Available via license: CC BY 4.0
Content may be subject to copyright.
Sustainability 2023, 15, 765. https://doi.org/10.3390/su15010765 www.mdpi.com/journal/sustainability
Review
Review on Urban Flood Risk Assessment
Cailin Li 1, Na Sun 1, Yihui Lu 2,*, Baoyun Guo 1, Yue Wang 1, Xiaokai Sun 1 and Yukai Yao 1
1 School of Civil and Architectural Engineering, Shandong University of Technology, Zibo 255000, China
2 Geographic Information Engineering, Shandong Provincial Institute of Land Surveying and Mapping,
Jinan 250102, China
* Correspondence: luyihui@shandong.cn
Abstract: Under the background of rapid urban development and continuous climate change, fre-
quent floods around the world have caused serious economic losses and social problems, which has
become the main reason for the sustainable development of cities. Flood disaster risk assessment is
an important non-engineering measure in urban disaster prevention and mitigation, and scientific
flood disaster risk assessment is the premise and foundation of flood disaster risk management. This
paper summarizes the current situation of flood risk assessment by analyzing the international lit-
erature in recent 20 years. The mechanism of flood disaster is mainly discussed. The flood disaster
assessment methods are summarized, including historical disaster statistics method, multi-criteria
index system method, remote sensing and GIS (Geographic Information System) coupling method,
scenario simulation evaluation method and machine learning method. Furthermore, the develop-
ment status of flood risk analysis and forecasting is summarized. Finally, the development trend
and direction of flood risk assessment are put forward.
Keywords: urban flood; risk assessment; disaster-causing mechanism; review; assessment indica-
tors; flooding forecast
1. Introduction
With the changing global climate and the impact of human activities on the environ-
ment, extreme rainfall events in urban areas are increasing. Rainstorm and flood disasters
have become one of the main natural disasters affecting the social and economic develop-
ment of urban areas and the safety of people’s lives and property [1,2]. Many scholars
have pointed out that with the acceleration of urban development, the frequency of urban
floods is increasing, and the impact on cities is also greatly enhanced [3,4]. As shown in
Figure 1, the frequency of global flood disasters has shown an upward trend in the past
three decades.
Flood disaster is considered to be one of the most frequent disasters in the world.
According to statistics, the proportion of rainstorm flood disaster is about 40% of the
global losses caused by natural disasters [5]. Globally, floods have caused huge economic
and social losses, and they continue to increase [6]. For example, rapid climate change has
led to more frequent floods in Canada and wider spread [7]. Between December 2010 and
January 2011, parts of Australia experienced widespread flooding, resulting in 37 deaths
and total economic losses of more than A $30 billion [8]. In the UK, annual flood damage
is estimated at £ 1.1 billion and, without additional adaptation measures, is expected to
rise to £27 billion by 2080 under the worst climate change scenario [9]. In 2021, China ‘s
flood disasters caused a total of 59.01 million people affected, 590 people died and disap-
peared, 152,000 houses collapsed, and direct economic losses of 245.89 billion yuan. From
14 June to 30 August 2021, severe floods occurred in Pakistan, resulting in 1162 deaths,
3554 injuries, more than 33 million people affected, and direct economic losses of more
than USD 10 billion. The global losses caused by floods cannot be underestimated. Figure
Citation: Li, C.; Sun, N.; Lu, Y.; Guo,
B.; Wang, Y.; Sun, X.; Yao, Y. Review
on Urban Flood Risk Assessment.
Sustainability 2023, 15, 765.
https://doi.org/10.3390/su15010765
Academic Editor: Miklas Scholz
Received: 9 November 2022
Revised: 18 December 2022
Accepted: 28 December 2022
Published: 31 December 2022
Copyright: © 2022 by the authors. Li-
censee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and con-
ditions of the Creative Commons At-
tribution (CC BY) license (https://cre-
ativecommons.org/licenses/by/4.0/).
Sustainability 2023, 15, 765 2 of 24
2 shows the number of deaths and missing persons caused by floods in the world in the
past 30 years.
Figure 1. Frequency of flood disasters from 1992 to 2021.
Figure 2. Number of dead, missing and affected by floods, 1992–2021.
Under the influence of human activities and climate change, the form and mechanism
of urban flood disasters have undergone dramatic changes. The change of underlying sur-
face caused by urbanization has affected the mechanism of runoff and confluence, and to
some extent destroyed the urban drainage and waterlogging system [10,11]. The construc-
tion of urban flood control and disaster reduction system is facing new pressures and
challenges, and the risk of urban flood disaster is on the rise [12]. Facing the severe flood
disaster situation, carrying out flood disaster monitoring and risk assessment has become
an urgent need to enhance the capacity of disaster prevention and mitigation [13]. In order
to alleviate the current increasing urban flood disaster problem, it is imperative to
strengthen the systematic research on urban flood disaster risk assessment.
Flood disaster risk assessment is a series of processes to analyze and evaluate the
vulnerability of disaster-causing factors and disaster-bearing bodies that may bring po-
tential threats or injuries to life, property, livelihoods, and the environment on which hu-
mans depend, and then determine the nature, scope, and loss of flood risks. It aims to
improve the accuracy of mastering the spatial distribution of flood disaster risks, and com-
prehensively evaluates the natural and social attributes of flood disasters [14]. The basic
process can be summarized as [15]. (1): risk identification, that is, to find out the risk
source of flood disaster; (2): Risk assessment, according to a certain definition of risk, gives
the quantitative analysis results, the results of the form of risk zoning map [16] and dam-
age statistics [17], provide the basis for risk management; (3): Risk analysis, including risk
analysis, vulnerability analysis and exposure analysis.
Flood disaster risk assessment is one of the foundations of disaster prevention plan
formulation, and also an important basis and means for formulating disaster prevention
Sustainability 2023, 15, 765 3 of 24
and mitigation policies and measures. Before the occurrence of disasters, corresponding
disaster prevention methods, measures and priorities can be formulated for various pos-
sible disaster situations. It is also possible to use multi-year disaster prevention and miti-
gation data to compile a disaster prevention expert system and formulate various emer-
gency plans to cope with various emergencies. According to the results of flood risk as-
sessment, the area is forecasted in advance, and the decision-makers are given sufficient
decision-making time, which can reasonably determine the flood disaster prevention
standards, optimize the order of rescue and disaster relief, optimize the implementation
order of disaster prevention system construction, and provide scientific basis for disaster
risk area management.
Therefore, it is necessary to systematically sort out and summarize the research of
flood disaster risk assessment to provide better theoretical and technical support for sub-
sequent researchers. The purpose of this paper is to review and summarize the research
progress of urban flood risk assessment technology from a scientific and professional per-
spective. In this context, this paper first systematically expounds the mechanism of urban
flood disaster, summarizes the methods of urban flood risk assessment, and expounds the
latest research status of each method, then summarizes the research status of flood risk
analysis and forecasting, and finally emphasizes the key research points in three aspects:
model construction, data utilization and discipline integration.
2. Mechanism of Urban Flood Disaster
Scientific and systematic understanding of the disaster-causing mechanism of urban
waterlogging disasters is helpful to find the source of urban waterlogging disasters, ex-
plore the driving force of their occurrence and development, and find corresponding so-
lutions in urban flood disaster management. As shown in Figure 3, global climate change,
urbanization, urban water system shrinkage and municipal facilities lag have seriously
affected the occurrence of urban flood disasters. Therefore, we expound the formation
mechanism of urban flood disaster from these three aspects.
Figure 3. Factors Affecting Urban Flood Disaster.
Sustainability 2023, 15, 765 4 of 24
2.1. Global Climate Change Leads to Frequent Extreme Rainfall
In 2021, extreme cold weather will break out in the United States, South Africa and
Brazil; after entering the summer, the United States, Canada, Kuwait and Algeria contin-
uously set high temperature records. In July, China’s Henan Province encountered ex-
treme heavy rainfall, which is rare in history. Zhengzhou broke the rainfall record and
rained nearly a year in a day. At the same time, Western Europe also suffered heavy rains,
resulting in hundreds of deaths. The frequent occurrence of extreme weather warns us
that global climate change is an indisputable fact, and climate change and its conse-
quences are considered to be one of the most important challenges facing contemporary
mankind [18]. According to the Fifth Assessment Report of the Intergovernmental Panel
on Climate Change (IPCC), the global average surface temperature increased by about
0.85 degrees from 1880 to 2012 [19], posing a huge challenge to the sustainable develop-
ment and utilization of global and regional water resources [20,21]. Global warming has
led to more intense and frequent atmospheric activities, changing the climate pattern in
the continental region, especially in the middle and high latitudes. Extreme precipitation
weather occurs frequently and has an increasing trend [22].
Extreme rainfall is the most direct driving factor for urban flood events, and climate
change directly leads to changes in global and regional precipitation. Many studies have
analyzed the temporal and spatial evolution characteristics and causes of global and re-
gional precipitation in a changing environment. Trenberth’s study showed that global
land surface rainfall increased by 2% since the 20th century [23]. Konapala et al. divided
the world into nine regions according to rainfall characteristics, and analyzed the charac-
teristics of precipitation and evaporation in nine regions of the world. The results showed
that the precipitation increased in four of the nine regions of the world, and the evapora-
tion decreased in the remaining five regions, but the average precipitation and evapora-
tion in nine regions showed an increasing trend [24]. Studies by Schuster et al. show that
the frequency and magnitude of heavy rainfall events in the midwestern United States are
increasing due to climate change [25]. Many scholars have analyzed the temporal and
spatial evolution characteristics and causes of changes in China’s annual average precip-
itation and extreme precipitation. Studies have shown that the annual average precipita-
tion in southwestern, northwestern and eastern China has increased significantly in the
past few decades. The extreme rainfall events in the North China Plain are more random,
and the maximum rainfall in history is much higher than the common heavy rainfall in
the North China Plain [26,27].
At present, the problem of extreme rainfall caused by climate change has attracted
wide attention around the world, and further in-depth research work is needed. There-
fore, we should strengthen international cooperation, realize low-carbon emission reduc-
tion, green and high-quality development, carry out ecological environment protection,
comprehensively enhance green and low-carbon energy conservation and environmental
protection awareness, reduce greenhouse gas emissions, slow down climate warming,
and reduce the impact of extreme weather.
2.2. Urbanization Leads to Vegetation Reduction and Underlying Surface Hardening
Urbanization is a common phenomenon worldwide. With the rapid urbanization
process, the number of urban population is growing rapidly. According to statistics, 52
percent of the world’s population lives in urban areas and the number of urban popula-
tions is projected to continue to grow, with nearly 67 percent of the world’s population
expected to live in urban areas in 2050 [28]. In the process of rapid urbanization, the trend
and speed of urban underlying surface hardening and imperviousness are increasing year
by year. Studies have shown that during the rapid urbanization of the four central cities
of Iowa in the United States from 1940 to 2011, the annual average total impervious surface
change rate of the city increased from 2.42%(1940–1961) to 4.17%(1990–2002).
Sustainability 2023, 15, 765 5 of 24
The development of urbanization will inevitably lead to the increase of urban con-
struction area, and the surface vegetation such as grassland and trees in the city will be
transformed into urban infrastructure construction land such as high-rise buildings and
asphalt roads. As a result, the underlying surface is gradually hardened, and the ability
and area to absorb rainwater are gradually reduced. The original ‘breathing ground’ be-
comes unable to infiltrate naturally, and a large amount of rainwater is retained on the
urban ground [29]. At present, many scholars have carried out research on the hydrolog-
ical effects of urbanization with the help of hydrological and hydrodynamic models. It is
generally believed that the expansion of urbanization has increased the runoff and peak
flow of urban watersheds, aggravated the degree of urban flood disasters, and reduced
the interaction process between the base flow of urban watersheds and groundwater and
surface water in the surrounding areas of cities [30–33].
