**Preface to "Water Resources Management Models for Policy Assessment"**

Water resources management models support a variety of research applications, including the assessment of water availability, allocation of water among competing uses, evaluation of system performance, identification of optimal system expansion, or definition of suitable operating strategies. System analysis tools, such as simulation and optimization, have been enriched with novel modelling concepts drawn from social sciences, economic analysis, conflict resolution, agent-based systems, or game theory, among others. This field has evolved from the traditional emphasis on cost–benefit analysis in water resource project investments to a wider scope that includes environmental implications, stakeholder concerns, social welfare, and human dimensions. We now face the challenge of developing integrated modelling frameworks to provide quantitative evidence to policymakers on water management issues.

This book compiles original research papers that apply a variety of techniques to identify and evaluate water resource management policies. The compilation presented here covers a wide range of topics and methodologies applied across the world, from a local to a continental scope. Open challenges in water resource management, such as quantitative assessment of policy impacts, trade-off analyses, understanding the water-energy-food-environment nexus, collaborative model development, stakeholder engagement, formalizing social interactions, or improving the theoretical understanding of complex adaptive systems, are outlined. Therefore, this book covers research areas that have emerged from the origins of water resource systems analysis, seeking to improve the way in which water policy is formulated and implemented.

> **Luis Garrote** *Editor*

### *Editorial* **Water Resources Management Models for Policy Assessment**

**Luis Garrote**

Department of Civil Engineering, Hydraulics, Energy and Environment, Universidad Politécnica de Madrid, 28040 Madrid, Spain; l.garrote@upm.es

Water resources management models support a variety of research applications, including the assessment of water availability [1], the allocation of water among competing uses [2], the evaluation of system performance [3,4], the identification of optimal system expansion [5], and the definition of suitable operating strategies [6]. System analysis tools, like simulation and optimization, have been enriched with novel modelling concepts drawn from social sciences [7], economic analysis [8], conflict resolution [9], agent-based systems [10], and game theory [11], among others. The field has evolved from a traditional emphasis on cost–benefit analysis in water resource project investments to a wider scope that includes environmental implications, stakeholder concerns, social welfare, and human dimensions [12].

This Special Issue of Water integrates a collection of research papers that develop or apply water resources management models for policy identification and assessment. Active research has been conducted to address the challenge of developing integrated modelling frameworks to provide quantitative evidence for policymakers on water management issues. The compilation presented here covers a wide range of topics and methodologies applied across the world, from a local to continental scope. It illustrates open challenges in water resources management, like quantitative assessment of policy impacts, trade-off analyses, understanding the water-energy-food-environment nexus, collaborative model development, stakeholder engagement, formalizing social interactions, or improving the theoretical understanding of complex adaptive systems. This issue is therefore a representation of research areas that have emerged from the origins of water resource systems analysis seeking to improve the way water policy is formulated and implemented.

The contributions to the Special Issue may be classified into four major topics: water availability and accessibility, management of water infrastructure, environmental concerns, and social and economic issues. Contributions in the first group focus on the estimation of water availability under different climate and policy scenarios. Two papers are focused on Europe and two are focused on China. The paper by Sordo-Ward et al. [13] presented a regional assessment of future water availability in Europe. They applied a high-resolution model to produce detailed maps of water availability in European rivers and evaluated model and scenario uncertainties under different climate projections. The work presented in [14] was specifically focused on the role of reservoir storage to enhance resilience to climate change. The authors studied 16 major river basins in Southern Europe and found that increased storage capacity attenuated the reduction of water availability and reduced its uncertainty under climate change projections. Li et al. [15] evaluated five spatial factors to obtain a water accessibility index in Southwest China. They produced a spatial pattern and compared water accessibility and water demand at the county level. As a result of their analysis, the authors provided policy recommendations to correct the imbalance. Finally, Wang et al. [16] studied the water-carrying capacity of the Chang-Ji region in Northeast China. They applied techniques such as the fuzzy comprehensive evaluation method, gray correlation analysis, and multiple linear regression models to evaluate water-carrying capacity under different social development plans, identified critical issues, and provided suggestions to allow for a sustainable development of the economy in the region.

**Citation:** Garrote, L. Water Resources Management Models for Policy Assessment. *Water* **2021**, *13*, 1063. https://doi.org/10.3390/ w13081063

Received: 7 April 2021 Accepted: 9 April 2021 Published: 13 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

The second topic deals with models intended to provide support for management policies for water infrastructure. The paper by Rubio-Martin et al. [17] presented an application of system dynamics for the strategic planning of drought management in a river basin located in Southeast Spain. The authors proposed a system state index that is used to trigger dynamic reservoir operating rules, policies, and drought management strategies. They argued that application of their decision support system may lead to a substantial reduction of the economic impact of droughts in the basin. Gabriel-Martin et al. [18] aimed at solving conflicts that arise in the operation of multipurpose reservoirs. Their technical contribution is a model that maximizes reservoir yield subject to constraints imposed by hydrological dam safety and downstream river safety. They produced a set of Pareto optimal configurations that may be used by policymakers to emphasize water availability or flood protection. Bejarano et al. [19] offered a computational tool intended to summarize data on sub-daily streamflow into manageable, comprehensive, and ecologically meaningful metrics, which can be used to qualify and quantify flow alteration. This tool may be used by policymakers to evaluate the potential ecological consequences of the hydrological alteration produced by water infrastructure. The contribution by Martin-Candilejo et al. [20] is focused on energy efficiency. They proposed a novel method to account for energy costs associated to water pumping in the design and operation of water supply systems.

Water quality is the major focus of the third topic, which deals with environmental concerns. Xie et al. [21] reported on the experience of implementing the nation-wide freshwater health evaluation in China. They proposed a new indicator framework combining ecosystem integrity with non-ecological performance with the objective of improving water governance. The result of their work is directly policy-relevant because it will be integrated into a new national standard. Salehi et al. [22] evaluated the pollutant discharge characteristics for 12 facilities in an industry sector in the United States. They applied principal component analysis to water quality parameters and developed water quality indexes to monitor water quality fluctuations. They characterized stormwater quality variations among studied facilities and seasons, concluding with suggestions for future changes for decision makers. The work by Duan et al. [23] focused on background pollutants and their influence on water quality management and assessment methods in China. The authors argue that it is unreasonable to use a uniform standard to evaluate water quality across the country. They defined a suitable pollutant yield coefficient by coupling an export coefficient model with a mechanistic model. Based on their results, they proposed a more reasonable sewage discharge limit and water quality evaluation method. Best management practices to control water pollution were analyzed in [24]. The authors evaluated the performance of three types of pollution control measures on dissolved nitrogen by coupling an improved watershed model with a multi-objective optimization algorithm. Their optimization model system could assist decision-makers in selecting the most appropriate measures for pollution control in a watershed. Wang et al. [25] proposed an index system to evaluate the degree of coordination between economic development and infrastructure construction in a sponge city in China. They studied the spatial statistical pattern of coordination and concluded that the problems due to inadequate coordination were prominent in the region. They suggested a stronger emphasis on the construction of green infrastructure.

The fourth topic is related to social and economic issues. Lima-Quispe et al. [26] discussed river basin planning in Bolivia from the wider perspective of regional planning. They tackled the problems of coordinating watershed planning with other planning units and integrating watershed management with water resources management. The authors proposed the novel technique of robust decision support to help stakeholders discern positive and negative interactions of interventions, use spatially explicit indicators, and identify adequate management strategies. Li et al. [27] explored the applicability of China's policy based on water saving contracts by risk assessment. Overall risk was found to be low, but they showed concern for some potential risk factors, such as audit, financing, and payment risk. Feria-Dominguez et al. [28] analyzed the impact of a severe drought on the Brazilian stock market. They found statistical evidence of financial impact caused by the declaration

of drought among agri-food firms, particularly in those companies that shell perishable products. Shen et al. [29] studied the impact of tourism on the sustainable development of a reservoir in China. They applied different analytical techniques to process hundreds of questionnaires filled by the local population. In their conclusions, they found that stakeholders were very critical of the consequences of tourism development in the region and provided suggestions to mitigate the negative impacts. Santasusagna Riu et al. [30] also used questionnaires to analyze the management of urban public services in the internal border area between two Spanish regions. Based on their analysis of the replies, they concluded that there are deficiencies to correct and suggested enhanced cooperation across the border to improve priority urban public services.

This Special Issue is a compilation of 18 contributions that offer a wide perspective of the potential of water resources management models for policy assessment. The papers focus on a diversity of topics, geographical locations, spatial scales, and methodologies that illustrate successful case studies of science inspiring policy. This work is offered as an asset for researchers and policymakers.

**Funding:** This research received no external funding.

**Acknowledgments:** The Guest Editor wishes to thank the authors for their relevant contributions to the Special Issue and to the anonymous reviewers for their constructive comments and their dedication. Special thanks to the editorial managers who worked hard to speed up the handling of the manuscripts submitted to this Special Issue. Their help is gratefully acknowledged.

**Conflicts of Interest:** The author declares no conflict of interest.

#### **References**


### *Article* **Blue Water in Europe: Estimates of Current and Future Availability and Analysis of Uncertainty**

#### **Alvaro Sordo-Ward 1 , Isabel Granados 1 , Ana Iglesias <sup>2</sup> and Luis Garrote 1, \***


Received: 29 December 2018; Accepted: 21 February 2019; Published: 26 February 2019

**Abstract:** This study presents a regional assessment of future blue water availability in Europe under different assumptions. The baseline period (1960 to 1999) is compared to the near future (2020 to 2059) and the long-term future (2060 to 2099). Blue water availability is estimated as the maximum amount of water supplied at a certain point of the river network that satisfies a defined demand, taking into account specified reliability requirements. Water availability is computed with the geospatial high-resolution Water Availability and Adaptation Policy Assessment (WAAPA) model. The WAAPA model definition for this study extends over 6 million km<sup>2</sup> in Europe and considers almost 4000 sub-basins in Europe. The model takes into account 2300 reservoirs larger than 5 hm<sup>3</sup> , and the dataset of Hydro 1k with 1700 sub-basins. Hydrological scenarios for this study were taken from the Inter-Sectoral Impact Model Inter-Comparison Project and included simulations of five global climate models under different Representative Concentration Pathways scenarios. The choice of method is useful for evaluating large area regional studies that include high resolution on the systems´ characterization. The results highlight large uncertainties associated with a set of local water availability estimates across Europe. Climate model uncertainties for mean annual runoff and potential water availability were found to be higher than scenario uncertainties. Furthermore, the existing hydraulic infrastructure and its management have played an important role by decoupling water availability from hydrologic variability. This is observed for all climate models, the emissions scenarios considered, and for near and long-term future. The balance between water availability and withdrawals is threatened in some regions, such as the Mediterranean region. The results of this study contribute to defining potential challenges in water resource systems and regional risk areas.

**Keywords:** climate change; water resources; water availability; uncertainty; WAAPA model; Western Europe

#### **1. Introduction**

Water management is challenged by climate change. By the 2070s, the percentage of the surface area under conditions of severe water stress is expected to increase from the current 19% to 35% in central and southern Europe [1]. Populations living under water stress conditions in regions from 17 countries of Western Europe are projected to increase by between 16 and 44 million [2]. It is also predicted that the runoff of certain rivers may diminish by up to 80% during the summers. Reservoirs may lose resources due to a decrease in rainfall and the frequency of droughts will increase. The consensus is that the effect of climate change will also exacerbate precipitation extremes with more pronounced drought and flood periods [3–5]. At the same time, future water demand is increasing due to climate and social changes. Higher temperatures lead to increased water demand for irrigation and

urban supply, hydroelectric potential of Europe may decrease 6% on average, and between 20 and 50% in the Mediterranean region. Advances in technology efficiency may only affect industrial demand [2]. In the Mediterranean region, impacts of climate change on water will certainly have a large influence on human water security and biodiversity [6]. There are several hundred local studies on the potential impacts of climate change on water resources in the Mediterranean, which apply many different approaches. Although the results are diverse and sometimes contradictory, a common element is that one of the primary impacts of climate change will be a reduction of water availability in the Mediterranean Region [1,2]. Furthermore, several authors showed that Global Climate Models (GCMs) were the main source of uncertainty when assessing the impacts of climate change on hydrologic processes [7,8]. Meanwhile, uncertainty associated with streamflow appeared to be more consistent with precipitation than temperature and showed higher sensitivity to the selection of GCMs than to the Regional Climate Models (RCPs) [9,10].

Water availability focuses on blue water, which is defined as water that runs off the landscape into streams, rivers, reservoirs, and groundwater [11]. However, the term "water availability" includes multiple aspects. A multitude of studies consider water availability to be directly linked to changes in average runoff, estimated as the net difference between precipitation and evapotranspiration [12,13]. In non-altered basins, water availability would be either null or extremely low because it would be determined by long term minimum values of flow. It is clear that hydraulic infrastructure plays an important role in making water available for users, mainly by the regulation and transportation of water resources. Even though the storage-based strategy proved to be very successful in the past [14], expanding infrastructure is not an option to increase availability in many regions due to social and environmental constraints [15]. As a result, increasing demand relies heavily on management. The emphasis is currently being placed on how to improve management of existing infrastructure and on socio-economic measures through demand management and water use efficiency [16,17]. The main factors to be considered in regulated water resource systems are the stream flow variability, storage capacity, and yield reliability. In this study, we define blue water availability as the amount of water that can be supplied at a certain point of the river network to satisfy a regular demand under specified reliability requirements [18,19]. Therefore, water availability is the combined result of natural processes, existing infrastructure, and policy. A wide range of techniques have been proposed to analyse water availability, from relatively simple stochastic processes relating these variables to highly complex models solving the water allocation problem [20–24], even including social and economic considerations [25]. In the water sector, institutions, users, technology, and the economy cooperate to achieve equilibrium between water supply and demand in water resource systems. In order to understand the process of reaching future goals for water under climate change, science has developed a set of tools to understand uncertainty [26–29], assess future impacts [30,31], and facilitate policy development [1,16,18,32]. However, most studies were developed using detailed water management and planning models, and were applied at the local scale. In systems and situations where limited information is available and regional or continental-scale studies are needed, it is generally better to obtain a global overview of the water supply systems' performance under different climate and policy scenarios, using simplified regional models rather than carrying out very detailed simulations with conventional models, which require very specific information on water demands and infrastructure [18,33,34]. These continental scale-models are conceived to estimate the maximum water availability and to provide technical and quantitative support to possible water policies in the short and long term. Then, these models and detailed water management and planning models should be considered as complementary tools.

Over forty percent of the total water withdrawal in Europe is used for agriculture. Southern countries use the largest percentages of abstracted water for agriculture. This generally accounts for more than two thirds of total abstraction. In northern member States, levels of water use in agriculture are much smaller, with irrigation being less important but still accounting for more than 30% in some areas [35]. Moreover, if the climate in a given region gets drier and warmer, water availability will decrease, and the issue will be exacerbated by increasing water demand [36]. For example, it is expected that areas of maize grain cultivation will expand up to 30–50% in Europe [37–40] with increases of up to 50% in net primary productivity in northern European ecosystems, as a result of a longer growing season and higher CO<sup>2</sup> concentrations [37]. As the projected impacts on productivity of crops and ecosystems included the direct effects of increased CO<sup>2</sup> concentration on photosynthesis, the variation in simulated results attributed to differences between the climate models were, in all cases, smaller than the variation attributed to emissions scenarios [37]. The objective of this study is to estimate future potential blue water availability in Europe and its associated uncertainty, which is induced by emissions scenarios and climate change models. This study first proposes a methodology to conduct climate change analyses in water resource systems, which is based on a high-resolution geospatial model and the use of information available in public databases. Second, the study evaluates distributed mean annual runoff and its uncertainty in main rivers within Western Europe in the baseline period and in two future periods. Third, the study analyses water availability changes and its uncertainty across Western Europe under different climate change scenarios and climate models. Finally, the study analyses the geographically distributed relationships at a continental-scale among the mean annual runoff, water availability, and water withdrawals under the baseline and future periods.

#### **2. Materials and Methods**

The methodological approach is detailed in Figure 1. The methodology is based on a high-resolution GIS-based model, named "Water Availability and Adaptation Policy Assessment (WAAPA)" which enables the estimation of water availability under many climate scenarios to produce a global picture of the situation [33]. The model assimilates climate and geospatial information seamlessly, accounts for reservoir storage (from an individual reservoir or from a system of reservoirs), and produces blue water availability estimates. The model computes net blue water availability for consumptive use of a river basin, taking into account the regulation capacity of its water supply system, and a set of management standards defined by water policy. The model estimates the water availability not only at the outlet of sub-basins (e.g., river intersections), but also at any desired point of the defined river network (e.g., each dam location), by accounting for the entire system of dams in the upstream basin. Basic components of WAAPA are reservoirs, inflows, and demands and they are linked to nodes of the river network. The joint reservoir operation model simulates the behaviour of a set of reservoirs that supply water for a set of prioritized demands, complying with specified ecological flows and accounting for evaporation losses.

**Figure 1.** General scheme of the applied methodology. The displayed procedure was applied to each defined sub-basin. Grey areas indicate the first path carried out, from the selection of the emissions scenario to the estimation of the water availability.

In this study, we evaluated the water availability of the joint reservoir operation model following a high resolution and global management scheme (Figure 2). For each selected sub-basin (derived from dam locations and river confluences), this scheme considers each reservoir individually and all reservoirs are jointly operated to supply a set of prioritized demands. It is assumed that any demand at a given point in the stream network can be supplied by any reservoir located upstream from it. It corresponds to a situation where there is little development of system interconnections, but a large development of water distribution networks, which are managed globally to supply all demands present in the analysed system. Water is first released (to satisfy demands) from the reservoirs located at low areas of the basin. If these reservoirs are full and receive more contributions, uncontrolled spills are released and water falls out of the system. On the other hand, if upstream reservoirs are full and receive more inflows, the extra water is collected by the downstream reservoirs. This management criterion is not totally real, because real systems usually are managed taking into account more conditions and constraints. The joint reservoir operation model maximizes water availability because it minimizes the excess storage. In each time step, the model performs the following operations:


**Figure 2.** Operation scheme of the high-resolution Water Availability and Adaptation Policy Assessment (WAAPA) model for each given point of the stream network (blue lines). Triangles represent dams, big coloured arrows represent inflows, small arrows represent reservoir evaporation, uncontrolled spills, and environmental flows, and grey dashed lines represent supplies from each reservoir to the basin demands (rectangles).

The result of the joint reservoir operation model is a set of time series of monthly volumes supplied to each demand, monthly storage values, monthly values of spills, environmental flows, and evaporation losses in every reservoir. Finally, we calculated the system performance by applying the Gross Volume Reliability performance index. This index is the ratio of total volume supplied to demand in the system and the total volume demanded by the system, during the analysed period [33]. In this study, water availability is estimated by considering only one demand present in the system under the hypothesis of 90% reliability.

To define the maximum amount of water that can be supplied at a certain point of the river network to satisfy a regular demand, a bipartition method is applied: Excessive values of demands are set (for example, similar to mean monthly runoff) and the simulation is carried out. The deficits are obtained and specified reliability requirements are checked. If the specified reliability requirements are not fulfilled, the demand is reduced by half and simulated again. If the specified reliability requirements are satisfied, half of the difference is added and simulated again, and so on until the deficit (or gain) is smaller than a pre-set tolerance (e.g., 0.1 hm3/year).

