The impacts of the wetness–dryness encountering of the upstream inflow and downstream salt intrusion on water supply in coastal urban agglomeration

Coastal urban agglomerations host a dense population, have developed economies, and usually are the economic growth poles of a nation or region (Ren et al., 2019; Wang et al., 2021). A steady and sustainable resource supply is the prerequisite of urban agglomeration development, with water resources being one of the strategic natural resources and foundational for socio-economic development. A safe water supply is vital for the domestic, productive, and ecological water demands of urban agglomerations (Wang et al., 2020a, 2020b; Saucedo-Ramirez et al., 2022). However, due to climate change and human-environmental system dynamics, the water supply for coastal urban agglomerations is challenged by the uncertainty of available water resources and the often-ineffective layout of water infrastructure (Ashofteh et al., 2017; Pienaar & Hughes, 2017; Saharwardi & Kumar, 2021). Coastal urban agglomerations are usually located downstream of large rivers, where the river meets the coastal region. The upstream inflows and downstream salt tides are the two chief factors influencing the available water resources within these regions (Gong et al., 2022), and the decrease of upstream runoff is also one of the causes of salt tides (Munia et al., 2018; Tang et al., 2020). Therefore, incorporating changes in upstream inflows and salt intrusion into water resource allocation models is crucial for developing effective and reasonable water supply schemes. This is particularly important given the great uncertainty of hydro-climate change on coastal urban agglomerations.

Allocation of water resources can be conducted at various spatial scales (e.g., basin, region, and city) and several simulation and modeling technologies have been proposed to obtain the optimal water supply scheme at various time scales (e.g., daily, monthly, and yearly). However, modeling of water allocation within coastal urban agglomerations considered with saltwater intrusion is still not sufficient. In addition, traditional water allocation models usually choose long time series for the runoff input and water supply schemes are chosen and analyzed under several frequencies (e.g., 97, 90, and 50%) (Fu et al., 2018; Li et al., 2018; He et al., 2019). Spatiotemporal variation of runoff is being intensified as the compound impacts of climate change and human activities continue to advance. Analysis of the wetness–dryness encountering of runoff in different rivers is a good alternative to promote a more comprehensive understanding of runoff characteristics within a basin and to help refine the input of available water resources for water resource allocation models. The copula function is widely used in the analysis of wetness–dryness encountering of runoff (Wang et al., 2022), and compared to other multivariate hydrological analysis methods (e.g., joint distribution, multiple normal distribution, and non-parametric methods), has the advantage of having the same adaptability to variables with different distribution characteristics (Liu et al., 2015, 2020a, 2020b; Fan et al., 2018). The multiple copula function can better portray the randomness and dependence of hydrological variables and simplifies the joint probability problem in multidimensional analysis. For example, it has been used to construct two- and three-dimensional runoff combinations of rivers and then to estimate the wetness–dryness encountering of the runoff as an input for the water allocation model (Chen et al., 2019; Wu et al., 2020).

The water evaluation and planning (WEAP) model is one of the most useful water modeling systems and has the advantage of analyzing and simulating different water systems for policy analysis of water resource systems at the regional- or basin-scale (Li et al., 2015). WEAP can provide an integrated water resource planning framework for water management under various scenarios and has been successfully applied to evaluate the sustainability of limited water resource management strategies in coastal zones (Kou et al., 2018). The Pearl River Delta is one of China's three largest coastal urban agglomerations with a dense population, developed economy, and complicated water networks. The uneven distribution of water resources and high dependence on upstream runoff poses a great challenge to water supply security within the Pearl River Delta. In addition, the water supply within this region is threatened by salt intrusion downstream during the dry period (Yao et al., 2016; Payo-Payo et al., 2022). Given these factors, it is necessary to work out strategic water supply schemes of the Pearl River Delta that consider the wetness–dryness encountering of the upstream inflow and downstream salt intrusion, for guaranteeing water supply security.

