidance; “alternative facility point” is the possible site
selection
for emergency priority andand
shelters, other
thoseconditions.
sites that meet the basic conditions of disaster shelters
can be used as alternative.
Int. J. Environ. Res. Public Health 2023, 20, 2127 4 of 16

Step 2: To construct the location models of the emergency shelter, analyze and screen
out the distance costs of the “service demand point” and “alternative facility point”, and
propose the priority of the site selection of emergency shelters.
Step 3: To make planning suggestions to optimize the spatial layout based on the site
selection priority and other conditions.

3.2. Model Construction

After World War II, with the emergence of the welfare state, Western governments
involved themselves extensively and deeply in public economic activities by providing
public facilities (services). It is against this background that the location theory of public
facilities was developed. This theory is indeed based on location theory, which has experi-
enced three stages: classical location theory, modern location theory, and modern location
theory. The location theory of public facilities belongs to the research content of modern
location theory [9].
By nature, the site selection of emergency shelters can be classified as the site selection
of public facilities. Public facilities refer to all kinds of public service facilities that provide
public services for residents. They have the characteristics of public (or quasi-public)
property rights. The goal of public facilities is not profit, but mainly the optimization of
social equity and interests. Therefore, non-profit and government investment in public
facilities determine that the location theory of public facilities is different from that of
traditional location theory. The location theory of public facilities and the traditional
location theory were compared in the following four aspects: theoretical basis, location
selection target, decision-maker, and research focus. As the results show, the theoretical
model of public facility location theory is more service-oriented and more equitable and
efficient. It is a government-led decision-making theory, and the research results are more
accurate. Therefore, this study chose the location theory model of public facilities as the
research basis.

3.2.1. Construction of Location Model of Fixed Emergency Shelters

The emergency shelters that have been built should meet all of the necessary shelter
requirements for victims and the site selection should adhere to the basic principles of
fairness and efficiency. The number of facilities is to be reduced on the premise that all
of the demand points have been covered; and a model for selecting shelter sites is to
be constructed. The construction of this model consists of two stages. First, to build a
collection of coverage. Second, to specify the minimum number of facilities to be used after
fully covering all demand points, and then to use the median problem model to minimize
the distance costs between the demand points and facility points by setting their distance.
The following is the mathematical model:
Model assumptions: 1 The location distribution of the candidate shelter facility points
and emergency shelter demand points are known, and there is a certain distance between
these points. 2 The distance from the candidate shelter facility to the emergency shelter
demand point is known, and the distance between the facility point and the demand point
is based on the path distance of the actual road. 3 The demand point of the emergency
shelter cannot be covered by two or more emergency shelter facility points, and there is
only one facility point among the demand points in the covered area.
 H = { Hi |i = 1, 2, . . . , m } represent the collection of demand points and
Let
K = K j | j = 1, 2, . . . , n represent the collection of alternative facility points; a maxi-
mum range of r = 3000 m between the fixed emergency shelter and the demand point was
set. dij is the actual distance (road travel distance)
 between the demand point Hi and the
candidate shelter facility point K j ; Ti = j dij ≤ r denotes the collection of candidate
service facilities for demand point i. If the candidate service facility K j is determined as the
service facility, x j = 1; otherwise, x j = 0.
Int. J. Environ. Res. Public Health 2023, 20, 2127 5 of 16

The model Is constructed as:

Minz = ∑ xj (1)
j∈K

Constraint conditions:
s.t. ∑ x j ≥ 1, ∀i ∈ H (2)
j∈ Ti

x j ∈ {0, 1}, ∀i ∈ K (3)

0 ≤ dij ≤ 3000 (4)
Equation (1) is the objective function, which is the minimum number of all facility
points in the emergency shelter. Equation (2) is the constraint condition, based on which
the radiation interval that meets the demand points of this emergency shelter is specified.
Through the constraint condition, it can be guaranteed that the x j values are integers, and
through the constraint condition calculated in Equation (4), it can be seen that the distance
between the site facility point j and the emergency shelter demand point i was less than
3000 m.
After modifying, the model can be expressed as:

Minz = ∑ dij x j (5)

j∈K

3.2.2. Construction of Location Model of Central Emergency Shelters

The central emergency shelter is the highest-level shelter, featured with large invest-
ment and construction costs. Its planning and construction are upgraded on the basis of
fixed emergency shelters. The selection of its location should maximize the site service
and focus on the use efficiency of the site. The model construction is divided into two
stages. First, on the basis of the set covering model, the minimum number of facility points
is selected when the demand points in the research area can be covered by all the facility
points. Second, the E facility points are selected, and the maximization coverage model is
used to obtain the solution; then, the facility points can cover the most demand points of
the emergency shelter by setting the corresponding constraint conditions, which are shown
in the mathematical model as follows: the mathematical model of the fixed emergency
shelters and the corresponding model assumption conditions.
 H = { Hi |i = 1, 2, . . . , m } represent the collection of demand points and
Let
K = K j | j = 1, 2, . . . , n represent the collection of alternative facility points; ai is the
demand of the demand point Hi (such as population size, sales volumes and so on). Set
a maximum range of r = 8000 m between the fixed emergency shelter and the demand
point. dij is the actual distance (road travel distance) between the demand point Hi to the
alternative shelter
 facility point K j ; E is the number of planned central emergency shelters
(e ≤ n); Ti = j dij ≤ r is the collection of alternative fixed emergency shelter facilities for
demand point i. If demand point Hi is covered by the emergency shelter, yi = 1; otherwise,
yi = 0. If the candidate service facility K j is determined as an emergency shelter, x j = 1;
otherwise, x j = 0.
The model is constructed as:

Maxz = ∑ ai yi (6)
i∈ D

Constraint conditions:
s.t. ∑ x j ≥ yi , ∀i ∈ H (7)
j∈ Ti

∑ xj = e (8)
j∈K
Int. J. Environ. Res. Public Health 2023, 20, 2127 6 of 16

x j , y j ∈ {0, 1}, ∀i ∈ H, ∀i ∈ K, (9)