#### *Case Study*

The area under analysis is composed of the major river basins in Western Europe. WAAPA model data are geographically referenced (Figure 3). Following, we present the data used to build the WAAPA model. We determined the topology of the model by dividing the area under study into a number of units of analysis, which are homogeneous sub-basins from the water management perspective. The sub-basins are related through the "drain to" relationship, and the analysis is applied to all possible basins, from the small headwater sub-basins to the largest basin draining to the sea. In this work, we divided western Europe into sub-basins (3839), based on the Hydro1k data set (1.538 sub-basins [41]), and the derived-from dam locations (2.301 sub-basins), which belong to 621 large basins draining to the sea. The total area under study is over 6,000,000 km<sup>2</sup> .

**Figure 3.** Case study: Western Europe. (**a**) Domain under analysis. Colours represent the 621 major river basins draining to the sea. (**b**) Information utilized for the estimation of withdrawals (domestic (hm3/km<sup>2</sup> ), agriculture (hm3/km<sup>2</sup> ) and industry (hm3/km<sup>2</sup> )) in present and future scenarios and for each analysed sub-basin.

Naturalized streamflow was obtained from the results of the application of the PCRGLOBWB model [42] to the Inter-Sectoral Impact Model Inter-Comparison Project [43]. The PCRGLOBWB model was run for the entire globe at 0.5◦ resolution, using forcing from five global climate models

Mean annual runoff at future scenario Mean annual runoff change

 

Mean annual runoff at reference scenario

1 100

  under historical conditions and climate change projections, corresponding to four Representative Concentration Pathways scenarios: RCP2.6, RCP4.5, RCP6.0, and RCP8.5. The following climate models were used as input: GFDL-ESM2NM (GFDL), HadGEM2-ES (HadGEM2), IPSL-CM5A-LR (IPSL), MIROC-ESM-CHEM (MIROC), and NorESM1-M (NorESM1).

Three time periods were considered: Reference (1960–1999), short term (ST, 2020–2059), and long term (LT, 2060–2099). Since runoff obtained from climate model input usually presents significant bias, average runoff values were corrected for bias using the UNH/GRDC (University of New Hampshire/Global Runoff Data Centre) composite runoff field, which combines observed river discharges with a water balance model [44], and is a reference of the current global surface runoff [34,44,45]. Following González-Zeas [45], we applied the bias-correction methodology based on the determination of a monthly correction factor. We calculated the monthly mean runoff series for the control scenario to obtain twelve representative statistical parameters: The ratios between the UNH/GRDC values (observed) and the simulated runoff. These multiplying factors were used to correct bias in the control and the projected series. The reservoir storage volume available for regulation in every sub-basin was obtained from the ICOLD World Register of Dams [46]. Dams in the register with more than 5 hm<sup>3</sup> of storage capacity were georeferenced and linked to the corresponding storage capacity and flooded area (2.301 dams). Environmental flows were computed through a hydrologic method. The minimum environmental flow was set to the 10% percentile of the marginal monthly distribution, according to Spanish legislation. In the absence of more advanced methods, the Spanish regulation for river basin plans establishes several hydrologic methods to define minimum environmental flows [31]. One of them is based on the percentile of the marginal distribution of monthly flows, defining a range between 5 and 15%.

In this study, we estimated current, short-, and long-term geographically distributed water withdrawals. Country-based data on current freshwater withdrawal were taken from the World Bank database. These data were spatially distributed using proxy variables: Population density for urban and industrial withdrawals and irrigated area for agricultural withdrawals. The population density was obtained from the Gridded Population of the World product of the Global Rural–Urban Mapping Project (GRUMP), available at the Centre for International Earth Science Information Network (Figure 3b) [47]. The area potentially under irrigation was estimated from the Global Map of Irrigated Area dataset [48]. Future withdrawals were estimated using the projections of population and gross domestic product (GDP) provided by IIASA. These projections were estimated following RCP scenario assumptions [38,39]. Projections of total freshwater withdrawal and industrial withdrawal were estimated from regressions based on World Bank data using per capita GDP projections [40].

#### **3. Results**

Figure 4 shows the comparison of streamflow change from reference (1960–1999) to climate change RCP4.5 scenarios, both for short (2020–2059) and long term (2060–2099), and over the five climate models. Figure 4 is dimensionless (percentage), and the values were obtained by applying Equation (1). The red shading represents a decrease (negative values) and green shading an increase (positive values) of the future mean annual runoff. The yellow shading represents no changes of mean annual runoff for future periods compared to the reference scenario.

$$\text{Mean annual runoff change} = \left(\frac{\text{Mean annual runoff at future scenario}}{\text{Mean annual runoff at reference scenario}} - 1\right) \times 100\tag{1}$$

Overall, the models produce a smooth picture of mean annual runoff change in Europe, with decreases in the South. Severe negative changes are projected in the Iberian Peninsula, from the Black Sea in the South almost to the Baltic Sea in the North, and predominantly positive changes are projected in western to central Europe and in northern Europe. A mixed pattern with higher variability in mean annual runoff is shown across central Europe and the Carpathians. The climate models that produce more annual runoff reduction are HadGEM2 and NorEsM1. However, it can be seen that the values

and spatial extent of the regions with reduced streamflow (in brownish colours) vary significantly from one climate model to another. This is remarkable considering that all simulations were performed with the same hydrologic model. As expected, in general, the changes are more intense in the long-term period. The region of neutral changes (represented in yellow) moves toward the north from low carbon (RCP2.6) to high carbon (RCP8.5) emissions scenarios (not shown).

**Figure 4.** Changes (percentage) of mean annual runoff in future scenarios (2020–2059 and 2060–2099) compared with the reference scenario (1960–1999), according to different climate models and for the emissions scenario RCP4.5. Red shading represents a decrease of the mean annual runoff and green shading an increase.

The results of potential water availability in historical conditions (1960–1999) for all climate models are shown in Figure 5. It shows the values of potential water availability as a function of mean annual runoff in all the analysed sub-basins. Small, blue dots represent results in intermediate sub-basins, while larger, darker blue dots represent results in the global basins. All models show a similar picture, with a large variation of water availability among basins as a consequence of differences in hydrologic regime and reservoir storage.

**Figure 5.** Mean annual water availability as a function of mean annual flow for the historical period (1960–1999) and for the different climate models. Small, blue dots represent results in intermediate sub-basins, while larger, darker blue dots represent results in the global river basins. Red line shows the value of 40% of mean annual runoff. (**a**) GFDL model, (**b**) HadGEM2, (**c**) IPSL, (**d**) MIROC and (**e**) NorEsM1.

The spatial distribution of changes (between the long term and reference periods) of potential water availability along the major rivers in Europe is presented in Figure 6, for all emissions scenarios and climate models analysed. Figure 6 is dimensionless, and the values were obtained by applying an equation similar to Equation 1, but using potential water availability instead of runoff. Red shading represents a decrease (negative values) of the future potential water availability and green shading an increase (positive values). The yellow shading represents no changes of potential water availability compared to the reference scenario. Although, in general, the climate models show a gradient of potential water availability changes with larger reductions in South Western Europe and larger increases in Northern Europe, values show important differences by comparing the results among climate models (same emissions scenario). By comparing the maps within each column (Figure 6), we visualize important differences in the results from one climate model to another, and by keeping each emissions scenario unaltered. The models that produce the most potential water availability reduction are HadGEM2 and NorEsM1, while IPSL and MIROC produce the least reductions. On the other hand, by comparing the maps within each row (Figure 6), we observe the different results obtained for the same climate model and different emissions scenarios. It can be seen that, in general, differences among the emissions scenarios (for each climate model) are smaller than those among different models (for each emissions scenario). The driest scenario is RCP8.5 for all analysed climate models.

Figure 7 shows, for each analysed sub-basin, the changes in the potential water availability in the long-term period with respect to the reference period (*y* axis), as a function of changes in the mean annual runoff in the long-term period with respect to the reference period (*x* axis), for all emissions scenarios and the climate model GFDL. The equations used to plot the results are similar to the proposed Equation 1 for the runoff variable (see Figure 4) and the proposed for the water availability variable (see Figure 6). Quadrant 1 (q.I) shows sub-basins where runoff decreases in the future and water availability increases. Both runoff and water availability increase in q.II, runoff increases and water availability decreases in q.III, and both runoff and water availability decrease in q.IV. In addition, basins with the same reduction of runoff experience different reductions in availability as a result of changes in the hydrologic variability and their different regulation capacity.

Figure 8 shows the spatial distribution of the ratio of the runoff, water availability, and water withdrawal for the model GFDL in emissions scenario RCP4.5 for the reference (1960–1999) and long-term period (2060–2099). The bottom row shows potential water availability as a fraction of runoff, the central one shows water withdrawal, also as a fraction of runoff, and the upper row shows the water withdrawal as a fraction of water availability.

**Figure 6.** Changes in potential water availability for the long-term scenario (2060–2099) compared to the reference scenario (1960–1999), according to all climate models and emissions scenarios analysed. Red shading represents a decrease of the potential water availability and green shading an increase (individual maps at full resolution are available as supplementary files).

**Figure 7.** Changes of mean annual water availability from historical period (1960–1999) to long-term period (2060–2099) as a function of changes in runoff for model GFDL and emissions scenario RCP2.6, RCP4.5, RCP6.0, and RCP8.5. Small, blue dots represent results in intermediate sub-basins, while larger, darker blue dots represent results in the global basins. q.I, q.II, q.III, and q.IV point out each quadrant. (**a**) RCP2.6, (**b**) RCP4.5, (**c**) RCP6.0 and (**d**) RCP8.5.

Figure 9 shows the uncertainty associated with the climate models and emissions scenarios, both for mean annual runoff and mean water availability, by calculating the coefficient of variation (CV, standard deviation divided by mean) in each calculation point, for each climate model (five), for each emissions scenario (four) and for the short term (ST) and the long term (LT). We represented the probability distribution function (Pdf) of the CVs in each case. Continuous lines represent the uncertainty for each climate model, obtained by comparing the four emissions scenarios for each climate model. The dashed lines represent uncertainty for each emissions scenario, obtained by comparing the CV of the five models for each emissions scenario. Figure 9a,c shows the uncertainty associated with runoff for the ST and LT, respectively. Comparatively, Figure 9b,d shows the uncertainty associated with availability for the ST and LT, respectively.

14

**Figure 8.** Spatial distribution of the per unit change of potential water availability between the historical period (1960–1999) and the long-term period (2070–2099) for the model GFDL, under the emissions scenario RCP4.5. (**Top row**) Withdrawal as a fraction of availability. (**Centre row**) Water withdrawal as a fraction of runoff. (**Bottom row**) Potential water availability as a fraction of runoff.

**Figure 9.** Climate model and emissions scenario uncertainties. Each continuous line represents the probability distribution function (Pdf) of the coefficient of variation (CV) corresponding to mean annual runoff and mean annual water availability in each calculation point, for each climate model (GFDL, green; HadGEM2, brown; IPSL, purple; MIROC, red; and NorEsH1, blue) and four emissions scenarios (RCP2.6, RCP4.5, RCP6.0, and RCP 8.5). Dashed line represents the Pdf of CV for each emissions scenario and the five climate models. (**a**) Runoff and short-term period (ST) and (**c**) long-term period (LT); (**b**) Water availability and ST and (**d**) LT.

#### **4. Discussion**

Both for short- and long-term periods, the models show similar spatial patterns of mean annual runoff changes in Europe, with decreases in the south, especially in the south-west and increases in the north. Our results agree with global and continental-scale studies that reported mean annual runoff projections [1,49,50]. These studies provide a coherent pattern of change in annual runoff, predicting with a high degree of confidence severe decreases (up to 40%) of surface runoff in areas already affected by water scarcity, like the Mediterranean region, and are consistent with the projected runoff increases in northern Europe (5–30%). However, it can be seen that the values and spatial extent of the regions with reduced streamflow vary significantly from one climate model to another. It suggests that there is an important climate model uncertainty, being the changes of mean annual runoff among emissions scenarios (and the same climate model) smaller than those among the different climate models (same emissions scenario).

In the short-term runoff, Figure 9a clearly shows that higher CV values are more frequent (amplitude of each Pdf curve) by comparing the results among models (and the same emissions scenario, dashed lines) than among emissions scenarios (and the same climate model, continuous lines). In addition, the uncertainties associated with the emissions scenarios are also similar among them (differences among continuous lines for the *y* axis). Although climate change models are the most robust tools available to generate consistent climate change projections, they are still a source of considerable uncertainties [10,51]. In this regard, Garrote [18] highlighted that the uncertainty has not been reduced with the progressive improvement of modelling tools; on the contrary, it seems to be increasing as a result of the evolving approach to generating emissions scenarios. On the other hand, results suggest that, because of the number of variables and complexity involved in the estimation of the future

climate, its estimation has an implicit uncertainty that should be acknowledged for the development of climate adaptation plans. In the long-term runoff (Figure 9c), the uncertainty increases (increasing the amplitude of the Pdf curves) for both climate models and emissions scenarios, although climate models remain more uncertain than emissions scenarios. Also, greater dispersion of uncertainty is found among models than among emissions scenarios. It could be partially explained by the increase of the differences between emissions scenarios for the long-term analysis. These results are consistent with several inter-comparison studies that also show considerable variability in the magnitude and timing of the projected runoff [9,49,50,52,53]. At this point, it is remarkable that all simulations in this study were performed with the same hydrologic model. Databases of climate scenarios are available from different research projects [54,55], including surface runoff among their output variables. As the characterization of the water cycle in the models used in these types of studies usually is very simple and results provide a low signal-to-noise ratio (especially in arid and semi-arid regions), varying the large-scale hydrological models incorporates an additional source of uncertainty [18,50,52]. Some authors state that hydrologic model uncertainties are less significant than those originating from climate change models [9,56].

Changes in potential water availability in short- and long-term scenarios according to all climate models and emissions scenarios were analysed. High resolution results showed similar future spatial patterns to mean annual runoff, with the differences among the emissions scenarios (for each climate model) being smaller than those among different models (for each emissions scenario). Figure 9b shows that the uncertainty associated with the emissions scenarios increases and their values draw near to the climate model uncertainties. Furthermore, the Pdfs of the uncertainty associated with the climate models for water availability remain similar to that for runoff. Similar behaviour is observed for the long-term period (Figure 9d). It suggests that the management of hydraulic infrastructures (mainly reservoirs in this study) plays an important role by decoupling water availability from hydrologic variability. This is observed for all climate models and emissions scenarios considered. Svensson et al. [57] reinforced the importance of the installation of reservoirs in several river basins in Europe in the last century, by attenuating the basins' drought conditions. For quantifying and summarizing purposes, Table 1 shows the emissions scenarios' and climate models' uncertainty for the 50% probability of exceeding CV values. Several local and regional studies agree that the propagation of the uncertainties affects water resource system performances [26,58–60]. Thus, the assessment (or projection) of the performance of a water resources system should be evaluated with extreme care. As previously stated, the reservoir operation model applied in WAAPA is highly simplified and was designed to maximize water availability. Thus, the reality of reservoir operation is much more complex. Usually, not all reservoirs in the basin are jointly managed to supply all demands. They are either managed individually to supply local demands or grouped in systems that are managed independently. Availability of storage volume for water conservation management is also variable according to local conditions, due to the need to allocate storage volume to flood control. Therefore, it is unlikely that upstream reservoirs are kept full to release space in downstream reservoirs. Normal operation would tend to balance storage in all reservoirs to prevent uncontrolled spills. In practice, the spatial pattern of water availability will differ from that obtained in WAAPA. WAAPA results should only be considered as an upper bound of the actual water availability that could be obtained in practice.

Results from the comparisons of the changes in potential water availability with changes in runoff clearly show how changes in the former are not proportional to changes in the latter, suggesting the inadequacy of methodologies that estimate availability as a fraction of mean annual runoff. As an example, in Figure 5, the red line shows the traditional value of 40% of the mean annual runoff adopted for water availability when no simulation of reservoir regulation is performed [61]. It can be seen that adopting this constant value as a proxy of water availability can be strongly misleading, since only those basins with very regular flow or very large reservoir storage can reach this value. In most basins, water availability is a smaller fraction of the mean annual runoff.


**Table 1.** Summary of the emissions scenarios' and climate models' uncertainty for the 50% probability of exceeding CV values.

As shown, availability and withdrawal are only a small fraction of runoff in most of Europe and their projected changes are small, except for the south-east and the south-west. However, the representation of the water withdrawal as a fraction of water availability (Figure 8, upper row) shows that these two variables have similar values in many regions of Europe, and that they are getting closer in the long-term scenario. It means that in many regions, water shortage struggles to satisfy the demand with a specific reliability could emerge or increase, both for the present and future periods. It can also be seen that the relationship between these variables is complex, and that it varies significantly among regions, depending on hydrologic regime, climate, reservoir storage, and socioeconomic factors.

Green water (not analysed in this study), similarly to blue water, is also expected to decrease in most of western Europe except for northern countries. However, changes in green water result from complex interplay of impacts on precipitation, temperature, and CO<sup>2</sup> concentration, which ultimately affects potential evapotranspiration, soil moisture conditions, and growing periods. Thus, patterns of expected changes differ for green and blue water [62]. Irrigation demands will also be affected, due to modified seasonal patterns and evapotranspiration demands [36,63].

Finally, some limitations of this study should be noted. We estimated the potential water availability (upper theory limit) by considering only one demand present in the system. System performance was evaluated as gross volume reliability. Potential water availability was obtained under the hypothesis of 90% reliability. The data used in this study were obtained from specific climate models and emissions scenarios, thus, the conclusions derived from this study are inextricably affected by the models' uncertainty. Additionally, we made a series of simplifying assumptions. We assumed variable geographic and temporal water withdrawals, both in the present and future climate, from indirect methods (GDP and population). We assumed that the reservoirs, whose sole purpose was hydropower generation, were not included in the systems to manage the water resources. We considered that the hydraulic infrastructure corresponding to each analysed sub-basin (determined from a given point in the stream network) was being jointly managed to supply global demands, while in some real cases it could have been divided in to several rather independent subsystems. Furthermore, in our model, there were no system interconnections nor a large-scale water distribution infrastructure. We did not consider other sources of uncertainty as, for instance, the observed climate data source or the hydrologic model applied and the inclusion of regional climate models (RCMs). It is expected that RCMs have less associated uncertainty than GCMs when a particular region is analysed, as they account for more detailed and specific regional characteristics.

#### **5. Conclusions**

This study presents the potential water availability changes under alternative climate change scenarios in western Europe. Results are geographically referenced at high resolution across the major European river basins. The study includes the estimation of the associated uncertainties, resulting from differences among climate change scenarios and climate models. The authors are not aware of similar studies conducted at such a high-resolution continental scale. In this study, we applied the WAAPA model on a high-resolution dataset to analyse water availability changes across western Europe. The proposed model and the applied methodology demonstrated their ability to perform regional studies covering extensive domains, while maintaining high resolution on the characterization of the systems. The climate models that produced the most reduction of mean annual runoff and potential water availability were HadGEM2 and NorEsM1, while IPSL and MIROC produced the least reduction. Overall, for both mean annual runoff and potential water availability, a gradually varying picture of change in Europe was observed, with a decrease in the south (especially in the south-west) and an increase in the north. Moreover, the region of neutral changes moves to the north, from low carbon (RCP2.6) to high carbon (RCP8.5) emissions scenarios. Climate model uncertainties for mean annual runoff and potential water availability were found to be higher than scenario uncertainties. This conclusion was derived by comparing the variability of the results obtained, while the PCRGLOBWB model was forced with different climate models under the same emissions scenario to that of the results from different emissions scenarios for the same climate model forcing. Thus, although climate change models are the most robust tools available to generate consistent climate change projections, they are still a source of considerable uncertainties and their results should be carefully used for operative purposes.

