Abstract

Rain–flood utilization refers to transforming some amount of rain or floodwater into ordinary water resources without decreasing flood control standards or damaging the ecological environment of rivers, which has gained widespread attention as it can alleviate water shortages and gain benefits. This paper put forward the evaluation method of rain–flood utilization availability at the distributed watershed scale. Based on the water node, some indices of rain–flood utilization availability were defined. Then the evaluation method and calculation process were unified. Finally, the status and potential of the rain–flood utilization of Hanjiang River Basin were analyzed. The results indicated that the rain–flood resource in the whole basin is 48.9 billion m3, the outflow is 29.9 billion m3, and the actual utilization is about 19.0 billion m3. The current available rain–flood amount and rain–flood utilization potential are 27.7 billion m3 and 11.0 billion m3, and the rain–flood utilization rate is 49.4%. Limited by regulation ability and the rain–flood resources, current rain–flood utilization has a clear threshold range. The potential utilization objects are mainly for a rainfall process of about two to ten years return period. The application in Hanjiang River Basin offers some practical information for assessing rain–flood utilization scientifically, and the premise for effectively guiding and formulating adaptive countermeasures for water resources management.

HIGHLIGHTS

  • New evaluation method of rain–flood resource utilization availability is proposed.

  • Experimental analysis has been performed, and reveals the spatiotemporal distribution of rain–flood utilization availability indexes.

  • Current rain–flood utilization rate of the demonstration area is 49.4%, still having some potential.

  • The work enriches the theory and offers some practical information for scientific assessment.

Graphical Abstract

Graphical Abstract
Graphical Abstract

INTRODUCTION

The provision of a safe, reliable and sustainable supply of water is a key guarantee for socioeconomic development and human living improvement (Navarro-Ortega et al. 2012; Zheng et al. 2016; Zeng et al. 2017). However, in arid and semi-arid regions, the uneven spatial and temporal distribution of water resources will cause water shortages to varying degrees (Mantua et al. 2010; Karthe et al. 2017; Ye et al. 2019). To further guarantee water security, large numbers of reservoirs and dams have been built for providing water resources, flood control, and power generation over the past century (Ren et al. 2019). Multi-purpose reservoirs are the main component measures to alleviate the water shortage, but there is a conflict between flood control and other water-use benefits (Wasimi & Kitanidis 1983; Rani & Moreira 2010; Anvari et al. 2014; Ye et al. 2019). During flood periods, to free up storage capacity for flood control, reservoirs often operate at lower flood-limited water level (FLWL) to release water as residual water. In areas of water shortage, this behavior often leads to water shortages in non-flood seasons (Liu et al. 2011; Ye et al. 2019). In China, affected by climate, the inter-annual distribution of precipitation in the basin varies greatly, and the proportion of water resources in the flood season reaches more than 70% (Lu et al. 2012). Especially in the upper reaches of the Hanjiang River Basin (HRB), which is the water source area of the South-to-North Water Diversion Project of China, the flood season has been divided into two parts, named the summer flood season and autumn flood season. For the safety of autumn flood prevention, rain–flood resources in summer often cannot be actively utilized. If the incoming water of autumn is insufficient, it will seriously affect the water resources supply-and-demand security of basins. In recent years, the incoming water from the upper reaches of the HRB has continued to decrease. From 2012 to 2016, Danjiangkou reservoir had an average inflow runoff of 25.1 billion m3 for five consecutive years, only 65% of the multi-year average runoff, bringing the conflicts between flood control and water supply tasks to increasing prominence. Under these conditions, rain–flood utilization has become a research hot topic in recent years.