0 ≤ dij ≤ 8000 (10)
The objective function (Equation (6)) allows the facility points to cover the maximum
emergency shelter demand points. Meanwhile, the demand of the demand points is also
considered. The constraint condition (Equation (7)) is that when demand point i is assigned
to cover, there is one and only one candidate shelter being used as the central emergency
shelter among all candidates within the constrained distance. The constraint condition
(Equation (8)) specifies that the number of selected central emergency shelters is E; the
constraint (Equation (9)) is to ensure that x j ,y j is an integer; and the constraint condition
(Equation (10)) is to determine that the distance from the emergency shelter demand point
i to the site facility point j is within 8000 m.
The location models of the fixed and central emergency shelters were applied to the
actual research, respectively. In fact, it was difficult to directly apply the models to the
location selection of emergency shelters. Certain mathematical methods can be used to
solve them, but the workload could be tremendous. Therefore, the facility location model
can be solved through GIS technology and can be visually represented from the map.
In the actual layout planning, the constraint conditions of different cities should also be
considered, and the theoretical model obtained by the location model should be further
adjusted and optimized.
Int. J. Environ. Res. Public Health 2023, 20, x FOR PEER REVIEW 7 of 17
4. Optimization of Spatial Layout of Emergency Shelters
4.1. Data Preparation and Analysis of Emergency Shelters
4.1.1. The Analysis of Demand Points
GIS was applied in the research area to produce a corresponding regular grid, and
GIS was applied in the research area to produce a corresponding regular grid, and
based on the combination of urban road network density and the accuracy of ArcGIS cal‐
based on the combination of urban road network density and the accuracy of ArcGIS
culation results, the city was divided into cell grids (300 m × 300 m). Then, the grid inter‐
calculation results, the city was divided into cell grids (300 m × 300 m). Then, the grid
section points
intersection werewere
points selected as theas
selected center pointspoints
the center of different areas, and
of different theand
areas, datathe
of data
the cell
of
intersection points in roads, non‐construction land, and water were removed
the cell intersection points in roads, non-construction land, and water were removed after screen‐
after
ing. Combined
screening. with thewith
Combined actualtheplanning layout of the
actual planning site of
layout andthe
with
sitereasonable
and with adjustment,
reasonable
all center points in the construction site that met the conditions were
adjustment, all center points in the construction site that met the conditions selected, with
were a total
selected,
of 4285
with (as shown
a total of 4285in Figure
(as shown 3).in Figure 3).

Figure 3. Distribution of
Figure 3. of the
the demand
demand points.
points.

4.1.2.
4.1.2. The
The Analysis
Analysis of
of Candidate
Candidate Facility
Facility Points
Points
For
For the selection of emergency shelters, existing
the selection of emergency shelters, existing resources
resources such
such as
as parks, greenbelts,
parks, greenbelts,
squares, stadiums, and school playgrounds can be utilized as candidate sites. By collecting
squares, stadiums, and school playgrounds can be utilized as candidate sites. By collecting
data related to various emergency shelters in Xuzhou including existing park, greenbelts,
data related to various emergency shelters in Xuzhou including existing park, greenbelts,
other green squares, sports sites, and school playgrounds in the city that can be used as
other green squares, sports sites, and school playgrounds in the city that can be used as
emergency shelters for usual disaster prevention, it was found that there was a total of
emergency shelters for usual disaster prevention, it was found that there was a total of
367 candidate sites, 294 of which (Figure 4) were finally screened out after analyzing their
367 candidate sites, 294 of which (Figure 4) were finally screened out after analyzing their
safety, floor space, and traffic.
safety, floor space, and traffic.
 For the selection of emergency shelters, existing resources such as parks, greenbelts,
squares, stadiums, and school playgrounds can be utilized as candidate sites. By collecting
data related to various emergency shelters in Xuzhou including existing park, greenbelts,
other green squares, sports sites, and school playgrounds in the city that can be used as
emergency shelters for usual disaster prevention, it was found that there was a total of
Int. J. Environ. Res. Public Health 2023,367
20, 2127
candidate sites, 294 of which (Figure 4) were finally screened out after analyzing 7their
of 16

safety, floor space, and traffic.

Distribution of
Figure 4. Distribution
Figure of the
the candidate
candidate facility points.

4.2. Optimization of Spatial Layout of Fixed Emergency Shelters

4.2. Optimization of Spatial Layout of Fixed Emergency Shelters
The selection of fixed emergency shelter sites is based on the site selection model that
The selection of fixed emergency shelter sites is based on the site selection model that
has been constructed. In other words, on the basis of the minimum number of facilities
has been constructed. In other words, on the basis of the minimum number of facilities
determined under the condition of the full coverage of demand points, the median problem
determined under the condition of the full coverage of demand points, the median prob‐
model was used to minimize the total cost between the facility points and the demand
lem
points model was used
by setting to minimize
distance the total
restrictions cost
[10]. between
First, a new the facility assignment
location points and the demand
model was
points by setting distance restrictions [10]. First, a new location assignment
built on the basis of the screened candidate emergency shelters, then the 294 screened model was
built on the
candidate basis of
shelters theloaded
were screened candidate
as facility emergency
points shelters,points
and all demand then the 294 screened
as request points.
candidate shelters were loaded as facility points and all demand points
In the model operation, the actual length of roads in the central urban area of Xuzhou as request points.
was
In the model
Int. J. Environ. Res. Public Health 2023, taken
20, x FORasPEER operation,
the REVIEW
travel cost, the
the actual length
impedance of roads in the
interruption wascentral urbanthat
set to 3000, areais,of
8the
Xuzhou was
maximum
of 17
taken as the
radiation travel
radius of cost,
fixedthe impedance
shelters was 3000interruption
m; and thewas set to 3000,
impedance that is, the maximum
transformation was linear
radiation radius of fixed shelters was 3000 m; and the impedance transformation
by default; based on the above steps, the solution was conducted. First, the minimization was lin‐
ear by point
facility
minimization default;
model based
facility point on the
was applied
model toabove
was the steps,
candidate
applied the solution
shelters.
to the candidate Afterwas conducted.
theAfter
shelters. solution First, the
and calculation,
the solution
the number
and of the
calculation, fixed emergency
number of fixedshelters for shelters
emergency earthquakes required required
for earthquakes was at least
was 79,
at and the
least
number79, and the number
of demand of demand
points covered points covered
was 4055, withwas 4055, with
a coverage ratea of
coverage
94.63%.rate
Theofoperation
94.63%.
results The operation
of the model results of theinmodel
are shown are5.shown in Figure 5.
Figure

Figure Analysis
Figure5.5.Analysis of of
thethe minimum
minimum facilities
facilities of fixed
of fixed emergency
emergency shelters.shelters.

