Comprehensive Zoning Strategies for Flood Disasters in China

Abstract

The frequency of global floods has increased, posing significant threats to economic development and human safety. Existing flood risk zoning studies in disaster prevention lack integration of the natural–economic–social chain and urban resilience factors. This study addresses this gap by constructing flood disaster risk and intensity indices using data from 31 provinces and 295 prefectural-level cities in China from 2011 to 2022. These indices incorporate natural (rainfall), economic (GDP), and social (population, built-up area) indicators to assess the flood likelihood and loss degree, providing comprehensive risk and intensity ratings. The study also examines the impact of resilience factors—environmental (green space), infrastructural (rainwater pipeline density), and natural resource (watershed areas)—on flood intensity. Findings reveal that high-risk regions are mainly in the Yangtze River Basin and southern regions, while high-intensity regions are primarily in the middle and lower Yangtze River and certain northwestern cities. Increasing rainwater pipeline density mitigates flood impacts in high-risk, high-intensity areas, while expanding green spaces and pipelines are effective in high-risk, low-intensity regions. This paper proposes a comprehensive flood hazard zoning mechanism integrating natural, economic, and social factors with urban resilience, offering insights and a scientific basis for urban flood management.

1. Introduction

The world is experiencing dramatic urbanization, and this rapid urbanization process is particularly vulnerable to climate change and related disasters [1,2]. Data from the National Climate Center (NCC) indicate that the duration and impact range of heavy rainfall events in China have been increasing from 2012 to 2022 [3]. Flood patterns in China generally show a spatial differentiation: heavier in the south than in the north, and more severe in the central-east compared to the west, especially in the eastern urban agglomerations. For instance, in April 2024, Guangdong Province experienced unprecedented heavy rainfall, breaking April records in several cities and causing the Beijiang River to flood on a large scale earlier than ever recorded [4]. Similarly, in June of the same year, Changsha encountered record-breaking hourly rainfall, leading to extensive road flooding and subway shutdowns. The AR6 Synthesis Report of IPCC states that global warming trends will continue to increase the likelihood of extreme rainfall events, thereby raising the risk of destructive flooding [5,6,7]. For example, under a 4 °C global warming scenario, climate change could elevate flood risk in Europe by 220% [8]. In China, the highest flood risk in the Pearl River Delta is projected to increase by 8.7% and 19.8% under 3 °C and 5 °C global warming scenarios, respectively [9]. The “amplification effect” of the compounded urban flood chain cannot be ignored, as it can disrupt urban services (e.g., transportation, sewage, communication, and electricity supply) and damage infrastructure, leading to significant adverse socioeconomic impacts [10,11]. Under the combined effects of natural and social factors, comprehensively and systematically mitigating urban flooding has become a new challenge.

Identifying the risk and intensity of urban flooding and understanding the spatial pattern of flood disasters at a macro level are crucial for effective flood prevention and mitigation [12]. Numerous researchers have conducted extensive studies on the changing patterns of flood disasters in China. Wan et al. [13] investigated the spatial and temporal characteristics of China’s flood disasters from 1950 to 2013 and found that disasters in the east were more severe than those in the west, and more significant in the south than in the north. Fang et al. [14] evaluated global heavy rainfall and flood disaster risk from 1985 to 2013 based on disaster system theory and identified the Yangtze River Basin as a high-risk area for China’s population and economy. Sun et al. [15] assessed urban flooding risk in developed regions of eastern China from 1961 to 2018, classifying cities into five risk levels by constructing disaster intensity, exposure, and urban adaptability indices, and found that the highest-risk cities were primarily located in the Guangdong Province. This body of work underscores the importance of comprehensive flood risk assessment and the development of robust zoning strategies to mitigate flood impacts, especially in rapidly urbanizing and climate-sensitive regions.

Flood hazard assessment is the cornerstone of integrated flood hazard risk management, which helps to understand the spatial distribution patterns of flood risk on a macro level. However, flood disasters are influenced by various factors including precipitation, topography, urban layout, population distribution, and rainwater pipeline density, making the causes highly complex [16,17,18]. Due to the dual natural and social attributes of flooding, effective mitigation of urban flooding through human intervention remains a significant challenge. Li et al. [19] analyzed the distribution of flooding points in eight major cities in China (Beijing, Guangzhou, Nanjing, Shanghai, Shenzhen, Tianjin, Wuhan, and Xi’an), finding that building density, green space proportion, and road density are critical factors influencing flood damage. Significant changes in urban surface hydrological characteristics caused by land use changes during rapid urbanization are major contributors to urban flooding [20,21]. Green spaces mitigate flooding to some extent through water infiltration and retention, supplementing urban drainage [15]. Several cases, such as in Beijing [22,23], Ningbo [24], and Wuhan [25], demonstrate that urban green space is an important strategy for alleviating urban flooding. Additionally, the condition of the drainage system is a crucial factor affecting urban flooding. Inadequate drainage density and high impervious areas in older urban districts make these areas susceptible to flooding events, exhibiting significant spatial aggregation effects [26,27]. Drainage pipes, as a key process in urban water management, are directly related to the formation of urban storm flooding [1,7,28]. Wu et al. [29] found that the construction of urban drainage networks lagged behind urban development, limiting their effectiveness in reducing urban storm flooding.