The current study aims to explore the impacts of wetness–dryness encountering of upstream inflow and downstream salt intrusion on the water supply scheme of the Pearl River Delta via a WEAP-based water allocation model. Different from the previous studies, the input of available water resources for the water allocation model is deduced from wetness–dryness encountering instead of long time series for the runoff, and the critical upstream runoff for saltwater suppression, reflecting downstream salt intrusion, is estimated and taken as the river flow requirement in the model. The objectives of this study are to: (1) develop a WEAP based water allocation model for the Pearl River Delta considering the wetness–dryness encountering of upstream inflow quantified by the multiple copula function and downstream salt intrusion; (2) work out reasonable water supply schemes of the Pearl River Delta under different scenarios of upstream inflow and downstream salt intrusion; and (3) reveal the response of the water supply scheme of the Pearl River Delta to the wetness–dryness encountering of the upstream inflow and downstream salt intrusion and based on which the corresponding policy recommendations are put forwards for the local government to reduce the water shortage risk of the Pearl River Delta in dry periods. This study has the potential to provide policy structure and solutions for water resource planning within coastal urban agglomerations in a changing environment.

Dependence between random variables can be modeled by the C function, while several functions such as NORM, POISS, EV, GEV, GAMMA, WBL, and GP are chosen to describe the marginal distributions of the random variables. Copulas have to be chosen and the dependence under the given marginal distributions is compared, to derive better multivariate joint distributions. Well-known copulas are the Gaussian copula, t-Copula, and Archimedean copulas including the Clayton Copula, Frank Copula, and the Gumbel Copula.

In the current study, the multidimensional copula functions were compared and chosen for the analysis of the wetness–dryness encountering of the monthly runoffs of the East River, North River, and West River. Parameters in the marginal distributions of the runoffs of the East River, North River, and West River were estimated by the Maximum Likelihood Method (MLE) and a goodness of fit test for the copulas was conducted by examining the Root Mean Square Error (RMSE), and Akaike Information Criterion (AIC).

Consequently, there are 125 groups of three-dimensional wetness–dryness combinations for the monthly runoff of the East River, North River, and West River. The three-dimensional wetness–dryness combinations are numbered sequentially to facilitate the distinction (Table 1).

The application of WEAP involves several steps. First, the time-span, spatial boundaries, system components, and problem structure are set. Simulations of the status quo benchmark can then be conducted, which provide the present-day water demand, water resources, and supply of the system. Alternative simulations of ‘future assumptions’ can subsequently be carried out, which are based on policies, costs, technological advances, and other factors that affect demand, supply, and hydrology. The water supply scheme is based on a set of alternative assumptions or policies. Finally, the adequacy of water and its sensitivity to the uncertainty of key variables are evaluated.

Two user-defined priority systems are used to determine the distribution priority from the water source to the demand site and the catchment basin (for irrigation purposes), for in-channel flow requirements and reservoir storage. The priorities range from 1 to 99, where 1 is the highest priority and 99 is the lowest.

Establishing a ‘status quo baseline’ requires users to ‘calibrate’ system data and assumptions to accurately reflect the actual operation of the system. The ‘status quo baseline’ includes monthly supply and demand data (including reservoirs, pipe networks, treatment plants, pollution generated, etc.) for the first year of the study. Scenario analysis is the core of WEAP modeling. A preplan is a self-consistent description of how the future system will evolve under a specific socio-economic background and a set of specific policy and technical conditions. Using WEAP, plans can be generated and compared to evaluate the water requirements, costs, and environmental impacts of different plans. The plan includes any factors that can change over time, including factors that may change due to specific policy interventions and factors that reflect different socio-economic assumptions. A sensitivity analysis can also be performed by varying the values of uncertainties within the range of possibilities and comparing the results.

In the current study, the upstream runoff was from the monthly runoffs of the West River, North River, and East River under the wetness–dryness encountering. The 5 groups of dryness combinations with higher joint probabilities were selected among the 125 groups of three-dimensional wetness–dryness combinations for the runoff of the West River, North River, and East River. The 5 groups of three-dimensional wetness–dryness combinations of the runoff were combined with high and low levels of the critical upstream runoff for saltwater suppression (i.e., high level of the critical upstream runoff is combined in scenarios 1, 3, 5, and 7 and the low level of the critical upstream runoff is combined in scenarios 2, 4, 6, and 8) and formed the 10 scenarios. With these scenarios, the water supply scheme under different combinations of upstream runoff and levels of critical upstream runoff for saltwater suppression were calculated with the WEAP model. This modeling enabled the future changing satisfactory rate of water demand within the Pearl River Delta to be detected and revealed the impact of the wetness–dryness encountering of the runoff of the West River, North River, and East River on the water supply scheme of the Pearl River Delta.