Therefore, rational planning of urban green space and transportation, increasing the
area of green plants, enhancing the permeability of underlying surface, and increasing the
use of low-impact development design are effective means to reduce urban rainwater sur-
face runoff, restore urban natural hydrological cycle, and reduce urban vulnerability [34].
2.3. The Shrinkage of Urban River Network and the Lag of Municipal Facilities
The single influencing factor of continuous precipitation or heavy precipitation is not
enough to cause urban waterlogging. Another key point is that the drainage of the city’s
drainage pipe network is not smooth and the drainage is not timely, resulting in an in-
creasing depth of water in the city and an increasing time of water retention, which ulti-
mately leads to the occurrence of urban waterlogging disasters [35].
Due to the planning and design requirements of rapid drainage of urban municipal
facilities and the timing of construction, municipal facilities are updated slowly and over-
loaded. This has led to the problems of urban water accumulation, slow construction of
municipal facilities in urban fringe areas, and lack of water storage capacity in pipeline
networks. When the rainfall is large, the pipe network system water collection and drain-
age speed is slow, rainwater overflow phenomenon [36]. In addition, urban municipal
facilities are not only an important means of urban waterlogging prevention, but also a
vulnerable point when urban waterlogging disasters occur. Especially lifeline projects,
once affected by disasters, may lead to a series of secondary disasters and cause huge
losses to the city. At the same time, the storage function of urban water system is neglected
due to the value of urban land. Urban construction often artificially interferes with river
regulation and storage, changes the natural catchment pattern and regulation and storage
function pattern, and reduces the anti-waterlogging function of urban rivers and ponds
[37].
Therefore, it is necessary to focus on a larger urban basin to rationally arrange mu-
nicipal water conservancy facilities, reduce the conflict between cities and cities in dense
urban clusters on the relief and utilization of water resources, promote the recovery of
natural hydrology, and promote the construction of urban sustainable flood control sys-
tem.
3. Flood Risk Assessment Method
The selection of flood risk assessment methods is generally based on the spatial scale
of evaluation and the completeness of basic data. The use of methods directly affects the
timeliness of analysis results and the accuracy of evaluation results. The current flood as-
sessment methods have some limitations. Aiming at these problems, this paper compre-
hensively divides the flood assessment methods into five aspects, namely, historical dis-
aster mathematical statistics method, multi-criteria index system method, remote sensing
and GIS coupling method, scenario simulation evaluation method and machine learning
method, and expounds its current situation, advantages and limitations.
Sustainability 2023, 15, 765 6 of 24
3.1. Historical Disaster Mathematical Statistics
The historical disaster mathematical statistics method generally uses the historical
flood event disaster data released by the government or relevant departments, such as the
number of affected people, the number of house damage, the inundation range, and the
facility damage data caused by a flood disaster event. According to the law of flood dis-
aster, the mathematical statistics method is used to analyze these data and predict the
frequency, depth and direct loss of flood disasters in different spatial locations [38]. This
method is an early method used in flood disaster assessment research. In the absence of
existing assessment data, it analyzes the law and trend of risk based on historical data,
and then assesses the current or future risk.
Some researchers use historical rainstorm and flood data to construct vulnerability
curves and conduct quantitative or qualitative flood risk analysis in the study area [39–
42]. On the basis of historical disasters, combined with land use type data [43] and hydro-
logical model [44], the spatial transformation and damage of flood risk are evaluated in
more detail. Benito integrated multidisciplinary knowledge and methods to assess flood
risk in the study area using long time series of historical flood data [38]. Liu used the
historical flood disaster data of the affected counties and cities as the basic unit to com-
prehensively analyze the flood disaster risk degree of the study area by Markov chain
[45]. Li proposed a quantitative assessment model for flood disaster losses by studying
the database and the spatial distribution method of socioeconomic data [46]. Denis stud-
ied flood risk management on the basis of historical flood disaster records [47].
The historical disaster statistics method has clear thinking and simple calculation,
and its evaluation results are in good agreement with the actual situation. However, the
data it relies on are usually of poor spatial accuracy and time lag, which is only suitable
for large-scale spatial areas such as cities and provinces [14]. In addition, this method re-
quires the study area to have more detailed and abundant historical disaster data [48],
which is more sensitive to missing data.
3.2. Multi-Criteria Index System Method
Multi-criteria index system is the most widely used flood risk assessment method.
According to the natural characteristics and socio-economic particularity of the study
area, it selects a number of direct and indirect index factors related to the formation of
urban flood disasters, constructs the flood risk assessment index system of the study area,
and applies certain mathematical models or methods (such as analytic hierarchy process,
set pair analysis, fuzzy comprehensive evaluation method) to evaluate the comprehensive
impact of different factors on the flood risk of the study area, so as to define the risk size
of the study area. Then select the corresponding evaluation methods for risk assessment,
such as analytic hierarchy process, set pair analysis, fuzzy comprehensive evaluation
method and so on.
Most researchers have selected indicators such as slope, land use, precipitation, and
population density to construct a multi-index system, and used the analytic hierarchy pro-
cess to quantify the index weights to conduct risk assessments in the study area [49–51].
In the risk assessment of flood disaster, Gilbert F. White took the human response to the
disaster as one of the basis for the first time, and incorporated the characteristics of popu-
lation and housing structure into the evaluation system [52,53]. Jiang determined the in-
dicators affecting the flood risk of each county from the four aspects of flood risk, expo-
sure, vulnerability, and regional disaster prevention and mitigation capabilities, and inte-
grated them into a flood risk index. On this basis, the flood risk zoning map of the Song-
hua River Basin was obtained [54]. When using this method for risk analysis, the determi-
nation of index weight is an important research issue. Therefore, researchers will use dif-
ferent methods such as subjective and objective weight determination to determine the
weight of the selected indicators. For example, Lyu used the analytic hierarchy process
Sustainability 2023, 15, 765 7 of 24
and the interval analytic hierarchy process [55], Cabrera used the analytic hierarchy pro-
cess and the maximum entropy model [56] to distribute the weights of the indicators. Feng
found that there are many factors involved in the process of estimating regional vulnera-
bility, and it is an effective way to use fuzzy comprehensive evaluation method to carry
out comprehensive research. Using fuzzy comprehensive method, we can understand the
vulnerability of the region in an all-round way [57]. In order to assess the housing risk,
agricultural risk and comprehensive risk in Malaysia, Jiang established a risk assessment
model library by using fuzzy comprehensive evaluation method, simple fuzzy classifica-
tion method and fuzzy similarity method [58]. Chen used the projection pursuit method
to evaluate the flood disaster in Sichuan Province [59]. With the deepening of research, it
has become more popular to combine with geographic information system technology to
study flood risk grade map and carry out risk zoning in the study area [60,61].
The multi-criteria index evaluation method can flexibly select indicators according to
regional characteristics and data availability, and can flexibly select index quantification
methods to evaluate flood risk. It is not only suitable for regions with different spatial
scales, but also can intuitively reflect the relationship between each index and flood risk.
However, when applying the multi-criteria index evaluation method, the difficulty lies in
the selection of indicators and the determination of weights [62]. The selection of indica-
tors is limited by factors such as data availability, characteristics of the study area, and
data accuracy [16], and the determination of weights is affected by the analysis method.
At present, the weight calculation method is mostly based on expert knowledge and ex-
perience [61], which makes the evaluation results of multi-index evaluation method sub-
jective.
3.3. Coupling Method of RS and GIS
The method based on the coupling of RS and GIS is to use remote sensing technology
to obtain information such as water area, inundation duration, and disaster-bearing bod-
ies in the disaster area, and then input these information into GIS tools for spatial analysis.
Based on time series Radarsat remote sensing images and other geographic infor-
mation data, Chubey constructed a prediction model that coupled Markov model and
spatial logistic regression model. Based on this model, they predicted the intensity and
spatial distribution of future floods in the Red River Basin in southern Manitoba [63]. Us-
ing IRS-1D and LISS-III data from 2002 to 2003, Chowdary evaluated the surface and sub-
surface flooding areas in the pre-monsoon and post-monsoon seasons of the Bihar flood-
prone areas in a GIS environment [64]. Barredo analyzed the spatial distribution of flood
disasters in Europe for many years by means of GIS data set, and obtained the trend map
of flood disasters in the study area [65]. Mateeul Haq used MODIS data in 2011, the study
area is Sindh Province, Pakistan, combined with his population distribution, vegetation
cover, etc., while using RS and GIS for loss assessment and flood warning [66]. Using
NOAA AVHRR images, Md. Monirul Islam extracted flood inundation frequency infor-
mation and used it as an indicator in flood risk assessment [67]. On the basis of regional
risk identification, Cheng selected appropriate evaluation indicators from the aspects of
disaster-causing factors, disaster-pregnant environment and hazard-affected bodies, and
established an evaluation index system to evaluate flood disaster risk from the composite
scenario of various indicators through scenario analysis technology [68]. Liu used GIS
technology and AHP method to obtain the disaster vulnerability assessment map and dis-
aster risk assessment map of the study area, and determined the high risk level distribu-
tion area of the area [69]. Liu used GIS technology and RS technology to find that the
fusion of multi-source remote sensing data has a good effect on the timely and accurate
assessment of flood disasters [70]. Based on long time series of AVHRR and MODIS im-
ages, Huang studied the inundation range of flood water bodies in the area and analyzed
the inundation frequency of flood water bodies. The flood risk model was constructed
from four aspects: hazard, vulnerability, exposure and disaster prevention and mitigation
capacity, and the expression of risk assessment results was expressed from two scales of
Sustainability 2023, 15, 765 8 of 24
grid and administrative unit [71]. Li used 10 years of MODIS images to extract the sub-
merged area of water body, analyzed the spatial and temporal changes, and evaluated the
flood loss [72]. Taking Quzhou City, Zhejiang Province as the research area, Xiao used
gridded geographical background data, spatial socio-economic data, combined with re-
mote sensing data of the research area, and used GIS to evaluate the flood risk of the re-
search area [73]. Considering the risk and vulnerability, Jiang selected suitable index fac-
tors, applied GIS to spatialize each factor, and combined with the analytic hierarchy pro-
cess to determine the weight, and obtained the flood disaster risk distribution map of
Zhejiang Province [74].
The use of RS technology can quickly obtain the flood risk information of the study
area, which is more practical for the study of large flood disasters. However, for small
floods, due to the short duration, remote sensing data often cannot accurately capture the
flood process, and there are great limitations in spatial scale and time resolution [75]. In
the future, using multi-source fusion data and GIS integrated application for flood risk
analysis will become a key research direction [76].
3.4. Scenario Simulation Evaluation Method
Based on hydrological and hydrodynamic models, scenario simulation assessment is
a method for dynamic simulation and assessment of disaster processes in the study area
by designing disaster scenarios of specific frequency and intensity (design rainfall) [35].
This method can intuitively and accurately give the spatial distribution characteristics of
urban flood disaster risk, which can provide some reference for managers’ disaster pre-
vention and mitigation and risk management decision-making, and provide data support
for disaster risk transfer [77].