Current studies related to flood-prone regions focus on various aspects. The first aspect is the change of research scale, covering national regions, provincial regions [30,31,32,33], river basins [34,35,36,37], prefectural cities [38], and specific flood-prone county-level and below regions [39]. Secondly, the study indicators have evolved, focusing on flood-causing factors [33,34,36,39,40] (e.g., average annual rainfall, maximum rainfall, and frequency of heavy rainfall), disaster-conceiving factors [34,35,36,37,38,39,40] (e.g., slope, distance from major water bodies, topographic moisture index, soil type, and the normalized difference vegetation index), and disaster-bearing factors [32,33,34,35,36,37,38,39,40] (e.g., affected population, direct economic loss, and affected land), encompassing multiple dimensions of natural, social, economic, and anthropogenic activities associated with flooding behavior.

The focus of research has shifted over time. Early 21st-century studies concentrated on large-scale areas such as national, provincial, and river basins, often relying on low-precision coarse-resolution data. This data limitation allowed for the study of wide areas but was deficient in spatial detail and data precision, resulting in the inadequate analysis of the causes of waterlogging at the micro level [30,31,32,33,34,35,36,37]. In recent years, studies have gradually focused on smaller-scale areas, such as prefectural cities and small watershed scales, allowing the use of high-precision data to delve into more detailed causes of waterlogging. However, these studies lack a macro perspective on flood risk management at a large scale [38,39]. The research content has also evolved, with a gradual shift from assessing the likelihood of flooding to evaluating the vulnerability of flood-related factors. These studies now focus more on disaster prediction and assessment, covering multiple dimensions such as natural vulnerability, social vulnerability, and economic vulnerability. These developments reflect the deepening and refinement of flood disaster research, aiming to understand and respond to the impacts of flood disasters more comprehensively by integrating various vulnerability factors.

Despite significant advancements in research scales, indicators, and content, there is still a lack of systematic exploration of flood hazard zoning and classification mechanisms across the macro-scale natural–economic–social chain. Most existing macro-scale studies are based on data from over a decade ago, while China’s urbanization process and disaster defense mechanisms have changed significantly in recent years. This study aims to fill this gap by integrating natural, economic, and social factors, along with urban resilience factors, into the zoning classification mechanism of flood disasters. By combining risk evaluation and intensity evaluation of flood disasters and utilizing data from the past decade, this study constructs a flood disaster risk index and a flood disaster intensity index and establishes a flood disaster risk-intensity partition. Additionally, this paper analyzes the influencing factors of flood disaster intensity and urban resilience in key regions. The innovations of this paper include: (1) updating the characteristics of the distribution pattern of flood disaster risk and intensity at the national scale in China; (2) constructing flood disaster zoning by integrating the two dimensions of risk and intensity; (3) providing new approaches for urban flood prevention and control through a zoning and classification mechanism that combines urban resilience factors with the entire chain of natural, economic, and social dimensions. This provides a scientific basis for optimizing resource allocation, improving policy formulation, enhancing urban flood prevention and control, and boosting urban resilience in comprehensive flood management.

2. Data and Methods

2.1. Data and Sources

The data utilized in this study are categorized into three main sections:

(1)

Flood Impact Data

This includes crucial indicators such as affected areas, affected populations, deaths, and direct economic losses, reflecting the specific impact of floods in various regions of China. These data are sourced from the China Flood and Drought Disaster Bulletin, which provides detailed statistical information on all types of natural disasters and offers comprehensive data support on flooding. Provincial flood impact data are directly obtained from these nationally published statistics. Since flood damage data for prefecture-level cities are not available directly, we employed a proportional method using the population size, built-up area, and gross domestic product (GDP) of each prefecture-level city in relation to the relevant provincial data. By calculating these proportional relationships, we derived the flood damage data for each prefecture-level city.

(2)

Disaster Resilience Data

This section emphasizes the construction of disaster resilience across various regions of China. Key indicators include GDP, the greening area within built-up urban areas, and the total length of urban rainwater pipes. The data are primarily sourced from the China Statistical Yearbook and the China Urban Construction Statistical Yearbook, covering more than 295 prefecture-level cities. These authoritative publications provide critical information regarding economic, environmental, and infrastructural development.

(3)

Rainfall data

This encompasses annual average rainfall data derived from the ERA5-Land dataset, jointly published by the European Union and the European Center for Medium-Range Weather Forecasts (ECMWF). This high-resolution meteorological dataset provides essential climatic information for flood risk assessment.

To ensure data completeness and uniformity in the study, Hong Kong, Macao, and Taiwan were excluded. The data span the period from 2011 to 2022, ensuring both the timeliness and integrity of the study.

2.2. Flood Hazard Evaluation Methods

Floods cause direct property damage and loss of life and can also trigger indirect impacts, such as health risks. To fully capture the comprehensive impact of flood disasters, a variety of indicators are used in the evaluation process. The primary indicators for flood disaster assessment in China include the number of deaths, the affected population (or the proportion of the affected population to the total population of the region), the affected area of crops (or the proportion of the affected area of crops to the area of cultivated land in the region), the number of collapsed houses, direct economic loss (or the proportion of direct economic loss to the region’s GDP from the previous year), and the proportion of the loss of water conservancy facilities to the direct economic loss.

