Integrated assessment of available water volume for sustainable sponge city construction – A case study in Xi'an, China

Over the past three decades, urban construction and development in China have experienced rapid growth, with the urbanization rate increasing from 26.41% in 1990 to 63.89% in 2020 (Zhou et al. 2023). This rapid urbanization and concentrated population have led to increasingly severe urban water-related issues in cities, particularly water shortage and frequent urban floods (Han et al. 2023). According to existing literature, 107 cities across China are currently facing water scarcity issues, with 79 cities experiencing quantity-induced water scarcity, predominantly concentrated in the North China Plain and Northwest China (Ba et al. 2023). In addition, nearly 98% of China's 654 leading cities are grappling with flooding and waterlogging due to the replacement of pervious vegetated surfaces with impervious pavements in the urban area (Li et al. 2023).

To address these urban water-related challenges, the concept of sponge city construction (SCC) was officially proposed as a national strategy for sustainable urban development in December 2013 during the National Working Meeting on Urbanization (Shang et al. 2023). While existing theories and technical measures such as best management practices (BMPs) (Rentachintala et al. 2022), low impact development (LID) (Qiao 2023), green infrastructure (GI) (Grabowski et al. 2022), and water sensitive urban design (WSUD) (Williams 2020) are widely applied and focused on restoring the local water ecological environment, SCC emphasizes the integration of the city with the watershed, emphasizing the concept of water cycle management (Ren et al. 2017; Su et al. 2019).

In the context of SCC, a ‘Sponge City’ refers to a city that possesses adaptability to environmental changes and can effectively respond to natural disasters. This concept is based on the city's ability to absorb, store, infiltrate, and purify water while also releasing stored water during times of water shortages (Ji & Bai 2021). However, the current implementation of SCC in China predominantly follows the ‘Sponge City Development Technical Guideline – Low Impact Development Rainwater System Construction,’ which falls short of providing comprehensive theoretical guidance. The Chinese government launched the SCC program to integrate this guideline with the concept of LID in urban areas, modernizing and expanding the traditional approach to water management. However, the lack of comprehensive theoretical guidance often leads to an incomplete control of flooding risks in cities.

Currently, research on SCC primarily focuses on exploring the concept of sponge city and its implications, as well as estimating its effectiveness. For instance, Nguyen et al. (2020) propose a new framework for the sponge city model, integrating four main sub-models to simulate the environmental, social, and economic aspects of different sponge city infrastructure options. Various assessment methods have been proposed, including cellular automata (Wang et al. 2021), robust decision-making (Liang et al. 2020), and analytical hierarchy process (Song 2022), among others. However, there is limited research in the literature regarding the potential implementation of SCC.

From an Earth perspective, it can be viewed as a large sponge that continuously provides clean water resources through the natural hydrological cycle process of storage, infiltration, and purification. Water cycle management aims to maintain the naturalness of this process to the fullest extent possible and design systems based on natural laws. It requires comprehensive consideration of both the natural and artificial hydrological cycles within a watershed to achieve sustainable urban development (Yang et al. 2019). SCC strives to create an urban water space that mimics a natural water cycle by making use of the natural conditions of urban watersheds and preserving, protecting, and enhancing the natural hydrological processes to the greatest extent possible (Kirshen et al. 2018). This core concept is the foundation for implementing SCC in a systematic and comprehensive manner (Pathirana et al. 2018). SCC goes beyond the limitations of ‘LID’ and emphasizes the provision of ‘Water Volume’ to transform irregular resources into stable resources.

The proposal and implementation of SCC have been ongoing over the past 10 years. However, significant progress in theoretical research remains elusive, as most efforts have focused on technological innovations and measures associated with BMPs, LID, GI, and WSUD (Yuan et al. 2023). These efforts, while significant, have not fundamentally achieved SCC's purpose and core concept for realizing sustainable development. Therefore, it is imperative to introduce the concept of available water volume (AWV) as a guiding principle to improve SCC.

In the context of urban watersheds, the internal water resources can be categorized into two types. The first is the ecological water demand (EWD), which is necessary to sustain the development of the natural ecological environment. The second category encompasses additional water resources that are available for human use under the premise of guaranteeing the EWD (Poff & Matthews 2013). Therefore, AWV refers to the volume of water resources within a city that can be absorbed, stored, and released for use by humans. By accurately calculating the AWV, it becomes possible to effectively prevent, control, and resolve water-related problems such as shortages and floods within urban watersheds.