Since the end of the 20th century, hydrological and hydrodynamic research has made
great progress, objectively promoting the application of scenario simulation in flood risk
assessment. One-dimensional pipe network hydraulic model is the first widely used nu-
merical model, because the required data is relatively simple and the model results are
more accurate, such as SWMM model [78], SIPSON model [79]. Based on the one-dimen-
sional hydraulic model, two-dimensional numerical models have gradually begun to be
promoted. These models can accurately simulate the evolution of surface floods during
the entire rainfall process because they can couple one-dimensional surface, river channel
and two-dimensional surface, such as commercial models MIKE [80], Info Works ICM
[81]. In order to achieve a more realistic simulation results, three-dimensional numerical
model has also been the attention of researchers, such as Delft3D model, FLOW-3D model,
water quality simulation and flow field simulation. Table 1 is the current mainstream
flood simulation model.
Table 1. Flood model.
Model
The Model Characteristic
The Input Parameters
Development Organiza-
tions
SWMM
Provides distributed hydrological
module, one-dimensional hydrody-
namic module.
Land use, terrain data, rain-
fall, rainfall intensity, pipe
network data, etc.
EAP United States Envi-
ronmental Protection
Agency
HEC-RAS
One-dimensional and two-dimen-
sional hydrodynamic modules are
provided.
Topographic and hydrologi-
cal data
U.S. Army Corps of Engi-
neers Hydrological Engi-
neering Center
PCSWMM
With SWMM as the core, it provides
pre-processing and post-processing
modules, and can simplify the calcu-
lation of two-dimensional surface.
Land use, terrain data, rain-
fall, rainfall intensity, pipe
network data, etc.
Canadian Hydraulic Cal-
culation Institute
Sustainability 2023, 15, 765 9 of 24
LISFLOOD-FP
Provide two-dimensional hydrody-
namic module.
Terrain and hydrological
data, pipe network data
University of Bristol
InfoWorks ICM
Highly integrated, fully functional, re-
alizes the coupling simulation of hy-
drology, hydrodynamics and water
quality, and has strong pre- and post-
processing functions.
Pipe network data, terrain
data
UK HR Wallingford
MIKE
It includes MIKE URBAN, MIKE
FLOOD, MIKE21 and other modules.
Each module is relatively independ-
ent and fully functional, and is widely
used in various projects.
Topographic and hydrologi-
cal data, roughness, waves,
tide levels, etc.
Danish Hydraulic Insti-
tute DHI
EFDC
Provide water quality module, can
simulate point source pollution, non-
point source pollution, organic matter
migration process.
River network data, pipe net-
work data, terrain data
Virginia Institute of
Oceanography
Delft3D
It is suitable for three-dimensional hy-
drodynamic water quality simulation,
which can simulate the hydrodynam-
ics of estuary and port.
Terrain and hydrological
data, grid data
Delft Hydraulic Research
Institute, Netherlands
FLO-2D
Provide two-dimensional hydrody-
namic module, one-dimensional cal-
culation embedded SWMM module.
Topographic and hydrologi-
cal data
FLO-2D Software, Inc.
FLOW-3D
CFD software, provides three-dimen-
sional hydrodynamic module, suita-
ble for analysis of three-dimensional
flow field.
Terrain data, hydrological
data
Flow Science, USA
Flood Simulation Model
Based on the unstructured grid, the
coupling of urban ground flooding
and pipeline is realized for the first
time.
Pipe network data, terrain
data
China institute of water
resources and hydro-
power research
HydroInfo
Numerical simulation of complex wa-
ter flow and transport process is pro-
vided.
Pipe network data, river net-
work data, terrain data
Dalian university of tech-
nology
HydroMPM
The numerical method is used to sim-
ulate the dynamic processes such as
water flow, water quality and sedi-
ment and their associated processes.
Hydrological data, river net-
work data
Zhujiang Institute of Wa-
ter Conservancy Science
GAST
The Godunov scheme is used to solve
the two-dimensional Saint Venant
equations, and the GPU parallel com-
puting technology is used to acceler-
ate the calculation.
Hydrological data, pipe net-
work data, terrain data
Xi’an university of tech-
nology
IFMS/Urban
Based on the self-developed GIS plat-
form, one-dimensional and two-di-
mensional coupling calculation is re-
alized.
Terrain and hydrological
data, river network data
China institute of water
resources and hydro-
power research
Based on the hydrological and hydrodynamic model, the researchers carried out sce-
nario simulation and flood risk assessment in the study area. Some scholars coupled one-
Sustainability 2023, 15, 765 10 of 24
dimensional pipe network model with two-dimensional surface model [82], and even pro-
posed two-dimensional flood simulation model based on empirical formula and artificial
intelligence [83]. For example, Xu pointed out that the scenario simulation method is the
trend of future risk assessment. Taking the sponge city demonstration area of Jinan City
as the research area, combined with the simulation results of the hydrological hydrody-
namic model, the risk zoning based on the submerged depth and time threshold and the
risk zoning based on the empirical formula were compared [38]. Suarez evaluated the
impact of floods and climate change on Boston’s urban transport system using scenario
simulations [84]. By designing different rainfall scenarios, Su constructed a two-dimen-
sional hydraulic simulation to simulate the spatial and temporal distribution of rainstorm
waterlogging, and then evaluated the rainstorm waterlogging in Xinluo District, Longyan
City, Fujian Province in combination with the vulnerability curve [85]. Bisht compared the
simulation effect of one-dimensional SWMM hydrological model with that of two-dimen-
sional Mike Urban hydrodynamic model in the study area, and found that the two-di-
mensional hydrodynamic Mike Urban model overcomes the limitations of one-dimen-
sional SWMM hydrological model in simulating flood inundation range and water depth,
and the simulation effect is better [86]. Zeng coupled the one-dimensional SWMM hydro-
logical model with the two-dimensional Lis Flood-FP hydrodynamic model, and realized
the simulation of the submerged range and submerged depth of rainstorm and flood in
the study area. The accuracy of the coupling model was verified by field investigation
[87]. In addition, some researchers constructed coupling models, designed simulated rain-
fall under different scenarios, and assessed the risk of precipitation inundation in the
study area. Jing analyzed the vulnerability of rainstorm flood disaster in Pudong area of
Shanghai by designing rainstorm runoff scenarios with different return periods [88].
Huang used Info Works ICM software to construct a rain and flood model for the
Donghao River Basin in Guangzhou, and combined with eight indicators to form an index
system to evaluate the flood risk of the study area under the 1-year, 5-year and 50-year
rainstorm scenarios [89]. Zhang used the MIKE hydrological and hydrodynamic model to
simulate the flood in the downstream area of Yindongnan Plain under the rainstorm sce-
narios of 5-year return period, 11-year return period, 20-year return period, 50-year return
period and 100-year return period, and dynamically evaluated the flood risk by combin-
ing the evaluation index system composed of flood depth, population density and GDP
per area [90].
Because of its advantages of intuitive and high-precision reflection of the influence
range and influence degree of disaster events (the characterization effect of disaster-caus-
ing factors), the scenario simulation assessment method has been widely used in related
disaster refined simulation scenarios at home and abroad. At the same time, this method
is also one of the main methods considered in future flood disaster risk assessment re-
search. However, the scenario simulation method has high requirements for the geo-
graphical data of the study area, the model is not universal, and greatly depends on the
constructed stormwater model. Due to the strictness of the data required for modeling
and the complexity of the modeling process, this method is not suitable for large study
areas and generally lacks consideration of the vulnerability of hazard-affected bodies [16].
3.5. Machine Learning
Machine learning method is a new method for flood risk assessment in recent years.
It relies on intelligent algorithms to learn the characteristics of flood risk and automatically
acquires the input-output relationship between driving factors and flood risk, providing
a more flexible, objective and rapid flood risk assessment method.
At present, some machine learning models have been applied to flood risk assess-
ment and have shown good evaluation results, such as random forest model, extreme
gradient boosting model [91], support vector machine [92], etc. For example, Thanh used
random forest and support vector machine to evaluate flood risk in urbanized areas, and
found that the overall accuracy and Kappa coefficient of the evaluation were high [93].
Sustainability 2023, 15, 765 11 of 24
Random forest model is more suitable for flood risk assessment than support vector ma-
chine model [94,95]. Studies have also shown that it is necessary to use machine learning
methods to develop reliable flood risk assessment maps. Shu conducted back propagation
neural network training on flood disaster data in the study area to generate flood risk
distribution map. The results are basically consistent with the results obtained by the
emergy method [96]. Ma introduced the XGBoost model for risk assessment and found
that this method is an effective way to obtain high-quality county-level flood risk maps
[97]. Phama uses a risk assessment method based on the combination of deep learning
network and analytic hierarchy process. Research shows that the combination of the two
can more accurately develop regional flood risk assessment maps [98] Many researchers
use different machine learning models in the same research area to compare the effects of
the models. Khosravi applied three index evaluation methods (VIKOR, TOPSIS and SAW)
and two machine learning models (NBT and NB) to the flood risk assessment of Ningdu
County. The results show that the NBT model performs best in the field of flood risk as-
sessment [99]. Li proposed a set of rainstorm waterlogging disaster prediction index sys-
tem by comprehensively considering disaster-causing factors, exposure and vulnerability,
and constructed the BP model and XGBoost model for rainstorm waterlogging disaster
prediction. The results show that when the combination of indicators that comprehen-
sively consider disaster-causing factors, exposure and vulnerability indicators and do not
reduce the dimension by principal component analysis is used as the rainstorm waterlog-
ging disaster prediction index system, the prediction accuracy of BP model and XGBoost
model is optimal [100].
With the development of computer technology and the progress of various machine
learning algorithms in recent years, machine learning methods have great potential in
flood risk assessment. However, the machine learning method relies heavily on the com-
pleteness and reliability of the sample data set, which directly affects the rationality of the
flood characteristics and risk assessment results learned by the model.
As shown in Table 2, there is a certain correlation between the five flood disaster risk
assessment methods. Different evaluation methods have their applicability. In the face of
different research objects, it is necessary to select the appropriate evaluation method ac-
cording to the characteristics of the research area, considering the requirements of the re-
search scale and the accuracy of the research results.
Table 2. Comparison of flood risk assessment methods.
Appraisal
Procedure
Appraisal
Unit
Methods and Principles
Method Advantages
Method Disadvantages
Historical
disaster
mathematical
statistics
Province,
city, county
For the randomness of flood
disasters, historical samples
are used to estimate the
probability of flood disasters.
The idea is clear, the
calculation is simple, and
the evaluation results are
in good agreement with
the actual situation.
The study area is required to
have more detailed and
abundant historical disaster
data. More sensitive to missing
data.
Multi-criteria
index system
method
Province,
city, county,
grid unit
Select indicators to build index
system, determine the weight
and establish a comprehensive
function of indicators to obtain
the risk index of the evaluation
area.
Suitable for different
spatial scales of the
region, and can
intuitively reflect the
relationship between the
indicators and flood risk.
The selection of indicators and
the determination of weights
greatly affect the evaluation
results.
Coupling
Method of
Remote Sensing
and GIS
Province,
city, county,
grid unit
The analysis method of remote
sensing and geospatial data
processing combined with GIS
analysis.
Using RS technology can
quickly obtain the flood
risk information of the
study area, which is
more practical for large-
For small floods, due to the
short duration, remote sensing
data often cannot accurately
capture the flood process, and
there are great limitations in
Sustainability 2023, 15, 765 12 of 24
scale flood disaster
research.
spatial scale and time
resolution.
