Stormwater Runoff Allocation and Basin-Wide Flood Mitigation Mechanism From the Perspectives of Efficiency and Equity

Abstract

Watershed flood management requires a systematic assessment of disaster-related risks and the application of resilience-oriented design to reduce exposure and vulnerability. However, how to enhance basin resilience by integrating land use planning for stormwater regulation, while balancing the interests of upstream, midstream, and downstream regions, remains a key challenge. To address this issue, this study constructs a basin-wide flood-mitigation resilience framework considering allocation efficiency and equity. First, a two-stage data envelopment analysis model is established to evaluate the economic efficiency and runoff management efficiency, forming an efficiency-oriented stormwater allocation method. Second, near–far telecoupling relationships are identified and a multi-regional input–output model is used to examine regional development imbalances, thereby developing an equity-oriented method. Finally, the allocation proportions are adjusted according to decision-makers' preferences. The results show that: 1) under the efficiency-oriented scheme, the upstream region is required to undertake 88.80% of stormwater runoff; 2) under the equity-oriented scheme, the downstream bears 78.25% of the rainstorm runoff; and 3) when the decision makers' preference is set to 0.2 (i.e., indicating greater emphasis on equity), inter-regional allocation disparities within the basin are minimized. This study responses to the challenges of spatial runoff allocation and cross-regional compensation, providing a practical and instructive approach for improving basin-wide flood-mitigation resilience.

Graphical abstract

Keywords

Flood Mitigation Resilience / Runoff Allocation / Stormwater Management / Efficiency / Equity / Basin-Wide Framework

Highlight

	· Develops a basin-wide flood-mitigation resilience framework that incorporates efficiency and equity
	· Clarifies the runoff compensation mechanism, enhancing the basin's overall flood prevention resilience
	· Addresses key technical challenges of spatial runoff allocation in water–land integrated basin planning

1 Introduction

In recent years, rapid urbanization has led to a sharp increase in the amount of impervious surfaces. Meanwhile, climate change has further intensified the severity of stormwater runoff in terms of its magnitude, intensity, and frequency^[1], resulting in an annual increase in flood disasters triggered by typhoons and heavy rainfall worldwide. As the impacts of stormwater runoff on cities and basins become more pronounced, enhancing the resilience of these systems to future environmental challenges has become ever more critical^[2–^3]. With the growing recognition of the resilience concept, the Rockefeller Foundation launched the "100 Resilient Cities" initiative^[4], while Europe and the USA subsequently advanced the notion of "Planning for Resilient Cities and Regions"^[5], aiming to strengthen cities' adaptive capacity to acute shocks and chronic stresses^[6]. Accordingly, resilient city building and basin-wide risk management urgently require systemic approaches to assessing disaster risks, coupled with design-oriented measures to reduce disaster exposure and lower the vulnerability of both cities and basins.

The flood risk induced by changes in stormwater runoff is characterized by significant uncertainty. Despite substantial investment, the structural efficiency and environmental impacts of traditional flood control projects are increasingly questioned. For example, the elevation of upstream levees may generate negative externalities for downstream areas. Since disaster losses often become apparent only ex post, controlling risks at the source has become a crucial pathway for reducing potential losses^[7].

At present, basin governance is shifting toward a resilience-oriented paradigm that emphasizes the coordination between hydraulic engineering and spatial planning. International practice in this field has exhibited multi-dimensional innovations. For example, the USA has adopted low impact development (LID) techniques to replace conventional pipe networks with small-scale water treatment facilities^[8]; Australia promotes water sensitive urban design (WSUD) to integrate urban water–land systems and to optimize flood detention^[9]; the sponge city program in Chinese mainland focuses on multi-scale stormwater management^[10] and the "runoff sharing–outflow control" strategy in Taiwan, China regulates land development and reserve flood storage space^[11]. Although these policy frameworks have established a dual framework compromising macro-scale total runoff control and micro-scale construction codes, two major gaps remain at the meso-scale within the cross-subcatchment runoff allocation mechanism^[12]: first, the lack of operational criteria to address distributive equity issues arising from unbalanced regional development; and second, the insufficient use of land-use planning to effectively balance disaster-prevention needs with development rights. In response to these challenges, an incremental stormwater runoff allocation mechanism can be established to realize in-situ risk control. By clearly defining runoff quotas for each subarea and setting up a compensation scheme for stormwater runoff management, this mechanism can provide quantitative guidance for engineering measures such as the construction of flood detention basins, thereby enhancing disaster-prevention resilience across the entire basin. Among these tasks, accurately determining the runoff quota for each subarea is not only a prerequisite for implementing stormwater management measures, but also a critical foundation for constructing a systemic flood control regime.