Equations (6)–(8) show the relationship between the upstream flow (⁠⁠) of West River and the daily excessive chlorinity hours (⁠⁠, and ⁠) of Pinggang, Guangchang, and Zhuzhoutou pumping stations, respectively.

It was found that as the upstream flow increased, the excessive chlorinity hours of the pumping station gradually decreased and trended toward zero. Moreover, the relationship of Q and the daily excessive chlorinity hours varied according to the different geographical locations of the pumping stations.

To guarantee the water supply security of the downstream cities in the Pearl River Delta, the daily excessive chlorinity at the pumping station should be less than 18 h. Therefore, the average monthly streamflow of Makou station should be equal to or more than 1,650 m³/s for Pinggang, 2,500 m³/s for Guangchang, and 1,400 m³/s for Zhuzhoutou, according to Equations (6)–(8). Three levels (i.e., high, medium, and low) of the critical upstream runoff for saltwater suppression were defined according to the distances between the three pumping stations and the Pearl River estuary. High critical upstream runoff (2,500 m³/s) guarantees water withdrawals for all three pumping stations; medium critical upstream runoff (1,650 m³/s) guarantees water withdrawals for pumping stations at Pinggang and the upstream stationsand low critical upstream runoff (1,400 m³/s) guarantees water withdrawals for pumping stations at Zhuzhoutou and stations upstream.

Scenarios 1 and 3 had the same critical upstream runoff for saltwater suppression, while the combination of the upstream runoff from West River, North River, and East River differed from ‘extreme dryness-extreme dryness-extreme dryness’, respectively, in scenario 1 to ‘extreme dryness-more dryness-more dryness’ in scenario 3. From scenarios 1 to 3, the monthly water demand satisfaction rates of Guangzhou, Shenzhen, Dongguan, Huizhou, and Foshan, whose water supplies are largely impacted by the upstream runoff from the North River and East River, increased by 7, 24, 34, 6, and 24% in the dry period in 2020, respectively, an average increase of 19%. The monthly water demand satisfaction rates of the cities whose water supply is largely impacted by the upstream runoff from West River, however, did not change with the change in the wetness–dryness combination of the upstream runoff of the North River and East River, because their water supplements are mainly from West River and the local reservoirs. This was similar to the differences between scenarios 2 and 4. When the combination of the upstream runoff transferred from ‘extreme dryness-extreme dryness-extreme dryness’ in scenario 1 to ‘more dryness-extreme dryness-extreme dryness’ in scenario 5, the monthly water demand satisfaction rates of Zhuhai, Zhongshan, Jiangmen, and Zhaoqing, whose water supplies are largely impacted by the upstream runoff from the West River, increased by 24, 15, 15, and 22% in the dry period in 2020, respectively, an average increase of 19%. The monthly water demand satisfaction rate of Guangzhou was also influenced by the wetness–dryness combination of the upstream runoff of the West River and exhibited an average increase of 3%, due to the West River water transfer project and the Nanzhou water withdrawal project. This is similar to scenarios 2 and 6. The combination of the upstream runoff transferred from ‘extreme dryness-extreme dryness-extreme dryness’ in scenario 1 to ‘more dryness-more dryness-more dryness’ in scenario 7, and the monthly water demand satisfaction rates of the Pearl River Delta increased by 9% in Guangzhou, 24% in Shenzhen, 24% in Zhuhai, 33% in Dongguan, 6% in Huizhou, 31% in Zhongshan, 31% in Foshan, 15% in Jiangmen, and 22% in Zhaoqing, respectively, with an average increase of 22%, which is similar to scenarios 2 and 8.