IfIfthe
theprinciples
principlesof of reasonableness
reasonableness and fairness
and fairness of seismic
of seismic evacuation
evacuation are takenare intotaken into
account,
account,ititisisnecessary
necessary to cover
to coverall demand
all demand pointspoints
in the in
area,
theand at the
area, andsame
at thetime, all of
same time, all of
the
thedemand
demandpointspointsareare
placed
placedin the
in emergency
the emergency seismic sheltershelter
seismic and evacuation area at the
and evacuation area at the
travel
travelcost
costrequired
required bybythethe
index. However,
index. However,the actual situation
the actual shows that,
situation shows in general,
that, in the
general, the
population
populationdensity
densityininthe central
the centralarea of the
area citycity
of the is higher, whilewhile
is higher, it is lower in the in
it is lower periph‐
the peripheral
eral areas of the city. The population distribution is relatively scattered, and in general,
areas of the city. The population distribution is relatively scattered, and in general, the
the distribution of urban population is unbalanced. Therefore, if the evacuation places are
distribution of urban population is unbalanced. Therefore, if the evacuation places are
evenly distributed throughout the area only considering the distance cost, it will lead to
evenly
low distributed
utilization throughout
of evacuation placesthe area only
in marginal considering
areas. the distance
Economically, this layout cost,
doesitnot
will lead to
low utilization
make sense. of evacuation places in marginal areas. Economically, this layout does not
make sense.
Therefore, in order to achieve the fairness and economy of fixed evacuation sites, the
minimumTherefore, in order
impedance model to was
achieveagainthe fairness
operated to and economy
confirm of fixed
the number andevacuation
location of sites, the
minimum
fixed impedance
evacuation sites basedmodel
on thewas again operated
minimization to confirm
facility point model.theThenumber
number of and
fixedlocation of
evacuation sites before
fixed evacuation siteseach operation
based on the of minimization
the minimum impedance modelmodel.
facility point was given Theand number of
the actual road distance was used as the operation cost. The number of facilities was re‐
duced from 79 to 73 by minimizing the number of facilities and the arithmetic results of
the number of fixed evacuation sites, that is, the coverage rate and shelter area, were ob‐
tained. The green dot blocks in the figure show the decreasing candidate points (Figure 6
and Figure 8). Table 2 summarizes the relationship between the number of fixed evacua‐
 the distribution of urban population is unbalanced. Therefore, if the evacuation places are
evenly distributed throughout the area only considering the distance cost, it will lead to
low utilization of evacuation places in marginal areas. Economically, this layout does not
make sense.
Therefore, in order to achieve the fairness and economy of fixed evacuation sites,
Int. J. Environ. Res. Public Health 2023, 20, 2127 the
8 of 16
minimum impedance model was again operated to confirm the number and location of
fixed evacuation sites based on the minimization facility point model. The number of fixed
evacuation sites before
fixed evacuation each operation
sites before of theofminimum
each operation impedance
the minimum model
impedance was given
model and
was given
the actual road distance was used as the operation cost. The number of facilities
and the actual road distance was used as the operation cost. The number of facilities was was re‐
duced from 79 to 73 by minimizing the number of facilities and the arithmetic
reduced from 79 to 73 by minimizing the number of facilities and the arithmetic results results of
the number of fixed evacuation sites, that is, the coverage rate and shelter
of the number of fixed evacuation sites, that is, the coverage rate and shelter area, were area, were ob‐
tained.
obtained. The green
The dotdot
green blocks inin
blocks the figure
the figureshow
showthethedecreasing
decreasingcandidate
candidatepoints
points(Figure
(Figure 66
and
and Figure 8). Table
Figure 8). Table22summarizes
summarizesthe therelationship
relationshipbetween
between the
the number
number of of fixed
fixed evacua‐
evacuation
tion sites and their coverage
sites and their coverage rate. rate.

Figure
Figure
Int. J. Environ. Res. Public Health 2023, 6.
20, x6. Minimization
Minimization
FOR impedance analysis
PEER REVIEWimpedance analysis for
for aa facility
facility point
point resistance of 72. 9 of 17

Table 2. The number and coverage rate of fixed evacuation facilities in Xuzhou.
Table
Number2. The number and coverage
of Facilities Number rate of fixed Point
of Demand evacuation
in thefacilities in Xuzhou.
Total Number Coverage Rate
(Unit) Covering Point (Unit)
Number of Facili‐ Number of Demand Point
Total Number Coverage Rate
ties 69
(Unit) 4019
in the Covering Point (Unit) 4285 93.79%
70
69 4019 4026 4285 4285 93.79% 93.96%
70
71 4026 4032 4285 4285 93.96% 94.10%
71
72 4032 4037 4285 4285 94.10% 94.21%
72
73 4037 4041 4285 4285 94.21% 94.31%
73
74
4041 4044 4285 4285
94.31% 94.38%
74 4044 4285 94.38%
75 4047 4285 94.45%
75 4047 4285 94.45%
76 4049 4285 94.49%
76 4049 4285 94.49%
77
77 4051 4051 4285 4285 94.54% 94.54%
78
78 4053 4053 4285 4285 94.59% 94.59%
79 4055 4055 4285 4285 94.63% 94.63%

The variation in the coverage rate with the number of sites obtained from the above
The variation in the coverage rate with the number of sites obtained from the above
analysis was plotted as a scatter plot and simulated as a curve. The results are shown in
analysis was plotted as a scatter plot and simulated as a curve. The results are shown
Figure 7.
in Figure 7.

Figure 7. Relationship between number of fixed sites and coverage.

Figure 7. Relationship between number of fixed sites and coverage

As can be seen from the figure, when the number of places was 73 (Figure 8), the
curve took a turn and the curvature had a significant decrease with a coverage rate of
94.31%. After that, the slope of the curve tended to be smooth again. Taking the rules of
Int. J. Environ. Res. Public Health 2023, 20, x FOR PEER REVIEW 10 of 17

Int. J. Environ. Res. Public Health 2023, 20, 2127 9 of 16

As can be seen from the figure, when the number of places was 73 (Figure 8), the
curve took a turn and the curvature had a significant decrease with a coverage rate of
94.31%. After that, the slope of the curve tended to be smooth again. Taking the rules of
Int. J. Environ. Res. Public Health 2023,fairness
20, x FOR and
PEEReconomy
REVIEW into account and combining the coverage rate of evacuation10
places,
of 17
the evacuation area, and other hazard reasons, the number of fixed evacuation places was
set at 73.

Figure 8. Minimization impedance analysis for a facility point resistance of 73.