In this study, we employ two main evaluation indices:

(1)

Flood risk index

This index comprehensively evaluates the frequency and scope of flood impacts. It includes data related to floods in each province, such as the number of deaths, the number of people affected, the affected area, and direct economic losses. The purpose of this index is to measure the overall risk level of flooding in the region by considering both the probability of occurrence and the severity of the impact. The index is constructed based on two primary dimensions: natural factors, such as average annual rainfall (M), and the relative importance of flooding among natural hazards, represented by the flood exposure risk weight (RA). Specifically, the flood exposure risk weight (RA) reflects the proportion of losses caused by floods relative to the total losses from all natural hazards, thereby indicating the relative significance of flood risk compared to other natural hazards. Average annual rainfall (M) is a critical measure of natural conditions and is widely recognized as a key factor influencing flood probability.

Although both RA and M are strongly associated with flood risk, their relationship is not simply linear. While M contributes to RA, with higher rainfall generally increasing the risk of flooding, RA is influenced not only by rainfall but also by other mediating factors such as land use type, the effectiveness of urban drainage systems, and regional flood management strategies. This indicates that the relationship between M and RA is complex and requires a comprehensive consideration of multiple interacting factors. In the construction of the flood risk index (R) in this study, both M and RA play a role. By incorporating both RA and M, the index captures the cumulative effect of disaster history (RA) as well as current and potential future climate risks (M), providing a more accurate reflection of actual flood risk. This approach integrates past and present data to better account for the multifaceted nature of flood risk.

(2)

Flood Intensity Index

This index evaluates the intensity and depth of flood disasters by considering the total number of disaster-bearing entities in each province, such as the resident population, GDP, and total area. The Flood Intensity Index mainly reflects the actual degree of impact of the disaster on the regional economy, society, and land resources.

In this study, both indices are derived by accumulating the relative contributions of multiple factors, and the standardized cumulative values are used as the evaluation index. Given the large scale of the assessment region, along with significant regional differences, varying levels of economic development, and disparate disaster defense capabilities, it is challenging to accurately quantify the relative weights of these factors. To avoid biases that could arise from the dominance of certain data, we reference previous literature [41,42,43,44,45], which suggests that when dealing with complex and multidimensional influencing factors, an equal weighting approach can help prevent unrealistic weighting schemes and ensure a balanced assessment based on similarities.

Following this approach, our study employs equal weights to integrate the contributions of different factors and develop new evaluation indicators. Specifically, M and RA are assigned equal weights in their influence on R, and all factors constituting RA and F are similarly given equal influence weights. Figure 1 illustrates the research framework diagram of this paper.

To eliminate the effect of magnitude and ensure comparability across different datasets, the commonly used maximum-minimum normalization method (Min-Max scaling) was employed. This method compresses the data into the interval [0, 1], as illustrated by the following conversion equation:

$$Y_{nrom}={\displaystyle \frac{Y-Y_{\min }}{Y_{\max }-Y_{\min }}},$$

(1)

where Y_nrom is the normalized data result, Y is the original data, and Y_min and Y_max are the minimum and maximum values in the dataset for a specific feature, respectively. This normalization process ensures that each feature in the original dataset is scaled to a range between 0 and 1.

To visually represent the severity levels, the normalized flood risk index and flood intensity index were further classified. The Natural Breaks Classification method was applied to these indices using Origin software (OriginPro 2024b), dividing the study area into four levels: Level 1 (highest), Level 2, Level 3, and Level 4 (lowest).

2.2.1. Flood Risk Index (R)

The flood risk index (denoted as R) is designed to indicate the likelihood of flooding and its potential threat to a given area. This index combines the normalized values of flood damage weights with the normalized values of local average annual rainfall. The following formula expresses the calculation:

$$R_i=R{A}_i+M_i,$$

(2)

where M_i denotes the normalized value of the average annual rainfall of city i, and RA_i denotes the normalized value of the flood damage weight of city i. The flood damage weight (RA_i) is calculated based on the specific damages caused by flooding in each city. This includes metrics such as the number of people affected, the number of deaths, the area of land affected, and the direct economic losses. To measure the relative importance of flooding among all natural disasters in each city, these specific flood damages are compared with the total damage caused by all of the natural disasters in the city. The calculation formula is given below:

$$R{A}_i={\displaystyle \frac{p_i}{P_i}}+{\displaystyle \frac{d_i}{D_i}}+{\displaystyle \frac{l_i}{L_i}}+{\displaystyle \frac{e_i}{E_i}},$$

(3)

where p_i, d_i, l_i, and e_i represent the number of people affected by floods, the number of deaths, the area of land affected, and the direct economic losses in city i, respectively, and P_i, D_i, L_i, and E_i represent the total number of people affected by natural disasters, the number of deaths, the area of land affected, and the direct economic losses in city i, respectively.

2.2.2. Flood Intensity Index (F)

The flood damage intensity index (denoted as F) is designed to quantify the impact and loss associated with flood events, reflecting the extent of the damage caused by floods in each city. The calculation of F is based on the following formula:

$$F_i={\displaystyle \frac{p_i}{P_f}}+{\displaystyle \frac{d_i}{p_i}}+{\displaystyle \frac{l_i}{L_f}}+{\displaystyle \frac{e_i}{E_f}}$$

(4)

where P_f, L_f, and E_f denote the total resident population, the size of the administrative area, and the gross product of city i, respectively

2.2.3. Evaluation Factors of Resilient Cities Affecting the Intensity of Flood Damage

To systematically assess the factors influencing flood damage intensity, this study establishes urban resilience indicators. These indicators encompass the area of urban green space, the total length of the water drainage system, and the construction status of urban infrastructure. The aim is to analyze the city’s adaptive capacity and resilience in the face of flooding, providing a scientific basis for enhancing disaster management, prevention, and control strategies.