Scenario
simulation
evaluation
method
city, county,
grid unit
Hydrological and
hydrodynamic models are
used to simulate the process of
disaster occurrence and
dynamically show the
temporal and spatial evolution
of disaster.
Intuitively and
accurately reflect the
scope and extent of the
impact of disaster events.
The requirements for
geographic data in the study
area are high, and the model is
not universal, which greatly
depends on the constructed
stormwater model.
machine
learning
Province,
city, county
Machine learning methods
such as support vector
machine and random forest
are used to evaluate flood risk
in the evaluation area.
Effectively improve the
accuracy and efficiency
of assessment.
Over-reliance on the reliability
of sample data.
4. Analysis of Urban Flood Risk Assessment and Flood Forecast
The flood risk assessment process mainly includes three parts: risk identification, risk
analysis and risk assessment, as shown in Figure 4. Among them, the selection of risk
analysis indicators is the key link, and the selection of appropriate evaluation indicators
to construct an index system is particularly important for improving the quality of flood
risk assessment [16]. According to the theory of disaster science, the disaster process is
determined by the circular feedback process composed of the three elements of disaster
environment, disaster-causing factor and disaster-bearing body. The stability of the dis-
aster environment, the risk of disaster-causing factors and the vulnerability of disaster-
bearing bodies determine the size of the disaster.
Figure 4. Flood disaster risk assessment process.
4.1. Flood Disaster Theory
In the process of flood disaster risk research, different scholars have given different
flood disaster risk expressions. Some scholars have given the definition of flood disaster
risk: combined with the characteristics of flood disaster, flood disaster risk is a compre-
hensive representation of Hazard × Vulnerability [101,102]. Meyer et al. proposed the ex-
pression ‘flood risk = Probability × Consequence’ [103]. Some other scholars have given
the expression of ‘flood risk = Hazard × Exposure × Vulnerability’, and believe that flood
risk is the result of the interaction of hazard, exposure and vulnerability (H-E-V frame-
work) [104,105]. In addition, under the framework of ‘H-E-V’, some scholars considered
the weakening effect of regional emergency capability on disaster risk, added the factor
Sustainability 2023, 15, 765 13 of 24
of disaster prevention and mitigation capability, and put forward the expression of ‘flood
risk = Hazard × Exposure × Vulnerability × Emergency (H-E-V-R framework)’ [106,107].
The existing flood disaster system theory is shown in Table 3.
Table 3. Summary of flood disaster theory.
System Components
Risk Expression
Researchers (Source)
Hazard-vulnerability
R = f(H,V)
Zhou et al. [102]; Xu et al. [101]
Hazard-Exposure-vulnerability
R = f(H,E,V)
Kron [104]; Li et al. [105]
Hazard-Exposure-vulnerability-Emergency
R = f(H,E,V,R)
Zhang et al. [106]; Du et al. [107]
4.2. Risk Result Analysis of Flood Disaster
The urban flood assessment framework adopted in the IPCC Fifth Assessment Re-
port is Hazard × Exposure × Vulnerability, that is, the ‘H-E-V’ framework. In order to
analyze the causes of flood disaster risk more comprehensively and deeply, and fully con-
sider the importance of disaster prevention and mitigation capacity to flood disaster as-
sessment, based on the ‘H-E-V-R’ framework, this paper sorts out the main research con-
tents of the four major elements of disaster-causing factor risk, disaster-pregnant environ-
ment sensitivity, disaster-affected body vulnerability, and disaster prevention and miti-
gation capacity, and summarizes the commonly used assessment indicators [108].
4.2.1. Risk of Disastrous Factors
Disaster-causing factors, that is, various abnormal factors produced by the disaster-
pregnant environment, are rare or extreme events that can adversely affect human life,
property or various activities in a natural or man-made environment and cause disaster
procedures. In general, the greater the intensity of the disaster-causing factor, the higher
the probability, the greater the risk, and the greater the possibility of serious disaster [109].
The disaster-causing factor is the main hazard source of urban rainstorm flood dis-
aster and an important factor inducing rainstorm flood disaster. Therefore, hazard analy-
sis of disaster-causing factors is usually the first step in disaster risk analysis, analyzing
the time (probability), intensity, scale and spatial location of disasters. The elements
needed for flood disaster risk analysis include rainfall, water depth, water flow rate, water
range, etc., to analyze and quantify the risk characteristics of disaster-causing factors such
as disaster intensity, impact range and duration of different frequencies [38].
The risk analysis of flood disaster-causing factors is mainly divided into two types:
(1) Simulation analysis of flow scale based on probability distribution theory. For exam-
ple, the linear matrix method of extreme value distribution and P-III probability distribu-
tion is used to divide the flood frequency in different magnitude regions, the flood fre-
quency is predicted by the binary Gumbel-Logistic model, and the annual flood peak flow
and annual flood volume are predicted by the Copula model. (2) Submergence simulation
analysis based on watershed runoff and runoff. In the geographic information system,
based on the digital elevation model (DEM), the submerged depth and range are simu-
lated by inputting rainfall, historical flow data, watershed characteristics, surface cover-
age characteristics and other data. There are SCS, TOPMDEL, SWMM and other common
models (Table 4).
Table 4. Flood disaster risk assessment index system.
Influencing Factor
Index
hazard factor
Submerged water depth, submerged duration, submerged range, flow velocity, precipi-
tation, flood frequency...
disaster environment
Ground height, slope, river system, topographic relief, land use type...
vulnerability
Population density, economic density, land use, labor index, urban lifeline, regional
GDP, population age composition ratio...
Sustainability 2023, 15, 765 14 of 24
preventing disasters and re-
ducing damages
Monitoring and early warning capacity, flood control and drainage capacity, disaster
relief capacity, disaster prevention publicity and education level, emergency shelter lay-
out distribution...
4.2.2. Environmental Sensitivity of Disaster
The sensitivity of hazard inducing environment refers to the sensitivity of the natural
environment in the disaster area to the disaster. The sensitive elements include river sys-
tem, topography, etc. Many scholars have studied the sensitivity of hazard inducing en-
vironment [1]. Under the same damage intensity, the more sensitive the study area, the
greater the risk of disaster [110]. Mo et al. used GIS technology to assess the sensitivity of
flood hazard-formative environments in Guangxi Province [111]. Su et al. analyzed the
relationship between the rapid growth rate of urban space and the frequency of flood dis-
asters in Jiangning Development Zone from the perspective of the characteristics of flood-
pregnant environment [112].
4.2.3. Carrier Vulnerability
Vulnerability of hazard-affected bodies refers to the degree of vulnerability of haz-
ard-affected bodies to disasters, including various aspects of human and social develop-
ment, such as transportation, education and culture, various disaster reduction engineer-
ing facilities, and various wealth accumulated by people [113]. Sanyal et al. used a combi-
nation of GIS and remote sensing to assess the vulnerability of residential areas in West
Bengal in the eastern Ganges Plain, India [114]. Ge et al. used analytic hierarchy process
and fuzzy evaluation theory to evaluate the vulnerability of flood disaster bearing bodies
in Nanjing [115].
4.2.4. Disaster Prevention and Mitigation Capacity
Disaster prevention and mitigation is the response measures taken by people to re-
duce the impact of flood damage, including emergency management capacity, emergency
team building, emergency supplies reserves, disaster reduction funding and resource
preparation [116]. A region’s disaster prevention and mitigation capacity is relatively
high, indicating that the region has invested more in disaster prevention and mitigation
management, and the economic losses and casualties caused by disasters will be lower.
In the study of flood disaster risk assessment, there are also many studies on disaster
prevention and mitigation capabilities. For example, Cao et al. used the analytic hierarchy
process and fuzzy comprehensive evaluation method to evaluate the disaster prevention
and mitigation capabilities of Ningbo City [117]. Wang applied the multi-level fuzzy com-
prehensive evaluation model to quantitatively evaluate the disaster reduction ability of
typhoon-flood-geological disaster chain [118]. According to the results of flood assess-
ment and the actual situation, Chen et al. put forward suggestions for disaster prevention
and mitigation [119]. Lu used AHP method to stratify the risk of flood disaster, and took
the ability of disaster prevention and mitigation as the first-level index to evaluate the
risk, so as to provide reference for the risk control and disaster relief of flood disaster in
Changzhou [120].
4.3. Flooding Forecast
Flood models and forecasts are key to managing and responding to extreme floods.
Comprehensive flood forecasting can provide the possibility of upcoming extreme events
and related risk loss instructions, so that people can carry out scientific disaster prevention
and mitigation. Longer forecast time can win the important opportunity of disaster relief
and prevention. There are many techniques for flood forecasting, among which the main
models are numerical weather prediction (NWP) model and ensemble forecasting model.
In flood forecasting, most models use precipitation data as an important input parameter.
Sustainability 2023, 15, 765 15 of 24
The following mainly classifies and summarizes the precipitation data used in flood fore-
casting technology and expounds the research status of two main forecasting models.
4.3.1. Common Precipitation Data for Flood Forecast
Heavy rainfall is an important cause of urban flood disasters, and its accuracy seri-
ously affects the accuracy of urban flood simulation and rainstorm flood forecasting re-
sults. According to the different data sources and processing methods used in the current
rainstorm flood forecasting research, this paper summarizes the commonly used precipi-
tation input data into five types. Table 5 is a summary of five precipitation data.
(1) Spatial grid precipitation data simulated by regional NWP model. The numerical
forecast simulation technology has been rapidly developed. The new generation of
regional convective-scale weather forecast model has significantly improved the abil-
ity to simulate heavy rainfall, and can predict regional rainstorms several days or
even a week in advance. However, the uncertainty of the model itself makes the pre-
cipitation intensity and precipitation area deviate from the actual observation
[121,122]. Therefore, when using this data, it is necessary to evaluate and optimize
the regional applicability of each parameter configuration of the model in advance,
so as to reduce the impact on the numerical precipitation forecast results [123].
(2) Spatial gridded precipitation products from global mesoscale circulation model. This
kind of numerical precipitation forecast product is a kind of data that is relatively
easy to obtain. At present, there are more than 20 meteorological operation centers in
the world that can provide precipitation forecast products with a resolution of 10 km
and 7 h. Using this type of data, it performs well in describing frontal precipitation
processes, but performs poorly in capturing convective precipitation characteristics
[124,125]. In practical applications, statistical downscaling or multi-model integrated
forecasting methods are usually used to correct the precipitation forecast results
[126,127] to predict the probability of rainstorm and flood in the future.
(3) Precipitation data obtained by radar or remote sensing interpretation. This kind of
data has high spatial and temporal resolution, and can predict surface precipitation
several hours in advance by interpreting high-altitude precipitation information.
However, the forecast results are easily affected by local climate and deviate from the
actual observation. In practical applications, mathematical statistical methods or
NWP models are often used to correct precipitation data to reduce the initial error of
flood simulation and forecast [128–130].
(4) Historical long sequence site observation data. Long time series and high accuracy
are the main advantages of this kind of data, but its spatial representativeness is low,
so it is unable to obtain enough regional spatial precipitation information. In practical
applications, such data are mainly used as the basis for the design of return period
precipitation, which is used to simulate typical rainstorm events or predict rainstorm
floods under different rainstorm scenarios, and provide scenario forecast sets for
rainstorm flood disaster prediction.
(5) Real-time station observation of precipitation data. With the increase of the density
of the national precipitation observation network, the current available station obser-
vation data has been significantly improved in terms of spatial representation com-
pared with the historical long series observation data. Stormwater model driven by
real-time observed precipitation data can provide forecasting information of near or
real-time rainstorm and flood disasters for urban areas with high accuracy. In prac-
tice, this data is usually not directly used for the input parameters of precipitation
forecast model, but used to correct other sources of precipitation interpretation or
forecast products.