While potential water availability and water withdrawal are only a small fraction of runoff in most of Europe for current and future scenarios (except in the south-east and the south-west of Europe), water withdrawal and water availability are similar in many regions of Europe, and they are getting closer in the long-term scenario (2060–2099). Thus, the balance between water availability and withdrawals is threatened in some regions. Furthermore, social factors, like management of hydraulic infrastructure, play an important role by decoupling water availability from hydrologic variability. This is observed for all climate models and emissions scenarios considered. Finally, although this study presents significant progress in terms of spatial scale and detail compared to previous studies, it is still only indicative of the importance of regional change, due to the assumptions and uncertainties discussed. Nevertheless, the results are useful for envisioning potential water resource system conflicts and contributing to the identification of regions where an in-depth analysis may be necessary.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4441/11/3/420/ s1.

**Author Contributions:** Conceptualization, L.G. and A.I.; methodology, A.S.-W. and A.I.; software, L.G.; investigation and formal analysis, A.S.-W. and I.G.; resources and data curation, I.G.; writing—original draft preparation, I.G.; writing—review and editing, A.S.-W.; visualization and supervision, L.G.; funding acquisition, A.S.-W. and A.I.

**Funding:** This research was partially funded by Universidad Politécnica de Madrid through the "Programa propio: ayudas a proyectos de I+D de investigadores posdoctorales" and the "ADAPTA" project. We also acknowledge the financial support of the European Commission BASE project (grant agreement no.: ENV-308337) of the 7th Framework Program (http://base-adaptation.eu).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

1. IPCC 2014. *Climate Change 2014: Impacts, Adaptation, and Vulnerability Part A: Global and Sectoral Aspects; Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change 2014*; Field, C.B., Barros, V.R., Dokken, D.J., Mach, K.J., Mastrandrea, M.D., Bilir, T.E., Chatterjee, M., Ebi, K.L., Estrada, Y.O., Genova, R.C., et al., Eds.; Cambridge University Press: Cambridge, UK; New York, NY, USA, 2017; pp. 1–32.


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Exploring the Role of Reservoir Storage in Enhancing Resilience to Climate Change in Southern Europe**

**Alfredo Granados 1, \*, Alvaro Sordo-Ward 1 , Bolívar Paredes-Beltrán 1,2 and Luis Garrote 1**

	- <sup>2</sup> Carrera de Ingeniería Civil, Facultad de Ingeniería Civil y Mecánica, Universidad Técnica de Ambato, Ambato 180206, Ecuador

**Abstract:** Recent trends suggest that streamflow discharge is diminishing in many rivers of Southern Europe and that interannual variability is increasing. This threatens to aggravate water scarcity problems that periodically arise in this region, because both effects will deteriorate the performance of reservoirs, decreasing their reliable yield. Reservoir storage is the key infrastructure to overcome variability and to enhance water availability in semiarid climates. This paper presents an analysis of the role of reservoir storage in preserving water availability under climate change scenarios. The study is focused on 16 major Southern European basins. Potential water availability was calculated in these basins under current condition and for 35 different climatic projections for the period 2070–2100. The results show that the expected reduction of water availability is comparable to the decrease of the mean annual flow in basins with large storage capacity. For basins with small storage, the expected reduction of water availability is larger than the reduction of mean annual flow. Additionally, a sensitivity analysis was carried out by replicating the analysis assuming variable reservoir volumes from 25% to 175% of current storage. The results show that increasing storage capacity attenuates the reduction of water availability and reduces its uncertainty under climate change projections. This feature would allow water managers to develop suitable policies to mitigate the impacts of climate change, thus enhancing the resilience of the system.

**Keywords:** climate change; reservoir performance; water availability; water resources

#### **1. Introduction**

Climate change, associated with the recorded rise of average temperatures, which are expected to continue increasing to a greater or lesser extent, may also influence other climatic variables such as precipitation, frost, or evapotranspiration [1]. All these changes may affect, in turn, the hydrological processes and consequently net water resources. This threatens the performance of water resource systems and their capability to supply demand and ecological need as presently planned. Therefore, it is necessary to assess both the impact on water resources and the behavior of water systems under such a scenario [2].

Many authors have devoted significant efforts to evaluate net water resources in climate change projections, on all scales from global to basin [3–9]. Their results show that climate change will affect, in varying ways and to different extents, each region of the planet. As a global result, it could be synthesized that there will be a reduction of water resources of between 10% and 30% [1]. This is an indicative value, useful for developing macro-policies and for raising awareness in the population.

With regard to Southern Europe, despite the dispersion of the various models, the general trend indicates that net resources will decrease, and that the variability of their distribution will increase, as shown in the results of the Prediction of regional scenarios and uncertainties for defining European climate change risks and effects (PRUDENCE) [10]

**Citation:** Granados, A.; Sordo-Ward, A.; Paredes-Beltrán, B.; Garrote, L. Exploring the Role of Reservoir Storage in Enhancing Resilience to Climate Change in Southern Europe. *Water* **2021**, *13*, 85. https://doi.org/10.3390/w13010085

Received: 10 November 2020 Accepted: 29 December 2020 Published: 1 January 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

and Climate change and its impacts at seasonal, decadal and centennial timescales (EN-SEMBLES) [11] projects. In Southern Europe, the prognosis is that traditional water scarcity problems will be aggravated. In many basins in this region, the available resources can hardly meet the existing water demands [12]. These are areas with a benign climate, which favors the implementation of agriculture and the development of tourism and services. All these activities require substantial amounts of water. Although these regions have scarce water resources, they are also resilient as they have long experience in dealing with water scarcity and are well adapted to its management [13].

The present study focuses on understanding the effect of reservoir storage capacity on water availability in Southern European basins under climate change. As above mentioned, these basins are typically characterized by scarce and highly variable water resources. The adopted strategy for water resources development in the last century relied on reservoir storage, as it is necessary to store water during the wet periods for its use in the dry ones. In Spain, for example, the existing 1350 large dams helped to increase water availability from 10% to between 40% and 50% of mean natural flow during the last century [14]. As storage capacity grew in parallel with water use, this water availability is used strictly enough to serve current water needs. Alternative adaptation and mitigation measures are being developed in the current century: controlling irrigation water rights, increasing water use efficiency through localized and drip irrigation, developing non-conventional resources, such as water reutilization and desalination, among others [15,16]. Despite these efforts, projections of climate change suggest less water resources with higher variability, which will negatively affect system performance, so water availability is expected to reverse its growing trend [17].

The analysis of reservoir storage capacity and its relationship with safe yield has been a topic of study since the beginning of the development of large hydraulic systems [18]. Initially, graphical methods were developed to determine the reservoir capacity needed to satisfy a given demand with required reliability, and their use was restricted to single reservoir models. Later methods introduced uncertainty of future inflows and attempted to estimate required reservoir size through statistical analysis of inflows, leading to the concepts of risk of failure and reliability. The development of computing allowed the stochastic generation of synthetic series and the disaggregated analysis of multiple reservoirs in a system [19]. Löf and Hardison [20] provided storage-reliability-yield (SRY) relations for assessing the required storage capacity in the USA. The study was later revisited by Vogel et al. [21], who concluded that areas with lower variability tend to be equipped with within-year storage systems while those with large variability required larger over-year storage facilities. Further developments introduced the concepts of resilience and robustness [22] to complete the reliability-yield analysis [23]. An alternative approach is the simulation of the water resources system behavior, which in conjunction with the power of computers allows the development of complex models that reproduce a simulated operation of the system [24–26]. These models are useful as decision-support tools for allocating water among users and assessing the effectiveness of structural and managerial actions [27,28] and their capabilities are even extended to groundwater resources and social and economic considerations [29].

Focus is slowly being placed on the impact of climate change on water availability and the role of reservoir storage to increase resilience. Wurbs et al. [30] highlighted the need to introduce climate change in the analysis of water availability and proposed a methodology to couple climatic and system behavior models. Garrote et al. [31] developed a simulation model specifically suited to account for the role of reservoirs in providing water availability in the context of climate change. Several authors [32–34] have argued in favor of adaptive reservoir management as an effective mitigation measure during climate change. Adaptive management requires a good knowledge of the interplay between reservoir storage and the reliability, resilience, and vulnerability of a water supply system subject to uncertain input [35]. Water availability deriving from reservoir systems may become increasingly unstable under climate change [36] and knowledge on how regulated water supply systems react to flow alterations is essential for system managers to design climate adaptation policies.

This paper looks beyond the impact of climate change on water availability to provide insight into the performance of reservoir storage systems and their effectiveness of adaptation and mitigation measures. With this purpose we include a regional analysis of the performance of reservoir storage and a sensitivity analysis of the reservoir-yield relations under less abundant resources and larger variability conditions in 16 representative European basins. The objective of the research is to check if reservoir storage enhances resilience to climate change. Given the uncertainty of climate projections, the adopted approach is to evaluate basin response under a large ensemble of plausible future scenarios and to evaluate if reservoir storage plays a role in determining the response to changes in hydrologic forcing. System response is quantified in terms of the elasticity of water availability to climate change, comparing changes in potential water availability with changes in mean annual flow. Elasticity is evaluated with the help of two new indices proposed in this work, which characterize the attenuation of changes and the reduction of uncertainty provided by reservoir storage.

#### **2. Data and Methods**

#### *2.1. Area under Analysis*

We present results for 16 major river basins in Europe, which are shown in Figure 1. Basin selection was based on a regional focus on Southern Europe, but including different climates, hydrologic regimes, and storage capacities to allow for a more effective comparison. The selected basins cover a large fraction of the Atlantic and Mediterranean divides of Southern Europe and are representative of the variety of conditions that can be found across the region. The main characteristics of the basins considered in this study are shown in Table 1. Basin areas range from 17,550 km<sup>2</sup> (Segura) to 115,910 km<sup>2</sup> (Loire). Reservoir Storage Volume V includes all reservoirs, except those managed exclusively for hydropower. The basin with largest storage volume is Guadiana, which includes two of the largest reservoirs in Southern Europe: Alqueva (4.15 km<sup>3</sup> ) and La Serena (3.21 km<sup>3</sup> ). Basin hydrology is very variable, with Specific Runoff ranging from 11 mm/year in the Segura basin to 563 mm/year in the Po basin. The most relevant characteristic for this study is specific storage, defined as the ratio of Storage Volume V in km<sup>3</sup> divided by Mean Annual Flow F in km3/year for the period 1960–1999. This ratio is usually called Residence Time (in years) and represents the regulation capacity of reservoirs in the basin. In the basins under study, it ranges across three orders of magnitude, from 0.01 years (Arno) to nearly 6 years (Segura).

The spatial support for the analysis is taken from the "Hydro1k" data set [37], derived from the Global 30 Arc-Second Elevation (GTOPO) 30 arc-second digital elevation model of the world. The dataset provides a digital elevation map and a set of topographically derived rasters at 1 km resolution, including streams and drainage basins divided into catchments. The original drainage basins in "Hydro1k" were processed to eliminate catchments which were too small (less than 1000 km<sup>2</sup> ), which were merged to neighboring catchments. The merging was always done with downstream areas and avoiding catchments including reservoirs. The reservoir storage volume in every catchment was obtained from the International Commission on Large Dams (ICOLD) World Register of Dams [38]. We selected dams in the register with more than 0.005 km<sup>3</sup> of storage capacity, excluding dams managed only for hydropower. The reservoirs were georeferenced and linked to the corresponding Hydro1k streams. All dams located in the same Hydro1k subbasin were grouped in an equivalent reservoir adding the storage volume and flooded area (to account for reservoir evaporation losses).

**Figure 1.** Southern European basins considered in this study.


**Table 1.** Basic characteristics of the basins analyzed in this study.

#### *2.2. Methodological Overview*

The methodological approach is presented in Figure 2. The analysis is structured in three steps: analysis of the forcing scenarios for water resources systems, analysis of system response in terms of potential water availability, and analysis of the sensitivity of system response to reservoir storage. The analysis of system forcing consists of the compilation of a large name of model runs producing monthly streamflow series in the basins under analysis, both for a historic control period and for a projected future period. Streamflow series for the control period were corrected for bias. Streamflow series for the future period were obtained under different climate scenarios. The scenarios were characterized in terms of the expected changes of mean, standard deviation, and coefficient of variation

of annual streamflow. The analysis of system response is focused on the estimation of the potential water availability allowed by current reservoir storage in the basins under analysis. Uncertainty of potential water availability is first characterized for the control and the future periods. Then, the elasticity of water availability to climate changes is explored by comparing changes in potential water availability to changes in mean annual flow, both for individual projections and for the distribution of all projections in each basin. The focus of the analysis is to explore how this elasticity is affected by reservoir storage and streamflow variability. The third step is focused on exploring the sensitivity to reservoir storage. The analyses of the previous step are repeated considering variable storage in each basin. The performance of the system is characterized by two new indices proposed in this study: the attenuation index and the uncertainty index. These indices describe how the performance of the water supply system is affected by changes in streamflow. The main conclusions of the study are obtained by comparing how these indices change as a function of reservoir storage for all basins.

**Figure 2.** Main methodological steps followed in the analysis. White circles show the figures that illustrate results from each step.

#### *2.3. Current and Future Runoff Scenarios*

The focus of the present study is the analysis of the role of reservoir storage to determine how water resources systems react to changes in hydrologic forcing. An effort was made to obtain a wide ensemble of scenarios that would represent the uncertainty linked to climate projections. Therefore, we chose to combine model results obtained under two sets of emission scenarios, the Special Report on Emission Scenarios (SRES) and Representative Concentration Pathways (RCP), in order to increase the size of the ensemble. Current and future runoff scenarios were compiled from three previous studies that include Southern Europe [39–41]. These studies were based on results from different climate models developed over the last 15 years under two sets of emission scenarios: SRES and RCP. They jointly describe the uncertainty that is currently challenging water managers.

The first set of scenarios was taken from the output of regional climate models from the PRUDENCE project [10]. The study by González-Zeas et al. [39] was based on the projections of surface runoff made by eight RCMs at 50 km resolution nested in a single global model, referred to as HadAM3H, in emission scenarios A2 and B2. They analyzed current (1960–1990) and future (2070–2100) time slices. The second set of scenarios was based on the results of the Regional Climate Models (RCMs) of the ENSEMBLES project [11]. The project produced many transient model runs for the time period from 1960 to 2100

using RCMs to characterize model uncertainty. The study by Garrote et al. [40] selected runoff output from four ENSEMBLES models at 25 km resolution under emission scenario A1B to study the major Mediterranean river basins of Europe. They worked with windows of analysis on the transient model runs in three time slices: historical (1960–1990), short term (2020–2050) and long term (2070–2100). In the first two sets of scenarios, monthly runoff time series were directly obtained from the "Total runoff" variable (*mrro*) produced by RCMs. The values of surface runoff flux available at the nodes of the native grid of the RCMs (50 km resolution in PRUDENCE and 25 km resolution in ENSEMBLES) were used to produce monthly runoff maps by interpolation at the finer grid provided by the Hydro1k dataset (1 km). The center of the RCM grid was taken as a point equal to the average for that cell. Interpolation was based on a weighted mean using the inverse of the distance squared as weight. These runoff maps were combined with the subbasin definitions of Hydro1k to obtain monthly streamflow values for each subbasin. The third set of scenarios was based on the results of the global hydrological model PCRaster GLOBal Water Balance (PCRGLOBWB) model [42] in the Inter-Sectorial Impact Model Intercomparison Project (ISIMIP) [43]. In ISIMIP, the PCRGLOBWB model was forced with five global climate models under historical conditions and climate change projections corresponding to four Representative Concentration Pathways scenarios: RCP-2., RCP-4., RCP-6. and RCP-8., corresponding to radiative forcing in the year 2100 of 2.6, 4.5, 6.0, and 8.5 W/m<sup>2</sup> , respectively. The study by Sordo-Ward et al. [41] used naturalized streamflow from PCRGLOBWB at 50 km resolution to analyze 1261 subbasins covering the entire territory of Western Europe. They considered two time slices in their analysis: historical (1960–1999) and long-term projection (2060–2099). The monthly streamflow time series in the subbasins were also obtained from monthly runoff maps derived from the runoff produced from the PCRGLOBWB model through interpolation at the Hydro1k 1 km grid.

A total of 16 model runs were compiled for the historical period (eight model runs from the PRUDENCE project, three model runs from the ENSEMBLES projects and five model runs from the PCRGLOBWB model). The windows of analysis in this period overlap for years 1960–1990. All these model runs produced different results in the basins under analysis. To assess the quality of these hydrological projections, the results obtained at the working scale of each model run were compared to a reference estimate of mean annual runoff under current conditions. The selected reference was the annual surface runoff layer (Global Composite Runoff Fields) of the University of New Hampshire Global Runoff Data Centre (GRDC) [44]. This data layer was produced by combining a database of observed river discharge information in more than 9900 gauging stations with a climate-driven water balance model to develop consistent runoff fields. The combination of direct readings from gauging stations with the water balance model preserves the spatial distribution of runoff generation and provides the best estimate of observed runoff over large domains. The mean values of the time series compiled for the historical period were compared with mean annual runoff produced by GRDC. The results are presented in Figure 3, which shows the scatterplot resulting from comparing catchment mean annual runoff produced from GRDC with that produced by model runs for the historical period. Model runs corresponding to the Special Report on Emissions Scenarios (SRES) (PRUDENCE and ENSEMBLES projects) show poor agreement. The models that performed best were Universidad de Castilla La Mancha (UCM) and Eidgenössische Technische Hochschule Zürich (ETHZ2), with coefficients of determination slightly lower than 0.6. This poor performance can be explained because runoff was obtained directly from RCM output. Model runs for the RCP scenarios (ISIMIP project) were produced by the hydrological model PCRGLOBWB. They show better performance, with coefficients of determination close to 0.7, but they reveal significant bias for low runoff. The discrepancies obtained in the comparison suggest that bias correction is necessary to overcome this very large model uncertainty. Using the monthly series of individual models without bias correction would imply significant distortion in the regulation provided by reservoirs in each basin. The ratio between reservoir storage capacity and mean annual flow would change for each model

run, affecting the evaluation of the regulation capacity provided by the reservoirs. For this reason, runoff derived from RCM results and from PCRGLOBWB was corrected for bias in each location. The chosen method for bias correction was linear scaling [45]. This method is justified by data availability, because GRDC only provides monthly long-term means of runoff. Therefore, all model projections for the historical period have the same mean.

**Figure 3.** Comparison of mean annual runoff in catchments produced from the Global Runoff Data Set (GRDC) dataset with that produced by model runs for the historical period.