The concept of resilient cities relates to the ability of urban systems and their socio-ecological and technological networks to maintain or rapidly recover their intended functions in the face of disturbances, as well as systems with the ability to adapt to change and adjust to limit the current or future adaptive capacity [46]. Indicator systems for assessing resilient cities usually cover a wide range of economic, social, infrastructural, and ecological aspects, such as rainfall, proportion of urban green space, urban water infiltration rate, density of drainage pipes, and local financial and technological expenditures [47].

In this study, the relationship between urban resilience indicators and the flood damage intensity index is explored using Spearman correlation analysis. The evaluation factors include:

(1) Environmental resilience: Urban green space promotes rainwater infiltration and groundwater recharge, reducing surface runoff and flooding. The percentage of green space in built-up areas is selected as an environmental resilience evaluation factor to assess its role in mitigating flood disasters.

(2) Infrastructure resilience: A robust water drainage system can efficiently remove water during heavy rainfall, mitigating the impact of flooding on urban life and economic activities. The total length of the water drainage system in built-up areas is chosen as an infrastructure resilience evaluation factor to measure the effectiveness of drainage facilities in reducing flood risks.

(3) Natural resource resilience: Urban water bodies, such as rivers and lakes, play a crucial role in flood regulation and precipitation storage. The percentage of the watershed area is used as a natural resource resilience evaluation factor to assess the ability of these water bodies to mitigate flooding.

By analyzing the correlation between these resilience evaluation factors and the flood damage intensity index, this study aims to quantitatively assess the resilience level of cities in the face of flood disasters. The findings will provide a scientific foundation for enhancing urban resilience and inform the development of more effective disaster management and prevention strategies.

3. Results

3.1. Flood Risk and Flood Intensity Zoning

This study constructs the flood disaster risk index (R) and flood disaster intensity index (F) for 31 provinces based on the data on disaster losses and the total number of disaster-bearing bodies from 2011 to 2022. The average values of these indices, denoted as R0 and F0, respectively, serve as key metrics for evaluating the national disaster situation. To visualize the trends and fluctuations of the data, we plotted the changes in the risk and intensity indices, as shown in Figure 2. Analysis of the past years’ data reveals significant inter-annual fluctuations in both indices. The overall trend of the flood risk index over the past decade remains relatively flat with large year-to-year variations. In contrast, the overall trend of the flood intensity index shows a significant downward trajectory with reduced annual fluctuations, indicating that China’s flood resilience has significantly improved over the past decade. This improvement suggests that the intensity of actual floods has decreased even when faced with similar risk levels. Notably, the peaks and troughs of the risk and intensity indices were consistent before 2015, while post-2015, particularly since 2018, they have become inversely related, further indicating enhanced resilience to flooding.

Heat maps based on the flood risk index and flood intensity index were generated for 31 provinces across China from 2011 to 2022, as shown in Figure 3. The results indicate that the southeastern coastal provinces, such as Shanghai, Jiangsu, Guangdong, Fujian, Shandong, and Zhejiang, exhibited greater volatility in flood risk and intensity overall. These provinces had higher index values in several years, particularly in 2013, 2016, and 2018, suggesting that they experienced more severe flooding events during these periods.

Similarly, the flood risk and intensity indices for central provinces, including Sichuan, Chongqing, Hunan, Hubei, Jiangxi, and Guizhou, also reached high levels in several years, notably in 2016 and 2017, highlighting elevated flood risks and impacts in these areas during that time. In contrast, provinces in the Northwest, such as Ningxia, Qinghai, and Xinjiang, consistently showed low levels of flood risk and intensity indices, indicating a relatively low frequency and impact of flooding throughout the 2011–2022 period.

Overall, the flood risk and intensity indices reveal significant spatial variations across China, with the southeastern coastal and central provinces generally exhibiting higher risk and intensity, while the northwestern provinces tend to have lower risk and intensity.

To delve deeper into the spatial characteristics, we classified the regions into four levels: Grade 1 (highest risk/intensity) to Grade 4 (lowest risk/intensity), as illustrated in Figure 4. The flood risk level areas exhibit distinct geographical patterns, with the Yangtze River Basin serving as a key dividing line. High-risk areas (Grade 1 and 2) are predominantly located in and south of the Yangtze River Basin. This pattern aligns with the trend of extreme rainfall conditions in China, which generally decrease from the southeast to the northwest. The southeast coast and areas south of the Yangtze River Basin experience high extreme precipitation, ranging from 350 to 500 mm annually. This supports the scientific validity of high-risk region delineation [48]. Conversely, lower-risk regions (Grade 3 and 4) are mainly situated north of the Yangtze River Basin, characterized by drier climates and less rainfall. For example, northern China, northwest China, and the Qinghai-Tibet region have relatively low extreme precipitation, with Turpan in Xinjiang recording the lowest extreme precipitation in the country at 12.6 mm [48].