The monthly water demand satisfaction rates of the cities in 2030 were not quite similar to those in 2020. First, different scenarios for the wetness–dryness combination of the upstream runoff impacted the water supplies and water demand satisfactions, which is like 2020. Second, due to the different development plans of each city, the water demand of each city would increase or decrease, respectively, compared to 2020, thus affecting the water demand satisfaction rates (Table 2). Third, Shenzhen and Dongguan will receive water supply from West River through the Pearl River Delta water division engineering in 2030, while Guangzhou will receive more water supply from West River in 2030. For example, compared to 2020, the monthly water demand satisfaction rate of Shenzhen in scenarios 1 and 5 was increased by 29–33%, in Dongguan by 34–35%, in Guangzhou by 2–1%, and in Huizhou by 3–4%. While the satisfaction in Jiangmen decreased from 4 to 2%, and in Zhaoqing from 8 to 5%. The results showed the influence by water demand and water division engineering. From 2020 to 2030, Shenzhen and Dongguan decreased their water demand and water division engineering began to supply water, thus their water demand satisfaction rates were greatly increased. On the contrary, because of the increased water demand in Jiangmen and Zhaoqing (Table 2), the water demand satisfaction rate would decrease in the same wetness–dryness combination scenario and monthly runoff. The water demand was also increased in Huizhou, and Huizhou did not get the water supply from water division engineering. However, due to the water division engineering from West River to Shenzhen and Dongguan, the pressure of water supply in ER was eased; therefore, the satisfaction in Huizhou has also been increased. This result has shown the regional impact of water division engineering on the water demand satisfaction rates of the cities in the Pearl River Delta.

As shown in Figure 12, the water transfer projects provided water supply, easing pressure during the dry period. However, as the runoff of the West River accounts for the largest proportion of the upstream runoff of the Pearl River Delta (Figure 8), the impact of the water transfer project on the water supply of the Pearl River Delta is weakened when there is an extreme dryness event impacting the upstream runoff of the West River. Changes in the wetness–dryness combination of the upstream runoffs from the West River, North River, and East River demonstrate that the upstream runoffs have a significant impact on the water demand satisfaction rates of the Pearl River Delta. The upstream runoffs of the West River, North River, and East River cannot be fully utilized in the flood period, even if the city reservoirs have the largest storage capacity before the dry period, water shortages will occur when there is extreme/more dryness of events impacting the upstream runoff, and the water demand satisfaction rate of the Pearl River Delta will decrease month by month in the dry period as the reservoir storage capacity cannot be effectively replenished. In particular, the water withdrawal of the cities along the river is directly influenced by the upstream runoff, and a reduction in the upstream runoff would cause a significant decrease in the water demand satisfaction rates. Our results demonstrate that water transfer projects are necessary to reallocate water resources in regions like the Pearl River Delta, in which the water supply is dominated by the upstream runoff and its wetness–dryness combination.

As shown in Table 5, the average saltwater suppression passing rate in the dry period is only 75.8% under extreme dryness and 85% under more dryness of the upstream runoff of the West River. Furthermore, the saltwater suppression passing rate is only 60–70% in January and February, suggesting that much of the critical upstream runoff for saltwater suppression cannot be satisfied when there is an ‘extreme’ or ‘more dryness’ event affecting the upstream runoff of the West River.

In the scenarios with a high level of critical upstream runoff for saltwater suppression, even if the average saltwater suppression passing rate is 70–80%, the monthly water demand satisfaction rates of Zhuhai, Zhongshan, and Jiangmen are largely impacted. For instance, comparing the scenarios with low-level critical upstream runoff (i.e., scenarios 2, 4, 6, 8, and 10) to their paired scenarios with high-level critical upstream runoff (i.e., scenarios 1, 3, 5, 7, and 9), the monthly water demand satisfaction rate of Zhuhai in 2020 decreased by 23, 23, 2, 2, and 2%, respectively. In Zhongshan, the water demand satisfaction rates decreased by 23, 23, 2, 2, and 2%, respectively, and for Jiangmen they decreased by 21, 23, 13, 15, and 15%, respectively. However, the impact of the critical upstream runoff for saltwater suppression on the monthly water demand satisfaction rate could be alleviated as the upstream runoff increases.

The current study proposes further implications for water resource management in the coastal urban agglomeration of the Pearl River Delta. The higher guarantee of water supply security is one of the important prerequisites for the sustainable development of the region, and water shortages will affect human life and normal social development, especially in the Pearl River Delta, which has a growing population and rapidly developing economics. Although uncertainty about water resources increases under changing conditions, the risks and stresses of water scarcity can be minimized by taking anthropogenic measures.