The number of model runs compiled for the long-term climate change projection was 35: eight model runs corresponding to the A2 scenario, four model runs corresponding to the B2 scenario, three model runs for the A1B scenario, five model runs for RCP-2 scenario, five model runs for RCP-4 scenario, five model runs for RCP-6 scenario and five model runs for RCP-8 scenario. The windows of analysis in the long-term projection overlap for the years 2070–2099. These projections were corrected for model bias by applying the same correction as in the corresponding model in the historical period. This ensemble of climate projections was put together from different projects developed over a 15-year period, running a range of global climate models under two sets of emission scenarios, and applying different methodologies. It can thus be considered a representative description of the range of scenarios that climate change science is projecting for the region. However, it should be noted that runoff projections derived from climate models are uncertain. Climate models provide a good overall representation of climate, but their performance degrades at the scale of individual grid boxes, indicating that they are not skillful at their smallest scale. The performance of RCMs generally improves after suitably removing bias. However, model errors still remain large, particularly for climatic variables relevant for hydrology, like precipitation or runoff [46]. Given this inherent uncertainty, a basic hypothesis of this work is that water management decisions based on the global analysis of a wide range of

projections produces better results than decisions based on a very detailed analysis of a reduced number of projections.

The average annual runoff obtained from GRDC in the period 1960–2000 was also used to characterize the basins under analysis. The relationship between Specific Runoff and reservoir Residence Time is plotted in Figure 4 for all Hydro1k basins in Southern Europe, highlighting the 16 basins under analysis. As can be seen in Figure 4, there is a clear relation between both variables, with larger values of storage corresponding to basins with lower values of specific runoff. The selected basins produce a good coverage of the possible range of behaviors found in the region, from basins with large specific water resources and low storage capacity like Arno, Po or Loire, to others in the opposite situation with very low water resources and large storage volumes as Júcar, Guadiana or Segura.

F/A V/F **Figure 4.** Relation between Specific Runoff (F/A) and reservoir Residence Time (V/F) for Hydro1K basins in Southern Europe. The 16 basins under study are highlighted using the same color coding and numbering as in Figure 1.

#### *2.4. Water Availability Analysis*

The study is based on the analysis of how climate change affects water availability in the different basins, and how this affect is modified by available reservoir storage. Potential Water Availability (PWA) is defined as the annual water demand that can be satisfied in a point of the drainage network with a given reliability. PWA depends on the mean and variability of the streamflow series, the storage available for flow regulation, the monthly distribution of the demand and the reliability indicator adopted in the analysis. In this study, PWA was estimated with the Water Availability and Adaptation Policy Analysis (WAAPA) model [31,47]. WAAPA simulates the operation of a complex water resources system with many reservoirs. The basic topological unit of WAAPA is the river network. The main components are inflows, reservoirs and demands, all linked to nodes in the network. WAAPA computes the amount of water supplied to demands from a system of reservoirs accounting for ecological flows and evaporation losses. Input data for WAAPA are monthly inflows in relevant points of the river network, monthly demand values, and reservoir data. Reservoirs are described by monthly maximum and minimum capacity, storage-area relationship, monthly rates of evaporation, and monthly required environmental flow. WAAPA applies an algorithm with simple operating rules, where

all reservoirs in the basin are jointly managed to satisfy the set of demands, drawing water preferably from reservoirs located upstream. This algorithm is applied to potential demands located in every node in the river network, and therefore water availability is obtained for the entire river network. The main results of WAAPA are time series of monthly volumes supplied to each demand, monthly storage values and monthly values of spills, environmental flows, and evaporation losses in every reservoir. From this output, demand reliability can be computed for the criterion of choice (volume reliability, time reliability at the monthly or annual scale, or more complex criteria).

WAAPA can obtain PWA for a given demand reliability criterion through an iterative scheme that changes local demand values until the reliability criterion is met with a given precision. In this study, PWA is estimated by considering only one type of demand in the system, with constant monthly distribution. This choice was made because the true monthly distribution of demands in each model node is unknown. Results therefore should be considered only approximate and could be fine-tuned if the ratio between urban and irrigation demand was known in every model node. Ecological flows were specified as the 10% percentile of the monthly marginal distribution of natural flows. System performance is evaluated as gross volume reliability. PWA is obtained for 92% volume reliability. This reliability level was chosen as an intermediate value between reliabilities required from urban demands (usually close to 100%) and those required from irrigation demand (usually close to 90%), assuming an approximate distribution of 20% urban demand and 80% irrigation demand, which is typical of Portugal, Spain, and Greece [48].

#### **3. Results and Discussion**

The WAAPA model was run for the European Mediterranean region for the 16 hydrologic scenarios corresponding to the historical period (1960–2000) and for the 35 hydrologic scenarios corresponding to climate projections for the long-term time horizon 2070–2100. The long-term time horizon was chosen for two reasons. Firstly, the results from PRU-DENCE project were only available for this time horizon. Secondly, the changes in the long-term time horizon are usually more accentuated than in the mid-term time horizon and the effects are more apparent. Results were obtained for all catchments in the Hydro1k dataset, but, for the sake of simplicity, we only present global results for the 16 basins under analysis. We first analyze the climate projections, then we present the results obtained for PWA in the basins. Average values of these results are summarized in Table 2 and presented and discussed in detail in the following section. Finally, the role of storage is studied through a sensitivity analysis.

#### *3.1. Climate Projections*

We first present the characterization of climate projections for the basins under study. Climate projections were taken from the runoff variable of RCM models in the PRUDENCE and ENSEMBLES projects (under SRES emission scenarios) and of the PCRGBLOBWB hydrologic model (under RCP emission scenarios). Mean and coefficient of variation of annual flows were computed for each basin during the historical period and during the long-term projection. Changes in the long-term projection were estimated taking the control period as a reference, applying the following expressions:

$$
\Delta \text{F} = \frac{\text{F}\_{\text{PROJ}} - \text{F}\_{\text{HST}}}{\text{F}\_{\text{HST}}} ; \,\Delta \text{SD} = \frac{\text{SD}\_{\text{PRO}\text{J}} - \text{SD}\_{\text{HST}}}{\text{SD}\_{\text{HST}}} ; \,\Delta \text{CV} = \frac{\text{CV}\_{\text{PRO}\text{J}} - \text{CV}\_{\text{HST}}}{\text{CV}\_{\text{HST}}} \tag{1}
$$

where F is Mean Annual Flow, SD is the Standard Deviation of the annual time series of streamflow, and CV is the Coefficient of Variation of the annual time series of streamflow (standard deviation of the annual time series divided by mean annual flow). The subindices HIST and PROJ refer to the historical period and to the long-term projection.


**Table 2.** Summary of the results of the analysis of changes in streamflow ∆F, ∆SD and ∆CV and Potential Water Availability, ∆PWA, in the basins analyzed in this study (Ave: average of values for the 35 projections; Std: standard deviation of the values for the 35 projections).

> The results are depicted in Figure 5, which compares the relative changes in Standard Deviation (∆SD) and Coefficient of Variation (∆CV) of annual flows versus changes in Mean Annual Flow (∆F) for the 35 available projections in the 16 basins under study. All projections are shown together in the left plots of Figure 5, showing for basins the same color codes as in Table 1 and Figure 1. The plots on the right show the mean value for each basin. A plot of each basin is available in the Supplementary Materials, showing individual projections. Projections under SRES emission scenarios are represented as plus signs and projections under RCP scenarios are represented as circles. The analysis of chart (a) of Figure 5 shows positive correlation between changes in Mean Annual Flow ∆F and Standard Deviation ∆SD. If the changes of F and SD were similar, the scatter plot of Figure 5a would be centered around the main diagonal (highlighted in grey). The mean values of changes are above the main diagonal for all basins, suggesting a relative increase of variability in future projections. The joint analysis of all projections for all basins in chart (c) of Figure 5 shows negative correlation between changes in Mean Annual Flow ∆F and Coefficient of Variation ∆CV: reduction of F and increase of CV. The general shape of the scatter plot is similar in all basins in Southern Europe. This has clear implications for water management since both factors will negatively impact water availability. This tendency is stronger for basins with larger residence times that, as seen in Figure 4, are located in water scarce regions, already facing strong hydrologic irregularities. The dispersion of results is stronger for basins with larger residence times, presenting an additional challenge for water management. The ensemble of projections, jointly considered, suggests that water managers should be ready to cope with less abundant and more variable water resources in the future. Given the large dispersion of results, water managers should also be ready to deal with greater year-on-year variability or extreme events than in the past. Figure 5 also shows that expected changes in CV are much larger than changes in F, with many basins reaching extreme values close to 2 (a 100% increase). The basins showing more extreme projections are Guadalquivir, Júcar and Guadiana.

ΔSD ΔCV ΔF ΔF vs Δ Δ Δ Δ Δ Δ Δ **Figure 5.** Relative changes in Standard Deviation (∆SD) and Coefficient of Variation (∆CV) of annual flow versus changes in Mean Annual Flow (∆F) for the 16 basins under study. (**a**) ∆F vs. ∆SD, all projections; (**b**) ∆F vs. ∆SD, mean values; (**c**) ∆F vs. ∆CV, all projections; (**d**) ∆F vs. ∆CV, mean values.

#### *3.2. Water Availability*

The WAAPA model was used to compute Potential Water Availability (PWA) for the historical period and for the long-term projection in the 16 basins under analysis. The results are shown in Figure 6, which presents the value of PWA obtained in each basin as a function of the relative rank of the corresponding projection. All 35 projections were used to prepare this figure, thus mixing projections under SRES and RCP emission scenarios. An individual plot of each basin is included in the Supplementary Materials, where the joint distribution is compared to the distributions of both sets of emission scenarios. The corresponding emission scenario is identified for each model run available in the long-term projection. These plots show that there is no clear correlation between the emission scenario and the projected PWA. Values corresponding to different emission scenarios are mixed and the most extreme scenarios (A2 and RCP-8) do not always produce the minimum values for PWA.

PWA is expressed as a fraction of Mean Annual Flow (F) in the historical period. Results for the historical period are shown in the upper chart (a) and results for the long-term projection are shown in the lower chart (b). If all model runs were assumed equiprobable, this plot would correspond to the empirical estimation of the probability distribution function of PWA expected in each basin. The results show that the relative value of PWA to F tends to be larger for basins with larger storage capacity, both in the historical and in the projection periods. This fact clearly illustrates the effectiveness of reservoir storage to increase water availability. The plots also show large uncertainty in the estimation of PWA. For the historical period, this result is remarkable because historical time series were corrected for bias with respect to the GRDC estimation of F and therefore all had the same Mean Annual Flow. The uncertainty in PWA reflects model uncertainty because the differences in PWA can only be attributed to the differences in the seasonal and interannual variability of the time series produced by each model run. Unfortunately, the skill of the models to reproduce current hydrological irregularity cannot be evaluated because there are no available regional data sets for Southern Europe on interannual naturalized streamflow variability.

**Figure 6.** Estimated cumulative probability distribution function of Potential Water Availability (PWA PWA) expressed as a fraction of current Mean Annual Flow (F) in the 16 basins under study. (**a**) historical period; (**b**) long-term projection.

Except in the Arno basin, PWA is expected to decrease significantly in the long-term projection with respect to the historical period, with average reductions between 15% and 35%. These reductions are the consequence of reduced F and increased CV. The most significant reductions are projected for the basins of South Western Europe: Guadiana and Guadalquivir (35% on average) and Duero (28% on average). The uncertainty of PWA in the long-term projection is larger than that in the historical period due to the additional variability introduced by emission scenarios. However, the large model uncertainty hinders the interpretation of results obtained for different emission scenarios.

The estimated changes in PWA are compared to estimated changes in F in Figures 7 and 8. Figure 7 shows the scatter plot of changes in both variables for the set of emission scenarios analyzed in all basins. A plot of each basin is available in the Supplementary Materials, showing individual projections. Projections under SRES emission scenarios are represented as plus signs and projections under RCP scenarios are represented as circles. Figure 8 shows the comparison of the estimated probability distributions of F and PWA. All 35 projections were used to prepare this figure, thus mixing projections under SRES and

RCP emission scenarios. An individual plot of each basin is included in the Supplementary Materials, where the joint distribution of PWA is compared to the distributions of both sets of emission scenarios. HIST

PROJ − HIST

Δ

ΔF Δ Δ Δ Δ Δ **Figure 7.** Comparison of the estimated changes in Mean Annual Flow (∆F) and the estimated changes in Potential Water Availability (∆PWA) for the 35 available projections in the 16 basins under study. (**a**) ∆F vs. ∆PWA, all projections; (**b**) ∆F vs. ∆PWA, mean values.

∆F <sup>∆</sup>PWA **Figure 8.** Estimated cumulative probability distribution function of changes in Mean Annual Flow (∆F, in gray) and changes in Potential Water Availability (∆PWA, in the color code for each basin) for the 16 basins under analysis.

In Figures 7 and 8, the changes of F are estimated from the first expression shown in Equation (1). The changes of PWA are similarly estimated from the comparison of values obtained in the long-term projection and the control period for the same model, applying the following expression:

$$
\Delta \text{PWA} = \frac{\text{PWA}\_{\text{PRO}\,\text{J}} - \text{PWA}\_{\text{HIST}}}{\text{PWA}\_{\text{HIST}}} \tag{2}
$$

where PWA is Potential Water Availability and the sub-indices HIST and PROJ refer to the historical period and to the long-term projection.

The mean values plotted in chart (b) of Figure 7 reveal that basins with small storage capacity (Po, Loire, Tiber, Garonne and Rhône) show a larger reduction of PWA than the reduction of F. The Arno basin is the exception, with no reduction of PWA despite a small average reduction of F. The basins with larger storage capacity, in general, show a smaller reduction of PWA than the reduction of F. Struma-Strymon, Vardar-Axios, Ebro, Maritsa-Evros, Guadalquivir, Tajo-Tejo, Júcar, Guadiana and Segura belong to this group. Duero-Douro is an exception, with larger reduction of PWA than of F. This may be explained because Duero-Douro shows the largest difference between change in F and change in SD. The wide scatter of changes in F and PWA in chart (a) of Figure 7 shows that there is no exact relation between changes in Mean Annual Flow (∆F) and changes in Potential Water Availability (∆PWA). For individual projections, changes in PWA may be larger, equal or smaller than changes in F. This is, in part, a consequence of changes in hydrologic variability, which may explain why negative changes in F produce positive changes in PWA and vice versa. However, the comparison of Figures 5 and 7 shows that changes in hydrologic variability alone cannot explain the diversity of behaviors seen in Figure 7. Hydrologic variability is measured in terms of Coefficient of Variation of annual flows and is therefore referred to interannual variability. Basins with small storage capacity show a behavior more exposed to changes in CV because, for them, water availability is almost directly determined by short-duration dry periods of the streamflow series. These dry periods show a large variability among model runs, which explains the variability observed in values of PWA. As basin storage grows larger, the reservoirs attenuate the effect of short-duration dry periods and the interannual variability becomes less important. Basins with storage capacity larger than mean annual flow show a much less sensitivity to changes in the coefficient of variation of mean annual flows.

Figure 8 is useful to analyze the effect of the uncertainty on emission scenarios. The estimated probability distributions shown in Figure 8 reveal a wide range of behaviors. The basins were classified in five groups (A1, A2, A3, B1 and B2), according to the relative value of the distributions of changes in F and PWA. Group A1 is integrated by basins where the distribution of expected reductions in PWA is to the left of the distribution of expected reductions in F, suggesting that the availability of reservoir storage tends to dampen the effect of climate change. Struma-Strymon, Vardar-Axios, Guadalquivir and Tajo belong to this group. In the second group, A2, the distributions of expected changes in F and PWA are very similar. This group is formed by Ebro, Júcar and Segura. The only basin in Group A3 (Douro-Duero) presents larger expected reductions in PWA than in F. In group B, the probability distributions of F and PWA cross each other. In group B1, the distribution of changes in PWA is to the left of the distribution of changes in F for low probability values. For high probability values, the distribution of changes in PWA is to the right of the distribution of changes in F. This results in larger uncertainty for changes in PWA than in F. This effect may be due to increased exposure to changes in variability due to lack of regulation storage. Arno, Po, Loire, Tiber, Garonne, Rhône and Maritsa-Evros belong to group B1. The only basin in group B2 is Guadiana, where the distribution of changes in PWA is to the right of the distribution of changes in F for low probability values and to the left for high probability values. Guadiana shows less uncertainty in changes of PWA than in changes in F, due to its large reservoir storage.

A remarkable effect shown in Figure 8 is that the uncertainty regarding changes in F (gray line) seems to grow as specific storage grows. The larger spreads of the estimated probability distributions appear in basins with larger specific storage, in the bottom row.

Basins with comparatively smaller storage capacity, in the first row, show much less uncertainty on changes in F. This shows that reservoir storage was developed where it was required: in basins with large hydrologic variability. Furthermore, the difference in uncertainty between changes in F and changes in PWA, which is large in basins with small storage, is progressively reduced as specific storage increases.

#### *3.3. The Influence of Storage*

The results obtained in Section 2.2 suggest that reservoir storage plays a relevant role in controlling how projected changes in Mean Annual Flow may be translated into changes in Potential Water Availability. However, the large variability of local conditions in the studied basins introduces uncertainties in the analysis. In this section we further explore the influence of reservoir storage on changes in water availability through a sensitivity analysis that discounts for local conditions. We repeated the water availability analysis but considering different storage volumes in each basin. Potential Water Availability was computed in current and future scenarios in the 16 basins, assuming changing reservoir volumes of 25%, 50%, 75%, 100%, 125%, 150% and 175% of current storage. Storage was proportionally reduced or increased in the same location of existing reservoirs. This choice was made for convenience, without any implications for projected future evolution of storage in the region. In fact, the most likely scenario in the future for European basins is a progressive reduction of available storage due to reservoir sedimentation, with very little additional storage being built. Figure 9 shows the scatter plots of changes in F versus changes in PWA for four basins covering a wide range of values of reservoir storage: Loire (V/F = 0.03), Ebro (V/F = 0.30), Guadalquivir (V/F = 0.73) and Guadiana (V/F = 3.35). Individual plots for all basins are included in the Supplementary Materials. Projections under SRES emission scenarios are represented as diamonds and projections under RCP scenarios are represented as circles. Results shown in Figure 9 reveal that the dispersion of the scatter plot gets reduced as the storage capacity is increased. This effect is more marked for the Guadiana basin, which has the largest reservoir storage.

In order to assess the global behavior, the values obtained for changes in F and PWA were classified according to the relative rank of the corresponding projection, obtaining an empirical estimate of their probability distributions, under the assumption that all scenarios analyzed are equally likely. The results are shown in Figure 10, which presents the estimate of the probability distribution of changes in Mean Annual Flow (∆F, blue line) and changes in Potential Water Availability (∆PWA) for different storage values in colored lines from brown (25% of current storage volume) to green (175% of current reservoir storage). The basins analyzed showed variable sensitivity to storage. Some basins, like Arno, Ebro, Maritsa-Evros or Guadiana, showed very little sensitivity to storage capacity because the distributions of expected changes of PWA are very similar. Other basins, like Po, Tiber, Rhône or Struma-Strymon, presented significant differences in behavior depending on the storage volume assumed. In some of these basins, the estimated probability distributions of changes in PWA for high storage values (green color) were located to the right, indicating less reductions of PWA. Po, Loire, Rhône, Duero-Douro, Ebro, Júcar, Guadiana and Segura show this behavior. For other basins, however, the probability distributions for high storage values were located to the left. Tiber, Struma-Strymon and Guadalquivir belong to this group.