Rain–flood resource utilization refers to transforming some amount of rain or floodwater into ordinary water resources on a basin scale without decreasing flood control standards or damaging the ecological environment of rivers (Croley & Raja Rao 1979; Liu et al. 2009; Wang et al. 2012a, 2012b). China's rain–flood development and utilization model is increasingly mature in practice, but its corresponding availability evaluation theory is still in the exploratory stage. In order to guide the effective utilization of regional rain–flood resources, although some scholars have proposed floodwater availability analysis methods, they are limited to the related concepts of flood resource quantity, available quantity, and utilization potential, lacking the systematicity of theoretical methods (Hu et al. 2010; Kim & Furumai 2012; Zou et al. 2015; Ye et al. 2019). Simultaneously, due to the differences in research purposes and research areas, the present analysis objects are mostly single basins, and the correlation between distributed sub-basins has not been coordinated. For example, Ye et al. (2019) put forward an assessment approach to the floodwater utilization potential of a basin and made an empirical analysis in the Nansi Lake Basin of China. But the method did not consider the direct relevance of the sub-basin and lacks the analysis of spatial distribution characteristics. Obviously, the rain–flood availability potential assessment system is still incomplete and there are no systematic analysis methods. Systematically and effectively measuring the indices of rain–flood availability and potential and drawing their distribution map are the basis for the practice of water resource utilization in the basin and for guiding and formulating countermeasures for water resources management.

This paper addresses the evaluation of rain–flood resource utilization availability at the distributed watershed scale. The distributed hydrological model (e.g. SWAT model) has been constructed for water node division and runoff restoration first. Second, based on the division of the rain–flood process, concepts of rain–flood resource amount (WF), available rain–flood amount (Wkly), and rain–flood utilization potential (R) have been defined. The calculation principle and method of rain–flood utilization availability index based on distributed water nodes are proposed. Third, taking the Hanjiang River Basin as a demonstration area, rain–flood utilization availability indices are calculated from top to bottom for each sub-basin, combined with the ArcGIS drawing platform to describe the distribution map of rain–flood availability. Last, results of the analysis are demonstrated, and summarizing remarks are drawn.

METHODOLOGY

Hydrological model and water node division

Hydrological model

A distributed hydrological model is needed in our methodology. In our study, the Soil and Water Assessment Tool (SWAT) was selected, which is a semi-distributed, eco-hydrologic model and easy to establish and operate (Arnold et al. 1998). Of course, other hydrological models can also be selected without specific requirements here. SWAT divides the basin into various sub-watersheds based on water node and constructs hydrology response units (HRUs), which are the basic computing unit. HRUs are the unique combination of soil, land use and slope characteristics, and are hydrologically homogeneous (Deng et al. 2019). The SWAT model simulates the land hydrology cycle based on a water mass balance, for which the equation is as follows (Wang et al. 2012b):
formula
(1)
where: Swi is the final soil water content, Swoi is the initial soil water content, Pi is the amount of precipitation, Qsfi is the amount of surface runoff, Ei is the evapotranspiration amount, Wsi is the water amount that enters the vadose zone from the soil profile, Qwi is the return flow amount and t is the time.

The SWAT model requires three basic inputs: a digital elevation model (DEM, 90 m × 90 m, https://glovis.usgs.gov/), a land-use cover map (LULC, http://westdc.westgis.ac.cn/), and a soil map (http://westdc.westgis.ac.cn/). These inputs are integrated for delineating the watershed into sub-basins and HRUs. Driving the model requires some daily time-series of meteorological data, such as precipitation, air temperature, pan evaporation and relative humidity. Our team has built the SWAT model in the Hanjiang River Basin and it performs with good accuracy, which is the Nash–Sutcliffe efficiency (NS) and percent bias (PBIAS) up to satisfactory levels (0.50 < NS < 0.65,|PBIAS| < 10%) (Nash & Sutcliffe 1970; Abbaspour et al. 2007; Moriasi et al. 2007; Deng et al. 2019), by using the sequential uncertainty fitting (SUFI-2) algorithm of the SWAT-CUP software for auto-calibration and uncertainty analysis (Abbaspour et al. 2007).

In this study, the hydrological model is only an auxiliary tool, which is mainly used for the calculation of the natural rain–flood process. Because the SWAT model is widely used, the specific construction process will not be described in detail. If necessary, refer to the literature of Deng et al. 2019.