The core of establishing an incremental stormwater runoff allocation mechanism lies in integrating spatial planning and water resource management concepts to develop a resilience-oriented, water–land integrated risk prevention and control model. However, commonly applied approaches to water–land integrated runoff allocation still face several limitations. For example, the cost–benefit analysis (CBA) method is highly subjective in evaluation, requires the unification of indicator dimensions, and has difficulty in specifying concrete allocation schemes^[13]. Hydrological–hydrodynamic modelling methods are mainly employed to simulate the spatial distribution of stormwater runoff and to reveal rainfall–runoff processes and flood routing characteristics in river channels, but they overlook economic and social factors, which may lead to inequitable runoff allocation and insufficient incentives for governance^[14]. By contrast, the data envelopment analysis (DEA) method overcomes the shortcomings of CBA and can serve as an alternative tool to evaluate the efficiency of water–land integrated runoff allocation; the near–far coupling approach helps clarify land-use transformation relationships among different areas within a basin, thereby explaining inequities in stormwater runoff allocation; meanwhile, multi-regional input–output (MRIO) models can further quantify the coupling between urbanization and the ecological environment, providing a quantitative basis for runoff allocation^[15–^16].

Ideally, stormwater runoff should be allocated by identifying its underlying causes and then assigning flow volumes to the corresponding areas. However, its allocation and the assignment of governance responsibilities within a basin are often shaped by spatially uneven development, e.g., excessive urbanization in the middle and lower reaches heightens the demand for ecosystem services provided by upstream areas. Consequently, trade-offs between efficiency and equity are difficult to avoid in the allocation process^[17]. Given the complexity of basin governance, allocation approaches that emphasize only one aspect are overly reductive and hard to effectively balance the interests of all subareas across the entire basin.

Thus, this study aims to construct a basin-wide flood-mitigation resilience framework with quantifiable flood-mitigation targets by integrating allocation efficiency and equity considerations and incorporating decision-makers' preferences in adjusting allocation ratios. This framework addresses the issues of allocation and compensation in basin-wide flood mitigation.

2 Allocation Efficiency and Equity

Over the long term, economic development has relied heavily on intensive ecological resource inputs, which has not only placed tremendous pressure on resource supply but also caused severe environmental degradation. Existing studies have attempted to apply ecological efficiency to measure the degree of coordination between economic development and environmental protection, thereby alleviating the conflict between human development and the environment^[18]. This study seeks to employ the DEA model for calculating ecological efficiency to explore uneven development within the basin^[19]. Then, it takes the results as the basis for allocating stormwater runoff, aiming to enhance the basin's regulatory resilience to floods. Related research has applied DEA to integrate the economic and environmental governance dimensions of ecological efficiency. Specifically, economic efficiency refers to the economic output generated per unit of resource input in economic activities (e.g., land use, economic resource allocation) of a basin; whereas environmental governance efficiency denotes the effectiveness of environmental governance per unit of environmental resource consumption (e.g., occupation of ecological space, increase in runoff volume)^[20–^21]. Building on this approach, this study takes the economic production and runoff governance efficiency of a basin as the basis for runoff allocation (Fig. 1), according to factors such as land-use patterns, current economic conditions, expected runoff increments, and disaster losses in upstream and downstream sub-catchments. Areas exhibiting lower efficiency imply that resources have not been fully utilized and therefore possess greater potential to undertake flood-mitigation responsibilities.