Given the urgent need for a theoretical guidance method for SCC, this paper introduces a new approach called AWV to guide the implementation of SCC. Focusing on the city as the construction object of SCC and the aim of alleviating urban water-related challenges, this method is developed by considering urban watershed as the fundamental research unit and water as the basic research element. Firstly, a comprehensive understanding of the water mass balance within the urban watershed is essential as it forms the basis for addressing water-related problems. Secondly, the water storage spaces within the urban watershed represent the potential water resources and serve as crucial areas for studying water-related issues. Additionally, particular attention is given to water resources that can be controlled and regulated by humans within the urban watershed. Finally, the AWV method is derived from the consideration of water mass balance, water volume, and EWD within the urban watershed.

To illustrate the application of this theoretical model in addressing the severe water shortage and frequent flooding within urban areas, the Xi'an urban area was selected as a case study in this paper, considering data reliability and the significance of the water-related issues. The establishment of the theoretical model for SCC carries fundamental significance in guiding SCC implementation and mitigating urban water-related challenges.

The function and significance of the AWV in urban watersheds are summarized in Table 1.

Table 1

Function and significance of the AWV in urban watersheds

Condition .	Natural ecological environment .	Water resources artificially .	Utilization potential of AWV (%) .	Disaster .
	Deteriorating	0	100	Natural environment deterioration
	Balance	0	100	–
	Healthy and sustainable		0–100	–
	Healthy and sustainable		0	–
	Healthy and sustainable		0	Floods

Condition .	Natural ecological environment .	Water resources artificially .	Utilization potential of AWV (%) .	Disaster .
	Deteriorating	0	100	Natural environment deterioration
	Balance	0	100	–
	Healthy and sustainable		0–100	–
	Healthy and sustainable		0	–
	Healthy and sustainable		0	Floods

View Large

According to Figure 4, the land use in the study area comprised seven categories: rivers and lakes, woodlands, grasslands, cultivated land, rural residential land, industrial land, and urban impermeable land. Details regarding the area, CN value, potential maximum retention depth ⁠, and initial abstraction depth for each land use are shown in Table 2. The elevation analysis was performed by integrating the ENVI software with the Google Earth software (Figure 4).

The surface water increment in the study area comprises two components: river water increment and lake water increment. To evaluate the river water increment, it is necessary to analyze and calculate the river sections (Figure 3). The annual mean monthly runoff, length, width, and slope of each river are shown in the Supplementary Material. Manning's equation was then applied to calculate the river water increment.

The lake water increment represents the volume of water resources stored in the lakes, with lake storage capacity information shown in the Supplementary Material. Consequently, the surface water increment in the study area can be calculated using Equation (3), and the results are shown in Table 3.

Table 3

Results of the AWV evaluation (million m³) for Xi'an

Month .	.	.	.	.	.	.	.	.
1	10.49	0	55.83	8,938.5	1.95	43.94	53.88	8,894.56
2	10.40	0	55.83	8,938.5	1.85	60.14	53.98	8,878.36
3	10.91	0	55.83	8,938.5	2.28	104.33	53.55	8,834.17
4	12.25	0	55.83	8,938.5	2.63	147.15	53.20	8,791.35
5	12.49	0	55.83	8,938.5	2.77	193.74	53.06	8,744.76
6	11.44	0	55.83	8,938.5	2.55	236.61	53.28	8,701.89
7	12.76	0	55.83	8,938.5	2.82	233.28	53.01	8,705.22
8	12.65	0	55.83	8,938.5	2.68	208.32	53.15	8,730.18
9	15.18	0	55.83	8,938.5	3.16	132.08	52.67	8,806.42
10	11.69	0	55.83	8,938.5	2.32	89.06	53.51	8,849.44
11	11.25	0	55.83	8,938.5	2.12	53.65	53.71	8,884.85
12	10.70	0	55.83	8,938.5	2.06	40.36	53.77	8,898.14

Month .	.	.	.	.	.	.	.	.
1	10.49	0	55.83	8,938.5	1.95	43.94	53.88	8,894.56
2	10.40	0	55.83	8,938.5	1.85	60.14	53.98	8,878.36
3	10.91	0	55.83	8,938.5	2.28	104.33	53.55	8,834.17
4	12.25	0	55.83	8,938.5	2.63	147.15	53.20	8,791.35
5	12.49	0	55.83	8,938.5	2.77	193.74	53.06	8,744.76
6	11.44	0	55.83	8,938.5	2.55	236.61	53.28	8,701.89
7	12.76	0	55.83	8,938.5	2.82	233.28	53.01	8,705.22
8	12.65	0	55.83	8,938.5	2.68	208.32	53.15	8,730.18
9	15.18	0	55.83	8,938.5	3.16	132.08	52.67	8,806.42
10	11.69	0	55.83	8,938.5	2.32	89.06	53.51	8,849.44
11	11.25	0	55.83	8,938.5	2.12	53.65	53.71	8,884.85
12	10.70	0	55.83	8,938.5	2.06	40.36	53.77	8,898.14