This range of behaviors illustrates the complex relations between hydrologic variability and reservoir storage in determining water availability in climate change scenarios, suggesting that a specific analysis for local conditions is required to translate projections of changes in mean annual flow into projections of changes in water availability in basins with significant storage capacity.

∆F <sup>∆</sup>PWA **Figure 9.** Comparison of the estimated changes in Mean Annual Flow (∆F) and the estimated changes in Potential Water Availability (∆PWA) for 25%, 50%, 75%, 100%, 125%, 150% and 175% (in rows, ordered from top to bottom) for four representative basins of the study: Loire (left column), Struma-Strymon (center-left column), Guadalquivir (center-right column) and Guadiana (right column).

∆

∆

∆F <sup>∆</sup>PWA **Figure 10.** Probability distribution of changes in Mean Annual Flow (∆F, in blue) and changes in Potential Water Availability (∆PWA) for different storage values (color-coded, from 25% to 175% of current storage volume) in the 16 studied basins.

The influence of reservoir storage on the elasticity of water availability was analyzed by computing an attenuation index IA, defined in the following expression:

A

$$\mathbf{I}\_{\mathbf{A}} = \Delta \mathbf{PWA} - \Delta \mathbf{F} \tag{3}$$

<sup>A</sup> ∆ − ∆ ∆ ∆ where ∆PWA is the change in Potential Water Availability and ∆F is the change in Mean Annual Flow. As seen in the previous sections, most changes in PWA and F are reductions and therefore ∆PWA and ∆F are negative. A positive value of this index indicates an attenuation of the effect of climate change: the absolute value of ∆PWA is smaller than the absolute value ∆F.

∆ ∆ ∆ ∆ A We explore how the attenuation index I<sup>A</sup> changes with reservoir storage. The results are shown in Figure 11, which presents the value of the attenuation index as a function of reservoir storage in each basin for all available projections (thin grey lines), the average values (solid lines in the color code corresponding to the basin) and average values plus and minus one standard deviation (dotted lines in the color code corresponding to the basin). The results show a large variability for individual projections, which translates into large uncertainty for water managers. The variability of the attenuation index appears to be progressively reduced as specific storage grows across basin locations (from top row to bottom row). This suggests that reservoir storage plays a relevant role in reducing uncertainty on the effects of climate change projections on water availability.

IA **Figure 11.** Values of the attenuation index I<sup>A</sup> for different relative storage volumes in the 16 studied basins. Current relative storage is marked with a vertical black line.

A U The effect of reservoir storage on the variability of the attenuation index I<sup>A</sup> is further explored by analyzing the uncertainty index IU, defined as:

$$\mathbf{I}\_{\tilde{\mathbf{U}}} = \sigma(\mathbf{\hat{A}} \mathbf{\hat{B}} \mathbf{\hat{N}} \mathbf{A}) - \sigma(\boldsymbol{\Delta} \mathbf{\hat{S}}) \cdot \boldsymbol{\Delta} \tag{4}$$

σ ∆ σ ∆ U where σ(∆PWA) is the standard deviation of the changes in Potential Water Availability for all projections and σ(∆F) is the standard deviation of the changes in Mean Annual Flow for all projections. I<sup>U</sup> index compares the variability of the projections of changes in water availability to that of the projections of mean flow. A negative value of this index indicates a reduction of the uncertainty of climate change projections: the variability of ∆PWA is smaller than the variability ∆F.

∆ ∆ A U A The summary of results found in the analysis of the I<sup>A</sup> and I<sup>U</sup> indices is shown in Figure 12. Chart (a) of Figure 12 compares the average of the values of the I<sup>A</sup> index obtained in the sensitivity analyses of reservoir storage for all basins. Chart (b) of Figure 12 represents the corresponding values of the I<sup>U</sup> index.

U The plots shown in Figure 12 indicate that increased reservoir storage results in larger values of the attenuation index I<sup>A</sup> in most basins and smaller values of the uncertainty index I<sup>U</sup> in all basins. The results for the attenuation index are less conclusive than those for the uncertainty index. Out of the 16 basins analyzed, I<sup>A</sup> is observed to decrease with increasing reservoir storage in four basins: Tiber, Garonne, Struma-Strymon and Guadalquivir. In the case of Garonne and Struma-Strymon, the decrease only covers the range from 25% to 100% of current reservoir storage. For storage volumes between 100% and 175% of current reservoir storage I<sup>A</sup> is increasing with increasing reservoir storage. The case of Tiber basin may be explained because reservoirs only cover 12% of the contributing area and therefore 88% of the flow is unregulated. Guadalquivir basin is exposed to the most extreme reduction of Mean Annual Flow (43% on average) and this may have an influence on the observed behavior. In the case of the uncertainty index, the reduction with increasing

storage is observed for all basins. These results are valid for the range of storage volumes explored in the sensitivity analysis in each basin individually and for all basins as a whole, regardless of basin size and location in Southern Europe, and therefore show a clear picture of the role played by reservoir storage in attenuating the impact of reduced streamflow on water availability and on reducing the uncertainty of climate change projections. It is unlikely that reservoir storage will be further increased in Southern Europe. Most basins already have an adequate amount of storage and additional storage capacity would not increase water availability in a scenario of decreasing resources. However, water managers should be aware that proper management of currently available storage will be helpful to address the challenges posed by climate change.

<sup>I</sup><sup>A</sup> <sup>I</sup><sup>U</sup> IA U **Figure 12.** Comparison of the values of average I<sup>A</sup> index and I<sup>U</sup> index obtained in the sensitivity analyses of reservoir storage in the 16 basins under study. (**a**) Average value of the attenuation index (IA) for the 35 available projections as a function of reservoir storage. (**b**) Value of the uncertainty index ( IU) as a function of reservoir storage.

#### **4. Conclusions**

A U A A Projected changes in hydrologic regime and water availability were analyzed in 16 basins in Southern Europe applying the WAAPA model to streamflow time series obtained from 35 climate projections under 7 emission scenarios. The analysis of climate projections concluded that a significant reduction of mean annual flow can be expected in most basins. The reduction in the mean is supplemented by a strong increase in the coefficient of variation, due to an increase of the variability of the projected series. This analysis is uncertain due to the very large variability introduced by the different models and emission scenarios examined. The overall result implies a corresponding reduction in potential water availability, with variable results across basins depending on hydrologic regime and reservoir storage. Basins with large storage values showed reductions of water availability comparable to the reductions of mean annual flow. Basins with small storage capacity showed a larger reduction of water availability than the reduction of mean annual flow. Although model and emission scenario uncertainties are larger than the expected reduction of water availability, a consistent picture emerges from the joint analysis of all projections, requiring significant adaptation measures to compensate for the projected reduction of water availability.

The influence of reservoir storage on basin response to climate change was studied through a sensitivity analysis where current reservoir storage was modified to examine its effects on water availability with values ranging from 25% to 175% of current storage values. The results showed very large variability, which illustrates the complex interplay between hydrologic regime, reservoir storage and water availability. Two indices were introduced to clarify the overall behavior: the attenuation index and the uncertainty index. The attenuation index compares the changes in water availability to the changes in mean

annual flow. Positive values of this index indicate an attenuation of the impact of climate change projection on water availability. The uncertainty index compares the variability of changes in water availability and in mean annual flow. Positive values of this index indicate a reduction of the uncertainty of climate change projections on water availability. The results of the sensitivity analysis showed that increasing reservoir storage attenuates the reduction of water availability and reduces the uncertainty of climate projection. The results are valid for each individual basin within the range of storage volumes examined and for the set of 16 Southern European basins analyzed in this work. The effect observed for reservoir storage is a positive factor for system managers since decisions become harder as uncertainty grows. This feature would allow water managers to develop suitable policies to mitigate the impacts of climate change, thus enhancing the resilience of the system.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/2073-4 441/13/1/85/s1: High resolution images of Figures 1–12 and individual plots for each basin in Figures 5–11.

**Author Contributions:** Conceptualization, L.G., A.G. and A.S.-W.; methodology, A.G.; data processing, B.P.-B.; software, L.G. and B.P.-B.; validation, A.S.-W., A.G. and L.G.; writing—original draft preparation, A.G.; writing—review and editing, L.G.; visualization, A.S.-W.; supervision, L.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Spanish Ministry of Science and Innovation, grant number PID2019-105852RA-I00: "Simulation of climate scenarios and adaptation in water resources systems (SECA-SRH)". B.P.-B. would like to acknowledge Universidad Técnica de Ambato for the financial support through its doctoral student mobility program (award No. 1886-CU-P-2018 Resolución HCU).

**Acknowledgments:** Data from the PRUDENCE and ENSEMBLES projects were used in this work. PRUDENCE was funded by the EU FP5 (EVK2-CT2001-00132) and ENSEMBLES was funded by the EU FP6 (contract No. 505539). Support from both projects is gratefully acknowledged. Moreover, the authors acknowledge the World Climate Research Programme's Working Group on Regional Climate, and the Working Group on Coupled Modelling, former coordinating body of CORDEX and responsible panel for CMIP5; and also thank the climate modelling groups for producing and making available their model output. Finally, the authors also acknowledge the Earth System Grid Federation infrastructure an international effort led by the U.S. Department of Energy's Program for Climate Model Diagnosis and Intercomparison, the European Network for Earth System Modelling and other partners in the Global Organisation for Earth System Science Portals (GO-ESSP).

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**


### *Article* **Evaluating Water Resource Accessibility in Southwest China**

#### **Tao Li 1,2 , Sha Qiu 1,2 , Shuxin Mao 3 , Rui Bao 1,2 and Hongbing Deng 1, \***


Received: 18 July 2019; Accepted: 14 August 2019; Published: 16 August 2019

**Abstract:** The accessibility, quantity, and quality of water resources are the basic requirements for guaranteeing water resource security. Research into regional water resource accessibility will contribute to improving regional water resource security and effective water resource management. In this study, we used a water resource accessibility index model considering five spatial factors to evaluate the grid-scale water resource accessibility and constructed the spatial pattern of water resource accessibility in Southwest China. Then, we analyzed the coupling coordination degree between county-level water resource accessibility and eco-socio-economic water demand elements. The water resource accessibility showed obvious regional differences, and the overall trend gradually decreased from Southeast to Northwest. The coupling coordination degree between county-level water resource accessibility and eco-socio-economic water demand elements was between 0.26 and 0.84, and was relatively low overall, whereas the counties (districts) with high coordination, moderate coordination, low coordination, reluctant coordination, and incoordination accounted for 0.92%, 5.31%, 21.06%, 59.71%, and 13.00% of total counties (districts), respectively. Therefore, the Southwest region needs to further strengthen the construction of its agricultural irrigation facilities, protect the water resources, and coordinate the relationship between water resource management and water demand elements to comprehensively guarantee regional sustainable development.

**Keywords:** water resource accessibility; spatial pattern; coupling coordination degree; water resource management; Southwest China

#### **1. Introduction**

Water resources are essential for maintaining the sustainable development of eco-socio-economic systems [1,2]. However, due to climate change, economic growth, population increase, and improper water resource management, many environmental problems, such as serious water pollution, the deterioration of the water environment, and an increased contradiction between water supply and demand, have become increasingly severe, creating strategic problems worldwide [3–7]. The United Nations (UN) estimates that more than one-third of the population on the planet will face a freshwater crisis by 2030 [8], and residents in regions with relatively abundant water resources will still have to spend time and energy to obtain water resources, which not only increases their living costs but also impacts health [9,10]. Therefore, whether people living in different regions can easily and fairly obtain sufficient water resources to meet their water demand has become a hot topic for academics and policymakers.

Accessibility is a broadly accepted concept in various scientific fields such as urban planning, transportation planning, and geography [11], and was initially used to measure potential interaction

opportunities in transportation networks [12]. As research progressed, accessibility was defined as the difficulty in reaching a destination from a given location [13]. There are two main manifestations of accessibility: The number of opportunities or benefits that can be obtained within a given time or distance, and the amount of spatial resistance that needs to be overcome to reach a destination [14,15]. Therefore, the measurement of accessibility depends primarily on the spatial distribution of potential destinations and the spatial resistance that needs to be overcome to reach each destination [16,17]. The former is mainly measured by quantity or quality, reflecting the attractiveness of the destination, and the latter is mainly measured by indicators such as time, distance, or cumulative cost, reflecting the convenience of reaching the destination [17,18].

The quantity, quality, and accessibility of water resources are the basic requirements for ensuring water resource security [19,20]. However, the existing water resource evaluations mostly focus on the assessment of water quality and water quantity [21–25], but less attention has been paid to water resource accessibility [10,26–29]. Water resource accessibility refers to the difficulty of obtaining water resources from water sources [30], which is the fundamental factor determining the quantity, quality, and the efficiency of the water supply. Water resource accessibility is essential for human well-being, economic development, and ecological maintenance [31] and includes both spatial accessibility and time accessibility [10,28], often measured by indicators such as water intake distance (Euclidean distance, cost distance, path distance, etc.) and water collection time (shortest distance time, shortest path time, self-report time, etc.) [32–34]. However, these indicators ignore the impact of water quantity and various spatial resistance factors. Therefore, some development space remains within the existing quantitative research for the examination of water resource accessibility, and the quantitative methods need to be further improved.

Southwest China is the source and upstream of many rivers, and is also an important ecological barrier zone. It plays a key role in maintaining the ecological and socio-economic security of East China, South China, and even Southeast Asia [35]. The region has abundant rainfall and a large amount of water resources. However, due to the uneven water distribution in time and space, coupled with the limited infrastructure and the influence of complex topography, the use of water resources is difficult and costly, resulting in serious seasonal, regional, and engineering water shortages [36]. With the rapid population and economy growth, the demand for water resources in Southwest China continues to increase, and the misalignment between supply and demand is becoming increasingly acute. Therefore, we used a water resource accessibility index model considering five factors—runoff, slope, relative height difference, water intake distance and land use resistance—to evaluate the grid-scale water resource accessibility in Southwest China, using the ArcGIS platform (Environmental Systems Research Institute, Redlands, California, America) to construct the spatial pattern of water resource accessibility. Then, we analyzed the coupling coordination degree between water resource accessibility and eco-socio-economic water demand elements. The aims are to improve the water resource accessibility evaluation method, identify the areas with relatively low water resource accessibility and the key regions that the coupling coordination degree between water resource accessibility and eco-socio-economic water demand elements is relatively low, which is important for improving the determination of regional water resource security levels, strengthening regional water resource management allocation, and effectively implementing water conservancy facilities planning and urban development planning.

#### **2. Materials and Methods**

#### *2.1. Study Area*

The study area is located in Southwest China, including Chongqing Municipality, Sichuan Province, Guizhou Province, Yunnan Province, and the Guangxi Zhuang Autonomous Region, between 97◦21′–112◦3 ′ E and 20◦53′–34◦18′ N and covers a total area of 1.362 million km<sup>2</sup> (Figure 1). Southwest China is one of the three karst-concentrated contiguous areas in the world; the terrain in the area is complex and diverse, and the landforms are mainly plateaus and mountains in which basins

and hills are widely distributed. The altitude difference is large, and the average elevation is as high as 1700 m. At the end of 2015, the region had a total resident population of 242.89 million, with an urbanization rate of 47.5%, the gross domestic product (GDP) was 8669.52 billion yuan, and a farmland irrigation area was 7.86 million ha [37–41].

**Figure 1.** Location of study area in Southwest China (Note: The digital elevation model is abbreviated as DEM).

Southwest China is located in a tropical and subtropical humid region and is dominated by tropical and subtropical monsoon climates, with sufficient heat and abundant rainfall; however, rainfall is unevenly distributed in space and time. The region has developed water systems, including the Yangtze River, Yellow River, Irrawaddy River, Nujiang River, Lancang River, Yuanjiang River and Pearl River, which are the most important water resource enrichment areas in China [32]. In 2015, the average precipitation in the whole study area was 1211 mm, the total water resources amounted to 813.59 billion m<sup>3</sup> , and the per capita water resource was 3350 m<sup>3</sup> . The annual water supply was 89.13 billion m<sup>3</sup> ; and the agricultural, industrial, domestic, and ecological water consumptions were 54.31 billion m<sup>3</sup> , 19.19 billion m<sup>3</sup> , 14.48 billion m<sup>3</sup> , and 1.15 billion m<sup>3</sup> , respectively; but the water resource development use rate was only 11.0% [42–46], which is related to the difficulty in using water resources in the region [36].

#### *2.2. Data Sources*

Four main types of data sources were used in this study. Administrative boundary vector data, water system vector data, and digital elevation model (30 × 30 m) were downloaded from Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences [47]. The land use type (30 × 30 m) and the normalized vegetation index (1000 × 1000 m) were derived from Resource and Environment Data Cloud Platform, Chinese Academy of Sciences [48]. The meteorological and hydrological data were sourced from National Meteorological Information Center [49], in which the runoff coefficient was derived from the water resource bulletin of each province (municipality and autonomous region) [42–46]. Socioeconomic data (GDP, population, food production) were derived from the statistical yearbooks of provinces (municipality and autonomous region) [37–41]. The spatial data coordinate system had a unified projection of WGS\_1984\_Albers, and the resolution after data resampling was 90 × 90 m.

#### *2.3. Methods*

#### 2.3.1. Basic Theory and Hypothesis

Water resource accessibility is closely related to a series of natural and human factors. These natural factors mainly include the water source, distance, relative height difference, slope, and land use type, whereas human factors mainly include funds (income, water fee), infrastructure (water supply pipelines, waterworks), and technology (irrigation technology, water treatment technology). In this study, rivers, lakes, and reservoirs were selected as the main water sources. We assumed that water users collect water from the nearest water source, and the water resource accessibility is mainly affected by natural factors such as runoff, water intake distance, relative height difference, slope, and land use type, and ignoring human factors. Considering the amount of water is still a problem in Southwest China, and incorporating water quality elements would increase the calculation and interpretation complexity, the water quality was not measured.

Different spatial resistance factors have significant impacts on the amount of accessible water and the difficulty in obtaining water. The runoff is positively correlated with the water resource accessibility: The greater the runoff, the greater the available water, and the higher the water resource accessibility. The other factors (slope, water intake distance, relative height difference, and land use resistance) are negatively correlated with the water resource accessibility: The larger the factors, the greater the water intake difficulty, and the lower the water resource accessibility. The calculation methods of each spatial element index in this study are as follows: (1) Runoff. We first selected the rainfall data of the main meteorological stations in the study area to interpolate the rainfall. Then, we calculated the runoff by combining the spatial distribution of the average runoff coefficient of each administrative unit and the rainfall. (2) Water intake distance. We used the Euclidean tool in ArcGIS software (Environmental Systems Research Institute, Redlands, California, America) to calculate the distance from each grid unit to the water source. (3) Slope. We used the Slope tool in ArcGIS software (Environmental Systems Research Institute, Redlands, California, America) to calculate the slope of each grid unit. (4) Relative height difference. We first used the Mask tool in the ArcGIS software (Environmental Systems Research Institute, Redlands, California, America) to extract the elevation of the water system in the study area. Then, we used the Euclidean Allocation tool to assign the water system elevation to the nearest grid cell. Finally, the Raster Calculator tool was used to calculate the relative height difference between the elevation value of each grid cell and the water system elevation. The spatial factor resistance was referenced from the literature [50] (Table 1).