Sustainability 2023, 15, 765 16 of 24
Table 5. Rainstorm and flood common precipitation data.
Type
Advantage
Disadvantage
Ways to Reduce Disadvantages
Spatial grid precipita-
tion data simulated by
regional NWP model
Significant increase in simu-
lated heavy precipitation ca-
pacity
The uncertainty of the model it-
self makes the precipitation in-
tensity and precipitation area
deviate from the actual observa-
tion
Evaluation and optimization of
regional applicability of model
parameters in advance
Spatial gridded pre-
cipitation products
output by global
mesoscale circulation
model
Easy to obtain and performs
well in describing frontal
precipitation processes
Poor performance in capturing
convective precipitation charac-
teristics
Correction of precipitation fore-
cast results by statistical
downscaling or multi-model en-
semble forecast
Precipitation data ob-
tained by radar or re-
mote sensing interpre-
tation
High spatial resolution to
predict surface precipitation
hours in advance
The forecasted precipitation re-
sults are easily affected by local
climate and deviate from the ac-
tual situation
Precipitation data are corrected
using mathematical statistics or
NWP models to reduce initial
errors in flood simulation and
forecasting
Historical long se-
quence site observa-
tion data
Long time series and high ac-
curacy
Low spatial representativeness,
insufficient regional spatial pre-
cipitation information available
In practical applications, such
data are mainly used as the ba-
sis for design of return period
precipitation
Real-time site observa-
tion precipitation data
It can provide near or real-
time rainstorm and flood dis-
aster forecast information for
urban areas with high accu-
racy.
The length of the forecast period
is limited by the catchment time
This data is often used to cor-
rect precipitation interpretation
or forecast products from other
sources
4.3.2. Research Progress of Numerical Weather Prediction Technology
Accurate NWP model is an important part of flood forecasting system, which can
provide reliable rainfall forecast. NWP is a method to predict the atmospheric motion state
and weather phenomena in a certain period of time in the future by numerically solving
the thermodynamic and hydrodynamic equations describing the weather process under
given initial conditions according to the actual situation of the atmosphere.
The history of NWP can be traced back to the early 20th century. However, due to
the limitations of computing power and observation technology at that time, it was not
until the emergence of computers in 1949 that it was possible to solve atmospheric physi-
cal equations by computers. In the same year, the first real NWP model appeared [125]. In
the second half of the 20th century, there have been a number of mesoscale NWP models
that can be used for global or regional operational forecasting, such as MM5 model,
UKMO model, ETA model, RAMS model and so on. In the 1990s, the rapid development
of high-altitude detection technology and computer technology provided driving data for
the NWP model to perform higher-resolution simulations, and also provided the possi-
bility for analyzing and processing these data to improve the development of more com-
plex physical process parameterization schemes. On this basis, a new generation NWP
model (such as weather research and forecasting model, WRF model [131]) is developed
to simulate the regional convective precipitation process more accurately, and it shows
good performance in simulating and forecasting heavy precipitation [125]. The rapid im-
provement of the new generation NWP model in the simulation and prediction of precip-
itation, especially heavy precipitation, has led more and more institutions and scholars to
apply the NWP model to the simulation and prediction of regional rainstorm and flood
disasters [132,133].
Sustainability 2023, 15, 765 17 of 24
Although the ability of the new generation NWP model to simulate regional-scale
heavy precipitation has been gradually recognized, precipitation is still one of the most
difficult variables for NWP model simulation and prediction due to the influence of model
structure error, initial error and chaotic properties of weather system. Rama Rao et al.
[134] used WRF model to simulate a heavy precipitation process in Guyana, South Amer-
ica. The results showed that WRF model could capture the main characteristics of this
precipitation process, but the simulated value of precipitation was smaller than the meas-
ured value. Efstathiou et al. [135] simulated a heavy rainfall process in the Chalkidiki area
of Greece based on the WRF model. The results showed that the WRF model could well
simulate the process and range of the heavy rainfall, but the output precipitation value
was too large, and the range and center position of the rainstorm were somewhat devi-
ated. Since the numerical weather prediction is actually an approximate simulation of the
atmospheric motion process, the lack of human understanding of the atmospheric motion
mechanism and the model’s description of the sub-grid process, coupled with the ampli-
fication effect of the nonlinear chaotic properties of the atmosphere itself on the model
structure error and initial error, there are many uncertainties in the precipitation results
output by the NWP model [123,125].
4.3.3. Ensemble Forecasts Research Progress
Due to the growth of combined numerical weather and climate prediction and the
expansion of high performance computing, flood ensemble prediction has gained signifi-
cant momentum in the past decade. Ensemble prediction can not only provide determin-
istic ensemble average prediction results, but also estimate the prediction error of the re-
sults. At the same time, it can also give the possibility of events we are concerned about
through probability prediction.
Studies have shown that ensemble forecast can greatly extend the validity of flood
forecast [136]. Even if the numerical model is extended to the forecast time of 5–6 d, the
precipitation forecast is only 2–3 d, and the forecast time of flood cannot be extended [137].
Burger obtained the precipitation and temperature forecast information from the ECM-
WF ensemble forecast system for a 10-year 12-h period and a forecast period of 1–5 days
in a small watershed area [138]. The hydrological model was used to obtain the probability
forecast runoff value, and it was compared with the measured flood process, which
proved the application advantages of ensemble forecast in flood warning. Bao constructed
a flood forecasting model based on the combination of hydrology and hydraulics driven
by TIGGE [139]. The flood forecasting model was driven by the forecast precipitation of
each member of the ensemble forecasting system, and the flood forecasting was obtained
with the same number of ensemble forecast members. The flood forecasting and early
warning of complex water systems were realized by probabilistic forecasting. The early
warning and forecasting of flood in Huaihe River from 2007 to 2008 showed that the flood
forecast period was extended by 3~5 days. Zhang et al. [140] and Xu used the ensemble
forecast system of NMC, ECMWF and NCEP in the Linyi Basin of the lower reaches of the
Huaihe River, and used the GRAPES mesoscale model downscaling to obtain a 6 h pre-
cipitation-driven hydrological model for early warning of floods [141]. It is feasible and
effective. Another study used TIGGE ensemble forecast as a driver to improve the soil
moisture simulation results [142]. In addition, He’s case study and uncertainty analysis of
multi-center and multi-model TIGGE ensemble forecast in flood forecasting in the Huaihe
River Basin of China show that the probability attribute of ensemble forecast is still re-
tained [143].
The development of ensemble forecast still faces many severe challenges. Countries
should make a lot of preparations in the application of key technologies of ensemble fore-
cast, so as to promote the early application of ensemble numerical forecast products in
actual flood forecasting.
Sustainability 2023, 15, 765 18 of 24
5. Summary and Discussion
Under the dual influence of urban expansion and climate change, the global urban
flood disaster problem is becoming more and more serious. Flood risk assessment can
provide a scientific basis for current and future urban flood management, especially for
flood disaster relief and prevention decisions. Because of the complex process of flood
disaster occurrence and development, there are still many problems to be solved. Based
on a detailed review of the current status of urban flood disaster research, we identified
the challenges in current research and outlined potential future research opportunities.
(1) Urban flood risk assessment is an important research hotspot. There are a lot of re-
searchers evaluating the loss of urban areas after floods. However, most studies only
conducted a large-scale assessment, and did not conduct a more detailed risk assess-
ment for small-scale (similar to community). In the context of global big data sharing,
it becomes easier to obtain finer data. To make full use of the advantages of big data
in the current context, to establish a more sophisticated small-scale assessment
model.
(2) Urban flood management can effectively regulate people’s development and flood
control behavior, reduce the impact of flood disaster, and urban flood risk assessment
can provide more accurate decision for flood management. However, whether the
current urban flood risk assessment research can provide sufficient, accurate and re-
liable assessment results to the flood management department, and whether the re-
sults can be applied by the urban flood management department is a question worth
considering. Blindly carrying out a large number of flood risk assessment studies can
not be applied to the actual urban management, which will only cause waste of per-
sonnel and resources.
(3) Under the background of global climate change and urbanization, multi-scenario
flood risk assessment is a hot and difficult topic in flood disaster management re-
search. By comparing and analyzing the five current mainstream flood assessment
methods, it is found that each method has certain limitations, which directly affects
the accuracy, exposure and vulnerability assessment of flood risk assessment, and
will affect the effectiveness of the entire flood risk assessment results. In the construc-
tion of flood risk assessment model, we should make full use of RS and GIS technol-
ogy to establish a more refined and dynamic flood simulation model suitable for ur-
ban areas, and form an urban flood simulation system with good human-computer
interaction function and multi-functions such as early warning, forecasting and de-
cision support [94].
(4) In the process of flood risk assessment, sensitivity and vulnerability analysis has al-
ways occupied an important position. However, in the sensitivity analysis, the spatial
and temporal resolution of data has been restricting the accuracy and timeliness of
urban flood risk assessment; in terms of vulnerability analysis, vulnerability assess-
ment and quantitative flood risk assessment have not been supported by sufficient
data. Nowadays, international flood risk management has begun to pay attention to
the impact of multi-dimensional characteristics of social economy, environment, cul-
ture and policy on urban comprehensive flood vulnerability. The development of ar-
tificial intelligence and big data provides data and technical support for flood risk
assessment, which is conducive to the study of big data refinement and multi-dimen-
sional flood risk assessment under flood disasters.
(5) The smooth progress of disaster risk assessment can provide a scientific basis for dis-
aster prevention and mitigation, risk prediction, disaster transfer and other decision-
making, and provide technical support for disaster management departments. The
occurrence of flood disasters is often not an independent event, it is bound to cause
other secondary disasters. Therefore, it is the trend of disaster risk assessment to im-
prove the accuracy of risk assessment and enhance the reliability of the results by
Sustainability 2023, 15, 765 19 of 24
multidisciplinary joint, from single disaster risk assessment to integrated flood risk
assessment.
6. Conclusions
Under the background of increasingly serious urban flood problems worldwide, this
paper combs the disaster mechanism and evaluation index framework of urban flood risk
assessment, summarizes five flood assessment methods, compares and analyzes the ad-
vantages and disadvantages of different research methods, and finally discusses the de-
velopment trend of urban flood risk assessment under urban flood management. The
main conclusions are as follows:
(1) In terms of the mechanism of urban flood disaster, the problem of urban flood is
becoming more and more serious. In addition to natural factors, it is mainly caused
by the unscientific development behavior of the city. Therefore, in the process of ur-
ban development, we should pay attention to the protection of ecological environ-
ment, that is, we should consider the social and ecological benefits while developing
the economy. The urban planning management department should scientifically
plan and improve the urban drainage pipe network system, and protect the urban
wetland and other green space resources with water storage function.
(2) In the future research of urban flood risk assessment methods, we should make full
use of the effective resources in the era of big data, and make full use of emerging
methods such as data mining and machine learning to improve the efficiency and
accuracy of flood risk assessment. At the same time, it is necessary to strengthen the
connection between different disciplines and different flood assessment methods,
and build a more adaptable flood assessment model to provide a more scientific de-
cision-making basis for disaster management departments.