View Large

For the subsurface water increment, assuming negligible water transfer between the surface and subsurface space, it can be further assumed that the subsurface runoff inflow and the subsurface runoff outflow are approximately equal, as represented by Equation (4). Therefore, in the absence of rainfall, the subsurface water increment in the study area is zero, and the corresponding calculation results are shown in Table 3.

The surface water volume in the study area is divided into two components: river water volume and lake water volume. The Feng River and Ba River have maximum transport flows of 1,030 and 1,384 m³/s, respectively (Peng et al. 2016). Under these conditions, no flood occurs in the study area. Given the variable slopes in the study area, which includes plains and mountains, it is essential to segment the river channel for an accurate calculation of the water volume. Table S3 provides details on the maximum transport flow, length, average width, and slope of each river section in Figure 4. Combined with Manning's equation, the calculated result for the river water volume is 47.15 million m³.

The lake water volume in the study area is 8.68 million m³, as indicated in Table S2. Applying Equation (5), the total surface water volume is determined to be 55.83 million m³, as shown in Table 3. Regarding the subsurface water volume, the predominant soil type in the study area is sandy loam with a saturated water content percentage of 15%. The plain and mountain areas cover 2,629 and 1,665 km², respectively, with average phreatic aquifer thicknesses of 10 and 20 m, respectively (Table 2). Substituting these values into Equation (6), the calculated subsurface water volume in the study area is 8,938.5 million m³, as shown in Table 3.

The surface EWD in the study area is divided into two components: rivers and lakes. For rivers, 10% was selected as the percentage of EWD in river runoff. Therefore, the EWD volume in the river is determined using data from the average annual monthly runoff, the length, width, and slope presented in the Supplementary Material, and Manning's equation. The calculated results are also shown.

For lakes, the EWD comprises two parts: evaporation and infiltration. Evaporation was calculated using the comprehensive formula of free water surface evaporation estimation. The lakes in the study cover an area of 5 km². The calculated parameters for surface EWD and subsurface EWD are shown in Table S4. Infiltration was calculated using the infiltration coefficient, which for sandy loam is 10⁻⁷ m/s. Therefore, the infiltration of the lakes is determined and presented in Table S4. By substituting these data into Equation (7), the surface EWD in the study area can be calculated, and the results are shown in Table 3.

As for subsurface EWD, vegetation in the study area covers 1,665 km² (Table 2), and the other parameter values in Equation (8) are referenced from Yao et al. (2020). Thus, the subsurface EWD in the study area can be calculated and is shown in Table 3.

As outlined in Sections 2.4.1 and 2.4.2, the surface and subsurface AWV can be derived by subtracting the surface and subsurface EWD from the respective surface and subsurface water volumes, as represented in Equations (9) and (11), respectively. Consequently, the results for and can be obtained and are shown in Table 3. It is evident that the surface and subsurface AWV are not fixed values but dynamic, changing over time.

Table 3 reveals that the monthly surface water increment in the surface space of the study area exceeds the surface EWD. Therefore, the ecological environment of surface water is deemed healthy and sustainable. At the same time, even after deducting the surface EWD, a significant quantity of surface water resources remains available for human use, as depicted in Table S5. However, compared to surface AWV, the quantity of water resources accessible for human use is very small. Therefore, the surface AWV in the study area exhibits substantial potential for storing and utilizing surface water resources.

Table 3 indicates that the monthly subsurface water increment in the subsurface space of the study area is less than the subsurface EWD. This signifies a deterioration in the ecological environment of subsurface water, aligning with the actual situation in the urban area of Xi'an. Over the past few decades, excessive groundwater exploitation has led to severe groundwater shortage and the formation of a groundwater funnel. Presently, the Xi'an government has implemented measures to prohibit groundwater exploitation. This underscores that the subsurface AWV of the study area remains untapped, harboring significant potential for utilization. Notably, Table 3 reveals that the subsurface AWV is considerably higher than the surface AWV, indicating a greater potential for utilization.