Different eco-socio-economic factors have different spatial impacts on water demand. Four factors at the county level—normalized difference vegetation index (NDVI), per capita GDP, population density, and grain yield per unit area—were selected to reflect the spatial characteristics of water demand, which were used to characterize ecological water demand, industrial water demand, domestic water demand, and agricultural water demand, respectively. Among them, the NDVI was obtained by mask extraction and zonal statistics of national data, per capita GDP, population density, and grain yield per unit area were calculated based on statistical data. All data for each indicator in this study were converted to dimensionless using the maximum difference normalization method before calculation.


**Table 1.** Resistance classification and assignment of different spatial elements.

<sup>1</sup> When the relative height difference is negative, the slope value is also negative, and the spatial resistance is smaller.

#### 2.3.2. Water Resource Accessibility Index Model

We selected the runoff in each grid unit to represent the attractiveness of a water source to water users, and selected the cost distance constrained by the three resistance factors of slope, relative height difference, and land use to reflect the spatial resistance. Among them, runoff is positively correlated with water resource accessibility: The greater the runoff, the higher the accessibility. The cost distance is negatively correlated with the water resource accessibility: The larger the cost distance, the greater the accessibility. The cost distance can be determined using the cost distance model in ArcGIS software (Environmental Systems Research Institute, Redlands, California, America), which requires the input of two raster layers: The target layer and the resistance layer. In this study, the target layer was the water source (water system), whereas the resistance layer was the resistance matrix of three spatial factors: Slope, relative height difference, and land use type. According to the research [18,51], the water resource accessibility index model is as follows:

$$A\_i = \mathcal{W}\_j \times fmin\left(\sum\_{j=i,m}^{i=1,n} D\_{ij} \times R\_i\right) \tag{1}$$

where *A<sup>i</sup>* refers to the water resource accessibility index, *W<sup>j</sup>* refers to the water source attraction capacity (runoff), *f* is a positive correlation function that reflects the relationship between the minimum cumulative resistance and the spatial resistance from the water users to the water source, and *Dij* and *R<sup>i</sup>* refer to the distance and space resistance from the water users to the water source, respectively.

#### 2.3.3. Coupling Coordination Degree Model

Coupling refers to the phenomenon by which two or more systems interact with each other to achieve synergy, and the coupling coordination degree refers to the degree of coordinated development between two or more systems [52]. Water resource accessibility and eco-socio-economic water demand elements are two closely related systems that restrict and promote each other. Thus, we used a coupling coordination degree model to express the degree of coordinated development between the two systems. The equations are as follows [53–55]:

$$\mathcal{C} = 2\sqrt{f(a) \times g(b)} / [f(a) + g(b)] \tag{2}$$

$$T = af(a) + \beta g(b) \tag{3}$$

$$D = \sqrt{\mathbb{C} \times T} \tag{4}$$

where *C* refers to the coupling degree, with a value in the interval [0, 1]; *f*(*a*) refers to the water resource accessibility; *g*(*b*) refers to eco-socio-economic water demand elements; *D* refers to the coupling coordination degree, with a value in the interval [0, 1] where the greater the *D* value, the higher the coupling coordination degree of the two systems, and vice versa; *T* refers to the comprehensive coordination index; α and β refer to the contribution of water resource accessibility and eco-socio-economic water demand elements to the coupling coordination degree, respectively. According to related research [53–55], we selected α = β = 0.5 and divided the coupling coordination degree into 10 stages (Table 2).



#### **3. Results**

#### *3.1. Spatial Pattern of Water Resource Accessibility*

The spatial distribution characteristics of five factors—relative height difference, slope, land use resistance, water intake distance, and runoff—were analyzed. The relative height difference varies obviously. Extremely high mountains, such as Minshan, Nushan, and Hengduan Mountains, are concentrated in West Sichuan (in some areas, due to the existence of plateau lakes, the relative height difference is a large negative value) and West Yunnan, and the relative height difference is large, whereas the relative height differences in the Sichuan Basin, Guangxi, and Guizhou are relatively small (Figure 2a). The steep slope areas in the study area are relatively large, mainly distributed in West Sichuan and Northwest Yunnan, whereas the Sichuan Basin, Southwest Guangxi, and East Yunnan have relatively flat terrain with relatively low spatial resistance (Figure 2b). The concentrated distribution of glaciers and marshes in West Sichuan leads to a relatively high resistance value of land use in the region. In the Sichuan Basin and Northwest Sichuan, grassland, green land, and farmland are widely distributed, so the spatial resistance is relatively low (Figure 2c). The water intake distance is closely related to the spatial distribution of the water system. The water systems in Southwest Yunnan, North-Central Sichuan, and Guizhou are sparse, and the water intake distance is relatively large (Figure 2d). Due to the differences in the precipitation, temperature, and underlying surface of the watershed, and the influence of human activities, the regional differences in runoff are significant. The runoff is relatively high in East Guangxi and relatively low in the Sichuan Basin, West Sichuan, and Central and North Yunnan, whereas the overall trend is decreasing from the Southeast to the Northwest (Figure 2e).

**Figure 2.** Spatial distribution of different factors affecting water resource accessibility: (**a**) Relative height difference, (**b**) slope, (**c**) land use, (**d**) water intake distance, and (**e**) runoff.

The grid-scale water resource accessibility has obvious regional differences, and the overall trend gradually decreases from Southeast to Northwest (Figure 3). The high-value area is mainly concentrated in Northeast Guangxi, whereas the low-value areas are mainly concentrated in West Sichuan and North-Central Yunnan, which is closely related to the spatial distributions of the water system, slope, elevation, runoff, and land use in Southwest China. In the Southeast, especially the Guangxi Zhuang Autonomous Region, the geomorphological type is a mountainous and hilly basin, and the terrain is relatively flat, whereas spatial resistance, such as the slope and relative height difference, is relatively low. The water system in the region is well-developed and the runoff is relatively high, with fewer constraints on access to water resources, so the water resource accessibility is relatively high. However, in the Northwest, especially in West Sichuan and Northwest Yunnan, the wide distribution of extremely high mountains, glaciers, and swamps results in a significant elevation difference, a large slope and land use resistance in the region, coupled with the relatively sparse water system and fewer water

resources, which considerably increase the difficulty in obtaining water resources, resulting in relatively low water resource accessibility.

**Figure 3.** The spatial pattern of the water resource accessibility in Southwest China at the grid-scale.

Taking the county-level administrative district as the statistical unit, the water resource accessibility at the grid cell was calculated, and the spatial pattern of water resource accessibility at the county level was determined (Figure 4). The water resource accessibility varies considerably between different counties (districts) in Southwest China. The counties (districts) with relatively high water resource accessibility are mainly concentrated in Northeast Guangxi, whereas the water resource accessibility in some counties (districts) in the Sichuan Basin, West Sichuan, and Central and North Yunnan is relatively low, and the overall trend is a gradual decrease from Southeast to Northwest. The maximum water resource accessibility value was 0.996 in the Qixing District of Guangxi Zhuang Autonomous Region, and the lowest value was 0.113 in Derong County, Sichuan Province.

**Figure 4.** The spatial pattern of water resource accessibility in Southwest China at the county scale.

#### *3.2. Spatial Distribution Characteristics of Di*ff*erent Water Demand Elements*

In the study area, the superior hydrological and climatic conditions provide a suitable environment for the growth of vegetation, which increases the vegetation index overall. Only a few counties (districts) in West Sichuan and the urban center areas of each province (municipality and autonomous region) have relatively low NDVI scores (Figure 5a). As a typical ethnic minority settlement in China, Southwest China has a developing economy, and there is a significant difference in per capita GDP between different counties (districts) (Figure 5b). Among them, the per capita GDP in Central Sichuan, West Chongqing, and the urban center area of each province (municipality and autonomous region) is relatively high, and the per capita GDP in West and Northeast Sichuan, East Guizhou, Northwest Guangxi, and most parts of Yunnan is relatively low. The counties (districts) with high population density in the study area are mainly concentrated in the Sichuan Basin and the urban centers of each province (municipality and autonomous region) (Figure 5c). These regions have rapid economic development and a high level of urbanization, providing superior conditions for human survival and development. The low-value areas are mainly distributed in the ethnic minority areas of West Sichuan, West Yunnan, Southeast Guizhou, and Northwest Guangxi. East Sichuan, West Chongqing, Northeast Yunnan, and East Guangxi have flat terrain, superior climate and hydrological conditions, and the grain yield per unit area is relatively high (Figure 5d). In West Sichuan, Northwest Yunnan, and Guizhou, widespread mountainous areas, water shortages, and extensive desertification seriously affect food production.

**Figure 5.** The spatial distribution of different eco-socio-economic water demand elements at the county level: (**a**) Normalized difference vegetation index (NDVI), (**b**) per capita GDP, (**c**) population density, (**d**) grain yield per unit area.

To more accurately reflect the water demand characteristics of each county (district) in Southwest China, the different water demand elements of each county-level administrative unit were weighted and summed according to the proportion of the water resource use structure for each province (municipality and autonomous region) in 2015; then, the spatial distribution of the comprehensive water demand elements in Southwest China was determined (Figure 6). The counties (districts) with high water resource demand are mainly concentrated in the Sichuan Basin, East Guangxi, and Central Yunnan, where the agricultural production level is relatively high, economic development is relatively fast, and the population is relatively concentrated. West Sichuan, Northwest Yunnan, and Guizhou, where agricultural production and the population density are low and economic development is relatively slow, have a lower water demand.

**Figure 6.** The spatial distribution of comprehensive water demand elements at the county level.

#### *3.3. Coupling Coordination Degree of Water Resource Accessibility and Water Demand Elements*

The coupling coordination degree of water resource accessibility and eco-socio-economic water demand elements in Southwest China was between 0.26 and 0.84, showing significant regional differences (Figure 7). The coupling coordination degree in Northeast Guangxi is relatively high overall, whereas the coupling coordination degree in West and North Sichuan and Northwest Yunnan, is relatively low. The overall trend is a decrease from Southeast to Northwest. The highest is Diecai District, Guangxi Zhuang Autonomous Region, and the lowest is Hongyuan County, Sichuan Province. According to the statistical results, the coupling coordination degree of water resource accessibility and eco-socio-economic water demand elements in the study area are mainly distributed in the coordination and transition stages, among which 5 counties (districts) show high coordination, accounting for 0.92%; 29 counties (districts) show moderate coordination, accounting for 5.31%; 115 counties (districts) show low coordination, accounting for 21.06%; 326 counties (districts) show reluctant coordination, accounting for 59.71%; and 68 counties (districts) show near incoordination, accounting for 12.45%; 2 counties show reluctant coordination, accounting for 0.37%; and 1 counties show near incoordination, accounting for 0.18%. The coupling coordination degree of water resource accessibility and eco-socio-economic water demand elements in Southwest China is relatively low overall.

**Figure 7.** Spatial distribution of coupling coordination degree between water resource accessibility and water demand elements.

The regional difference in the coupling coordination degree between water resource accessibility and eco-socio-economic water demand elements in Southwest China are mainly due to the spatial differences in hydrological conditions, topography, and economic development level. East Guangxi has a relatively flat terrain, superior natural conditions, relatively low spatial resistance, and abundant precipitation, so the water resource accessibility is relatively high, which guarantees good growth and the efficient production of rice and other crops. East Guangxi is also the economic development center of Guangxi Zhuang Autonomous Region, and the population is highly concentrated, resulting in high water resource demand. Therefore, the water resource accessibility and the eco-socio-economic water demand elements show a relatively high coordinated development. The Sichuan Basin is one of the most important grain production bases in China. It has a relatively dense population and rapid economic development. However, due to the small amount of water resources and relatively low water resource accessibility, which lead to limited eco-socio-economic sustainable development to a certain extent, the coupling coordination degree between the two is relatively low. In Northwest Yunnan and West and North Sichuan, the wide distribution of extremely high mountains has caused large altitude differences and steep slopes, coupled with a relatively sparse water system and a low amount of water resources, so water resource accessibility in the region is extremely low, which means that it is difficult to meet the water resource demand of the eco-socio-economic elements. Therefore, the coupling coordination degree between the water resource accessibility and eco-socio-economic water demand elements shows near incoordination.

#### **4. Discussion**

#### *4.1. Evaluation Method of Water Resource Accessibility*

Water resource accessibility refers to the difficulty in obtaining water resources from water sources and is affected by multiple factors such as water quality, water quantity, distance, elevation, slope, land use, capital, infrastructure, and technology. Among them, the quantity and quality of water resources are decisive factors for the availability of water resources, whereas the distance, altitude, slope, land use, and others are factors affecting the convenience of obtaining water resources. The existing quantitative analyses of water resource accessibility often involved a single indicator or a few factors. For example, Jeff et al. [9] only considered the linear distance to the water source. Smiley [56] selected the four elements of water quality, water cost, water reliability, and water intake burden, and measured water resource accessibility through questionnaires and statistical analysis. Yu et al. [28] selected four factors including the slope, relative height difference, distance, and runoff to comprehensively analyze the accessibility of river water resources in the Hanjiang River Basin. Li et al. [29] constructed a grid-scale water accessibility evaluation model based on the length, runoff and viewshed value. Li et al. [57] constructed a water accessibility index by selecting indicators such as distance, altitude, ditch density, road density, and culvert number to study the water resource accessibility of freshwater wetland. In this study, we evaluated water resource accessibility by considering the five factors of runoff, slope, relative height difference, water intake distance, and land use resistance, and analyzed the spatial pattern of the water resource accessibility in different grid units. Among them, water quantity represents the attractiveness of the water source to the water users, and the other factors reflect the spatial resistance.

However, the water resource accessibility evaluation in this study still has some room for improvement. First, our evaluation only considered the impact of water quantity and ignored the water quality. Thus, in future research, the quality of water resources should be measured in terms of water quality requirements for different water demand elements. Second, we assumed that water users obtain water from the closest water source, but in practice, multiple water sources provide water for users. Therefore, it is necessary to weight the multiple water sources within a certain range in future research. Third, we ignored the impact of socio-economic factors, such as water supply facilities, water treatment technology, irrigation technology, and water fees, on water resource accessibility, which affects the accuracy of the evaluation results to a certain extent; thus, future research needs to comprehensively measure multiple factors.

#### *4.2. Water Resource Accessibility and Regional Eco-Socio-Economic Development*

Water resources are an important basis for supporting the development of eco-socio-economic systems, whereas social and economic development provides the necessary funds and conditions for ensuring the sustainable development and use of water resources [58], which affect and restrict each other. The limited water resource accessibility of spatial units not only threatens the supply of drinking water and irrigation water, but also threatens the sustainable and healthy development of the ecosystem [20,59,60]. Therefore, spatially accessible water resources are essential for an adequate freshwater supply [28].

In Southwest China, the topography and geomorphology are particularly complex, the spatial-temporal distribution of water resources is uneven, and the water supply facilities lack expansion and improvement potential, resulting in serious seasonal, regional, and engineering water shortages. The use of water resources is difficult and costly, which seriously restricts regionally sustainable eco-socio-economic development. Therefore, water resource management in Southwest China must fully consider the characteristics and formation of water resources in the region and should adopt different water resource development and use models. The analysis of the coupling coordination degree between water resource accessibility and eco-socio-economic water demand elements can be used to effectively identify the areas where the coupling coordination degree between the two is

relatively low. This can provide a decision-making basis for strengthening regional water resource management and allocation, for effectively implementing water conservancy facilities planning, and for urban development planning, thus ensuring the coordinated and sustainable development of various systems.

In West Sichuan and Northwest Yunnan, there are many extremely high mountains, resulting in significant altitude differences and steep slopes, and the water sources are far away from water users, so the water resource accessibility is low and agricultural irrigation is difficult. Therefore, these areas require more investment to improve irrigation conditions and increase irrigation efficiency to meet crop water requirements. In the Sichuan Basin, Central Yunnan, Northwest Guizhou, and Southeast Guangxi, the population density is relatively high; it is necessary to continuously strengthen water resource protection and infrastructure construction in densely populated areas to ensure a safe and adequate supply of drinking water. The economic development level of most counties (districts) in Southwest China is low, somewhat lagging behind the water resource accessibility level, whereas the water resource accessibility in West Sichuan and Northwest Yunnan is relatively low, which restricts economic development to some extent. Therefore, according to the spatial pattern of water resource accessibility, rationally adjusting the industrial structure and developing a circular economy is necessary, which contribute to promoting rational and rapid economic development and ensuring the coordinated development of water resource accessibility and social and economic elements. In Northwest Yunnan and West and Northeast Sichuan, relatively poor natural conditions and scarce water resources result in relatively low water resource accessibility, which considerably limits vegetation growth. Therefore, these regions should be the focus for ecological conservation and restoration.

#### **5. Conclusions**

This paper applied a water resource accessibility index model, considering five spatial factors of runoff, slope, relative height difference, water intake distance and land use resistance, which enabled the quantitative analysis of the spatial distribution characteristics of water resource accessibility on a grid-scale in Southwest China. The results show that due to the large spatial distribution differences of different spatial elements, the spatial differences in water resource accessibility in Southwest China are relatively significant, and the overall trend is a decrease from Southeast to Northwest.

Due to the differences in hydrological conditions, topography, and economic development level, the coupling coordination degree between water resource accessibility and eco-socio-economic water demand elements in Southwest China has obvious regional differences, and the overall distribution characteristics are higher in the Southeast and lower in the Northwest. The proportion of counties (districts) with moderate coordination or higher was only 6.23%, mainly concentrated in the Northeast part of Guangxi. The counties (districts) with near incoordination, low incoordination, and moderate incoordination accounted for 13.00%, mainly concentrated in West Sichuan and Northwest Yunnan. The coupling coordination degree between the two is relatively low overall.

The water resource accessibility and the eco-socio-economic system in Southwest China have not achieved coordinated or sustainable development. The insufficient water resource support capacity in the region has restricted the development of the region to a certain extent, and the rapidly increasing population and economic development have increased water supply stress to a certain extent. Therefore, it is necessary to continuously coordinate the relationship between water resource management and regional development.

**Author Contributions:** Conceptualization, H.D.; Data curation, T.L., S.Q., S.M. and R.B.; Formal analysis, T.L.; Investigation, T.L., S.Q., S.M. and R.B.; Methodology, T.L., S.Q. and H.D.; Project administration, H.D.; Visualization, T.L.; Writing—original draft, T.L.; Writing—review & editing, T.L., S.Q., S.M., R.B. and H.D.

**Funding:** This research was funded by the National Key Research and Development Program of China (No. 2016YFC0502106).