4.3. Optimization of Spatial Layout of Central Emergency Shelters

The central emergency shelter is the highest‐level shelter with the highest planning
and construction cost. Accordingly, the main considerations in its planning and construc‐
tion are economy and efficiency. This was selected on the basis of the optimized fixed
emergency shelters. The constructed location model of the central earthquake emergency
shelter was selected for the solution, that is, the minimization facility point model was run
first, and then the maximum coverage model was run [11]. First, the minimized facility
pointFigure
model8.in
Figure the network
Minimization
Minimization analysis analysis
impedance
impedance locationfor
analysis assignment
for model
a facility point in ArcGIS
resistance of 73.10.5 was used
to process the data and the demand points in the research area that could be covered by
4.3. Optimization of Spatial Layout of Central Emergency Shelters
the facility points of the
4.3. Optimization centralLayout
of Spatial shelters were selected.
of Central Emergency In the model operation, the actual
Shelters
length of theTheroads
central inemergency
the central shelter
urban is theof
area highest-level
Xuzhou was shelter
used withthethe
as with highest
travel costplanning
and 27 and
The central emergency shelter is the highest‐level shelter the highest planning
construction
candidate cost.
points that met Accordingly, the main considerations in its planning and construction are
and construction cost.the requirements
Accordingly, the of
maincentral shelters were
considerations sorted
in its as the
planning facility
and construc‐
points.economy
Then, and efficiency. This was selected on the basis of the optimized fixed emergency
tion are the
shelters. The
minimization
economy constructed
facility point
and efficiency.
location model
model
This was was run
selected
of the central
with
the the
onearthquake
basisimpedance
of interrup‐fixed
the optimized
emergency shelter was
tion of 8000
emergency m. The operation
shelters. results
The constructed of the model are shown
location modelfacility in Figure
of the point
central 9 The
earthquake minimum
emergency
number selected for
of facility the solution,
pointsforwas that
12, is, the minimization
and thethat number of demand points model
covered was run first, and
was 4131,
shelter was selected the solution, is, the minimization
then the maximum coverage model was run [11]. First, the minimized facility point facility point model was run
model
with first,
a coverage rate of 96.41%.
in theand then the
network maximum
analysis locationcoverage modelmodel
assignment was run in [11].
ArcGIS First,
10.5the minimized
was facility
used to process
The
point central
model emergency
in the shelter
network is often
analysis used as
location the command
assignment model center
in of urban
ArcGIS 10.5earth‐
was used
the data and the demand points in the research area that could be covered by the facility
quake to relief,
process which
the not only
data and possesses
the demand the points
basic functions
in the of fixedarea
research emergency
that couldshelters,
be but by
covered
points of the central shelters were selected. In the model operation, the actual length of
also serves
the as the placeoffor receiving and distributing rescueInsupplies and the command
the facility
roads inpoints
the centralthe central
urban shelters
area were was
of Xuzhou selected.
used as thethe model operation,
travel cost and 27 the actual
candidate
placelength
for health
of theand epidemic
roads in the prevention.
central urban Inarea
its planning
of Xuzhou and construction,
was used as the the construc‐
travel cost and 27
points that met the requirements of central shelters were sorted as the facility points. Then,
tion cost should
candidate be minimized
points that while
metpoint considering
the requirements the efficiency
of with
central [12]. On the basis of the 12
the minimization facility model was run theshelters
impedance wereinterruption
sorted as the facility
of 8000 m.
candidate
points. central
Then, emergency
the shelters
minimization for earthquakes
facility point model selected
was9run above,the different numbers
The operation results of the model are shown in Figure Thewith
minimum impedance
number interrup‐
of facility
of facilities
tion of were
800012,set as
m.and
TheE operation
(E is a positive integer less thanare12), and the maximization cov‐
points was the numberresults of demandof thepoints
model covered shown
was in Figure
4131, with9 aThe minimum
coverage rate
eragenumber
model was run to solve them, respectively, as shown in Figures
of facility points was 12, and the number of demand points covered was 4131, 10 and 11.
of 96.41%.
with a coverage rate of 96.41%.
The central emergency shelter is often used as the command center of urban earth‐
quake relief, which not only possesses the basic functions of fixed emergency shelters, but
also serves as the place for receiving and distributing rescue supplies and the command
place for health and epidemic prevention. In its planning and construction, the construc‐
tion cost should be minimized while considering the efficiency [12]. On the basis of the 12
candidate central emergency shelters for earthquakes selected above, different numbers
of facilities were set as E (E is a positive integer less than 12), and the maximization cov‐
erage model was run to solve them, respectively, as shown in Figures 10 and 11.

FigureFigure 9. Minimum
9. Minimum facilities
facilities analysis
analysis of theof the central
central emergency
emergency shelters.
shelters.

The central emergency shelter is often used as the command center of urban earth-
quake relief, which not only possesses the basic functions of fixed emergency shelters, but
also serves as the place for receiving and distributing rescue supplies and the command
place for health and epidemic prevention. In its planning and construction, the construc-
tion cost should be minimized while considering the efficiency [12]. On the basis of the

Figure 9. Minimum facilities analysis of the central emergency shelters.

Int. J. Environ. Res. Public Health 2023, 20, 2127 10 of 16

Int. J. Environ. Res. Public Health 2023,12

20,candidate
x FOR PEERcentral
REVIEWemergency 11 of 17
shelters for earthquakes selected above, different numbers
Int. J. Environ. Res. Public Health 2023, 20, x FOR PEER REVIEW 11 of 17
of facilities were set as E (E is a positive integer less than 12), and the maximization coverage
model was run to solve them, respectively, as shown in Figures 10 and 11.

Figure
Figure 10. Analysis
Analysis of
of the
the maximum
maximum coverage
coverage of E == 7.
7.
Figure 10. Analysis of the maximum coverage of E = 7.

Figure 11. Analysis of the maximum coverage of E = 8.