For the Xi'an urban area, the rainy season is concentrated in August, September, and October. Therefore, typical rainfall events representing moderate rain, heavy rain, and torrential rain were selected for analysis and calculations.

Rainfall event 1: Moderate rain occurred from 5:00 to 13:00 in October, lasting 9 h, with a rainfall of 13.7 mm.

Rainfall event 2: Heavy rain occurred from 11:00 on September 10 to 1:00 on September 11, lasting 15 h, with a rainfall of 24.4 mm.

Rainfall event 3: Torrential rain occurred from 19:00 on August 2 to 0:00 on August 3, lasting 6 h, with a rainfall of 59.9 mm.

Figure 5 illustrates that the rainfall runoff generated by rainfall event 1 was 8.85 million m³, which is lower than the utilization potential of 44.14 million m³ for the surface AWV in October. Therefore, this portion of the rainfall runoff could be collected and stored in the surface space of the study area and used artificially. The rainfall infiltration generated by rainfall event 1 was 49.97 million m³, significantly less than the utilization potential of 8,849.44 million m³ for the subsurface AWV in October. Therefore, this portion of rainfall infiltration would fully penetrate the subsurface space.

Similarly, the rainfall runoff generated by rainfall event 2 was 24.99 million m³, which was less than the utilization potential of 40.65 million m³ for the surface AWV in September. Therefore, this part of rainfall runoff could be collected and stored in the surface space of the study area for artificial use. The rainfall infiltration generated by rainfall event 2 was 79.77 million m³, which was markedly lower than the utilization potential of 8,806.42 million m³ for the subsurface AWV in September. Therefore, this part of rainfall infiltration would completely penetrate the subsurface space.

The rainfall runoff generated by rainfall event 3 was 99.08 million m³, exceeding the utilization potential of 43.18 million m³ for the surface AWV in August. In this scenario, 43.18 million m³ of rainfall runoff could be collected and stored in the surface space of the study area for artificial use. However, the remaining 55.9 million m³ of the rainfall runoff poses a direct flood risk. The rainfall infiltration generated by rainfall event 3 was 158.16 million m³, significantly less than the utilization potential of 8,730.18 million m³ for the subsurface AWV in August. Therefore, this portion of rainfall infiltration would completely penetrate the subsurface space.

From a temporal perspective, the evaluation of AWV involves two components. One aspect involves analyzing the water regulation relationship of urban watersheds by calculating water increment, water volume, EWD, and AWV in units of month, quarter, or year. The other aspect involves short-term rainfall analysis, suitable for examining urban flooding problems. At the same time, these two components can be combined to comprehensively address water volume relationships and flood control issues.

According to the calculations in Sections 3.2 and 3.3, the AWV of the surface and subsurface space in the urban area of Xi'an remains underutilized, indicating substantial potential for transforming Xi'an into a sponge city.

For the surface space in the urban area of Xi'an, the monthly utilization potential of AWV is shown in Table S5. If fully realized, these potentials could provide 527.75 million m³ of water resources annually for the urban population. The annual artificial water consumption in the urban area of Xi'an is 394.20 million m³, suggesting that the surface AWV could effectively address the water shortage issues in the study area. Given the concentrated rainy season in August, September, and October, collected rainfall runoff could supplement water resources during these months. In cases of insufficiency, cross-regional water transfer could be employed to achieve a 100% utilization rate of AWV. During the non-rainy season, cross-regional water transfer and sewage reuse could contribute to achieving a 100% utilization rate of AWV. The calculation method in Section 3.3.3 indicates that a flood disaster is likely when the rainfall of an event exceeds approximately 32.8 mm. The large increase in impervious surfaces in urban areas contributes to increased rainfall runoff, due to the failure of rainfall to rapidly infiltrate into the ground, and this can be mitigated by enhancing surface infiltration rates. Specific measures include the use of permeable paving materials, the creation of wetlands, and the construction of rainwater gardens, etc. Additionally, increasing the surface AWV can be achieved by building new lakes and reservoirs to collect and store more rainfall runoff, preventing floods.

Turning to the subsurface space in the urban area of Xi'an, Table 3 illustrates the monthly utilization potential of AWV. Notably, the monthly utilization potential of subsurface AWV exceeds 8,700 million m³, significantly higher than the utilization potential of the surface AWV. This also shows that the subsurface AWV can accommodate all rainfall infiltration. Despite this high potential, the ecological environment of subsurface water in the urban area of Xi'an is currently deteriorating. Therefore, the primary task is to supplement groundwater to achieve a healthy and sustainable development of the ecological environment of subsurface water. Specific measures to enhance this include using permeable materials, constructing new wetlands and rainwater gardens to increase rainfall infiltration, and directly recharging groundwater.