Flood-affected regions with high intensity (Grade 1 and 2) are concentrated in central China, the Yangtze River Basin, and parts of northwest China (Gansu, Qinghai, and Ningxia) and northeast China (Heilongjiang Province). For instance, in 2020, 28 provinces experienced varying degrees of flooding, with Anhui, Sichuan, Jiangxi, and Hubei in the Yangtze River Basin suffering substantial economic losses totaling CNY 163.93 billion, accounting for 61.4% of the national total. Regions with lower intensity (Grade 3 and 4) are primarily located in North China, Northeast China, and coastal areas. Interestingly, although Guangdong and Fujian provinces along the southeast coast are high-risk areas, their actual damage intensity is lower (Grade 3), indicating strong flood resilience.

The analysis of flood disaster risk and intensity zoning reveals some regional consistency. However, due to variations in rainfall, topography, and geomorphology, the potential risk and actual intensity zones do not completely overlap. The systematic and complex nature of flood damage means that focusing solely on risk or intensity does not provide a full picture. For example, Hubei Province has a flood damage risk level of Grade 2 but a damage intensity level of Grade 1. Ignoring the intensity level could result in overlooking Hubei’s significant flood damage.

Therefore, to develop comprehensive and accurate flood prevention and control strategies, this study combines the flood disaster risk and intensity indices to identify high-risk and high-intensity areas. This approach helps to avoid neglecting key regions, supports the formulation of scientifically sound prevention and control policies, and optimizes resource allocation and response plans.

3.2. Integrated Provincial Flood Risk-Intensity Zoning

In this study, we developed a comprehensive provincial flood hazard zoning framework by integrating the flood risk index (R) and the flood intensity index (F). The average values of the flood risk index (R) and flood intensity index (F) across provinces, recorded at 0.47 and 0.48, respectively, served as the reference points for this zoning classification. Provinces exhibiting a flood risk index (R) exceeding 0.47 were classified as high-risk areas, whereas those with an index below 0.47 were classified as low-risk areas. Similarly, provinces with a flood intensity index (F) greater than 0.48 were designated as high-intensity areas, and those with an index below 0.48 were categorized as low-intensity areas. Consequently, the study area was stratified into four distinct categories: high-risk and high-intensity (H-H), low-risk and high-intensity (L-H), low-risk and low-intensity (L-L), and high-risk and low-intensity (H-L), as depicted in Figure 5.

The high-risk, high-intensity (H-H) regions are predominantly located in Central China and the Yangtze River basin, encompassing provinces such as Anhui, Guangxi, Guizhou, Hainan, Hubei, Hunan, Jiangxi, Sichuan, and Chongqing. Notably, Jiangxi records the highest flood intensity index, while Chongqing has the highest flood risk index. This region demonstrates a strong correlation between flood risk and intensity, particularly in the middle and lower reaches of the Yangtze River. Historically, this area has been susceptible to flooding, necessitating robust enhancements in monitoring, early warning, and response mechanisms.

High-risk, low-intensity (H-L) regions are concentrated in the economically advanced southeastern coastal provinces of China, including Zhejiang, Fujian, and Guangdong. Despite facing higher flood risks, the disaster intensity remains relatively low owing to their enhanced risk management capabilities. It is advisable to further strengthen disaster monitoring and early warning systems in these regions.

Low-risk, high-intensity (L-H) regions include Gansu, Qinghai, and Heilongjiang. These regions, characterized by lower economic development, exhibit a low risk of flooding but high damage intensity in the event of a flood. There is an urgent need to bolster disaster early warning systems and enhance rescue operations, as well as improve overall disaster resilience and post-disaster recovery capabilities.

Low-risk, low-intensity (L-L) regions encompass North China and most of Northwest and Northeast China. In these areas, both flood risk and intensity are below the national average. However, this relative safety should not obscure the fact that specific locations with unique topographical and geomorphological features may still be vulnerable to sudden flooding events.

Using the flood partition data, along with the flood disaster risk index and flood disaster intensity index, we plotted the changes in risk and intensity indices for each partition from 2011 to 2022, as shown in Figure 6. Comparing these results to the national averages, we observed that high-risk, high-intensity regions demonstrated a certain decreasing trend in flood disaster risk, with a significant reduction in the flood disaster intensity index. High-risk, low-intensity regions also showed a notable downward trend in both the risk and intensity indices. In contrast, low-risk, high-intensity regions experienced a significant increase in the flood disaster risk index, with a slight rise in the intensity index. Meanwhile, low-risk, low-intensity regions displayed a relatively flat trend in the risk index, while the intensity index showed a marked decrease.

These findings suggest that, despite the overall reduction in flood risk and intensity, high-risk, high-intensity, and high-risk, low-intensity regions remain more vulnerable than the national average. Therefore, these subregions should be prioritized for targeted interventions to enhance their risk management, prevention, and control capabilities.

To further discuss the changes in flood hazard zoning, we divided 2011–2016 into the first stage and 2017–2022 into the second stage according to the time dimension, studied the changes in hazard zoning of each province in these two stages, and plotted the stage changes (see Figure 7). The analysis results show that during 2011–2022, the change of flood hazard zoning in each province shows obvious regional characteristics and stage differences. Several provinces such as Anhui, Guangxi, Guizhou, Hubei, Hunan, Jiangxi, Chongqing, Fujian, Hainan, Sichuan, Zhejiang, Yunnan, Gansu and Guangdong are in the high-risk or high-intensity region in both phases, indicating that these provinces continue to face higher flood risks and disaster intensities, and need to focus on and strengthen disaster prevention and mitigation measures.