Water node division

Our theory is based on the study of water nodes and the entire basins above the nodes, so it needs the water node division first. Fortunately, the SWAT model has made a preliminary division of the watershed. On this basis, we only need to combine the distribution of rain–flood resource utilization projects (RUP) and gauges (GA) in the basin, such as reservoirs and hydrometry stations and so on, and to screen out the nodes to be studied. The schematic diagram of water node division is illustrated in Figure 1.

Figure 1

Schematic diagram of water node division.

Figure 1

Schematic diagram of water node division.

In this application, there are ten reservoirs with strong rain–flood regulation capacity in the Hanjiang River Basin, and hydrometry gauges are arranged at the downstream of the reservoirs for discharge observation. Based on this information, Hanjiang River Basin has been divided into 12 sub-watersheds, each of which has an exit water node named G.

Derivation of rain–flood utilization availability indices

Derivation of rain–flood utilization availability indices can be done through a single watershed. First, there are some rain–flood utilization availability indices that need to be defined.

  • (1)
    Rain–flood resource amount (WF), that is the amount of rain or flood (named rain–flood) resource, is also the sum of the local natural runoff formed by the precipitation and the inflow of the basin during the same flood period. It can be expressed as:
    formula
    (2)
where: WF is the rain–flood resource amount; I(t) is the local natural rain–flood in the basin during the flood period of t; II(t) is the inflow of the basin during the flood period of t; ts, te are the start and end of the rain–flood process.

  • (2)
    Rain–flood utilization amount (Wly(x)), that is, under the rain–flood control capacity (FCC) of x, the water storage variables of the basin during the flood season, including the water consumption and the amount of water transferred outside the watershed, are recorded as:
    formula
    (3)
    where: Ws(x) represents the amount of water released or discharged from the node during the flood season under the FCC of x. If the regulatory FCC is taken as the current regulatory capacity of x0, the formula of represents the current rain–flood utilization amount, also known as the current rain–flood utilization capacity. If the Ws(x) is set to the actual discharge or outflow (Wsc) at the node during the flood season, then represents the actual rain–flood utilization, which satisfies .
  • (3)
    Available rain–flood amount (Wkly(x)), which is defined as the exploited rain–flood that can meet the necessary water requirements of the river channel (that is, production and living, ecological and environmental water use, also known as the minimum discharge (Wb)) under the FCC of the water node, recorded as:
    formula
    (4)
    where: Wqu(x) is the uncontrollable rain–flood due to the FCC, and Wb is the minimum discharge of the node. Wc is the unavailable utilization rain–flood.
  • If x → +∞ is taken as a limit, the unusable rain–flood amount of Wqu(x) will be established as 0, and then the extreme available rain–flood amount (Wkly_max) is equal to WFWb.

  • (4)
    Rain–flood utilization potential (Rx), which is defined as the increasing amount of rain–flood utilization compared with the actual condition under the meeting of the necessary water demand of the river, as the FCC increases, recorded as:
    formula
    (5)

If Rx < 0, it indicates water over-exploitation in the basin above the water node, and the rain–flood utilization potential is zero.

If the uncontrollable rain–flood under the FCC of x0 is known, then the current rain–flood utilization potential is defined as R1:
formula
(6)
If the rain–flood control project is perfect, that is, taking x → +∞, then the uncontrollable rain–flood of Wqu(x) is equal to 0, and the extreme rain–flood utilization potential is set as R2:
formula
(7)
  • (5)
    Rain–flood utilization rate(rx), which is the ratio of rain–flood utilization to the rain–flood available extreme amount, and characterizes the degree of rain–flood utilization. The formula is as follows:
    formula
    (8)

Taking the water node as the object, the interrelationship between the indices is given in Figure 2 (for example). There are four lines in the figure. Line① indicates the natural rain–flood process at the node, which is denoted as QF. Line② represents the rain–flood process at the node under the current FCC of x0, which is denoted as Q(x0). Line③ denotes the rain–flood process at the node under the current FCC of x, which is denoted as Q(x). Line④ shows the minimum discharge process at the node, which is denoted as Qb. During the flood period of [te,ts], the rain–flood resource amount is the area enclosed by line① and the abscissa, which is written as WF = A + B + C + D. The current rain–flood utilization amount is the area enclosed by line① and line②, recorded as . The available rain–flood amount is the area enclosed by line① and line③, and recorded as Wkly(x) = A + B. Extreme rain–flood utilization is the area enclosed by line① and line④, and recorded as Wkly_max = A + B + C. Rain–flood utilization potential is the area enclosed by line② and line③, recorded as Rx = B. Extreme rain–flood utilization potential is the area enclosed by line② and line④, recorded as R2 = B + C.