However, allocation schemes based solely on efficiency tend to overlook distributive inequities arising from spatially uneven development. To achieve allocation equity, it is first necessary to clarify the mechanisms through which various elements within the basin interact with and to identify the root causes of inequity^[15]. For instance, the middle and lower reaches depend on upstream areas for the supply of water, food, forest products, and other resources, while upstream areas, by obtaining industrial goods and daily necessities produced in the middle and lower reaches, can alleviate the pressure on the exploitation of their own resources^[16]. Accordingly, this study employs MRIO–based telecoupling relationships to capture the direct and indirect environmental impacts induced by land resource consumption embedded in production processes^[22]. This uncovers intra-basin development imbalances and provides a basis for stormwater runoff allocation (Fig. 1). Under this equity-oriented allocation approach, the governance responsibility for stormwater runoff will not shift away from its area of generation: resource-consuming areas are required to compensate resource-supplying areas, so as to balance development needs and, in turn, promote equity in basin governance.

Overall, the efficiency-oriented allocation approach distributes runoff according to the actual efficiency of economic production and runoff management in each region, resulting in a shift of treatment responsibility away from the locations where the runoff is generated. In contrast, the equity-oriented allocation approach enables runoff to be handled locally through a compensation mechanism, without shifting the management responsibility away from the generating location.

3 Methodology

3.1 Study Area and Data Sources

According to the Comprehensive Governance Master Plan for the Dajia River Basin, the Dajia River Basin is located in the central–western part of Taiwan, China, with a drainage area of 1,244 km² and a main channel length of 124 km. Land development is primarily concentrated in the middle and lower reaches. The basin has an average channel gradient of 1/60 and is classified as a steep, high-energy river. The Dajia River Basin passes through 13 management units and, based on geographical conditions and governance boundaries, can be divided into upstream, midstream, and downstream (Fig. 2). The data used in this study were derived from the following sources: 1) temperature and precipitation data for 2018 obtained from the CODiS climate observation data inquiry service; 2) land use data for 2008 and 2018 taken from the Current Land Use Survey of Taiwan, China; 3) records of flood disaster losses compiled from the White Paper on Disaster Prevention and Protection (2015–2019); and 4) other socio-economic indicators (i.e., labor force, energy use, changes in investment, GDP, and governance financial inputs) extracted from the 2018 Statistical Bulletin of each unit.

3.2 Research Framework

To address flood risk prevention and control in water–land integrated basin planning, this study develops a basin-wide flood-mitigation resilience framework that jointly considers efficiency and equity (Fig. 3). The framework consists of three main components: 1) incremental stormwater runoff simulation, 2) allocation efficiency and fairness, and 3) decision-makers' preferences.

For incremental stormwater runoff simulation, temperature and precipitation data were first processed using a downscaling method (to a grid resolution of 5 km) in conjunction with different Representative Concentration Pathways (RCPs) proposed by the Intergovernmental Panel on Climate Change (IPCC), to obtain temperature and precipitation projections for the next 200 years. Second, by comparing land use changes between 2008 and 2018, the incremental stormwater runoff induced by urbanization was estimated. Subsequently, the combined incremental stormwater runoff resulting from both climate change and urbanization was used as the basis for the subsequent model analysis.

In terms of allocation efficiency and equity, an improved two-stage DEA model was first employed to link the basin's economic efficiency with its runoff management efficiency to derive a composite ecological efficiency. On this basis, an efficiency-oriented stormwater runoff allocation model was constructed, and its results were used to allocate the reduction in stormwater runoff. Then, after identifying near- and far-distance coupling relationships, the MRIO model was applied for quantitative analysis to obtain multi-regional inequity in ecological space. The land consumption patterns of each analytical unit generated from the MRIO analysis served as the compensation basis for promoting allocation equity.

Finally, drawing on the results of allocation efficiency and equity, the study weighed the runoff allocation volumes derived from different orientations from the perspective of decision-makers' preferences, and treated the resulting trade-offs as decision references. By further integrating the two sets of allocation results, the study clarified the attribution of runoff-management responsibilities and the associated compensation relationships, thereby informing an optimal allocation scheme for stormwater runoff that enhances regional flood-mitigation resilience.