To further examine the changes in flood hazard zoning, we divided the period 2011–2016 as the first stage and 2017–2022 as the second stage, analyzing the changes in hazard zoning for each province across these two periods and mapping the stage-specific changes (see Figure 7). The results reveal that from 2011 to 2022, flood hazard zoning in each province exhibited distinct regional characteristics and stage differences. Provinces such as Anhui, Guangxi, Guizhou, Hubei, Hunan, Jiangxi, Chongqing, Fujian, Hainan, Sichuan, Zhejiang, Yunnan, Gansu, and Guangdong remain in high-risk or high-intensity zones in both stages, indicating persistent high flood risks and disaster intensities. These regions require focused attention and enhanced disaster prevention and mitigation measures.

Notably, Fujian, Hainan, Sichuan, and Zhejiang continue to face high risk, but have shown reductions in disaster intensity. This trend warrants further study to gain insights from their management strategies that have successfully reduced flood intensity. Meanwhile, Shanghai and Hebei, initially categorized as high-risk, high-intensity regions in the first stage, transitioned to low-risk, low-intensity (L-L) regions in the second stage, likely due to improvements in flood control infrastructure or emergency response capabilities, suggesting a decline in flood risk and intensity.

Beijing, Jiangsu, Inner Mongolia, Ningxia, Shandong, Shaanxi, Tianjin, and Xinjiang remained in the low-risk, low-intensity (L-L) category throughout both stages, indicating stable, low-frequency flood occurrences and impacts, alongside effective flood prevention measures. Conversely, Henan, Heilongjiang, Jilin, Liaoning, Qinghai, and Shanxi shifted from low-risk, low-intensity (L-L) in the first stage to low-risk, high-intensity (L-H) in the second stage, suggesting that while the overall risk remains low, the intensity of flood damage has increased in these provinces.

Overall, the stage change maps illustrate varying trends in flood risk and intensity across provinces, underscoring the importance of continuous monitoring and the ongoing optimization of disaster management strategies, particularly in high-risk or high-intensity areas.

3.3. Indicators of Urban Resilience Affecting the Intensity of Flood Damage

To investigate the critical factors influencing the intensity of flood damage, this study conducts a detailed analysis of nearly 300 urban resilience indicators. To refine the focus and facilitate comparative analysis, we specifically examine high-risk, high-intensity (H-H) regions and high-risk, low-intensity (H-L) regions. This comparative approach aims to elucidate why the extent of disaster damage in these two types of regions differs significantly despite their similar exposure to high flood risk. The objective is to accurately identify and assess the characteristics of flood risk and intensity across various regions, as well as the factors influencing these characteristics. This analysis seeks to enhance the precision and applicability of the study, thereby providing a scientific basis for improved risk management and decision-making, and informing the development of more targeted and effective disaster prevention and mitigation strategies. The spatial distribution of risk, intensity, and the corresponding indicators for evaluating urban resilience in high-risk, high-intensity (H-H) and high-risk, low-intensity (H-L) regions are illustrated in Figure 8.

Despite both being categorized as high-risk areas, there is a notable variation in their flood intensity performance. In the high-risk, low-intensity (H-L) regions, the percentage of green space in built-up areas and the density of the rainwater pipe significantly exceed those in the high-risk, high-intensity (H-H) regions (see Figure 5a,b). Specifically, the high-risk, low-intensity regions exhibit a built-up area green space average percentage of 35.85% and a rainwater pipe density of 4.31 km/km², whereas the high-risk, high-intensity regions have a green space average percentage of 39.38% and a rainwater pipe density of 6.41 km/km². On a temporal scale (refer to Figure 9), the percentage of built-up green space in high-risk, low-intensity (H-L) areas reached 38.02% as early as 2013, whereas high-risk, high-intensity areas had not achieved this level by 2021. Similarly, the rainwater pipe density in high-risk, high-intensity regions reached 4.39 km/km² by 2018, compared to 4.69 km/km² in the high-risk, low-intensity regions by 2011.

Analysis of flood intensity zones reveals notable discrepancies in flooding intensity ratings among prefecture-level cities within the same province, attributed to variations in urban resilience indicators. For instance, within Fujian Province, categorized as a high-risk, low-intensity region, Quanzhou exhibits a higher flood intensity rating compared to Putian. This difference primarily stems from Putian’s significantly higher rainwater pipe density (6.45 km/km²) compared to Quanzhou (3.15 km/km²), despite both cities having equivalent ratings for built-up area green space percentage and share of water bodies. A similar pattern is observed in high-intensity, high-risk areas; for example, in Jiangxi Province, Fuzhou City and Ganzhou City, despite having identical rankings for rainwater pipe density and share of water bodies, show a disparity in flood intensity due to Fuzhou City’s greater built-up area green space percentage (44.12%) compared to Ganzhou City (41.03%). The data results show a stronger intensity of flood damage in Ganzhou.

To further analyze prefecture-level cities, we examined the flood disaster risk index, intensity index, and resilience indicators (including the rainwater network ratio and green space density in built-up areas) for Fuzhou and Ganzhou (high-risk, high-intensity areas) and Putian and Quanzhou (high-risk, low-intensity areas) from 2011 to 2022. The results, presented in Figure 10, indicate some correlation between the cities’ resilience indicators and their flood risk and intensity.