Figure 2

Schematic diagram of the relationship among rain–flood utilization availability indices.

Figure 2

Schematic diagram of the relationship among rain–flood utilization availability indices.

Evaluation method and calculation process

Based on the above analysis, the evaluation method and calculation process of rain–flood resource utilization availability can be obtained, including the following steps.

Step 1: Prepare basic data, including hydrometeor, DEM, land use and soil, watershed, RUP, GA, etc.

Step 2: Construct model and divide water nodes. The main process includes the establishing of a distributed hydrological model in the study area, dividing the sub-basin, and determining the distributed water nodes of the basin based on the water system pattern and the regulation and control project considering the rain–flood utilization. Then use the distributed hydrological model to calculate the natural runoff of the water nodes, and delineate the rain–flood periods.

Step 3: Calculate the basic constraint water volume. First, based on the defined rain–flood periods, the minimum discharge of the water nodes is calculated, which is the lower limit constraint. Second, calculate the actual rain–flood utilization of the nodes over the years, and then estimate the current rain–flood utilization capacity, which is an upper bound constraint. Due to uncertain factors such as administrative intervention and deviations in water and rain judgments during the actual operation of water conservancy projects, it is difficult to regulate and operate water conservancy projects to operate impartially in accordance with a given dispatching rule. Moreover, rainfall has also changed over the years. The rain–flood utilization amount under the condition of current regulation capacity x0 is inconsistent. In order to truly reflect the current level of rain–flood utilization in the river basin, the current rain–flood utilization capacity is considered to be approximately the largest actual amount of rain–flood utilization in the past ten years, recorded as:
formula
(9)
where: Wly(k) represents the actual rain–flood utilization in the k-th year of the water node during the flood season.
Finally, the statistical survey method is used to calculate the water demand for production and living (W1), the Tennant method (Tennant 1976) is used to calculate the ecological water demand (W2), and the 7Q10 method (Xu et al. 2016) is used to calculate the environmental water demand (W3), then comprehensive production, living, and ecological-environment water demand, to obtain the minimum discharge of the node (Wb), which is a lower bound constraint. The formula for Wb is as follows:
formula
(10)
Under the current conditions, restricted by the regulated utilization capacity, unavailable utilization rain–flood (Wc) is mainly composed of Wb and uncontrollable rain–flood due to the FCC (Wqu), which can be calculated using the following formula:
formula
(11)
Then Wc takes the outsourced value of the Wb and Wqu, as follows:
formula
(12)

Step 4: Evaluate and calculate the rain–flood resource utilization availability indices. First, formulate the principles of distributed and coordinated calculation, including ① the safety principle, that is, the utilization of rain–flood resources is based on the premise of ensuring the safety of flood control, and the flood characteristics and engineering regulation capabilities of the river basin must be considered in coordination. ② Systematic criteria, that is, comprehensive consideration of the hydraulic connection among the distributed water nodes in the basin, starting from the research area and based on the top-down systematic measurement principle. ③ Coordination criteria, that is, to fully consider the differences in rain–flood utilization methods, capabilities, and targets in the sub-basin, reasonably coordinate the basic water requirements for production, living, and the ecological environment in the upstream and downstream rivers, and avoid excessive development of water resources. Second, rain–flood resource utilization availability indices are calculated for the distributed water node, and combined with the ArcGIS mapping platform, the indices distribution map is drawn.