3.3 Research Methods

3.3.1 Efficiency-Oriented Allocation Method

3.3.1.1 Two-Stage DEA Model

DEA model evaluates the relative efficiency among multiple decision-making units (DMUs), which in this study refer to the management units, under multiple inputs and outputs. This method has the advantage of being free from subjective weighting, is unaffected by correlations or multicollinearity among input and output variables. It is therefore suitable for comprehensive indicator evaluation and efficiency comparisons across decision-making units^[23]. In this study, the two-stage DEA model (Fig. 3) was implemented and solved using Lingo software.

In this two-stage DEA model, the first-stage economic-efficiency subsystem and the second-stage runoff management efficiency subsystem were each specified with their own sets of input and output indicators. The incremental stormwater runoff outputted in the first stage was taken as an input indicator in the second stage. The weight assigned to this intermediate unexpected output (i.e., the incremental stormwater runoff) is kept consistent across the two stages. Imposing a unified set of weights established an internal connection within the model, preventing the overall system efficiency from being decoupled from the efficiencies of the individual subsystems. Since the economic-efficiency subsystem constitutes a prerequisite for the whole model, its efficiency scores must be held constant when solving the runoff management-efficiency subsystem, and are incorporated as constraints in the latter. The output indicators in the second stage were used to measure flood-related losses, and the corresponding efficiency scores were negatively related to resilience: when the DEA-derived efficiency value is higher, the losses increase, reflecting relatively poorer runoff management performance; conversely, when the efficiency value is lower, disaster losses decrease, indicating that the DUM exhibits stronger resilience. In the two-stage DEA efficiency analysis, this study, for the first time, introduced a division model (see supplementary material for the detailed equations).

3.3.1.2 Stormwater Runoff Allocation

Building on the two-stage DEA model, this study developed a dynamic allocation mechanism under which the temporal linkage of production technology parameters (i.e., technical parameters related to resource conversion capacity and production processes) is explicitly specified, and the first-stage technology parameters are used to characterize the efficiency frontier of the second stage^[24]. Within the flood-resource allocation framework, this study assumed that the adjustment capacity of each management unit can only be adjusted through proportional scaling of its existing production level. Under the condition that unexpected inputs are freely disposable, a bi-objective optimization model was constructed: Objective 1 was to maximize desirable outputs (i.e., GDP growth, control of disaster losses), and Objective 2 was to minimize input variables (i.e., resource input, governance cost). The constraints included: 1) a total-balance constraint that ensured the full allocation of the total amount of flood reduction; and 2) unit-level equity constraints that prevented excessive allocation to specific units. A stepwise solution strategy was adopted, whereby the efficiency threshold associated with Objective 1 was achieved first, followed by the optimization of Objective 2. In this way, a resilience-oriented allocation scheme for each basin subregion was quantitatively derived (see supplementary material for the equation).

3.3.2 Equity-Oriented Allocation Method

3.3.2.1 Near–Far Coupling

Based on the resource services provided by land, this study constructed near–far coupling relationships (Fig. 4) and converted these resource-service linkages into land-consumption relationships^[25]. Near coupling refers to the consumption of various land resources within a management unit by its built-up areas; far coupling denotes the situation that when a management unit's land supply cannot meet local demand, it must obtain resource services from other units within the basin; and spillover effect describes cases where certain management units possess sufficient supplies of various resources and, beyond satisfying their own development needs, are able to export resources to external systems. Drawing on the per capita land use conditions in Taiwan, China, this study estimated the land consumption status of each management unit within the Dajia River Basin.

3.3.2.2 MRIO Model

On the basis of clarifying land-consumption relationships within the basin, this study employed the MRIO model for quantitative analysis. The MRIO model can capture the economic interdependencies among sectors and regions^[26], and the main advantage lies in its ability to trace the direct and indirect environmental impacts caused by the final consumption of goods and services. The specific formulation is given in Eq. (1):

(1)\mathrm{\Delta }L=\sum _{r\ne s}{L}^{rs}-\sum _{r\ne s}{L}^{sr},

where ∑_r≠sL^rs denotes the amount of land required in region r to satisfy the consumption of other regions, i.e., the outflow, and ∑_r≠sL^sr denotes the inflow. When △L > 0, region r is characterized by a net outflow of land resources, meaning that it must provide services for other regions and bear the transferred stormwater runoff burden. Conversely, when △L < 0, region r experiences a net inflow of land resources, indicating that it depends on other regions to support its development and should therefore assume compensatory responsibilities toward those regions. In this way, the MRIO model clarifies the runoff management responsibilities of each management unit (including both local runoff and the portion associated with external compensation), which are not reallocated among management units. This helps reduce the downstream flood management pressure and supports the enhancement of basin-wide flood-mitigation resilience under the equity-oriented allocation scheme.