(3) In terms of the reliability of flood risk assessment results and the division of risk ar-
eas, it is necessary to give full play to the role of cross-fusion of different methods
and models. Different methods and models were used for risk assessment in the same
research area to verify the reliability of the methods and models. At the same time, it
is necessary to further improve the theoretical system of flood disaster risk zoning,
which can provide a scientific basis for urban resource allocation and emergency
plans of relevant departments.
Author Contributions: Conceptualization, C.L.; Methodology, N.S.; Software, C.L.; Validation, Y.L.
and C.L.; Investigation, C.L. and B.G.; Resources, C.L.; Data curation, Y.Y. and X.S.; Writing—orig-
inal draft, N.S.; Writing—review & editing, Y.L. and Y.W.; Visualization, Y.L. and B.G.; Supervision,
N.S.; Project administration, C.L.; Funding acquisition, C.L. and B.G. All authors have read and
agreed to the published version of the manuscript.
Funding: This research was supported by Shandong Provincial Natural Science Foundation(No.
ZR2022MD039), the Key Laboratory Open Foundation for Geo-Environmental Monitoring of Great
Bay Area (Shenzhen University)，Ministry of Natural Resources of the People's Republic of China
(SZU51029202003).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments: This work was supported by Shandong Provincial Natural Science Foundation
(No. ZR2022MD039), in part by the Open Fund of Key Laboratory of Geospatial Technology for the
Middle and Lower Yellow River Regions (Henan University), Ministry of Education under Grant
GTYR202004, in part by the Key Laboratory Open Foundation for Geo-Environmental Monitoring
of Great Bay Area (Shen-zhen University) through the Ministry of Natural Resources of the People’s
Republic of China under Grant SZU51029202003.
Conflicts of Interest: The authors declare no conflict of interest.
Sustainability 2023, 15, 765 20 of 24
References
1. Zhang, Y. Research on the Risk Zoning of Urban Rainstorm and Flood Disasters Considering Spatial Distribution Characteris-
tics. Master’s Thesis, Nanjing Normal University, 2021.
2. Brody, S.; Blessing, R.; Sebastian, A.; Bedient, P. Examining the impact of land use/land cover characteristics on flood losses. J.
Environ. Plan. Manag. 2014, 57, 1252–1265.
3. Hammond, M.J.; Chen, A.S.; Djordjević, S.; Butler, D.; Mark, O. Urban flood impact assessment: A state-of-the-art review. Urban
Water J. 2015, 12, 14–29.
4. Dewan, A.M.; Yamaguchi, Y.; Rahman, Z. Dynamics of land use/cover changes and the analysis of landscape fragmentation in
Dhaka Metropolitan, Bangladesh. GeoJournal 2012, 77, 315–330.
5. Zhou, Y.; Peng, Tao.; Shi, R. Research progress on risk assessment of rainstorm flood disaster in China. Rainstorm Disaster 2019,
38, 494–501.
6. Kundzewicz, Z.W.; Kanae, S.; Seneviratne, S.I.; Handmer, J.; Nicholls, N.; Peduzzi, P. Flood risk and climate change: Global and
regional perspectives. Hydrol. Sci. J. 2014, 59, 1–28.
7. Whitfield, P.H.; Burn, D.H.; Hannaford, J.; Higgins, H.; Hodgkins, G. A.; Marsh, T.; Looser, U. Reference hydrologic networks
I. The status and potential future directions of national reference hydrologic networks for detecting trends. Hydrol. Sci. J. 2012,
57, 1562–1579.
8. Chief Scientist. Understanding floods: Questions and answers. Aust. J. Emerg. Manag. Off. Qld. Chief Sci. 2011, 26.
9. Foresight. Future Flooding: Executive Summary; Office of Science and Technology: London, UK, 2004.
10. Jianyun, Z.; Yintang, W.; Ruimin, H.E.; Qingfang, H.; Xiaomeng, S. Discussion on the urban flood and waterlogging and causes
analysis in China. Adv. Water Sci. 2016, 27, 485–491.
11. Jianyun, Z.; Xiaomeng, S.; Guoqing, W.; Ruimin, H.; Xiaojun, W. Development and challenges of urban hydrology in a changing
environment: I: Hydrological response to urbanization. Adv. Water Sci. 2014, 25, 594–605.
12. Wang, Y.; Wang, D.; Huang, G.; Ma, J. Urban flood disaster assessment and risk management. J. Water Conserv. Water Transp.
Eng. 2019, 6, 139–142.
13. Li, J.L.; Cao, L.D.; Pu, R.L. Progresses on monitoring and assessment of flood disaster in remote sensing. J. Hydraul. Eng 2014,
45, 253–260.
14. Xu, H.; Ma, C.; Lian, J.; Xu, K.; Chaima, E. Urban flooding risk assessment based on an integrated k-means cluster algorithm
and improved entropy weight method in the region of Haikou, China. J. Hydrol. 2018, 563, 975–986.
15. Li, C.; Tian, J.; Shen, R. Advances in flood risk assessment. J. Catastrophol. 2020, 35, 131–136.
16. Huang, G.; Luo, H.; Lu, X.; Yang, C.; Wang, Z.; Huang, T.; Ma, J. Urban flood risk analysis and zoning methods. Water Resour.
Prot. 2020, 36, 1–17.
17. Yin, Z.; Xu, S.; Yin, J.; Wang, J. Scenario simulation and risk assessment of urban rainstorm waterlogging disaster based on
small scale. Acta Geogr. Sin. 2010, 65, 553–562.
18. Leal Filho, W.; Balogun, A.L.; Olayide, O.E.; Azeiteiro, U. M.; Ayal, D. Y.; Muñoz, P. D. C.; Li, C. Assessing the impacts of climate
change in cities and their adaptive capacity: Towards transformative approaches to climate change adaptation and poverty
reduction in urban areas in a set of developing countries. Sci. Total Environ. 2019, 692, 1175–1190.
19. Church, J.; Clark, P.; Cazenave, A. The physical science basis. Contribution of working group I to the fifth assessment report of
the intergovernmental panel on climate change, 1535, 2013.
20. Solomon, S.; Qin, D.; Manning, M.; Averyt, K.; Marquis, M. Climate Change 2007—The Physical Science Basis: Working Group I
Contribution to the Fourth Assessment Report of the IPCC; Cambridge University Press: Cambridge, UK, 2007.
21. Kingston, D.G.; Taylor R G. Sources of uncertainty in climate change impacts on river discharge and groundwater in a head-
water catchment of the Upper Nile Basin, Uganda. Hydrol. Earth Syst. Sci. 2010, 14, 1297–1308.
22. Liang, S.; He, S.; Wang, X.; Li, M. Disaster-causing mechanism analysis and research prospect of urban waterlogging disaster.
J. Inst. Disaster Prev. 2020, 22, 77–83.
23. Trenberth, K.E. Changes in precipitation with climate change. Clim. Res. 2011, 47, 123–138.
24. Konapala, G.; Mishra, A.K.; Wada, Y.; Mann, M. E. Climate change will affect global water availability through compounding
changes in seasonal precipitation and evaporation. Nat. Commun. 2020, 11(1), 1-10.
25. Schuster, Z.T.; Potter, K.W.; Liebl, D.S. Assessing the effects of climate change on precipitation and flood damage in Wisconsin.
J. Hydrol. Eng. 2012, 17, 888–894.
26. Zhang, Q.; Singh, V.P.; Li, J.; Jiang, F.; Bai, Y. Spatio-temporal variations of precipitation extremes in Xinjiang, China. J. Hydrol.
2012, 434, 7–18.
27. Jiang, F.; Hu, R.J.; Wang, S.P.; Zhang, Y. W.; Tong, L. Trends of precipitation extremes during 1960–2008 in Xinjiang, the North-
west China. Theor. Appl. Climatol. 2013, 111, 133–148.
28. Heilig, G.K. World Urbanization Prospects: The 2011 Revision; United Nations, Department of Economic and Social Affairs (DESA),
Population Division, Population Estimates and Projections Section: New York, NY, USA, 2012; Volume 14, p. 555.
29. He, S.; Wang, X. Risk management of urban waterlogging disasters in a changing environment: Challenges and responses. J.
Henan Polytech. Univ. Soc. Sci. 2019, 20, 48–53.
30. Hou, J.; Han, H.; Qi, W.; Guo, K.; Li, Z.; Hinkelmann, R. Experimental investigation for impacts of rain storms and terrain slopes
on low impact development effect in an idealized urban catchment. J. Hydrol. 2019, 579, 124176.
Sustainability 2023, 15, 765 21 of 24
31. Damodaram, C.; Giacomoni, M.H.; Prakash Khedun, C.; Holmes, H.; Ryan, A.; Saour, W.; Zechman, E. M. Simulation of com-
bined best management practices and low impact development for sustainable stormwater management 1. JAWRA J. Am. Water
Resour. Assoc. 2010, 46, 907–918.
32. Jacobson, C.R. Identification and quantification of the hydrological impacts of imperviousness in urban catchments: A review.
J. Environ. Manag. 2011, 92, 1438–1448.
33. Qin, H.; Li, Z.; Fu, G. The effects of low impact development on urban flooding under different rainfall characteristics. J. Environ.
Manag. 2013, 129, 577–585.
34. Ni, L.; Zeng, J. Urban rainstorm waterlogging disaster mechanism and urban environment disaster evolution. Archit. Cult. 2015,
6, 116–118.
35. Xia, J.; Dong, B.; Li, Q.; Liu, L.; Zhou, M.; Wang, X. Hydrodynamic mechanism analysis and disaster reduction countermeasures
of urban flood disaster in recent years. China Flood Drought Manag. 2022, 32, 66–71.
36. Huang, Y.; She, L.; Lai, W. Urban underground space flood emergency response: Problems, causes and countermeasures—
Based on the disaster investigation report of the State Council of Zhengzhou ‘7·20’ heavy rainstorm. China Emerg. Rescue 2022,
5, 4–9.
37. Haiping, L. Cause analysis and countermeasures of urban flood disaster. Heilongjiang Hydraul. Sci. Technol. 2022, 50, 70–72.
38. Van Steenbergen, N.; Ronsyn, J.; Willems, P. A non-parametric data-based approach for probabilistic flood forecasting in sup-
port of uncertainty communication. Environ. Model. Softw. 2012, 33, 92–105.
39. Yang, P.; Jin, J.; Zhao, D.; Li, J. Quantitative Study of Urban Vulnerability Based on Historical Rainstorm and Flood Disaster
Data—A Case Study of Beijing. Sci. Geogr. Sin. 2016, 36, 733–741.
40. Wang, Z.; Tong, Y.; Jing, X. Quantitative analysis and calculation of flood risk in the lower Yellow River. Yellow River, 2013, 35,
14–16.
41. Qin, N.; Jiang, T. Flood risk zoning and assessment in the middle and lower reaches of the Yangtze River based on GIS. J. Nat.
Disasters 2005, 14, 5–11.
42. Zhao, S.; Zhang, Q. Spatio-temporal risk assessment of crop flood in three provinces of Northeast China. J. Catastrophol. 2013,
28, 54–60.
43. De Moel, H.; Aerts, J.C.J.H.; Koomen, E. Development of flood exposure in the Netherlands during the 20th and 21st century.
Glob. Environ. Change 2011, 21, 620–627.
44. Bell, V.A.; Kay, A.L.; Jones, R.G.; Moore, R. J.; Reynard, N. S. Use of soil data in a grid-based hydrological model to estimate
spatial variation in changing flood risk across the UK. J. Hydrol. 2009, 377, 335–350.