This article takes the Hanjiang River Basin as an example, and selects a built SWAT distributed hydrological model to conduct a study. The rainy period in the flood season is from June to October. The flow chart of the analytical process is illustrated in Figure 3.

Figure 3

Flow chart of the analytical process.

Figure 3

Flow chart of the analytical process.

STUDY AREA AND DATA

Study area

Hanjiang River originates in the south of Qinling Mountain in the southwest of Shanxi Province, and flows east across the southern part of Shanxi Province into Hubei Province, which is accompanied by the development of unbalanced tributaries (Wang et al. 2012a, 2012b). The area of Hanjiang River Basin (HRB) is 159,000 km2, which is one of the largest tributary regions of the Yangtze River Basin. The terrain shows high in the west and low in the east, and its elevation varies as 1 ∼ 3,535 m. The upper reaches (106.0° ∼ 110.5°E longitude, 31.5° ∼ 34.5°N latitude and 157 ∼ 3,508 m asl), named the UHRB, are the main water source area of the inter-basin water transfer project with a drainage area of about 61,703 km2 (Deng et al. 2019). The main rain–flood resource utilization projects (RUP) in the HRB are Shiquan, Ankang, Pankou, Huanglongtan, Danjiangkou, Sanliping, Yahekou, Wangfuzhou, Cuijiaying and Xinglong, and Danjiangkou is the main core project.

The HRB belongs to the East Asian subtropical monsoon region with an annual mean temperature of 12 ∼ 16 °C and an average annual rainfall of 700 ∼ 1,800 mm, of which 70% ∼ 80% is concentrated in the period from May to October. Maximum continuous four-month precipitation accounts for about 55% ∼ 65% of the annual total. The annual mean runoff of the HRB is about 462 × 108 m3, accounting for 60% ∼ 70% of the total runoff of the whole basin with large inter-annual variability. Usually, late June to late July is the summer flood season and late August to mid-October is the autumn flood season with abundant water resources (Deng et al. 2019).

Data collection

The daily data of 96 rainfall gauges and 26 meteorological stations across the whole HRB were collected as the inputs to construct the SWAT model. In addition, the daily discharge series of 16 hydrometry stations were chosen as the modeling reference dataset and water node devision. The time series of the data is from 1956.01.01 to 2016.12.31, which were supplied by the Bureau of Hydrology (BOH), Changjiang Water Resources Commission (CWRC) and Climatic Data Center, National Meteorological Information Center, China Meteorological Administration (http://data.cma.cn/). Figure 4(b) displays the spatial distribution of hydrological and meteorological gauges.

Figure 4

Location and layout of the HRB and the spatial distribution of gauges.

Figure 4

Location and layout of the HRB and the spatial distribution of gauges.

CASE APPLICATION AND RESULT DISCUSSION

Rain–flood resource and basic constraint

Rain–flood resource amount

Based on the above methods, the rain–flood resource amount (WF) in each sub-watershed from 1956 to 2016 is calculated. In multi-year average statistics, the WF of the HRB is 48.9 billion m3, the outflow is 29.9 billion m3, and the actual utilization is about 19.0 billion m3. From the perspective of spatial distribution, we can see from Figure 5(a), the rain–flood resources are mainly distributed in the basins of SQ (BSQ) and PK (BPK) and the interval basin from SQ to AK (SQ-AK), for which the WF is 9.25 billion m3, 4.58 billion m3, and 7.24 billion m3 respectively, owing to SQ-AK and PK both being rainstorm areas, so there are many more rain–flood resources, and the runoff depth being 489 and 512 mm respectively. While in the lower reaches, rain–flood resources are concentrated in the basins below XY, for which the WF in the area of XY to HZ (XY-HZ) and HZ to Outlet (HZ–Outlet) is 5.0 billion m3 and 4.30 billion m3, and the runoff depth is 238 and 301 mm, respectively.

Figure 5

(a) Spatial distribution of annual average WF; (b) inter-annual changes of WF at the HZ node.

Figure 5

(a) Spatial distribution of annual average WF; (b) inter-annual changes of WF at the HZ node.