All four cities exhibit high flood disaster risk indices, with notable peaks in certain years (e.g., 2016 and 2018), highlighting their vulnerability to flooding. However, there are clear differences in flood disaster intensity: Fuzhou and Ganzhou generally show higher intensity indices, with fluctuations that are closely synchronized with the risk index. In contrast, Putian and Quanzhou have lower and more stable intensity indices, even during years when the risk index rises significantly.

The resilience indicators for Fuzhou and Ganzhou, measured by the proportion of stormwater networks in built-up areas and green space density, remain relatively low and have not shown significant improvement over the study period. Conversely, Putian and Quanzhou have notably better resilience indicators, particularly after 2017, when both green space density and stormwater network coverage increased substantially.

Overall, it can be concluded that high-risk, high-intensity regions (such as Fuzhou and Ganzhou) are characterized by high flood risk and intensity, coupled with lagging resilience measures, as evidenced by low green space density and limited stormwater pipeline coverage. This suggests that insufficient resilience measures may exacerbate the severity of flooding. In contrast, high-risk, low-intensity regions (Putian and Quanzhou) demonstrate that higher levels of urban resilience, including increased green space density and more extensive stormwater networks, can effectively mitigate the impacts of flooding even in the face of high risk.

To further investigate the relationship between urban resilience indicators and flooding intensity, this study employed the Spearman correlation analysis to examine the associations between the percentage of built-up area green space, the percentage of built-up area rainwater pipes, the share of water bodies, and the flood intensity index. The results are summarized in

Table 1. Correlation Analysis Between Flood Intensity Index and Evaluation Indicators.

City Zoning Category	Evaluating Indicator	Proportion of Green Space	Proportion of Rainwater Pipes	Share of Water Bodies
High-risk and high-intensity	Intensity index	−0.03	−0.11 *	0.07 *
	p value	0.28	0.00	0.03
High-risk and low-intensity	Intensity index	−0.18 *	−0.16 *	−0.05
	p value	0.00	0.00	0.29

Notes: A two-tailed test of significance was used. * Significant correlation at the 0.05 level.

.

In high-risk, high-intensity areas, the flood intensity index showed a significant negative correlation with the proportion of storm drains in built-up areas (correlation coefficient of −0.11) and a positive correlation with a share of water bodies (correlation coefficient of 0.07). Conversely, in high-risk, low-intensity areas, the flood intensity index exhibited a significant negative correlation with both the proportion of built-up green space (correlation coefficient of −0.18) and the proportion of built-up water drains (correlation coefficient of −0.16), with no significant correlation observed with share of water bodies. These findings indicate that both green space and water infrastructure within built-up areas have considerable mitigating effects on flood intensity, particularly in high-risk, low-intensity regions where enhancements in these factors have markedly reduced flooding severity. Notably, in high-risk, high-intensity areas, the positive correlation between the share of water bodies and flood intensity suggests that increased water area may be less effective in managing and regulating flood events.

4. Discussion

The occurrence and evolution of floods represent a complex, systematic process influenced by a confluence of factors including geography, climate change, and socio-economic conditions. Existing research predominantly focuses on flood risk assessment and trend analysis across various spatial scales in China, ranging from provincial to watershed, and even prefectural and municipal levels. Common analytical methodologies include spatio-temporal analysis, fuzzy logic, hierarchical analysis, multi-criteria analysis, and gray modeling. However, results and recommendations often vary due to differences in study timeframes and data sources. Generally, the literature emphasizes the roles of natural factors and human activities in flood dynamics.

In this study, we developed a comprehensive evaluation index for flood risk and intensity by incorporating natural, economic, and social factors, and quantitatively analyzed the determinants of flood damage intensity in conjunction with urban resilience indicators. By updating and examining flood disaster data from the past decade, our study elucidates the spatial distribution characteristics of flood events in China and contrasts these with previous findings.

While earlier studies have identified regions such as the southeast coast of China (e.g., Shanghai, Fujian and Guangdong) and the middle and lower reaches of the Yangtze River (e.g., Hubei, Hunan and Anhui) as high-risk areas [30,31], our findings reveal that the southeast coast (specifically Zhejiang, Fujian, and Guangdong) is now categorized as a high-risk, low-intensity region. This shift is attributed to significant improvements in urban resilience, including increased green space and enhanced rainwater pipe network density. Conversely, our study highlights low-risk but high-intensity areas, such as Qinghai, Gansu, and Heilongjiang, which, despite having a lower flood risk, experience high flood intensity and, thus, warrant increased attention.

Table 2. Comparison of spatial zoning studies of flood disasters in China.