45. Liu, X.; Shi, P. Theory and Practice of Regional Flood Risk Assessment Model. J. Nat. Disasters 2001, 2, 66–72.
46. Li, J.; Ding, Z.; Huang, S.; Hu, Y. Flood disaster loss assessment model based on spatial distributed socio-economic database. J.
China Inst. Water Resour. Hydropower Res. 2003, 02, 27–33.
47. Cœur, D.; Lang, M. Use of documentary sources on past flood events for flood risk management and land planning. Comptes
Rendus Geosci. 2008, 340, 644–650.
48. Ruisong, Q. Research on Risk Assessment of Rainstorm Waterlogging Disaster in Typical Coastal Cities; East China Normal University,
2012.
49. Lyu, H.M.; Sun, W.J.; Shen, S.L.; Arulrajah, A. Flood risk assessment in metro systems of mega-cities using a GIS-based model-
ing approach. Sci. Total Environ. 2018, 626, 1012–1025.
50. Burton, I.; Kates, R.W.; White, G.F. The Environment as Hazard, 2nd ed.; Guilford: New York, 1993.
51. Feng, L.; Wu, Z. Fuzzy Comprehensive Evaluation of Regional Vulnerability. Geogr. Geo-Inf. Sci. 2001, 17, 63–66.
52. Bohle, H.G.; Downing, T.E.; Watts, M.J. Climate change and social vulnerability: Toward a sociology and geography of food
insecurity. Glob. Environ. Change 1994, 4, 37–48.
53. Jiang, X.; Fan, J.; Zhang, J.; Tong, Z.; Liu, X. Risk assessment of rainstorm and flood disasters in Songhua River based on GIS. J.
Catastrophol. 2009, 24, 51–56.
54. Hu, H.; Zhang, Y. Preliminary Study on Development and Application of Rapid Pre-Assessment Model for Loss Risk of Rainstorm Disaster
Personne. In proceedings of Innovation Driven Development to Improve Meteorological Disaster Prevention Capability–S16 the
Second Urban Meteorological Forum–Disaster, Environment, Impact and Response; Nanjing, China, 22 October 2013; pp. 66–
82 (In Chinese).
55. Cabrera, J.S.; Lee H S. Flood risk assessment for Davao Oriental in the Philippines using geographic information system‐based
multi‐criteria analysis and the maximum entropy model. J. Flood Risk Manag. 2020, 13, e12607.
56. Stefanidis, S.; Stathis, D. Assessment of flood hazard based on natural and anthropogenic factors using analytic hierarchy pro-
cess (AHP). Nat. Hazards 2013, 68, 569–585.
57. Jiang, W.; Deng, L.; Chen, L.; Wu, J.; Li, J. Risk assessment and validation of flood disaster based on fuzzy mathematics. Prog.
Nat. Sci. 2009, 19, 1419–1425.
58. Chen, Y.; Ding, J.; Zhao, Y. Flood Disaster Assessment of Sichuan Province Based on Projection Pursuit Principle. J. Hydraul.
Eng. 2010, 41, 220–225.
59. Jin, J.; Wei, Y.; Yang, X. Neural Network Model for Flood Disaster Assessment Based on Genetic Algorithm. J. Catastrophol. 1998,
4, 6–11.
60. Parthasarathy, B.; Sontakke, N.A.; Monot, A.A.; Kothawale, D. R. Droughts/floods in the summer monsoon season over differ-
ent meteorological subdivisions of India for the period 1871–1984. J. Climatol. 1987, 7, 57–70.
Sustainability 2023, 15, 765 22 of 24
61. Zou, Q.; Zhou, J.; Zhou, C.; Song, L.; Guo, J. Comprehensive flood risk assessment based on set pair analysis-variable fuzzy sets
model and fuzzy AHP. Stoch. Environ. Res. Risk Assess. 2013, 27, 525–546.
62. Dutta, D.; Herath, S.; Musiake, K. A mathematical model for flood loss estimation. J. Hydrol. 2003, 277, 24–49.
63. Chubey, M.S.; Hathout, S. Integration of RADARSAT and GIS modelling for estimating future Red River flood risk. GeoJournal
2004, 59, 237–246.
64. Chowdary, V.; Chandran, R.V.; Neeti, N.; Bothale, R. V., Srivastava, Y.; Ingle, P. Assessment of surface and sub-surface water-
logged areas in irrigation command areas of Bihar state using remote sensing and GIS. Agric. Water Manag. 2008, 95, 754–766.
65. Barredo, J.I. Major flood disasters in Europe: 1950–2005. Nat. Hazards 2007, 42, 125–148.
66. Haq, M.; Akhtar, M.; Muhammad, S.; Paras, S.; Rahmatullah, J. Techniques of remote sensing and GIS for flood monitoring and
damage assessment: A case study of Sindh province, Pakistan. Egypt. J. Remote Sens. Space Sci. 2012, 15, 135–141.
67. Islam, M.M.; Sado, K. Development priority map for flood countermeasures by remote sensing data with geographic infor-
mation system. J. Hydrol. Eng. 2002, 7, 346–355.
68. Cheng, X.; Dai, M.; Hao, D.; Wu, Q. Regional flood risk assessment based on scenario analysis—A case study of Chaohu basin.
Resour. Environ. Yangtze Basin 2015, 24, 1418–1424.
69. Liu, J.; Li, J.; Liu, J.; Cao, R. Flood disaster comprehensive risk assessment based on GIS/AHP integration—A case study of
Huaihe River Basin. J. Nat. Disasters 2008, 17, 110–114.
70. Liu, Y.; Wang, S.; Yan, S.; Zhou, Y.; Wei, C.; Xu, F. Study on Key Techniques of Flood Disaster Monitoring and Evaluation
System Based on Network Supported by Remote Sensing and GIS. Natl. Remote Sens. Bull. 2001, 1, 53–57
71. Dapeng, H. Research on Risk Assessment of Flood Disaster in the Upper Reaches of Hongze Lake; Graduate School of Chinese Academy
of Sciences, 2009.
72. Li, B.; Yan, Q.; Zhang, L.; Liang, Y. MODIS time series monitoring and analysis of flood disaster characteristics in the middle
reaches of the Yangtze River. Geomat. Inf. Sci. Wuhan Univ. 2013, 38, 789–793.
73. Ke, X. Flood disaster risk assessment based on GIS-Taking Quzhou City, Zhejiang Province as an example. J. Agric. Catastrophol.
2018, 8, 4–7.
74. Jiang, W.; Cheng, C.; Zhang, Y.; Zhao, H. Flood disaster risk assessment of Zhejiang Province based on GIS/AHP integration.
Bull. Surv. Mapp. 2019, 2, 125–130.
75. Lingtong, F. Risk Assessment of Flood Disaster in Zhengzhou Based on RS and GIS; Zhengzhou University,2020.
76. Wu, M. Research on Multi-Dimensional Coupling Simulation and Risk Assessment of Urban Waterlogging in Jinan; Shandong Univer-
sity, 2020.
77. Renn, O. White Paper on Risk Governance: Toward an Integrative Framework//Global Risk Governance; Springer: Dordrecht, 2008; pp.
3–73.
78. Randall, M.; Sun, F.; Zhang, Y.; Jensen, M. B. Evaluating Sponge City volume capture ratio at the catchment scale using SWMM.
J. Environ. Manag. 2019, 246, 745–757.
79. Djordjević, S.; Prodanović, D.; Maksimović, Č.; Maksimović, Č.; Ivetić, M.; Savić, D. SIPSON—Simulation of Interaction between
Pipe flow and Surface Over-land flow in Networks. Water Sci. Technol. 2005, 52, 275–283.
80. Kadam, P.; Sen, D. Flood inundation simulation in Ajoy River using MIKE-FLOOD. ISH J. Hydraul. Eng. 2012, 18, 129–141.
81. Zheng, J.; Zhang, C.; Li, Y.; You, Q.; Liu, J.; Wang, L.; Wu, Y. Research on prevention and control of waterlogging in subway
stations based on InfoWorks ICM model. J. Beijing Norm. Univ. Nat. Sci. 2019, 55, 648–655.
82. Wu, X.; Wang, Z.; Guo, S.; Liao, W.; Zeng, Z.; Chen, X. Scenario-based projections of future urban inundation within a coupled
hydrodynamic model framework: A case study in Dongguan City, China. J. Hydrol. 2017, 547, 428–442.
83. Zeng, Z.; Lai, C.; Wang, Z.; Wu, X.; Huang, G.; Hu, Q. Rapid simulation of urban rainstorm and flood based on WCA2D and
SWMM models. Adv. Water Sci. 2020, 31, 29–38.
84. Suarez, P.; Anderson, W.; Mahal, V.; Lakshmanan, T. R. Impacts of flooding and climate change on urban transportation: A
systemwide per-formance assessment of the Boston Metro Area. Transp. Res. Part D Transp. Environ. 2005, 10, 231–244.
85. Su, B.; Huang, H.; Zhang, N. Dynamic risk assessment method for urban waterlogging based on scenario simulation. J. Tsinghua
Univ. Sci. Technol. 2015, 55, 684–690.
86. Bisht, D.S.; Chatterjee, C.; Kalakoti, S.; Upadhyay, P.; Sahoo, M.; Panda, A. Modeling urban floods and drainage using SWMM
and MIKE URBAN: A case study. Nat. Hazards 2016, 84, 749–776.
87. Zeng, Z.; Wang, Z.; Wu, X.; Lai, C.; Chen, X. Simulation of rainstorm waterlogging based on SWMM and LISFLOOD model. J.
Hydroelectr. Eng. 2017, 36, 68–77.
88. Lijie, L. Vulnerability Assessment of Rainstorm and Flood Disaster in Pudong, Shanghai Based on Scenario; Shanghai Normal Univer-
sity,2009.
89. Huang, G.; Luo, H.; Chen, W.; Pan, J. Scenario simulation and risk assessment of urban flood disaster in Donghaochong Basin
of Guangzhou. Adv. Water Sci. 2019, 30, 643–652.
90. Zhang, Q.; Xu, Y.; Lei, C.; Wang, Y.; Han, L. Flood risk assessment based on dynamic simulation in the lower reaches of south-
east coastal reservoirs. J. Lake Sci. 2016, 28, 868–874.
91. Tong, J.; Zhang, H.; Liu, H.; Huang, J.; Hao, Y. XGBoost-based waterlogging risk assessment and influencing factors analysis of
core cities in Yangtze River Delta. Water Resour. Hydropower Eng. 2021, 52, 1–11.
92. Mojaddadi, H.; Pradhan, B.; Nampak, H.; Ahmad, N.; Ghazali, A. H. B. Ensemble machine-learning-based geospatial approach
for flood risk assessment using multi-sensor remote-sensing data and GIS. Geomatics. Nat. Hazards Risk 2017, 8, 1080–1102.
Sustainability 2023, 15, 765 23 of 24
93. Nguyen, T.S.; Nguyen, T.T.T.; Thanh, B.X.; Da Chau, T. Remote sensing and GIS for urbanization and flood risk assessment in
Phnom Penh, Cambodia. Geocarto Int. 2022, 37, 6625–6642.
94. Liu, Y.; Zhang, S.; Liu, L.; Wang, X.; Huang, H. Review of urban flood simulation research from the perspective of smart city.
Prog. Geogr. 2015, 34, 494–504.
95. Wang, Z.; Lai, C.; Chen, X.; Yang, B.; Zhao, S.; Bai, X. Flood hazard risk assessment model based on random forest. J. Hydrol.
2015, 527, 1130–1141.