3.3.3 Allocation Method Incorporating Decision-Makers' Preferences

Efficiency-oriented allocation generally places greater emphasis on market-based mechanisms, whereas equity-oriented allocation reflects a stronger role of government intervention. To balance the two, this study constructed an allocation model from the perspective of decision-makers' preferences. The overall runoff allocation in the Dajia River Basin is expressed as follows:

(2)f\left(U_{x1},U_{x2}\right)=\sum \left({\alpha }_1{U}_{x1}+{\alpha }_2{U}_{x2}\right),x=1,2,\dots ,n,

where α₁ and α₂ represent the weights assigned to allocation equity and allocation efficiency, respectively, with α₂ also serving as the decision-makers' preference parameter in this study, while U_x1 and U_x2 denote the allocation volumes derived from the equity-oriented and efficiency-oriented methods, respectively.

On the left-hand side of Fig. 5 is the utility function of decision-makers' preferences, where α₁/α₂ represents the slope of the utility curve. When α₁ < α₂ (e.g., Point B), allocation efficiency dominates; in this case, the allocation outcome may promote the development of a given region at the expense of others, thereby leading to inequitable allocation. When α₁ > α₂ (e.g., Point C), the government strengthens equity in runoff allocation, but this may come at the cost of overall basin development and result in reduced efficiency. When α₁ = α₂ (i.e., Point A), allocation efficiency and allocation equity are considered equally important. In addition, the closer the allocation point lies to α₁ = 0.5, the smaller the divergence in runoff allocation. By adopting different settings of decision-makers' preferences, effective technical support can be provided for constructing resilient basins and cities.

4 Empirical Analysis

4.1 Quantifying Incremental Stormwater Runoff

Drawing on existing research^[15], this study employed the TCCIP-AR5 model developed by the Disaster Prevention and Protection Technology Center in Taiwan, China to simulate the mean daily precipitation over the next 200 years (baseline year 2018) under different development scenarios, including RCP2.6 (mitigation scenario), RCP4.5 and RCP6.0 (stabilization scenarios), and RCP8.5 (high-emission scenario), to obtain the mean change in stormwater runoff. Subsequently, based on land use changes between 2008 and 2018, the incremental stormwater runoff induced by urbanization is further calculated^[28]. The calculation is given by:

(3)Q=\left(A_1-A_2\right)\times q\times (55\mathrm{\%}-10\mathrm{\%}),

where Q denotes the total increase in stormwater runoff, A₁ and A₂ represent the imperviousness before and after land use change, respectively, and q denotes the runoff volume per hectare of land under natural conditions. The simulated results of incremental stormwater runoff (Table 1) show that climate change is the dominant factor driving its increase. Among the upstream areas, Unit 1, Unit 12, and Unit 13 exhibit the largest increments, whereas the midstream and downstream areas of Unit 2 and Unit 11 show relatively slighter increases. If the runoff generated in the upstream areas is not effectively managed, it is likely to concentrate in the middle and lower reaches along topographic gradients, thereby exacerbating the overall difficulty of runoff management.

4.2 Equity-Oriented Runoff Allocation and Compensation Mechanism

Based on the near–far coupling and MRIO analysis (Tables 2, 3), this study finds that the total land-service outflow provided by the Dajia River Basin to external systems amounts to 246,504 hm². Under the equity-oriented allocation assumption, the proportional shares of runoff management responsibility differ across river sections: the upstream region undertakes only 1.39%, the midstream region 4.88%, while the downstream region bears the highest share, reaching 23.44%. Overall, the Dajia River Basin functions as a net provider of land services to external systems; accordingly, external systems should compensate 70.29% of the incremental runoff, with this portion to be coordinated by the higher-level or central governments. The remaining share is allocated within the basin according to each management unit's local responsibility and its proportion of land inflows.