Figure 11. Analysis of the maximum coverage of E = 8.
Figure 11. Analysis of the maximum coverage of E = 8.
As the solution showed (Table 3), when E = 10, the number of demand points covered
As the
was 4119 andsolution showed
the coverage (Table
rate 3),
was 3), when E = 10,E the
96.13%; number of demand points covered
As the solution showed (Table whenwhen = 9,number
E = 10, the these were 4097 andpoints
of demand 95.61%; when
covered
was
E = 4119
8, and
these the
were coverage
4046 andrate was
94.42%; 96.13%;
when Ewhen
= 7, E = 9,
these these
were were
3919 4097
and and 95.61%;
91.46%. when
Therefore,
was 4119 and the coverage rate was 96.13%; when E = 9, these were 4097 and 95.61%; when
E = 8, facility
these were 4046 and 94.42%;as when E = 7, these were 3919 and 91.46%. Therefore,
Eeight points
= 8, these were were
4046 andselected
94.42%; whenthe central emergency
E = 7, these shelters,
were 3919 which ensures
and 91.46%. the
Therefore,
eight facility
economy and points were
fairness of selectedshelters.
building as the central emergency shelters, which ensures the
eight facility points were selected as the central emergency shelters, which ensures the
economy and fairness of building shelters.
economy and fairness of building shelters.
Table 3. The number and coverage rate of central evacuation places in Xuzhou.
Table 3. The number and coverage rate of central evacuation places in Xuzhou.
Table 3. Theofnumber
Number Facilitiesand coverage rate
Number of of central
Demand Pointevacuation
in the places in Xuzhou.
Total Number Coverage Rate
Number(Unit)of Facili‐ NumberCovering
of Demand Point in
Point (Unit)
Number of Facili‐ Number of Demand Point in Total Number Coverage Rate
ties (Unit)
4 the Covering2584
Point (Unit) Total Number
4285 Coverage Rate
60.30%
ties (Unit)
5 the Covering3143
Point (Unit) 4285 73.35%
4 2584 4285 60.30%
46 2584
3556 4285
4285 60.30%
82.99%
57 3143
3919 4285
4285 73.35%
91.46%
58 3143
4046
4285
4285
73.35%
94.42%
6 3556 4285 82.99%
69 3556
4097 4285
4285 82.99%
95.61%
710 3919
4119 4285
4285 91.46%
96.13%
7 3919 4285 91.46%
811 4046
4129 4285
4285 94.42%
96.36%
812 4046
4131 4285
4285 94.42%
96.41%
9 4097 4285 95.61%
9 4097 4285 95.61%
10 4119 4285 96.13%
10 Evaluation of Spatial
5. Rationality 4119Layout of Emergency4285 Shelters 96.13%
11 4129 4285 96.36%
11
5.1. Rationality Evaluation Indexes 4129
of Spatial Layout of 4285Shelters
Emergency 96.36%
12 4131 4285 96.41%
12 paper, four evaluation 4131
In this 4285
indices, namely, the service 96.41%
area ratio, service population
ratio,
5. service capacity
Rationality ratio,
Evaluation ofand service
Spatial overlap
Layout rate were used
of Emergency to evaluate the rationality
Shelters
5.
ofRationality Evaluation
the spatial layout of Spatial
of emergency Layoutinof
shelters Emergency
the Shelters
central urban area of Xuzhou.
5.1. Rationality Evaluation Indexes of Spatial Layout of Emergency Shelters
5.1. Rationality
The service Evaluation
area ratioIndexes of Spatial
is the ratio Layout
of the total of Emergency
service Shelters
area of the public service facility
to theIn total
this paper,
area offour
the evaluation
research indices,
area. It namely,
is a the service
reflection of the area
spaceratio, service
service popula‐
capacity of
In this paper, four evaluation indices, namely, the service area ratio, service popula‐
tion ratio,
public service
facilities. capacity ratio, and service overlap rate were used to evaluate the ration‐
tion ratio, service capacity ratio, and service overlap rate were used to evaluate the ration‐
ality of the spatial layout of emergency shelters in the central urban area of Xuzhou.
ality of the spatial layout of emergency shelters in the central urban area of Xuzhou.
The service area ratio is the ratio of the total service area of the public service facility
The service area ratio is the ratio of the total service area of the public service facility
to the total area of the research area. It is a reflection of the space service capacity of public
to the total area of the research area. It is a reflection of the space service capacity of public
facilities.
facilities.
Int. J. Environ. Res. Public Health 2023, 20, x FOR PEER REVIEW 12 of 17
Int. J. Environ. Res. Public Health 2023, 20, x FOR PEER REVIEW 12 of 17
Int. J. Environ. Res. Public Health 2023, 20, 2127 11 of 16

The service
The service population
service population ratio
ratioisis
populationratio the
isthe ratio
ratioofof
theratio the
ofthe number
numberofof
thenumber of people
people in
in the service area
of a The
public service facility to the total population in the people
research area. thethe
inThis is
service
service
a
area
area
reflection of
of a public
aofpublic service
service facility
facility to to
the the
totaltotal population
population in in
thethe research
research area.
area. This
This is ais a reflection
reflection of
of the
the the population
population
population
capacity
capacity
capacity and
and
and service
service
service
capacity
capacity
capacity
of
of public
public
of public
facilities.
facilities.
facilities.
The
The service
service capacity
capacity ratio
ratio is
is the
the ratio
ratio of
of the
the total
total population
population in need
in need
need ofof public
of public service
public service
service
The in
facilities service
the capacityarea
research ratiotoisthe
thetotal
ratiodesign
of the total population
population in
capacity of public service fa‐
facilities in
facilities in the
the research
researcharea
areatotothe thetotal
totaldesign
designpopulation
population capacity
capacityof of
public
publicservice
servicefa‐
cilities.
cilities. This
This is aa reflection
is is reflection of
ofofthe
the space
space service efficiency of public service facilities.
facilities. This a reflection the spaceservice
serviceefficiency
efficiencyofofpublic
publicservice
servicefacilities.
facilities.
The service
The service overlap
service overlap ratio
ratioisis
overlapratio the
isthe ratio
ratioofof
theratio the
ofthe area
areaofof
thearea of the
the overlapping
overlapping part
part of the cover‐
The the overlapping part of of
thethe cover‐
coverage
age
age of
of each
each public
public service
service facility
facility and
and other
other public
public service
service facilities
facilities to
to the
the total
total area
area of the
of the
the
of each public service facility and other public service facilities to the total area of
coverage
coverage of
of all
all shelters.
shelters. This
This is
is a
a reflection
reflection of
of the
the overlapping
overlapping spatial
spatial layout
layout of
of public
public ser‐
ser‐
coverage of all shelters. This is a reflection of the overlapping spatial layout of public
vice
vice facilities.
facilities.
service facilities.
5.2. Rationality
5.2. Rationality Evaluation
Evaluation of of Spatial
Spatial Layout
Layout of of Emergency
Emergency Shelters
Shelters
Shelters
According
According to
to the
the built
built road
road network,
network, the
According to the built road network, the accessibilitythe accessibility
accessibility of
of emergency
of emergency
emergency shelters
shelters in
in Xu‐
in Xuzhou
shelters Xu‐
zhou
was was
zhoufirst first
wasevaluated evaluated after
after the
first evaluated after the
optimizedoptimized layout.
layout.layout.
the optimized The
The maximum maximum
The maximum service service distance
distance
service of fixed
distance of
of
fixed emergency
emergency
fixed emergencyshelters shelters
was 3000
shelters was
was 3000
m,3000 m,
and m, and
thatand that
of central of central
that of emergency emergency shelters
shelters was
central emergency was
8000was
shelters 8000
m. Seventy-
8000 m.m.
Seventy‐three
three
Seventy‐three selected
selected fixed
selected fixed
fixed emergency
emergency shelters
shelters were
emergency were
were analyzed,
analyzed,
shelters and
and the
and the service
analyzed, the service
area area
function
service areaoffunc‐
the
func‐
tion
ArcGISof the ArcGIS
network network
analysis can analysis
be used can
to be used
calculate to calculate
the service the service
area of
tion of the ArcGIS network analysis can be used to calculate the service area of the shelter.the area of
shelter.the shelter.
Based on
Based
the roadon the road
network network
model, themodel, the
coverage coverage
area of area of
facilities facilities
within within
the
Based on the road network model, the coverage area of facilities within the service radius the
service service
radius radius
can be
can
more be more
accurately accurately simulated
simulated on a on a realistic
realistic road road
network network
based based
on on
the
can be more accurately simulated on a realistic road network based on the traffic distance. the traffic
traffic distance.
distance. For
For
For fixed
fixed and and
fixed and central
central shelters,
shelters,
central the
the impedance
the impedance
shelters, impedance waswas set
set to
was to
setthe the
the road
to road network
network
road network travel
travel costcost
travel in
in the
cost in
the analysis
analysis
the analysis setting
setting of the
of theofnew
setting new
the service service area
area layer
new service layer
area attribute,attribute, and
and theand
layer attribute, the
default default interruption
interruption
the default was set
interruption
was
to
was set
3000 to
to 3000
set and 8000,
3000 and 8000,
8000, respectively.
respectively.
and The
The solution
The solution
respectively. results are
solution results
shown
results are
areinshown
Figures
shown in Figures
in 12 and 13.
Figures 12 and
12 The
and
Figure
same 13.
13. The
Figuremethod Thewassame
same method
applied
method to was
solveapplied
was to
to solve
the completed
applied solve the completed
emergency
the completed emergency
shelters and theshelters
emergency resultsand
shelters are
and
the results
graphically are graphically
represented represented
(Figure 14).
the results are graphically represented (Figure 14). (Figure 14).