**Acknowledgments:** We thank Yuebo Su for help and advice in data processing, article writing and modification.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

**Ge Wang 1,2,3,4, Changlai Xiao 1,2,3,4 , Zhiwei Qi 1,2,3,4 , Xiujuan Liang 1,2,3,4, \*, Fanao Meng 1,2,3,4 and Ying Sun 1,2,3,4**


2

**Abstract:** In view of the large spatial difference in water resources, the water shortage and deterioration of water quality in the Chang-Ji Economic Circle located in northeast China, the water resource carrying capacity (WRCC) from the perspective of time and space is evaluated. We combine the gray correlation analysis and multiple linear regression models to quantitatively predict water supply and demand in different planning years, which provide the basis for quantitative analysis of the WRCC. The selection of research indicators also considers the interaction of social economy, water resources, and water environment. Combined with the fuzzy comprehensive evaluation method, the gray correlation analysis and multiple linear regression models to quantitatively and qualitatively evaluate the WRCC under different social development plans. The developmental trends were obtained from 2017 to 2030 using four plans designed for distinct purposes. It can be seen that the utilization of water resource is unreasonable now and maintains a poor level under a business-as-usual Plan I. Plan II and Plan III show that resource-based water shortage is the most critical issue in this region, and poor water quality cannot be ignored either. Compared with Plan I, the average index of WRCC in Plan IV increased by 51.8% and over 84% of the regions maintain a good level. Strengthening sewage treatment and properly using transit water resources are more conducive to the rapid development of Chang-Ji Economic Circle.

**Keywords:** fuzzy comprehensive evaluation method; water resource carrying capacity; gray correlation analysis; multiple linear regression models; water environment capacity

Liang, X.; Meng, F.; Sun, Y. Water Resource Carrying Capacity Based on Water Demand Prediction in Chang-Ji Economic Circle. *Water* **2021**, *13*, 16. https://dx.doi.org/10.3390/w13010016

**Citation:** Wang, G.; Xiao, C.; Qi, Z.;

Received: 5 December 2020 Accepted: 21 December 2020 Published: 24 December 2020

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/ licenses/by/4.0/).

**1. Introduction**

With the development of the urbanization process, the demand for water resources has increased significantly, but the pollution of water resources has caused serious problems. These changes pose a potential threat to water resource carrying capacity in many regions [1]. The water resources carrying capacity (WRCC) refers to the ability of water resources to withstand the largest population, socioeconomic, and ecological environment requirements under the premise of maintaining sustainable development. Studies of the WRCC can provide helpful information about how the socioeconomic system is both supported and restrained by the water resources system, such as intuitively measuring regional development potential [2], etc. Since the 21st century, people began to pay attention to the research of WRCC. Through continuous improvement of related influencing factors, a relatively mature evaluation system has initially formed [3–5]. Nowadays, the research on the WRCC has changed from a simple natural factor to a water-ecological-economic factor [6].

The research on the WRCC's theory in the international context is focused more on the relationship between carrying capacity and sustainable economic and social development [7]. However, the research on the limitation of water resources security to the WRCC is relatively late. According to the latest analysis of the obstacle degree for the WRCC system in Northeast China, the agricultural water pollution index emerged as the main factor that is restricting the steady rise of the WRCC since 2004. Following 2014, with the upsurge of industrialization, the percentage of industrial wastewater discharge has increased significantly [8]. There are many old industrial bases in the Chang-Ji Economic Circle, and the percentage of industrial wastewater discharge and the risk of water pollution has increased year by year. Furthermore, the data published in the Water Resources Bulletin in the study area over the years show that the surface water quality is bad. So, the water quality should be included in the analysis of WRCC in the Chang-Ji economic circle. In China, there are few studies on the water resources carrying capacity of the Chang-Ji economic circle [9]. Moreover, most of the studies in the Chang-Ji Economic Circle are carried out in administrative regions, and few studies take into account the constraints of water pollution on the WRCC, which cannot directly and accurately measure the development potential of the entire Chang-Ji Economic Circle [10,11]. Therefore, this study takes into account the quantity and quality of water resources and combines the comprehensive carrying capacity of social economic development and ecological environment and selects appropriate methods to evaluate the WRCC of the Chang-Ji Economic Circle.

At present, there are plenty of methods to evaluate the WRCC, such as the traditional trend analysis method [12], the principal component analysis method [13], the fuzzy comprehensive evaluation method (FCE) [14], the multiobjective analysis method [15], the artificial neural network method [16], and the system dynamics (SD) method [17,18]. Zhang et al. [13] applied the principal component analysis method to evaluate the temporal scale variation tendency of the WRCC. However, there was still some uncertainty when integrating the method WRCC index standardization, the method of principal component determination, and the weights of contribution rates. Multiobjective analysis is influenced by its limitations and is more suitable for smaller areas of research. The SD model can effectively simulate and predict through negative feedback adjustment [19]. However, this method requires a large number of parameter settings and data simulations [20], which cannot achieve rapid evaluation. In fact, various WRCC prediction methods are based on the further evaluation of their influencing factors. The FCE method which is widely used by scholars can analyze the WRCC from all aspects [21,22], making the research results more reliable. For example, Zhang et al. [23] used fuzzy set pair analysis theory to evaluate the WRCC in Dagong Yellow River Diversion Irrigation District from 2013 to 2017. The study qualitatively measured the water resources carrying capacity of the ecological irrigation area, and the evaluation results can provide a scientific basis for optimal allocation of water resources in the Dagong Yellow River diversion irrigation district. At the same time, through the comparative analysis of some indicators, the FCE method can solve the defects of the parameters that are difficult to grasp and easily lead to unreasonable conclusions. Moreover, while assessing the regional WRCC, it is necessary to predict future development trends based on the status quo. Accurate trend analysis has become an important part of reasonable evaluation. However, most scholars use the simple linear equations to predict related influencing factors. To make up for the shortcomings of large errors in traditional trend analysis, this study quantitatively predicts relevant influencing factors by the method of gray correlation analysis (GCA) combined with multiple linear regression (MLR) models. Then, on the basis of quantitatively predicting the development trend of social economy, water resources, and water environment evaluation indicators, the FCE method is used to make a reliable assessment of the WRCC.

#### **2. Study Area and Data Sources**

#### *2.1. Study Area*

The Chang-Ji Economic Circle includes the whole of Changchun City, Jilin City, Jiutai District, Shuangyang District, and Yongji County, as well as parts of Nongan County, Gongzhuling City, and Yitong County (Figure 1). The study area is located in the center of Jilin Province, accounting for 7.96% of the total province's area. The total population of the region is about 648 million and the urbanization rate is 59.64%. There are many water systems in the area, and river networks are dense, all belonging to the second Songhua River system. The large reservoirs include Fengman Reservoir, Shitoukoumen Reservoir, and Xinlicheng Reservoir. The total water storage capacity is as high as 134.1 × 10<sup>8</sup> m<sup>3</sup> .

**Figure 1.** The location of the study area and the distribution of the river system and the water bodies.

#### *2.2. Data Sources*

The study area involves Jilin City and Changchun City, and the data used include socioeconomic data, water supply and demand data, and water environment data. The data

sources are Jilin statistical yearbook (2010–2017), Changchun statistical yearbook (2010–2017), Jilin water resource bulletin (2010–2017), and Changchun water resource bulletin (2010–2017). In addition, water quota references the industry water quota standards of Jilin province and some environmental data comes from the website of Jilin province environmental protection bureau.

#### **3. Establishment and Prediction of Water Demand Model**

The analysis of the balance between supply and demand of water resources can reflect the overall level of water use in a region and is an important factor affecting the WRCC. In recent years, ecological and environmental problems have become increasingly prominent, and ecological water demand will also become an important issue in the analysis of the balance between water supply and demand. This study collects water supply and demand data for the past eight years in the study area and focuses on ecological water demand to predict the water demand in the next 13 years.

#### *3.1. Model Related Methods*

Since the study area involves 15 different administrative districts, if the forecast is made one by one according to the different water demand industries, the amount of data required is large and it is difficult to maintain consistency. The combination of the GCA and MLR model can make up for this deficiency. Using the GCA method, the correlation coefficients between the six water demands and the total water demand are calculated respectively. The six water requirements are ecological and environmental water demand, urban public water demand, domestic water demand, forestry, animal husbandry and fishery water demand, farmland irrigation water demand, and industrial water demand. One selects indicators with high correlation coefficients and establishes MLR models of each indicator for total water demand, then obtains total water demand at different planning levels.

#### 3.1.1. The GCA Method

The GCA is to find the important factors that affect the target value by looking for the correlation between various factors between random sequences. This method is mainly to determine the correlation between the influencing factors and the target value, find the main characteristics of the problem, and intuitively quantify the complex influencing relationship. The main calculation method is to perform dimensionless processing on all data, set the processed target sequence to *x*0, and set the factor sequence to *x<sup>i</sup>* (*i* = 1, 2, . . . , *n*); perform a difference between the two sequences. Calculate to find the maximum difference and minimum difference between the two poles. Among them, the difference sequence is expressed as:

$$\Delta\_{0i}(k) = |\mathbf{x}\_0(k) - \mathbf{x}\_i(k)|\tag{1}$$

Find the correlation coefficient between each factor and the target value *εi*(*k*):

$$\varepsilon\_{i}(k) = \frac{\min \min |\mathbf{x}\_{0}(k) - \mathbf{x}\_{i}(k)| + \rho \max \max |\mathbf{x}\_{0}(k) - \mathbf{x}\_{i}(k)|}{|\mathbf{x}\_{0}(k) - \mathbf{x}\_{i}(k)| + \rho \max \max |\mathbf{x}\_{0}(k) - \mathbf{x}\_{i}(k)|} \tag{2}$$

Equation (1), ∆*oi*(*k*) is the absolute value of the difference between two sequences.

Equation (2), *ρ* for the resolution coefficient, the empirical value is 0.5; max max |*x*0(*k*) − *xi*(*k*)| is the maximum difference between the two levels; min min|*x*0(*k*) − *xi*(*k*)| is the minimum difference between the two levels; the greater the calculated value of *εi*(*k*), the greater the correlation between the factor and the target value.

Finally, according to the correlation coefficient, determine the degree of correlation between the influencing factors and the target sequence *r<sup>i</sup>* :

$$r\_i = \frac{1}{n}(\varepsilon\_i(1) + \varepsilon\_i(2) + \dots + \varepsilon\_i(n))\tag{3}$$

Similarly, among all influencing factors, *r<sup>i</sup>* with the larger value has a higher degree of correlation with the target sequence.

#### 3.1.2. The MLR Model

The MLR model is a predictive method that participates in the analysis of the linear relationship between two or more independent and dependent variables. The method of establishing MLR model is as follows:

Suppose there is a linear relationship between the dependent variable *y* and the independent variable *x<sup>i</sup>* (*i* = 1, 2, . . . , *n*):

$$y\_i = \beta\_0 + \beta\_1 \mathbf{x}\_1 + \beta\_2 \mathbf{x}\_2 + \dots \dots + \beta\_n \mathbf{x}\_n + \varepsilon \tag{4}$$

In Equation (4), *β<sup>i</sup>* is the regression coefficient, *i* = 1, 2, . . . , *n*; generally *n* > 2; *ε* is the random factor of error, and it follows the N(0, *σ* 2 ) distribution.

$$\begin{aligned} \text{Let } Y = \begin{pmatrix} y\_1 \\ y\_2 \\ \vdots \\ y\_m \end{pmatrix}, X = \begin{pmatrix} 1 & \mathbf{x}\_{11} & \cdots & \mathbf{x}\_{1n} \\ 1 & \mathbf{x}\_{21} & \cdots & \mathbf{x}\_{2n} \\ \vdots & \vdots & \cdots & \vdots \\ 1 & \mathbf{x}\_{m1} & \cdots & \mathbf{x}\_{mn} \end{pmatrix}, \beta = \begin{pmatrix} \beta\_0 \\ \beta\_1 \\ \vdots \\ \beta\_n \end{pmatrix}, \varepsilon = \begin{pmatrix} \varepsilon\_0 \\ \varepsilon\_1 \\ \vdots \\ \varepsilon\_n \end{pmatrix}, \text{the MLE} \end{aligned}$$

models in matrix form are available as follows:

$$Y = X\beta + \varepsilon \tag{5}$$

#### *3.2. Model Establishment and Error Analysis*

Combining the water supply and demand data from 2010 to 2017, the GCA method is used to obtain the correlation between the total water demand and the other basic water demand. The correlations obtained are ecological and environmental water demand (0.819), urban public water demand (0.652), domestic water demand (0.610), forestry, animal husbandry and fishery water demand (0.670), farmland irrigation water demand (0.709), and industrial water demand (0.744). Three indicators with a correlation degree greater than 0.7 are selected as the main influencing factors of total water demand.

With the help of SPSS statistical analysis software, a multivariate regression matrix with three factors of ecological environment (*X*1), farmland irrigation (*X*2) and industrial water demand (*X*3) as independent variables and total water demand (*Y*) as dependent variable was constructed. Solve the regression model as follows:

$$Y = -2.408X\_1 + 2.082X\_2 + 1.087X\_3 - 7.243\tag{6}$$

The MLR model is tested for errors based on each water demand and total water demand from 2010 to 2017, as shown in Table 1. According to the analysis results, the error values are all within 1%, which meets the prediction accuracy requirements. Therefore, the established MLR model is suitable for the prediction of the total water demand in this study area.


#### **4. Establishment of a Rapid Evaluation Model**

In order to achieve efficient and accurate evaluation of WRCC, the study selects the FCE method to establish corresponding evaluation system. Measuring the size of WRCC by setting various plans provides a basis for regional water resources development and utilization.

#### *4.1. Evaluation Method*

First, establish the set of influencing factors of WRCC *U* = (*u*1, *u2*, . . . , *un*), and the corresponding comment set *V* = (*V*1, *V2*, . . . , *Vn*). Second, the membership degree rij of the comment set *v<sup>j</sup>* is determined by a single factor fuzzy evaluation, which means to evaluate the single factor u, and the single factor evaluation set *r<sup>i</sup>* = (*ri*1, *ri*2, ri3) is obtained. Then, the evaluation index is quantified separately, and the corresponding membership degree *rij* is obtained, thus establishing the fuzzy relation matrix R with the amount of m evaluation factors. Finally, by analyzing weights *A* = (*α*1*, α*2, . . . , *αn*) of different influencing factors, the fuzzy operation *B* = *A* × *R* is used to synthesize the weight vector of the influencing factors and the fuzzy evaluation matrix to obtain the FCE's results.

The degree of membership usually indicates a certain degree of deviation, using two sets of formulas to calculate the membership function. *x*<sup>1</sup> and *x*<sup>3</sup> are the critical values of *V*1, *V2*, and *V*2, *V*3, respectively, and the grade of V<sup>2</sup> is the interval midpoint value of *x*<sup>2</sup> [14], where *x*<sup>2</sup> = *x*1+*x*<sup>3</sup> 2 . The first set of formulas is applicable to the case where the larger the index *U<sup>i</sup>* is, the better the system is. The second set of formulas is applicable to the case where the smaller the index *U<sup>i</sup>* is, the better the system is. The two sets of calculation formulas are as follows:

First set of formulas:

$$V\_1(\mu i) = \begin{cases} 0.5 \left( 1 + \frac{\mu\_i - \ge\_1}{\mu\_i - \ge\_2} \right) \prime \ u\_i \ge \ge\_1 \\\ 0.5 \left( 1 - \frac{\mu\_i - \ge\_1}{\ge\_2 - \ge\_1} \right) \prime \ x\_2 < u\_i < \ge\_1 \\\ 0 \qquad \restriction \ u\_i \le \ge\_2 \end{cases} \tag{7}$$

$$V\_2(\mu i) = \begin{cases} 0.5 \left( 1 - \frac{\underline{u\_i} - \underline{x\_1}}{\underline{u\_i} - \underline{x\_2}} \right) \prime \; u\_i \ge \underline{x\_1} \\\ 0.5 \left( 1 + \frac{\underline{x\_1} - \underline{u\_i}}{\underline{x\_1} - \underline{x\_2}} \right) \prime \; \mathbf{x\_2} \le u\_i < \mathbf{x\_1} \\\ 0.5 \left( 1 + \frac{\underline{u\_i} - \underline{x\_3}}{\underline{x\_2} - \underline{x\_3}} \right) \prime \; \mathbf{x\_3} < u\_i < \mathbf{x\_2} \\\ 0.5 \left( 1 - \frac{\underline{x\_3} - \underline{u\_i}}{\underline{x\_2} - \underline{u\_i}} \right) \prime \; u\_i \le \mathbf{x\_3} \end{cases} \tag{8}$$

$$V\_3(\mu i) = \begin{cases} 0, \ u\_i \ge \chi\_2 \\ 0.5 \left( 1 - \frac{\underline{u\_i} - \underline{\chi\_3}}{\underline{\chi\_2} - \underline{\chi\_3}} \right), \ \mathfrak{x}\_3 < \mathfrak{u}\_i < \mathfrak{x}\_2 \\ 0.5 \left( 1 + \frac{\underline{\chi\_3} - \underline{\mu\_i}}{\underline{\chi\_2} - \underline{\mu\_i}} \right), \ \mathfrak{u}\_i \le \mathfrak{x}\_3 \end{cases} \tag{9}$$

Second set of formulas:

$$V\_1(\mu i) = \begin{cases} 0.5 \left( 1 + \frac{u\_i - \mathbf{x}\_1}{u\_i - \mathbf{x}\_2} \right), \ u\_i \le \mathbf{x}\_1 \\\ 0.5 \left( 1 - \frac{u\_i - \mathbf{x}\_1}{\mathbf{x}\_2 - \mathbf{x}\_1} \right), \mathbf{x}\_1 < u\_i < \mathbf{x}\_2 \\\ 0 \end{cases} \tag{10}$$

$$\mathbf{V}\_{2}(\boldsymbol{u}\boldsymbol{i}) = \begin{cases} 0.5 \left( 1 - \frac{\boldsymbol{u}\_{i} - \boldsymbol{x}\_{1}}{\boldsymbol{u}\_{i} - \boldsymbol{x}\_{2}} \right), \ u\_{i} \le \boldsymbol{x}\_{1} \\\ 0.5 \left( 1 + \frac{\boldsymbol{x}\_{1} - \boldsymbol{u}\_{i}}{\boldsymbol{x}\_{1} - \boldsymbol{x}\_{2}} \right), \boldsymbol{x}\_{1} < \boldsymbol{u}\_{i} \le \boldsymbol{x}\_{2} \\\ 0.5 \left( 1 + \frac{\boldsymbol{u}\_{i} - \boldsymbol{x}\_{3}}{\boldsymbol{x}\_{2} - \boldsymbol{x}\_{3}} \right), \boldsymbol{x}\_{2} < \boldsymbol{u}\_{i} < \boldsymbol{x}\_{3} \\\ 0.5 \left( 1 - \frac{\boldsymbol{x}\_{3} - \boldsymbol{u}\_{i}}{\boldsymbol{x}\_{2} - \boldsymbol{u}\_{i}} \right), \boldsymbol{u}\_{i} \ge \boldsymbol{x}\_{3} \end{cases} \tag{11}$$

$$V\_3(\mu i) = \begin{cases} 0, \ u\_i \le \varkappa\_2 \\ 0.5 \left( 1 - \frac{\nu\_i - \varkappa\_3}{\varkappa\_2 - \varkappa\_3} \right), \varkappa\_2 < \nu\_i < \varkappa\_3 \\ 0.5 \left( 1 + \frac{\varkappa\_3 - \nu\_i}{\varkappa\_2 - \varkappa\_i} \right), \nu\_i \ge \varkappa\_3 \end{cases} \tag{12}$$

The corresponding membership degree *rij* is calculated separately, and matrix *R* corresponding to different horizontal years of each administrative region in the study area is obtained as follows:

$$R = \begin{bmatrix} r\_{11} & r\_{12} & \dots & r\_{1n} \\ r\_{21} & r\_{22} & \dots & r\_{2n} \\ \vdots & \vdots & & \ddots & \vdots \\ r\_{m1} & r\_{m2} & \dots & r\_{mn} \end{bmatrix} \tag{13}$$

The vectorization evaluation result *B* is obtained from the membership matrix and the weight, and the formula is as follows:

$$B = A \times R = (a\_1 a\_2 \dots a\_n) \cdot \begin{bmatrix} r\_{11} & r\_{12} & \dots & r\_{1n} \\ r\_{21} & r\_{22} & \dots & r\_{2n} \\ \vdots & \vdots & & \ddots & \vdots \\ r\_{m1} & r\_{m2} & \dots & r\_{mn} \end{bmatrix} = (b\_1 b\_2 \dots b\_m) \tag{14}$$

To facilitate the comparative evaluation, the vectorization result is quantified by the following formula, and the corresponding comprehensive score value *λ* is obtained, where *k* = 1, and the higher λ indicates that the regional WRCC is higher.

$$\lambda = \frac{\sum\_{i=1}^{n=3} b\_i^k \times \lambda\_i}{\sum\_{i=1}^{n=3} b\_i^k} \tag{15}$$

#### *4.2. Evaluation of Index Selection*

The key to correctly evaluate the regional WRCC is to properly select fuzzy evaluation indicators. To better reflect the status of regional WRCC, it is particularly important to select regionally representative evaluation indicators. Many factors influence the WRCC. Most studies only consider the impact of the amount of water resources, social economy, and ecological environment on WRCC [24–26] but ignore the impact of water quality on regional WRCC. This tends to make the calculation of the WRCC too large and fails to reflect the real situation in the study area. In fact, due to the different degrees of water pollution, the actual water supply and availability of water are smaller than themselves. Combined with the actual situation of the uneven distribution of water resources and the outstanding water quality problems in the study area, this study incorporates the water environment capacity (WEC) into the evaluation of WRCC. We considered the factors of water resources (including water quality and water quantity), as well as the socioeconomic and ecological environment to calculate WRCC. This lays a foundation for truly establishing a new realm of harmonious development of economic, social, and humanities based on the principle of sustainable development (Figure 2).