4.3 Efficiency-Oriented Runoff Allocation

The DEA efficiency analysis (Table 4) shows that the mean economic efficiency is 0.890, indicating relatively high resource utilization efficiency within the basin. The mean runoff management efficiency is relatively low at 0.61, implying a low overall natural disaster risk and a comparatively safe state in the basin. The average values of composite ecological efficiency in the midstream and downstream regions are markedly higher than that in the upstream region, exhibiting an overall pattern of "midstream > downstream > upstream." Assuming that 29.7% of the incremental stormwater runoff needs to be allocated in the second stage of the DEA model (based on the preceding equity-oriented analysis), Unit 12 bears the largest share of runoff treatment, followed by Unit 1 and Unit 13. These results indicate that the management units differ in terms of economic efficiency, runoff management efficiency, and composite ecological efficiency, and therefore exhibit heterogeneous capacities in stormwater-runoff allocation.

From the perspective of the stormwater runoff allocation results (Table 5), there are pronounced differences in runoff management shares among the regions. The upstream region undertakes the primary management tasks, whereas the downstream region receives the smallest allocation. The main reason is that the upstream region exhibits a relatively higher efficiency value of the DEA stage 2 and greater disaster losses, indicating low performance in runoff management and more residual runoff. Consequently, it shows a higher potential for flood mitigation. Meanwhile, the incremental stormwater runoff is predominantly concentrated in the upstream region, making it the key zone for absorbing and regulating runoff.

4.4 Decision-Makers' Preferences

According to the analysis results (Table 6), the upstream region bears the majority of the efficiency-oriented share of stormwater runoff, whereas the downstream region bears the majority of the equity-oriented share. The marked differences between these two allocation methods across upstream and downstream regions indicate that focusing on a single objective is insufficient for balancing the interests of all management units within the basin. As shown in Fig. 6, when the decision-makers' preference parameter α₂ ranges from 0.2 to 0.3 (i.e., when equity is given greater emphasis), the allocation disparities among management units are minimized, enabling overall optimal allocation for the entire basin. At the same time, aside from a slight decline in the standard deviation when α₂ lies between 0.1 and 0.2, the standard deviation increases with the rising weight on allocation efficiency in all other cases, implying that an excessive emphasis on efficiency exacerbates intra-basin inequity. Specifically, as α₂ increases, the runoff volume undertaken by Unit 1 and Unit 12 rises markedly, whereas that of Unit 3 and Unit 4 correspondingly decreases. Therefore, decision-makers need to flexibly adjust their strategies in light of the basin's actual development needs and the interests of different stakeholders. For example, under periods of high flood risk, efficiency-oriented allocation should be prioritized to rapidly mitigate flood peaks; during phases where ecological protection is prioritized, emphasizing equity-oriented allocation is more appropriate to safeguard upstream ecosystems.

5 Conclusions and Prospects

Enhancing flood-mitigation resilience is an important strategy for coping with flood risk; at its core, it aims to balance disaster-prevention capacity across regions and promote sustainable development through systematic planning and optimized allocation. Integrating the perspectives of allocation efficiency and equity, this study develops a basin-wide flood-mitigation resilience framework, to address the problem of allocating incremental stormwater runoff among upstream, midstream, and downstream regions. The framework not only clarifies runoff compensation mechanisms, but also provides an operational and quantitative basis for flood risk management, thereby contributing to the formation of more adaptive and sustainable resilience-oriented models for urban and basin governance. Considering both the efficiency and equity, this study proposes a total runoff volume control and optimized allocation method for future stormwater risk prevention. This method fills the gap in the quantitative allocation of runoff at the meso-scale and provides important support for subsequent floodplain zoning, the design of conveyance capacity for flood-control infrastructures, and the allocation of stormwater management responsibilities across various land use zones.