In an extreme dryness year, the reservoirs’ water storage decreased, and the impact of the water transfer project on the water supply of the Pearl River Delta was weakened, mainly because of the significant decrease of the upstream inflow. New water transfer projects require long-term construction and involve administrative communication issues across basins and regions. Therefore, the water shortage problem in the Pearl River Delta during the extreme dryness period cannot be properly solved by the construction of more new water transfer projects in a short period. Due to the shortage of land resources in the Pearl River Delta, it is also unrealistic to build more reservoirs in the short term (Hu et al., 2019). However, the capacity of the existing reservoirs can be expanded as much as possible to store more water during flood periods and supply more water resources during extremely dry years. In addition, the most effective management strategy in the near future should be to strengthen the rational allocation of water resources between the upstream and downstream of the Pearl River Basin by administrative methods, so as to supply more water for the Pearl River Delta during dry periods (Huang & Zhang, 2020; Liu et al., 2020a, 2020b; Ren et al., 2021). Moreover, cities at the Pearl River Delta estuary should accelerate the construction of emergency backup water source projects, so as to gradually form the strategic reserve system of water resources and reduce the water shortage risk, particularly in an extreme dryness year.

In an extreme or more dryness event, the critical upstream runoff for saltwater suppression could not be satisfied. Under the premise of a serious reduction of upstream runoff, for satisfying the rigid water demand in coastal cities during the saltwater intrusion, it is necessary for water resources management departments to build effective saltwater intrusion forecasting systems (Lu et al., 2021), and draw as much fresh water as possible into reservoir storage before the saltwater arrives, in particular to cities at the Pearl River Delta estuary. Moreover, the coastal cities should enhance water conservation measures on the basis of China's strictest water resource management system, to reduce the water demand gap under the influence of saltwater intrusion (Wang et al., 2020a, 2020b; Bai et al., 2021).

In short, it is suggested that, considering both water demand and supply sides, the local government should strengthen the unified planning and management of water resources, guarantee the capital investment in the construction of an emergency backup water source project, scientific and technological research on the development of an alarm forecasting system and water-saving renovation, so as to ensure the safety of water supply in the Pearl River Delta during the dry period.

In the current study, a WEAP-based water resource allocation model was developed to investigate the water supply schemes of the coastal urban agglomeration of the Pearl River Delta and reveal the impacts of the wetness–dryness encountering of the upstream inflow and downstream salt intrusion. The following main conclusions were drawn:

• (1)

  The upstream runoff is highly correlated with the daily excessive chlorinity hours of the pumping stations in the downstream of the Pearl River Delta. The estimated high, medium, and low levels of the critical upstream runoff for saltwater suppression are 2,500, 1,650, and 1,400 m³/s, respectively, according to a correlation analysis based on the excessive chlorinity hours and the average monthly streamflow of Makou station in the West River.

• (2)

  The West River, North River, and East River have stronger synchronous wetness–dryness encountering, and the water shortage of the Pearl River Delta is highly dependent upon the dryness combination of the upstream runoff during the dry period. Particularly for the cities at the Pearl River Delta estuary, the water shortages would be aggravated due to saltwater intrusion during the dry period. It is essential for the local government to tap the regulation potential of existing reservoirs of the Pearl River Delta and operate the unified regulation and management of water resources across the entire Pearl River Basin.

• (3)

  The water transfer projects could alleviate the water supply pressure of the Pearl River Delta when there is an extreme or more dryness encounter of upstream runoff, however, the impact of the water transfer projects is weakened and much of the critical upstream runoff for saltwater suppression cannot be satisfied when there is an extreme dryness event of the West River. It is necessary to develop the emergency backup water source project, particularly in cities at the Pearl River Delta estuary, and advance the effectiveness of the saltwater intrusion forecasting system.

The current study reveals the impacts of the wetness–dryness encountering of the upstream inflow and downstream salt intrusion on water supply in the Pearl River Delta, based on which reasonable water allocation considering regional and industrial water distribution priorities can be further studied in the Pearl River Delta, according to the different aims of development, function positioning, and the level of economic and social development for each city.

The authors would like to express their gratitude to all of the reviewers for their valuable recommendations. This study was financially supported by the National Natural Science Foundation of China (Grant No. 51979043, 51861125203, U1911204), and the Natural Science Foundation of Guangdong Province (Grant No. 2021A1515010723).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.