service area
area of 3000
3000 m.
Figure 12. Fixed
Figure 12. Fixed emergency
emergency shelters
shelters with
with the
the service
service area of
of 3000 m.
m.

service area
Figure 13. Central emergency shelters with the service area of
of 8000
8000 m.
m.
Figure 13. Central emergency shelters with the service area of 8000 m.
The above analysis of the reasonableness of the layout of the candidate central emergency
shelters in Xuzhou was carried out by arithmetic and calculation of the service parameters of
the evaluation criteria statistics (Figure 15). The results are shown in Tables 4 and 5.
 Int. J. Environ. Res. Public Health 2023, 20, x FOR PEER REVIEW 13 of 17

Int.
Int. J. Environ.
J. Environ. Res.Res. Public
Public Health
Health 2023, 2023, 20, 2127
20, x FOR PEER REVIEW 13 of 17 12 of 16

Figure 14. Built emergency shelters with the service area of 3000 m.

The above analysis of the reasonableness of the layout of the candidate central emer‐
gency shelters in Xuzhou was carried out by arithmetic and calculation of the service pa‐
Figure 14. Built
rameters
Figure emergency
14.ofBuilt shelterscriteria
the evaluation
emergency with thestatistics
shelters service
with thearea of 3000
(Figure
service m.The
15).
area results
of 3000 m.are shown in Tables 4
and 5.
The above analysis of the reasonableness of the layout of the candidate central emer‐
gency shelters in Xuzhou was carried out by arithmetic and calculation of the service pa‐
rameters of the evaluation criteria statistics (Figure 15). The results are shown in Tables 4
and 5.

Figure15.
Figure Optimized
15.Optimized layout
layout of emergency
of emergency shelters.
shelters.

Statistical
Table4.4.Statistical
Table table
table of service
of service indices
indices of established
of established emergency
emergency sheltersshelters in Xuzhou.
in Xuzhou.

Service Permanent Service

Permanent Service Service
Figure 15.
Service Area Optimized
Total layout
Service
Service of emergency
Total
Area shelters.
Service Service Service Service Service
District Name
(ha)
District
Area(ha) ratio
Population Residents
Residents in the Popula‐
Population Overlap Capacity
Area Area Area ra‐ Population
(Person) Area (Person) Ratio OverlapRate
Capacity Ratio
Table 4.Name in the Area
Statistical table of service indices of established emergency tion
shelters in Ra‐
Xuzhou.
(ha) (ha) tio (Person) Rate Ratio
Gulou District 38.10 77.40 49.23% 243,483 425,143
(Person) 57.27%
tio 76.59% 9.40%
Permanent Service
Yunlong Gulou Service 41.18%
Total Service Service Service Service
49.45 120.07
District 38.10 77.40 49.23%
193,704 243,483 425,143
471,566
Residents 57.27%
41.08% 76.59%45.98%
Popula‐ 9.40% 5.01%
District DistrictArea Area Area ra‐ Population Overlap Capacity
Name in the Area tion Ra‐
Quanshan
53.52
Yunlong
101.92 (ha)49.45 (ha)
52.51% tio 330,488
(Person) 619,784 53.32%Rate Ratio
District 120.07 41.18% 193,704(Person)
471,566 tio41.08% 45.98%77.37%
5.01% 6.45%
District
Tongshan
Gulou
2.13 Quanshan
128.49 38.1053.52 77.40 49.23%1410243,483
1.66% 425,143 57.27% 76.59% 9.40%
District District 101.92 52.51% 330,48884,932619,784 53.32% 1.66% 77.37%4.74%6.45% 0.00%
District
Economic Yunlong
Tongshan 49.45 120.07 41.18% 193,704 471,566 41.08% 45.98% 5.01%
Development 13.18 District
145.31 2.139.07%128.49 1.66%17,081 1410 188,32084,932 1.66%
9.07% 4.74% 1.69%
0.00% 5.63%
District
District
Quanshan
Economic 53.52 101.92 52.51% 330,488 619,784 53.32% 77.37% 6.45%
Total 156.39 District
573.19 27.28% 786,166 1,789,745 43.93% 53.27% 6.38%
Develop‐
Tongshan 13.18 145.31 9.07% 17,081 188,320 9.07% 1.69% 5.63%
ment Dis‐ 2.13 128.49 1.66% 1410 84,932 1.66% 4.74% 0.00%
District
trict5. Statistical table of service indices of optimized emergency shelters in Xuzhou.
Table
Economic
Total 156.39 573.19 27.28% 786,166 1,789,745 43.93% 53.27% 6.38%
Develop‐ Service17,081 Permanent Service Service Service
Service AreamentTotal 13.18 145.31 9.07% 188,320 9.07% 1.69% 5.63%
District Name Dis‐Area Service Area Population Residents in the Population Overlap Capacity
(ha) (ha) Ratio
trict (Person) Area (Person) Ratio Rate Ratio
Gulou District 71.64 Total77.40 156.39 92.57%
573.19 27.28% 457,811786,166 1,789,745
425,143 43.93% 53.27% 6.38%
107.68% 96.83% 40.05%
Yunlong
94.76 120.07 78.92% 371,170 471,566 78.71% 89.52% 123.28%
District
Quanshan
76.48 101.92 75.03% 472,234 619,784 76.19% 91.81% 149.28%
District
Tongshan
66.67 128.49 51.89% 44,071 84,932 51.89% 76.12% 298.75%
District
Economic
Development 79.99 145.31 55.05% 103,671 188,320 55.05% 82.24% 216.22%
District
Total 389.54 573.19 67.96% 1,448,957 1,789,745 80.96% 86.67% 130.62%
 Tongshan
66.67 128.49 51.89% 44,071 84,932 51.89% 76.12% 298.75%
District
Economic
Develop‐ 79.99 145.31 55.05% 103,671 188,320 55.05% 82.24% 216.22%
ment
Int. District
J. Environ. Res. Public Health 2023, 20, 2127 13 of 16
Total 389.54 573.19 67.96% 1,448,957 1,789,745 80.96% 86.67% 130.62%