From the perspective of inter-annual change, rain–flood has certain fluctuation characteristics. Take the HZ node as an example for further analysis. From Figure 5(b), the rain–flood resource amount mainly depends on precipitation, showing the trend of increasing first and then decreasing. There was a more obvious mutation in 1990, that is, the overall reduction of rain–flood resources after 1990. In multi-year average statistics, the WF of HZ is 44.6 billion m3 within a maximum value of 90 billion m3 and a minimum value of 20.7 billion m3. The corresponding average annual outflow amount is 28.3 billion m3, and the actual rain–flood utilization amount is about 17.5 billion m3.

Basic constraint of rain–flood utilization

Current rain–flood utilization capacity (Upper bound)

In multi-year average statistics from 1956 to 2016, the current rain–flood utilization amount (Wly) in the HRB is 30.6 billion m3, and Figure 6(a) shows its spatial distribution. In the UHRB, the high values are mainly distributed in the BSQ, and the interval basin from AK to BH (AK–BH) and BH to DJK (BH–DJK), for which the value is 2.05 billion m3, 1.89 billion m3, and 5.0 billion m3 respectively. In the middle and lower reaches, the high values are mainly concentrated below XY, such as the Wly in XY–HZ and HZ–Outlet being 2.03 billion m3 and 1.88 billion m3 respectively. The low-value areas are mainly located in the Nanhe River and the lower reaches of Duhe River, for which the value is 0.60 billion m3 and 0.26 billion m3.

Figure 6

Spatial distribution of annual average (a) Wly; (b) Wly(x0).

Figure 6

Spatial distribution of annual average (a) Wly; (b) Wly(x0).

Based on the analysis of Wly, the current rain–flood utilization capacity (Wly(x0)) in different sub-areas is further analyzed (see Figure 6(b)). Table 1 shows the annual average statistics of Wly(x0) of typical nodes from 1956 to 2016. It is indicated that the current rain–flood utilization capacity of the whole basin is 30.6 billion m3. In the UHRB, the Wly(x0) of SQ is 3.35 billion m3, while that of SQ–AK and AK–BH is 4.18 billion m3 and 2.46 billion m3 respectively. Affected by DJK reservoir regulation, the value of Wly(x0) in BH–DJK is largest, reaching 5.27 billion m3. In recent years, with the launch of RUP such as Wang Fuzhou, Cuijiaying, and Xinglong in the middle and lower mainstream, rain–flood utilization capacities in the area have been strengthened. Among them, the current rain–flood utilization capacity of XY–HZ is 3.83 billion m3, and that of HZ–HK is 2.68 billion m3.

Table 1

Annual average statistics of rain–flood utilization levels of typical nodes (1956 ∼ 2016)

RiverNodeWly (billion m3)WF (billion m3)Wly(x0) (billion m3)RiverNodeWly (billion m3)WF (billion m3)Wly(x0) (billion m3)
Hanjiang SQ 2.05 9.25 3.35 Duhe PK 1.61 4.89 3.06 
AK 3.84 16.5 7.53 HLT 1.87 4.58 3.38 
BH 5.73 21.8 9.99 Nanhe SLP 0.354 5.11 3.69 
HJG 12.6 33.0 18.95 Baihe XDP 1.14 1.48 0.96 
XY 13.2 35.3 20.4 Tanghe GT 0.749 2.41 2.07 
HZ 17.1 44.6 27.9 Hanjiang Outlet 19.0 48.9 30.6 
RiverNodeWly (billion m3)WF (billion m3)Wly(x0) (billion m3)RiverNodeWly (billion m3)WF (billion m3)Wly(x0) (billion m3)
Hanjiang SQ 2.05 9.25 3.35 Duhe PK 1.61 4.89 3.06 
AK 3.84 16.5 7.53 HLT 1.87 4.58 3.38 
BH 5.73 21.8 9.99 Nanhe SLP 0.354 5.11 3.69 
HJG 12.6 33.0 18.95 Baihe XDP 1.14 1.48 0.96 
XY 13.2 35.3 20.4 Tanghe GT 0.749 2.41 2.07 
HZ 17.1 44.6 27.9 Hanjiang Outlet 19.0 48.9 30.6 