The conclusions are as follows: 1) Under different allocation methods, the allocation of stormwater runoff increment varies. Under the efficiency-oriented scheme, the upstream region is required to undertake 88.80% of the incremental runoff, whereas under the equity-oriented scheme, the upstream region only needs to undertake 1.39% of the total runoff volume. 2) When the decision-makers' preference parameter is set to 0.2, the standard deviation within the basin is minimized, indicating the smallest inter-regional development disparity. Overall, basin hydrological processes are complex and dynamic, and urbanization introduces additional uncertainties into the allocation of stormwater runoff. Governments and relevant institutions should strengthen the regulation of stormwater-runoff management and enhance the coordination of spatial planning. These measures play a crucial role in improving urban resilience and the capacity to cope with flooding.

The following areas could be explored in future research. First, while this study has addressed the equity issue of "who should compensate," it has not yet resolved the operational issue of "how this should be done." Follow-up research could establish a compensation negotiation mechanism among management units based on natural basin units, and translate the quantified ecological responsibility of the MRIO model into actionable fiscal transfer payments or ecological credit quotas. Second, although this study quantifies runoff allocation at the meso-scale, a substantial gap remains between this and the precise estimation of the volume to be managed. Future work could develop a dynamic resilience-based allocation model that couples climate change scenarios with land use evolution forecasts, thereby enhancing the accuracy and forward-looking nature of runoff allocation schemes. Third, subsequent research may also explore pathways for stormwater utilization. Through engineering measures such as flood detention wetlands and groundwater recharge, runoff could be converted into usable water resources, thereby improving the marginal benefits of management interventions.

References

[1] Silvis, V. G. , & Nicholas, R. E. (2025) Climate change and flooding in the Susquehanna River watershed: An exploratory analysis of county hazard mitigation plans. International Journal of Disaster Risk Reduction, 116, 105143.

[2] Miralles-Wilhelm, F. , Matthews, J. H. , Karres, N. , Abell, R. , Dalton, J. , Kang, S.-T. , … & van Wesenbeeck, B. (2023) Emerging themes and future directions in watershed resilience research. Water Security, 18, 100132.

[3] Rentachintala, L. R. N. P. , Reddy, M. M. , & Mohapatra, P. K. (2022) Urban stormwater management for sustainable and resilient measures and practices: A review. Water Science and Technology, 85 ( 4), 1120– 1140.

[4] Rockefeller Foundation. (n.d.) 100 Resilient Cities.

[5] Zhang, X. , & Li, H. (2018) Urban resilience and urban sustainability: What we know and what do not know?. Cities, 72, 141– 148.

[6] Woldearegay, K. , Grum, B. , Hessel, R. , van Steenbergen, F. , Fleskens, L. , Yazew, E. , … & Haftu, M. (2024) Watershed management, groundwater recharge and drought resilience: An integrated approach to adapt to rainfall variability in northern Ethiopia. International Soil and Water Conservation Research, 12 ( 3), 663– 683.

[7] Gao, G. , Li, J. , Feng, P. , Liu, J. , & Wang, Y. (2024) Impacts of climate and land-use change on flood events with different return periods in a mountainous watershed of North China. Journal of Hydrology: Regional Studies, 55, 101943.

[8] Ben-Daoud, A. , Ben-Daoud, M. , Moroșanu, G. A. , & M'Rabet, S. (2022) The use of low impact development technologies in the attenuation of flood flows in an urban area: Settat city (Morocco) as a case. Environmental Challenges, 6, 100403.

[9] Wu, W. , Jamali, B. , Zhang, K. , Marshall, L. , & Deletic, A. (2023) Water sensitive urban design (WSUD) spatial prioritisation through global sensitivity analysis for effective urban pluvial flood mitigation. Water Research, 235, 119888.

[10] Yu, K. (2019) Sponge City—Philosophy and methodology. Construction Science and Technology, ( Z1), 10– 11.

[11] Water Resource Agency in Taiwan, China. (2018). Technique specifications for runoff allocation and outflow control.

[12] Wang, H.-W. , Castro, D. S. C. , & Chen, G.-W. (2024) Managing residual flood risk: Lessons learned from experiences in Taiwan. Progress in Disaster Science, 23, 100337.