96. Xiaolong, S.; Yufeng, R.; Duan, Z.; Liu, X.; Hua, X.; Lei, H. Flood risk assessment in Ya’an, Sichuan, China based on the emergy
theory. J. Water Clim. Change 2022, 13, 247–259.
97. Meihong, M.; Gang, Z.; He, B.; Li, Q.; Dong, H.; Wang, S.; Wang, Z. XGBoost-based method for flash flood risk assessment. J.
Hydrol. 2021, 598, 126382.
98. Thai, P.B.; Chinh, L.; Van Dao, D., Van Phong, T.; Nguyen, H. D.; Van Le, H.; Prakash, I. Flood risk assessment using deep
learning integrated with multi-criteria decision analysis. Knowl.-Based Syst. 2021, 219, 106899.
99. Khosravi, K.; Shahabi, H.; Pham, B.T.; Adamowski, J.; Shirzadi, A.; Pradhan, B.; Prakash, I. A comparative assessment of flood
susceptibility modeling using multi-criteria deci-sion-making analysis and machine learning methods. J. Hydrol. 2019, 573, 311–
323.
100. Li, H.; Wu, J.; Wang, Q.; Yang, C.; Pan, S. Research on prediction model of rainstorm waterlogging disaster in Shanghai based
on machine learning method. J. Nat. Disasters 2021, 30, 191–200.
101. Xu, Z.; Chen, H.; Ren, M.; Cheng, T. Research progress of urban flood disaster mechanism and risk assessment in China. Adv.
Water Sci. 2020, 31, 713–724.
102. Zhou, C.; Wan, Q.; Huang, S.; Chen, D. Flood risk zoning based on GIS. Acta Geogr. Sin. 2000, 55, 15–24.
103. Meyer, V.; Scheuer, S.; Haase, D. A multicriteria approach for flood risk mapping exemplified at the Mulde river, Germany.
Nat. Hazards 2009, 48, 17–39.
104. Kron, W. Keynote lecture: Flood risk = hazard × exposure × vulnerability. Flood Defence, 2002; pp. 82–97.
105. Li, C.; Cheng, X.; Li, N.; Du, X.; Yu, Q.; Kan, G. A framework for flood risk analysis and benefit assessment of flood control
measures in urban areas. Int. J. Environ. Res. Public Health 2016, 13, 787.
106. Zhang, H.; Zhang, J.; Han, J. The assessment and regionalization of flood/waterlogging disaster risk in middle and lower reaches
of Liao River of Northeast China. In Proceedings of the Fifth Annual IIASA-DPRI Forum on Integrated Disaster Risk Manage-
ment, 2005; pp. 103–118.
107. Du, K.; Liang, Z.; Wang, J.; Li, B. Research on risk zoning of urban flood disaster. Water Resour. Power 2018, 36, 61–64.
108. Zhou, S.; Zhai, G.; Shi, Y.; Lu, Y. Urban natural disaster risk assessment. J. Catastrophol. 2020, 35, 180–186.
109. Liu, Q.; Wang, H. Risk assessment and zoning of flood disaster in Tongzhou District of Beijing. Water Resour. Plan. Des. 2022,
10, 76–82.
110. Wan, X.; Yin, W.; Sun, P.; Tan, B. Zoning evaluation of rainstorm and flood disaster risk based on cloud model. J. Nat. Disasters
2017, 26, 77–83.
111. Mo, J.; Lu, J.; Li, Y.; Chen, Y. GIS-based environmental sensitivity assessment of flood disasters in Guangxi. J. Catastrophol. 2010,
25, 33–37.
112. Su, W.; Yang, G.; Chen, S. The impact of urban spatial expansion on regional flood disaster environment Resour. Sci. 2012, 34,
933–939.
113. Fan, G.; He, Y.; Gu, J. GIS-based rainstorm disaster and its risk assessment in Zhejiang Province. Chin. Agric. Sci. Bull. 2012, 28,
293–299.
114. Sanyal, J.; Lu, X.X. Remote sensing and GIS‐based flood vulnerability assessment of human settlements: A case study of Gan-
getic West Bengal, India. Hydrol. Process. Int. J. 2005, 19, 3699–3716.
115. Ge, P.; Yue, X. Spatio-temporal variability of vulnerability of flood hazard-affected bodies—A case study of Nanjing. J.
Catastrophol. 2013, 28, 107–111.
116. Cao, L.; Li, J.; Xu, L.; Yuan, Q.; Wang, M.; Huang, D. Preliminary Assessment of Flood Disaster Prevention and Mitigation
Capacity in Ningbo. J. Ningbo Univ. Nat. Sci. Eng. Ed. 2014, 27, 84–90.
117. Gao, W.; Shen, Q.; Li, M.; Hou, J. Time series monitoring and analysis of flood inundation range based on mul-ti-source remote
sensing data. Geomat. Spat. Inf. Technol. 2018, 41, 8–14.
118. Wang, T.; Wu, S.; Gao, J.; Wei, B. Regional typhoon-flood-geological disaster chain disaster reduction capacity assessment. J.
Catastrophology,2022, 12, 1–13.
119. Chen, C.; Sun, F. Evaluation method of rainstorm and flood disaster based on entropy weight-grey correlation analysis. J. Tsing-
hua Univ. Sci. Technol. 2022, 62, 1067–1073.
120. Lu, S.; Wu, J.; Zhou, S.; Wang, L.; Wang, Y.; Xing, Z. disaster risk assessment based on AHP-fuzzy mathematics method. Ind.
Saf. Environ. Prot. 2022, 48, 50–54.
121. Cuo, L.; Pagano, T.C.; Wang Q J. A review of quantitative precipitation forecasts and their use in short-to medium-range stream-
flow forecasting. J. Hydrometeorol. 2011, 12, 713–728.
122. Shih, D.S.; Chen, C.H.; Yeh, G.T. Improving our understanding of flood forecasting using earlier hydro-meteorological intelli-
gence. J. Hydrol. 2014, 512, 470–481.
Sustainability 2023, 15, 765 24 of 24
123. Chu, Q.; Xu, Z.; Chen, Y.; Han, D.Evaluation of the ability of the Weather Research and Forecasting model to reproduce a sub-
daily extreme rainfall event in Beijing, China using different domain configurations and spin-up times. Hydrol. Earth Syst. Sci.
2018, 22, 3391–3407.
124. Gao, Y.; Fu, J.S.; Drake, J.B.; Liu, Y.; Lamarque, J. F. Projected changes of extreme weather events in the eastern United States
based on a high resolution climate modeling system. Environ. Res. Lett. 2012, 7, 044025.
125. Bauer, P.; Thorpe, A.; Brunet, G. The quiet revolution of numerical weather prediction. Nature 2015, 525, 47–55.
126. Zhu, C.; Park, C.K.; Lee, W.S.; Yun, W. T. Statistical downscaling for multi-model ensemble prediction of summer monsoon
rainfall in the Asia-Pacific region using geopotential height field. Adv. Atmos. Sci. 2008, 25, 867–884.
127. Siddique, R.; Mejia, A.; Brown, J.; Reed, S.; Ahnert, P. Verification of precipitation forecasts from two numerical weather pre-
diction models in the Middle Atlantic Region of the USA: A precursory analysis to hydrologic forecasting. J. Hydrol. 2015, 529,
1390–1406.
128. Wardah, T.; Bakar, S.H.A.; Bardossy, A.; Maznorizan, M. Use of geostationary meteorological satellite images in convective rain
estimation for flash-flood forecasting. J. Hydrol. 2008, 356, 283–298.
129. Yucel, I.; Onen, A.; Yilmaz, K.K.; Gochis, D. J. Calibration and evaluation of a flood forecasting system: Utility of numerical
weather prediction model, data assimilation and satellite-based rainfall. J. Hydrol. 2015, 523, 49–66.
130. Gerard, A.; Martinaitis, S.M.; Gourley, J.J.; Howard, K. W.; Zhang, J. An Overview of the Performance and Operational Appli-
cations of the MRMS and FLASH Systems in Recent Significant Urban Flash Flood Events. Bull. Am. Meteorol. Soc. 2021, 102,
E2165–E2176.
131. Skamarock, W.C.; Klemp, J.B.; Dudhia, J.; Gill, D. O.; Barker, D. M.; Wang, W.; Powers, J. G. A Description of the Advanced Research
WRF Version 2; National Center For Atmospheric Research, Mesoscale and Microscale Meteorology Division: Boulder, CO, USA,
2005.
132. Hong, S.Y.; Lee, J.W. Assessment of the WRF model in reproducing a flash-flood heavy rainfall event over Korea. Atmos. Res.
2009, 93, 818–831.
133. Soares, P.M.M.; Cardoso, R.M.; Miranda, P.; de Medeiros, J.; Belo-Pereira, M.; Espirito-Santo, F. WRF high resolution dynamical
downscaling of ERA-Interim for Portugal. Clim. Dyn. 2012, 39, 2497–2522.
134. Rama Rao, Y.V.; Alves, L.; Seulall, B.; Mitchell, Z.; Samaroo, K.; Cummings, G. Evaluation of the weather research and forecast-
ing (WRF) model over Guyana. Nat. Hazards 2012, 61, 1243–1261.
135. Efstathiou, G.; Melas, D.; Zoumakis, N.; Kassomenos, P. A. Evaluation of WRF-ARW model in reproducing a heavy rainfall event
over Chalkidiki, Greece: The effect of land-surface features on rainfall. In Advances in Meteorology, Climatology and Atmospheric
Physics; Springer: Berlin/Heidelberg, 2013; pp. 65–71.
136. Tracton, M.S.; Kalnay, E. Operational ensemble prediction at the National Meteorological Center: Practical aspects. Weather
Forecast. 1993, 8, 379–398.
137. Buizza, R.; Hollingsworth, A.; Lalaurette, F.; Ghelli, A. Probabilistic predictions of precipitation using the ECMWF Ensemble
Prediction System. Weather. Forecast. 1999, 14, 168–189.
138. Bürger, G.; Reusser, D.; Kneis, D. Early flood warnings from empirical (expanded) downscaling of the full ECMWF Ensemble
Prediction System. Water Resour. Res. 2009, 45(10).
139. Bao, H.; Zhao, L. Development and application of an atmospheric-hydrologic-hydraulic flood forecasting model driven by
TIGGE ensemble forecasts. Acta Meteorol. Sin. 2012, 26, 93–102.
140. Zhang, W.C.; Xu, J.W.; Liu, Y.H. A catchment level early flood warning approach by using grand-ensemble weather predictions:
Verification and application for a case study in the Linyi Catchment, Shangdong, China. In Proceedings of the Third THORPEX
International Science Symposium (TTISS), 2009; pp. 14–18.
141. Xu, J.; Zhang, W.; Zheng, Z.; Jiao, M.; Chen, J. Early flood warning for Linyi watershed by the GRAPES/XXT model using TIGGE
data. Acta Meteorol. Sin. 2012, 26, 103–111.
142. Liu, J.G.; Xie, Z.H. Improving simulation of soil moisture in China using a multiple meteorological forcing ensemble approach.
Hydrol. Earth Syst. Sci. 2013, 17, 3355–3369.
143. He, Y.; Wetterhall, F.; Bao, H.; Cloke, H.; Li, Z.; Pappenberger, F.; Huang, Y. Ensemble forecasting using TIGGE for the July–
September 2008 floods in the Upper Huai catchment: A case study. Atmos. Sci. Lett. 2010, 11, 132–138.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual au-
thor(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.