#### 4.2.1. Evaluation Indicators *U*1, *U*2, and *U*<sup>3</sup>

Based on the surface water and groundwater supply capacity, actual water supply, and water demand in this region, the WRCC is evaluated from the perspective of the quantity of water resources. Per capita available water resource (*U*1) = available water resources/total population (m3/person); per capita water supply quantity (*U*2) = actual water supply/total population (m3/person); water resource utilization rate (*U*3) = water demand/available water resources (%).

**Figure 2.** The research system of water resource carrying capacity (WRCC) index evaluation.

#### 4.2.2. Evaluation Index U<sup>4</sup>

Chemical oxygen demand (COD) pollution receiving capacity (*U*4) = the water environment capacity of COD at this stage/the maximum water environment capacity of COD (%). The maximum water environment capacity of COD means the maximum pollution capacity that the region can received.

According to the results of the Water Resources Bulletin, Class IV, V, and above Class V water quality river sections are shown in the rivers in the study area. The quality of regional water resources is not optimistic [27,28]. In fact, the quality of water supply directly affects the availability of actual water supply, and the WEC can reflect the two main capabilities of water body dilution and natural purification. Therefore, based on the preliminary understanding of the water quality of the Chang-Ji Economic Circle, this study uses the water environmental capacity as the water quality evaluation index in the evaluation system. Since the flow rate of the river in the study area is stable as well as small, the calculation method of the overall standard is used.

The WEC is calculated as follows [29]:

$$W = 86.4 Q\_0 (\mathcal{C}\_s - \mathcal{C}\_0) + 0.001 KVC\_s + 86.4 q\mathcal{C}\_s \tag{16}$$

where *W* is the initial value for WEC of the water body, *Cs* is the standard water quality of the water body, *Q*<sup>0</sup> is the incoming water flow, *C*<sup>0</sup> is the upstream background concentration of the incoming water, *K* is the water quality degradation coefficient, *Q* is the side flow of the side stream, *V* is the flow rate, and *q* is the side inflow flow.

The overall standard calculation method usually does not consider the location of the pollution source, so the calculation results tend to be too large, which is nonconservative. Therefore, in order to conform to the reality, an uneven coefficient is introduced for correction [29]; the method is as follows:

$$\mathcal{W}' = \alpha \mathcal{W} \tag{17}$$

where *W*′ represents the corrected WEC, and *α* denotes the uneven coefficient.

Because the rivers in the area belong to the small and the middle rivers [30], their flow is small and slow, and the uneven coefficient takes an empirical value of 0.8.

#### 4.2.3. Evaluation Index *U*<sup>5</sup>

The Chang-Ji Economic Circle is a gathering place for old industrial cities. The development of its industry can reflect the social and economic conditions of the region. So, we select evaluation index *U*<sup>5</sup> to effectively reflect this. Water demand of industrial

output value of 10,000 Yuan (*U*5) = industrial water demand/industrial output value (m3/10,000 yuan). The industrial water demand and the industrial output value is obtained by the GCA and MLR model.

#### 4.2.4. Evaluation Index *U*<sup>6</sup>

Lou [31] and Wang [9] confirmed that the modulus of water demand can reflect the level of regional economic development. Thus, we select the Modulus of water demand as the evaluation index *U*<sup>6</sup> that effectively reflects the ecological environment of the study area. Modulus of water demand (*U*6) = total water demand/land area (10,000 m3/km<sup>2</sup> ). The total amount of water demand is obtained by the GCA and MLR model and the land area used the government published data. The required water modulus reflects the restriction of the ecological environment on the WRCC.

By selecting the above six indicators, it can effectively reflect the balance of supply and demand of water resources in the region, the amount of water resources, the quality of water resources, and the impact of socioeconomic conditions and ecological environment on the regional WRCC.

Based on the above-mentioned evaluation indicators *U*<sup>1</sup> to *U*6, the impact degree of WRCC is analyzed, and its comment set *V* = (*V*1, *V*2, *V*3) is established (as shown in Table 2 below). The WRCC of *V*<sup>1</sup> to *V*<sup>3</sup> is gradually weakened. For this evaluation, the second set of formulas applies to *U*1, *U*2, and *U*4; however, *U*3, *U*5, and *U*<sup>6</sup> apply to the first set of formulas. The weight is determined according to the influence degree of the influencing factors on the WRCC. According to expert analysis, the corresponding weights are obtained from empirical values, that is, *A* = (α1, α2, α3, α4, α5, α6) = (0.2, 0.2, 0.3, 0.1, 0.1, 0.1). Then, according to the score value (*λ*<sup>1</sup> = 0.95, *λ<sup>2</sup>* = 0.5, and *λ<sup>3</sup>* = 0.05), the water carrying capacity of each area is analyzed and evaluated, and the score value directly reflects the WRCC in the area.

**Table 2.** Evaluation standard of WRCC grading indicators.


#### *4.3. Reasonably Evaluate Regional WRCC*

The evaluation value of WRCC is statistically analyzed and divided into three levels: an evaluation value of WRCC greater than 0.6 is an area with good WRCC (I), an area evaluation value of water carrying capacity ranging between 0.3 and 0.6 is medium (II), and an area evaluation value of WRCC less than 0.3 is poor (III).

#### *4.4. The Establishment of 4 Different Plans*

#### 4.4.1. Plan I

Under the current conditions, Plan I only considers the WRCC of self-produced water, does not increase the local water supply, or expand the capacity of water transfer outside the region, and predicts the carrying capacity of water resources in different years.

#### 4.4.2. Plan II

On the basis of Plan I and considering the project "Carrying Water from Songhua River to Changchun" to increase the local water supply, we predict the water carrying capacity of different years. According to the government's economic development plan of the Chang-Ji Economic Circle, the cumulative water supply capacity of the design diversion project in Changchun City is 3.25 × 10<sup>8</sup> m<sup>3</sup> . After the completion of the water supply project in the central city of Jilin Province, the cumulative water intake will be 5.83 × 10<sup>8</sup> m<sup>3</sup> in 2020. Furthermore, the cumulative water intake will be 6.92 × 10<sup>8</sup> m<sup>3</sup> in vision level year 2030.

#### 4.4.3. Plan III

Based on Plan II, Plan III strengthens water governance and considers industrial, agricultural, and domestic water conservation. By reducing water usage quotas, increasing the reuse rate of reclaimed water and treating sewage as the most important measures.

According to the "Standards for Local Standard Water Use in Jilin Province" [32], the plan will appropriately reduce the industrial water quota, where each administration increases the amount of water reuse by 0.05 billion m3/per year.

#### 4.4.4. Plan IV

Based on Plan III, Plan IV increases an appropriate amount of transit water. The inflow water of Songhua River is 62.02 × 10<sup>8</sup> m<sup>3</sup> [27]. The increase in water supply is 40% of the inflow water of Songhua River, while the increase in actual water supply is 20%.

#### **5. Results**

Due to the inconsistent development speed of various regions, even within the same city, there are differences in regional WRCC. Most of the research only stays at the holistic research within the scope of the region, while ignoring the research of small administrative units [22,33]. Zhou et al. [11] compared with the temporal dynamic process of index change in the water environment carrying capacity and thought that it is urgent to carry on spatiotemporal dynamic change analysis in the WRCC considering spatial heterogeneity and spatial evolution. As a result, this article uses the smallest administrative unit to analyze change trend of WRCC from both time and space perspectives.

#### *5.1. Results of Each Program from the Perspective of Time* 5.1.1. Plan I

According to the results of Plan I (Figure 3), the study area is still generally in the middle area of WRCC (II). In 2020, the WRCC of Fengman District is the largest (0.606), while Chaoyang District is poor (0.287). By 2030, both Lvyuan District (0.298) and Chaoyang District (0.27) are in areas with poor WRCC (III). The comparison shows that the comprehensive score of WRCC in 2030 is decreasing compared with 2020, but the rate of decline is slow. It is comprehensively reflected that under the condition of not changing the status quo, the WRCC of the study area will continue to weaken. According to existing research, it likely due to the uneven distribution of the regional water resources [34], which makes the contradiction between water supply and demand increasingly prominent and the development potential decreases [35].

**Figure 3.** *Cont*.

**Figure 3.** WRCC evaluation value of each region in different Plan. (**a**) Plan I (**b**) Plan II (**c**) Plan III (**d**) Plan IV.

#### 5.1.2. Plan II

Under the influence of the open-source program, U<sup>1</sup> and U<sup>2</sup> increase in different degrees and U<sup>3</sup> decreases accordingly. The evaluation results have the highest degree of membership to V2, and the WRCC in the area has been improved. There is no poor (III) in the near and long term (see from Figure 3). In the next 13 years, the WRCC will show a downward trend after a short period of improvement. It shows that the fewer water resources chiefly caused its long-term overloaded status [36] and the construction of water diversion project will promote the improvement of WRCC. However, relying solely on the construction of drinking water projects will not alleviate the problem of resource constraints for a long time in the future. Thus, water conservation should be promoted while transiting water [11].

#### 5.1.3. Plan III

With the construction of the water diversion project and the popularization of water saving and sewage treatment policy, the WRCC of the study area will improve greatly compared with 2017. This plan can improve the utilization rate of water resources and Alleviate the tight of the water supply and demand. Unlike the previous plan, the WRCC of plan III will continue to grow in the future. It can be seen in Figure 3 that the WRCC of the whole study area has improved significantly compared to Plan I, and the average growth rate is 23%. In most areas of Jilin, there are areas with good (I) and medium (II) WRCC, and the difference in WRCC of each administrative region will gradually reduce. It shows that saving water and improving water quality are also important factors for enhancing the WRCC [9,37].

#### 5.1.4. Plan IV

Plan IV not only maintains the water resource utilization rate at a high level but also greatly increases the water resource availability. So, the WRCC of the whole region has been significantly improved compared with Plan I. Moreover, the difference in WRCC of each administrative region has gradually decreased. The WRCC of most areas in the region is good (I), and the WRCC in Changchun and surrounding areas has increased significantly to a relatively high level (Figure 3 shows). As of 2030, the WRCC of the eight administrative regions will increase by more than 50% over 2017. It shows great potential for regional development and utilization. The shortage and uneven spatial and temporal distribution of water resources has seriously restricted the sustainable development of regional society and economy [38]. Plan IV is more in line with the principle of sustainable development of society and meets the development goals of combining water quality, water quantity, water ecology, and water environment, which can be used as a recommended plan.

#### *5.2. Comparative Analysis of the Plan from the Perspective of Space*

The predicted levels of WRCC in 2020 and 2030 at 15 observation locations in the Chang-Ji Economic Circle were analyzed in four plans, the development potential of water resources was evaluated (Figures 4 and 5), and the evaluation level was tested. According to Figures 4 and 5, the WRCC at each observation location exhibited a continuously increasing trend from Plan I to IV. These findings are consistent with the measures used in the design of the plan, which provides a certain level of reliability and reference to the present research.

#### 5.2.1. WRCC Spatial Distribution in 2020

Figure 4 shows that the level of WRCC in the Fengman District of Jilin City, which has unique natural resource surrounding the Songhua River, will be good (I) during each plan in 2020. As the administrative center of Changchun City, Chaoyang District has a relative shortage of water resources and poor water quality. The level is predicted to improve from III to II, and the value to increase from 0.287 to 0.579. Nanguan, Changyi, and Chuanying District will retain the high level of II and will always have a certain development potential; the values for the three areas will change by 0.268, 0.136, and 0.153, respectively. The rest of the region will change from II to I and may adapt to social and economic development.

#### 5.2.2. WRCC Spatial Distribution in 2030

Figure 5 shows that the initial level of WRCC in Chaoyang and Lvyuan District are both at III, and the potential of water resources exploitation is small. The WRCC level Chaoyang changes from III to II, and it is greatly improved from III to I for Lvyuan District. Additionally, their carrying capacity is greatly improved. Changyi and Chuanying District will maintain the high level of II, and the value will change by 0.176 and 0.192, respectively. The rest of the region will change from II to I, which will gradually increase the security effect of the economy and society. The different plans to the subareas can provide a scientific reference to rational distribution of economic development, elaborate management of water environment as well as regional sustainable development in the future [35].

**Figure 4.** The level division of WRCC evaluation value in each plan (2020).

**Figure 5.** The level division of WRCC evaluation value in each plan (2030).

Throughout each plan in 2020 and 2030, plan III will greatly improve the WRCC of the whole region, with the eastern part of the study area having a significantly higher WRCC than the western part. Some areas in Jilin have strong WRCC, and each administrative region has a certain amount of development and utilization potential. Plan IV shows that the WRCC of the whole district is obviously improved, compared to the Plan I. By this time, the WRCC of the whole district will be strong, with most areas being at level I. Furthermore, the water resources would be able to provide certain guarantees for social and economic development.

The measures of water-saving, sewage treatment, water diversion projects, and transit water utilization mean that WRCC is constantly changing. Based on the above-mentioned various measures, the results of Plan IV show that over 84% of the regions have a relatively large development potential. As regional development progresses, most regions would develop slowly if no measures are taken (Plan I). These results further validate the intuition and visualization of the WRCC classification and provide a basis for the government to rationally allocate water resources [10,22].

#### *5.3. Limitations and Future Research Directions*

This study selects the GCA, MLR, and FCE combined model to evaluate the WRCC of the Chang-Ji Economic Circle. This combined model makes up for the shortcomings of traditional indicator evaluation methods and achieves qualitative and quantitative evaluation. Moreover, the coupling of the GCA and MLR model reduces the interference of human factors and reduces the error value of the predicted evaluation index. Nevertheless, this study still has certain limitations.


#### **6. Conclusions**

The study established a hybrid model to analyze the WRCC of the Chang-Ji Economic Circle. First of all, in order to make up for the shortcomings of traditional trend analysis, the GCA and MLR coupling model can predict the changing trend of WRCC's influencing factors [28,41,42]. Accurate and quantitative evaluation index trend prediction can increase the credibility of the evaluation results of the WRCC [9]. Then, using the FCE model, the WRCC of each region is evaluated. Finally, based on the coupling results of the hybrid model, the future WRCC of the districts and counties in the Chang-Ji Economic Circle are compared in terms of time and space. It is worth noting that according to the actual situation of the Chang-Ji Economic Circle, the impact of water quality on regional WRCC is considered. The water environmental capacity is taken as a new evaluation index and the WRCC evaluation system is proposed based on water quantity and quality, social economy, and ecological environment. This makes up for the shortcomings of the existing evaluation indicators and can more realistically reflect the status quo of regional development. The presented research results allowed us to draw the following conclusions.

Four different water intake plans are considered to assess the WRCC of the study area in 2020 and 2030. The study aims to eliminate potential problems in the societal development of Chang-Ji Economic Circle through various plans and improve the affordability of economic development. Considering the spatial heterogeneity and spatial evolution, the spatial and temporal dynamic changes of WRCC are analyzed. From the perspective of time changes, the WRCC of Plan I and Plan II remains at a medium level. Affected by the constraints of supply and demand, the WRCC will continue to decline. The improvement of the WRCC in Plan III was better than the abovementioned scenarios, yet the potential development potential of the region is still hindered. Water saving measures and sewage treatment can relieve the pressure of WRCC. In order to achieve sustainable development of the region, Plan IV comprehensively considers the advantages of the above plans and increases the amount of transit water to make up for the shortage of resource-based water shortage. The WRCC of the whole region is generally good, and water resources can support the rapid development of the social economy in Plan IV. From the perspective of space changes, the WRCC of Plan I in Chaoyang and Lvyuan District will become a poor level and lack the potential of water resources exploitation in the future. The improvement thought Plan II will still not be sufficient compared with Plan III. It shows that the WRCC in the eastern area of Chang-Ji Economic Circle is significantly higher than others. The WRCC

of the whole district Plan IV is significantly improved in comparison with the current conditions. Plan IV is proposed as the final recommendation through comprehensive analysis and research. Strengthening sewage treatment and proper use of transit water resources are more conducive to the rapid development of Chang-Ji Economic Circle.

**Author Contributions:** Conceptualization, G.W.; data curation, C.X. and X.L.; formal analysis, G.W., Z.Q., F.M., and Y.S.; funding acquisition, C.X. and X.L.; methodology, G.W.; project administration, C.X. and X.L.; supervision, X.L.; writing—original draft, G.W. All authors have read and agreed to the published version of the manuscript.

**Funding:** The study was financially supported by Natural Science Foundation of China (No. 41572216), the China Geological Survey Shenyang Geological Survey Center "Chang-Ji Economic Circle Geological Environment Survey" project (121201007000150012), the Provincial School Co-construction Project Special-Leading Technology Guide (SXGJQY2017-6), and the Jilin Province Key Geological Foundation Project (2014-13).

**Institutional Review Board Statement:** "Not applicable" for studies not involving humans or animals.

**Informed Consent Statement:** "Not applicable" for studies not involving humans.

**Data Availability Statement:** Data is contained within the article or supplementary material.

**Acknowledgments:** We would like to thank the anonymous reviewers and the editor.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Nomenclature**


COD Chemical oxygen demand

#### **References**


*Article*