The frequent occurrence of flood disasters in the study area can be attributed to two primary factors. Firstly, there is an increase in extreme rainfall events due to global climate change. Secondly, urban construction and development contribute to the continuous increase of the impervious area of the urban surface. This insufficient surface space leads to an inability to accommodate the increasing rainfall runoff, resulting in flood disasters. The first factor, involving extreme rainfall events, is beyond artificial control. However, the second factor offers an opportunity to mitigate flood disasters by augmenting the AWV in the study area, allowing for better management of rainfall runoff. Specific measures include the expansion and construction of lakes and reservoirs in the surface space of the study area (Song 2022).

The study area faces a severe water shortage problem. Statistical data indicate that the amount of surface water resources available for human use and the annual artificial water consumption amount to 104.29 and 394.20 million m³, respectively. The available surface water falls short of meeting the demand for artificial water consumption. Full utilization of the surface AWV can address the water resources shortage problem in the study area, mitigating over-exploitation of groundwater and alleviating the strain on surface water supply to a certain extent (Liu et al. 2022a). This indicates that there are enough surface water resources in the study area which can be used by humans. The increased water resources in the surface AWV are sourced from three avenues: rainfall runoff, inter-regional water diversion, and reclaimed water reuse (Ba et al. 2023). Therefore, improving the utilization rate of the surface AWV offers a solution to the water storage challenge in the study area.

Based on AWV calculation for the three rainfall events, the error between the calculated runoff value and the measured value is less than 5%, indicating the reliability of the method. This suggests the feasibility of implementing a sponge city concept in Xi'an, corroborating the results obtained by Jia et al. (2023) and Liu et al. (2022b). Moreover, Li et al. (2023) demonstrated that urban waterlogging does not occur when rainfall is 17.4 mm, consistent with the results of our analysis for rainfall event 1. Therefore, the AWV method is deemed feasible and can provide guidance for SCC.

However, this study has some limitations. The proposed method overlooks the exchange between surface water and groundwater. The intricate processes of rainwater infiltration, movement in loess soil, and recharge and interchange with groundwater remain unclear. Further research should improve the method by integrating urban stormwater management models, such as the System for Urban Stormwater Treatment and Analysis Integration (SUSTAIN) Model and Stormwater Management Model (SWMM). In addition, potential errors arise from limited hydrological data, and the exact results are subject to further observation through in situ testing. Subsequent method improvement should consider the development of water resource databases and the incorporation of long-term climate change impacts into the model (Liu et al., 2022a).

This study introduces a novel theoretical framework for evaluating AWV based on water balance, water volume and EWD in urban watersheds. AWV serves as a guiding metric for SCC, encompassing the total dynamic water volume of urban watersheds capable of absorbing, storing, releasing, and utilizing the water resources by humans. The Xi'an urban area of China was selected as a case study to assess the SPC using this methodology. The results show monthly values for surface and subsurface AWV ranging between 53.06 and 53.98 million m³ and between 8,701.89 and 8,898.14 million m³, respectively, with peak values occurring in February and December. For the surface space of Xi'an urban area, the full utilization of the surface AWV potential would provide an annual supply of 527.75 million of water resources, surpassing the annual artificial water consumption of 394.20 million m³ and effectively addressing the water shortage problem. During the rainy season, with rainfall exceeding 32.8 mm, flooding is likely to occur. To this end, strategies such as the application of permeable materials, and the construction of wetlands, rainwater gardens, reservoirs, and lakes can be adopted to mitigate flooding conditions. Although the subsurface space of Xi'an urban area possesses substantial potential for subsurface AWV utilization, the ecological environment of the subsurface water is deteriorating. The primary focus should be directed at revitalizing the ecological environment of the subsurface water through measures such as the application of permeable materials, construction of new wetlands and rainwater gardens, and direct groundwater recharge. Conclusively, the theoretical framework of AWV presents a comprehensive approach to solving and mitigating water shortage and flood problems in the Xi'an urban area, rendering AWV a vital theory and calculation method with leading significance for the SCC.

This work was supported by the Scientific Research Program funded by the Education Department of Shaanxi Provincial Government (No.22JT022), the Project Supported by the Natural Science Basic Research Plan in Shaanxi Province of China (No. 2022JQ-391) and the New Style Think Tank of Shaanxi Universities. The authors also thank Dr Mawuli Dzakpasu for carefully proofreading the manuscript.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.