Duhe River is a first-level tributary of the Hanjiang River, and its mainstream is built with two rain–flood resource utilization projects named Pankou and Huanglongtan. Affected by reservoir regulation, the rain–flood utilization capacity is relatively higher, for which the Wly(x0) of HLT is 3.38 billion m3. In contrast, although Nanhe River has SLP reservoir, the reservoir's capacity for flood control is relatively small, and its capacity for regulating flood is relatively low, so the Wly(x0) of SLP is only 960 million m3.

The Tanghe River and Baihe River are in the lower reaches of the HRB, and the Yahekou reservoir is also built on the Tanghe River. Owing to the low water resources, the high water-demand, and the well-developed water conservancy facilities in the region, the rain–flood utilization capacity is high, especially in the Tanghe River: the Wly(x0) of GT is 2.07 billion m3, accounting for 85.9% of the rain–flood resource amount.

Unavailable utilization rain–flood (Lower bound)

Analysis shows that (as shown in Table 2), in order to meet the necessary water demand downstream for production, living and ecological environment, the total water demand for the entire basin is 10.43 billion m3 and the minimum discharge (Wb) of the nodes is 0.35 ∼ 9.8 billion m3. The spatial distribution of Wb is shown in Figure 7(a).

Table 2

Annual average statistics of unavailable utilization rain–flood (1956 ∼ 2016)

RiverNodeWb (billion m3)Wqu (billion m3)Wc (billion m3)RiverNodeWb (billion m3)Wqu (billion m3)Wc (billion m3)
Hanjiang SQ 2.19 5.90 5.96 Duhe PK 1.03 1.46 1.83 
AK 4.10 8.99 9.33 HLT 1.13 1.66 2.04 
BH 5.50 11.8 12.3 Nanhe SLP 0.35 0.55 0.63 
HJG 7.70 14.3 15.7 Baihe XDP 0.54 0.75 1.02 
XY 8.28 15.2 16.7 Tanghe GT 0.36 0.58 0.76 
HZ 9.80 17.2 19.7 Hanjiang Outlet 10.43 18.8 21.31 
RiverNodeWb (billion m3)Wqu (billion m3)Wc (billion m3)RiverNodeWb (billion m3)Wqu (billion m3)Wc (billion m3)
Hanjiang SQ 2.19 5.90 5.96 Duhe PK 1.03 1.46 1.83 
AK 4.10 8.99 9.33 HLT 1.13 1.66 2.04 
BH 5.50 11.8 12.3 Nanhe SLP 0.35 0.55 0.63 
HJG 7.70 14.3 15.7 Baihe XDP 0.54 0.75 1.02 
XY 8.28 15.2 16.7 Tanghe GT 0.36 0.58 0.76 
HZ 9.80 17.2 19.7 Hanjiang Outlet 10.43 18.8 21.31 
Figure 7

(a) Spatial distribution of Wb; (b) inter-annual changes of Wc at the HZ node.

Figure 7

(a) Spatial distribution of Wb; (b) inter-annual changes of Wc at the HZ node.

Based on Wb, the unavailable utilization rain–flood (Wc) is further calculated, shown in Table 2. In multi-year average statistics, the Wc in the HRB is 21.31 billion m3, and the uncontrollable rain–flood due to the FCC (Wqu) is 18.8 billion m3. Figure 7(b) shows the inter-annual change of Wc at the typical node of HZ. Obviously, in years of abundant water (such as 1964, 1983, 2011, etc.), unavailable utilization rain–flood is mainly limited by the rain–flood control capacity, which has a high value of Wqu, while in the dry years (such as 1959, 1972, 1999, etc.), unavailable utilization rain–flood is mainly constrained by the necessary water demand, and it has a low value of Wqu.

Rain–flood availability and potential assessment

Available rain–flood amount

Based on the methods, the current available rain–flood amount (Wkly) and the extreme available rain–flood amount (Wkly_max) are calculated. The results are presented in Figures 8 and 9.