5.2.1. Analysis
5.2.1. of Service
Analysis of Service Index
Index Parameters
Parameters
The analysis
The analysis was
was based
based on
on the
the statistical
statistical comparison
comparison of of the
the service
service indices
indices of
of the
the built
built
emergency shelters and the emergency shelters after layout optimization.
emergency shelters and the emergency shelters after layout optimization. The overall The overall
comparison showed
comparison showed thatthatthe
theservice
serviceareaarearatio
ratioof of
emergency
emergency shelters before
shelters andand
before afterafter
op‐
timization was 27.28% and 67.96%m respectively, and the service population
optimization was 27.28% and 67.96%m respectively, and the service population ratio was ratio was
43.93% and
43.93% and 80.96%,
80.96%, respectively.
respectively. It can be
It can be seen
seen that
that the
the number
number of of emergency
emergency shelters
shelters
before optimization
before optimization was
was too
too small
small andand far
far from
from meeting
meeting thethe demand
demand for for shelters
shelters in
in the
the
central urban
central urban area
area of
of Xuzhou,
Xuzhou, and
and the
the service
service area
area and
and population
population of of emergency
emergency shelters
shelters
after optimization have been greatly improved. The comparison of the service area ratio
and the service population ratio of emergency shelters before and after optimization are
represented by
represented by bar
bar graph
graph plots
plots as
as shown
shown in in Figures
Figures 16 16 and
and 17,
17, respectively.
respectively.

area ratio
Figure 16. Bar chart of the service area ratio before
before and
and after
after optimization.
optimization.

ratio before
Figure 17. Bar chart of the service population ratio before and
and after
after optimization.
optimization.

From the comparison of the service overlap rate, the service overlap rate of emergency
shelters before optimization was high, while their service area and population ratio were
extremely low. The number of optimized emergency shelters increased and the service
overlap rate also increased. However, the service area and population were substantially
improved, with a 43.09% increase in the former and 38.05% in the latter. The reason for
the high service overlap rate is that the built emergency shelters are too concentrated,
mainly in Quanshan District and Gulou District, and they are close to each other; another
reason is that the population in the central urban area is large and concentrated and enough
emergency shelters must be arranged to meet the needs of the residents. Meanwhile, for the
random distribution and fixed places of the emergency shelters, the phenomenon of service
overlap will inevitably arise when screening is carried out. In order to balance equity and
efficiency, it shows that the optimized emergency shelter is more reasonable.

5.2.2. Macro Analysis of the Central Urban Area of Xuzhou

Analysis of Table 5 shows that the service shelter population of emergency shelters in
the central urban area of Xuzhou reached 80.96%. In general, these shelters covered a wide
population with a reasonable layout and can meet the index of covering more than 80% of
the population. In terms of the service area ratio, the ratio in the center was 67.96%, which
can satisfy about 70% of the demands in the research area. In terms of the service area ratio
of each district, the ratios of Quanshan District, Gulou District, and Yunlong District were all
above 70%. In contrast, those of Tongshan District and the Economic Development District
Int. J. Environ. Res. Public Health 2023, 20, 2127 14 of 16

were relatively low, between 50% and 60%, which correspond to the actual situation of low
population density and low construction density of Tongshan District and the Economic
Development District. In general, it shows that the planning layout of the emergency
shelter meets the demand. In terms of the service capacity ratio, this was generally 130.62%,
which indicates that there is a surplus in the service capacity of shelters and can meet the
demand of urban residents for shelters. In terms of the service overlap rate, the average
value was 86.67%, while the coverage rate reached 94.31%, where it can be seen that the
optimized spatial layout of the emergency shelters was more reasonable.

5.2.3. Analysis of Each District

For each district in Xuzhou, the service area ratios in Quanshan District, Gulou District,
and Yunlong District in the central urban area were relatively high due to the large number
and the dense distribution of emergency shelters.
In terms of the service population ratio, those for the Economic Development District
and Tongshan District were relatively low compared with other districts, both between
50% and 60% because of their relatively small population size, small population density,
low development intensity, and a large number of rural populations. The population in
these areas is sparse and relatively scattered, and the area beyond the service radius fails
to enjoy the corresponding services. Furthermore, there are still some deficiencies in the
distribution of emergency shelters after layout optimization.
In terms of the service capacity ratio of each district, those of all the other districts,
except for Gulou District, exceeded 1, which means that the design capacity of the other
four districts is enough to meet the evacuation and shelter requirements of the population
in the area. However, the design capacity of Gulou District cannot fully accommodate
the shelter population, because the effective shelter area of the emergency shelters built in
Gulou District is small and needs to be further improved.

5.3. The Supporting Strategy for Emergency Shelters

Site selection should be in accordance with the relevant government laws and regula-
tions as well as local standards. At present, China has promulgated a number of laws and
regulations related to urban disaster prevention and reduction. Local governments at all
levels should constantly supplement and improve the planning, construction, and design
standards of earthquake emergency shelters in accordance with the rules and regulations
promulgated by the state and the conditions of their own regions. Meanwhile, the site
selection of emergency shelters should closely match the policies and relevant standards
formulated by the government.