No.	Literature Cited	Year	Scope of the Study	Discussion	Whether to Discuss Urban Resilience Factors	Whether to Quantify the Factors Influencing the Intensity of Flood Damage	Main Findings
1	Dapeng Huang et al. (2012) [32]	2001–2011	China’s provincial scale	Assessment of flood vulnerability with discussion of economic and demographic factors	NO	NO	The eastern coastal provinces of Jiangsu, Zhejiang and Shandong have high flood vulnerability due to their developed economies and dense populations.
2	Chengjing Nie et al. (2012) [33]	1980–2009	China’s provincial scale	Spatial and temporal flooding variability and its influencing factors are discussed	NO	NO	Frequent floods and serious impacts in the middle and lower reaches of the Yangtze River (e.g., Hubei, Hunan, and Jiangxi) and the Pearl River Delta (e.g., Guangdong)
3	Minlan Shao et al. (2014) [30]	2004–2009	China’s provincial scale	The assessment of flood risk focuses on rainfall and economic losses.	NO	NO	The southeastern coastal areas (e.g., Fujian and Guangdong) and the middle and lower reaches of the Yangtze River (e.g., Hubei and Anhui) are high-risk areas.
4	Naiming Xie et al. (2014) [31]	2004–2010	China’s provincial scale	Analyses of regional meteorological disaster losses in China, focusing on meteorological disasters and economic losses.	NO	NO	East China (e.g., Shanghai, Jiangsu and Zhejiang) and South China (e.g., Guangdong and Guangxi) have suffered greater losses from meteorological disasters.
5	Yu Chen (2022) [34]	1960–2019	China’s provincial scale	Mapping of flood hazard zones, focusing mainly on geographic information and integrated multi-factor analyses	NO	NO	The middle and lower reaches of the Yangtze River (e.g., Jiangsu and Anhui) and the Pearl River Delta (e.g., Guangdong) are high-risk areas.
6	This study	2011–2022	31 provinces and 295 prefecture-level cities in China	Flood hazard zoning based on the whole natural–economic–social chain and incorporating urban resilience factors	YES	YES	Updating high-risk areas. Southeast coastal cities were found not to be high-intensity areas, and the region’s urban resilience capacity, such as the percentage of green space area and the density of the rainwater pipe, was found to be substantially better than before.

compares this paper with earlier related studies.

Future research should delve deeper into the interactions between flood-related factors, including the effects of climate change on precipitation patterns, urbanization on land-use changes, and socio-economic development on disaster response capabilities. Additionally, the data used in this study, primarily derived from historical disaster records, may suffer from temporal lag and incompleteness. The estimation methods for prefecture-level cities also present opportunities for improved accuracy in future assessments. Refining flood disaster prevention and control measures to develop more precise strategies for different regions is imperative. This study represents a significant advancement in constructing a flood disaster assessment index system and zoning strategy, providing a valuable scientific basis and reference for flood disaster management in China. Nonetheless, the complexity and variability of flood events necessitate ongoing improvements and deeper research to enhance prevention and control strategies.

5. Conclusions

This study examines the spatial and temporal characteristics of flood disasters in China over the past decade, incorporating a comprehensive natural–economic–social framework and urban resilience factors. By constructing flood disaster risk and intensity indices that account for population, natural environment, and economic loss, we have classified flood disaster areas into four categories: high-risk and high-intensity, high-risk and low-intensity, low-risk and high-intensity, and low-risk and low-intensity. Spearman’s correlation analysis was employed to explore the relationships between the proportion of green space, rainwater pipe coverage, and watershed density in prefecture-level cities, and the flood intensity index, from the perspectives of buffering, drainage, and water storage. The main findings of this study are as follows:

(1) Regional Characteristics of Flood Disasters: Flood disasters in China exhibit distinct regional patterns. High-risk areas are predominantly located in the Yangtze River Basin and its southern regions (excluding Yunnan Province) and along the southeast coast. High-intensity areas are primarily found in the central Yangtze River Basin and parts of the northwest.

(2) Focus Areas of Risk and Intensity: High-risk, high-intensity regions are chiefly concentrated in the middle and lower reaches of the Yangtze River, while high-risk, low-intensity areas are situated in the Pearl River Basin and the southeastern coastal regions. Additionally, low-risk, high-intensity regions, such as Gansu, Qinghai, and Heilongjiang, require attention due to their high flooding intensity despite lower risk levels.

(3) Influence of Urban Resilience Factors: In high-risk, high-intensity and high-risk, low-intensity regions, the proportion of rainwater pipes in built-up areas shows a significant negative correlation with flood intensity, indicating that expanding the rainwater pipe network can substantially mitigate flooding. In high-risk, low-intensity areas, the proportion of green space in built-up areas is significantly negatively correlated with flood intensity, suggesting that increasing green space helps reduce flood severity. Conversely, in high-risk, high-intensity areas, watershed density is positively correlated with flood intensity, implying that the watersheds in these areas may be ineffective in flood storage and regulation.

Based on these findings, the following recommendations are proposed:

(1) Enhance urban green space: Increasing the proportion of green space in urban areas is crucial for reducing flood severity and should be a key component of urban planning and development, especially in high-risk, low-intensity areas.

(2) Optimize drainage system: Expanding the coverage of rainwater pipes and optimizing urban drainage systems, alongside strengthening emergency response capabilities, are effective measures for flood mitigation in critical areas.

(3) Rational planning of waters: Although watershed density does not significantly impact flood intensity, effective planning and management of waters are essential to ensure their water storage and regulation functions, particularly in high-risk, high-intensity regions.

(4) Implement Comprehensive Flood Prevention Measures: Adopting a multifaceted approach that integrates increased green space, optimized drainage systems, and rational water management will enhance urban flood prevention capacity and reduce flood-related risks and losses.

In conclusion, this study offers a novel perspective and methodology for flood risk assessment and prevention strategies in China. Its findings have significant theoretical and practical implications, and future research should build upon this work to further develop comprehensive assessment and mitigation strategies for flooding.