[13] Islam, A. R. M. T. , Al Mamun, A. , Rahman, M. N. , Akter, M. Y. , Chisty, M. A. , Alam, G. M. M. , … & Sohel, M. S. (2024) A cost-benefit analysis of flood early warning system: Evidence from lower Brahmaputra River Basin, Bangladesh. International Journal of Disaster Risk Reduction, 104, 104380.

[14] Anand, J. , Gosain, A. K. , & Khosa, R. (2024) Hydrological and hydrodynamic modelling for flood management: A case study of the Yamuna River Basin in Delhi. Journal of Hydrology: Regional Studies, 56, 101960.

[15] Su, Q. , & Chang, H.-S. (2021) Exploring the spatial development of watersheds and the allocation of responsibility for stormwater runoff from the perspective of ecological efficiency based on the DEA method. Natural Hazards Review, 22 ( 4), 04021041.

[16] Liu, J. T. , Li, Y. P. , Huang, G. H. , Wang, S. G. , & Wang, P. P. (2024) A multi-perspective input-output model for the energy-water nexus in Kazakhstan. Sustainable Production and Consumption, 49, 115– 129.

[17] Huang, X. , Shen, J. , Li, S. , Chi, C. , Guo, P. , & Hu, P. (2024) Sustainable flood control strategies under extreme rainfall: Allocation of flood drainage rights in the middle and lower reaches of the yellow river based on a new decision-making framework. Journal of Environmental Management, 367, 122020.

[18] Xing, Z. , Wang, J. , & Zhang, J. (2018) Total-factor ecological efficiency and productivity in Yangtze River Economic Belt, China: A non-parametric distance function approach. Journal of Cleaner Production, 200, 844– 857.

[19] Guo, Z. , Tang, S. , Wang, D. , & Zhang, J. (2022) Spatial distribution and factors influencing ecological efficiency of the Yellow River Basin in China. Frontiers in Environmental Science, 10, 815898.

[20] Zhu, W. , Duan, C. , & Chen, B. (2024) Energy efficiency assessment of wastewater treatment plants in China based on multiregional input–output analysis and data envelopment analysis. Applied Energy, 356, 122462.

[21] Liu, X. , Wang, X. , Lu, F. , Liu, S. , & Chen, K. (2022) Evaluation of the governance efficiency of water environmental governance efficiency in Yangtze River Delta from the perspective of multivariate synergies. International Journal of Environmental Research and Public Health, 19 ( 4), 2347.

[22] Niu, C. , Wang, X. , Chang, J. , Wang, Y. , Guo, A. , Ye, X. , Wang, Q. , & Li, Z. (2023) Integrated model for optimal scheduling and allocation of water resources considering fairness and efficiency: A case study of the Yellow River Basin. Journal of Hydrology, 626, 130236.

[23] de Oliveira, M. S. , Steffen, V. , de Francisco, A. C. , & Trojan, F. (2023) Integrated data envelopment analysis, multi-criteria decision making, and cluster analysis methods: Trends and perspectives. Decision Analytics Journal, 8, 100271.

[24] Zhu, Q. , Wu, J. , Li, X. , & Xiong, B. (2017) China's regional natural resource allocation and utilization: A DEA-based approach in a big data environment. Journal of Cleaner Production, 142, 809– 818.

[25] Manning, N. , Li, Y. , & Liu, J. (2023) Broader applicability of the metacoupling framework than Tobler's first law of geography for global sustainability: A systematic review. Geography and Sustainability, 4 ( 1), 6– 18.

[26] Yu, Y. , Feng, K. , Hubacek, K. , & Sun, L. (2016) Global implications of China's future food consumption. Journal of Industrial Ecology, 20 ( 3), 593– 602.

[27] Liu, G. , Wei, L. , Gu, J. , Zhou, T. , & Liu, Y. (2020) Benefit distribution in urban renewal from the perspectives of efficiency and fairness: A game theoretical model and the government's role in China. Cities, 96, 102422.

[28] Liao, C.-H. (2016). The spatial patterns and influenced factors of urban impervious surface ratio [Doctoral dissertation]. Cheng Kung University.

RIGHTS & PERMISSIONS

© Higher Education Press 2025