5.4. Innovation Points

This paper constructed a location selection model system of urban emergency shelters
that can apply the Network Analyst extension module in ArcGIS (minimize resistance and
maximize coverage). The optimization of the layout of emergency shelters in downtown
Xuzhou was completed. The rationality evaluation index was constructed to verify the
scientificity and effectiveness of the location model, providing guidance for the planning
and construction of shelters in the future.

6. Conclusions
1. There are many service blind areas and service overlap areas in the emergency shelter
service areas. Emergency shelters are basically places with other urban functions that
are “borrowed temporarily”. In this sense, their layout follows their own principles.
For example, city parks have a different service radius according to their level; pri-
mary and secondary schools are distributed according to the service requirements of
residential areas; and city squares, parking lots, sports fields are allocated according
to the layout of the city’s functional needs. The layout of these sites does not need
to consider the layout of other sites. For instance, schools may happen to be located
Int. J. Environ. Res. Public Health 2023, 20, 2127 15 of 16

next to a park or square. However, if they are also used as emergency shelters, the
service area would overlap. Therefore, the government should determine the service
areas of emergency shelters according to the accessibility evaluation of people, and
supplement more emergency shelters in the service blind areas. Meanwhile it should
find out and reduce the service overlapping areas by re-siting in order to avoid waste.
2. The layout of the emergency shelter failed to match the demand for shelter. The
analysis of the layout of emergency shelters points out some existing problems such
as the confusion in the scale level of shelters, uneven construction in each district,
and insufficient number of places built in the city. In China, the selection of urban
emergency shelters generally follows the principle of dual use for daily life and disas-
ter relief. With urban parks, squares, parking lots, sports fields, school playgrounds,
and other sites as shelters, although these sites are combined to some extent with
the distribution of residential areas in their layout, their service population is often
calculated on the basis of the usual population. In the event of a disaster, each area
has different disaster risks. Although old urban communities, urban villages, etc.
have about the same population as new residential areas, these communities are
often more severely damaged after a disaster and have a greater need for disaster
avoidance. Local governments should improve urban emergency shelter systems and
focus on disaster damage assessment in cities to improve the sustainability of urban
disaster reduction.
3. Regular evacuation drills should be organized for residents to prevent disasters. With
the knowledge of the public, all construction work can play a role. In many cities in
China, residents know very little about how to respond to disasters, which requires
the government and communities to develop perfect emergency plans. All residents
should follow the procedures for self-rescue, and arrive at the emergency shelter
along the evacuation route. Through long-term training, residents should be familiar
with the evacuation plans and evacuation routes for different disasters around their
homes, so that they can save themselves and each other in an orderly manner during
disasters, minimize casualties, and play the life-saving role of emergency shelters and
emergency evacuation routes.
Although this paper has made some achievements in the location optimization and
layout of public facilities in urban emergency shelters, there are still some limitations which
are hoped to be further improved in the future. (1) This study took the central urban area
of Xuzhou as the research object to explore the layout optimization of urban emergency
shelters. Xuzhou, as an inland city with little relief, is featured with flat terrain. In this
sense, this study is of certain instruction to cities with flat terrain. However, the special
cases of mountain cities, coastal cities, and cities across rivers were not considered in this
study. We expect to study urban emergency shelters for different types of cities in the
future. (2) Currently, the site selection of emergency shelters is mainly focused on parks,
schools, sports venues, and other places and deals with common public emergencies and
natural disasters such as earthquakes, fires, floods, and epidemics. However, it is far from
competent to deal with other man-induced disasters such as war, toxic chemical leakage,
nuclear leakage, and other disasters. Future research should cover more types of disasters
and enrich the types of emergency shelter site selection to greater meet the shelter needs
of residents.

Author Contributions: W.Z. contributed to the conception of the study; H.X. performed the data
analyses and wrote the manuscript; W.K. helped perform the analysis with constructive discussions.
All authors have read and agreed to the published version of the manuscript.
Funding: Basic scientific research fees of central universities (approval no. 2021XCZX10).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Int. J. Environ. Res. Public Health 2023, 20, 2127 16 of 16

Data Availability Statement: The data used to support the findings of this study are available from
the corresponding author upon request.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Zhou, A.; Chen, L.; Zhu, H.; Chen, S. Reasonability of spatial distribution for urban emergency shelter in central district of Beijing
at community scale. J. Saf. Environ. 2021, 21, 1662–1669. [CrossRef]
2. Zhao, L. Theory and Method of Urban Emergency Shelter Planning and Construction; Science Press: Beijing, China, 2014.
3. Yu, L.; Zhang, X.; Zhu, Q. Rationality Analysis of Layout of Urban Earthquake Emergency Shelters Based on Remote Sensing
Image, ArcGIS and WVD Methods: A Case Study of Linfen City. In Proceedings of the International Conference on Smart City
and Systems Engineering, Changsha, China, 25–26 November 2016; Volume 11, pp. 59–63.
4. Chen, M.; Lv, F. Research on Layout Optimization of Emergency Evacuation Sites in Ancient Suzhou City Based on Social
Network Analysis. Urban Stud. 2022, 29, 1–8+15.
5. Yu, S.; Chen, W.; Zhang, N.; Liu, C.; Wang, S.; Wang, Y. Study on the Layout of Urban Earthquake Emergency Shelters based on
Spatiotemporal Big Data: A Case Study of Xixia District, Yinchuan City. North China Earthq. Sci. 2022, 40, 8–15.
6. Li, X.; Liu, C. Accessibility and service of Shengyang’s urban parks by network analysis. Acta Ecol. Sin. 2009, 29, 1554–1562.
7. Xing, X.; Zhou, D.; Luo, Y. Study on geological disasters of karst in Xuzhou and its protection measures. Chinses J. Geol. Hazard
Control 2014, 25, 51–58. [CrossRef]
8. Zi, J.; Lian, W. Karst geological disaster and geotechnical engineering problem of Xuzhou. Earth Environ. 2005, 33, 330–352.
9. Liu, S. The Location Choice and Spatial Layout of Urban Emergency Shelters: A Case Study of Nanjing. Phd Thesis., Nanjing
Normal University, Nanjing, China, 2012.
10. Hakimi, S.L. Optimum distribution of switching centers in a communication network and some related graph theoretic problems.
Oper. Res. 1965, 13, 462–475. [CrossRef]
11. Daskin, M.S. A new approach to solving the vertex p-center problem to optimality: Algorithm and computational results.
Commun. Oper. Res. Soc. Jpn. 2001, 65, 428–436.
12. Zhang, H. Study on the Optimized Design of the Layout of Emergency Refuge in the Main Urban Area of Baotou. Phd Thesis,
Inner Mongolia University of Science and Technology, Baotou, China, 2021. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.