Figure 8

Boxplot of Wkly at the typical nodes (1956 ∼ 2016).

Figure 8

Boxplot of Wkly at the typical nodes (1956 ∼ 2016).

Figure 9

Boxplot of Wkly_max at the typical nodes (1956 ∼ 2016).

Figure 9

Boxplot of Wkly_max at the typical nodes (1956 ∼ 2016).

It is indicated that, from the multi-year average perspective, the Wkly and Wkly_max in the HRB are 27.7 billion m3 and 38.5 billion m3 respectively. The upper reaches (UHRB) is the key area for rain–flood utilization, for which the Wkly and Wkly_max at the node of DJK are 17.3 billion m3 and 25.3 billion m3, accounting for 62.5% and 65.7% of the whole basin, respectively. Among the main tributaries, the high value of Wkly is distributed in the Duhe River while the low value is distributed in the Nanhe River, e.g., the Wkly of HLT node in the Duhe River is 3.06 billion m3, and the Wkly_max is 3.98 billion m3, which are 3.64 and 3.52 times that of SLP node respectively.

From the multi-year average perspective, the available rain–flood amount is further allocated to each sub-basin, and the spatial distribution map is drawn, shown in Figure 10. Obviously, in the upper reaches, the available rain–flood amount in the area of BH–HLT–DJK (Baihe River) is relatively the largest, for which the values of Wkly and Wkly_max are 5.06 billion m3 and 5.53 billion m3 respectively. The second is the watershed of SQ–AK (Duhe River), for which the values of Wkly and Wkly_max are 3.87 billion m3 and 5.33 billion m3 respectively. But in the interval section of AK to BK, due to the lack of large- and medium-size water conservancy projects, the available rain–flood amount is relatively small, the values of Wkly and Wkly_max being 2.32 billion m3 and 3.88 billion m3 respectively.

Figure 10

Spatial distribution of (a) Wkly and (b) Wkly_max.

Figure 10

Spatial distribution of (a) Wkly and (b) Wkly_max.

Rain–flood utilization potential

Based on the available rain–flood amount, the rain–flood utilization potential was further calculated. Table 3 presents the statistics of rain–flood utilization potential from 1956 to 2016 at the typical nodes, and Figure 11 shows the spatial distribution of the multi-year average rain–flood utilization potential.

Table 3

Annual average statistics of rain–flood utilization potential (1956 ∼ 2016)

RiverNodeR1 (billion m3)
R2 (billion m3)
MeanMaxMinMeanMaxMin
Hanjiang SQ 1.27 3.35 0.00 4.91 16.7 0.00 
AK 3.46 7.53 0.00 8.66 29.2 0.05 
BH 3.98 10.5 0.00 10.4 36.2 0.00 
HJG 5.37 20.9 0.00 12.8 47.0 0.00 
XY 6.31 22.6 0.00 14.2 51.3 0.00 
HZ 9.09 28.1 0.00 18.6 70.2 0.00 
Duhe PK 11.1 43.0 0.00 21.6 81.2 0.00 
HLT 1.03 2.53 0.00 1.81 5.57 0.00 
Nanhe SLP 1.05 2.60 0.00 1.95 5.67 0.00 
Baihe XDP 0.39 0.94 0.00 0.65 2.17 0.00 
Tanghe GT 0.22 1.94 0.00 0.63 6.02 0.00 
Hanjiang Outlet 11.0 43.0 0.00 21.5 81.2 0.00 
RiverNodeR1 (billion m3)
R2 (billion m3)
MeanMaxMinMeanMaxMin
Hanjiang SQ 1.27 3.35 0.00 4.91 16.7 0.00 
AK 3.46 7.53 0.00 8.66 29.2 0.05 
BH 3.98 10.5 0.00 10.4 36.2 0.00 
HJG 5.37 20.9 0.00 12.8 47.0 0.00 
XY 6.31 22.6 0.00 14.2 51.3 0.00 
HZ 9.09 28.1 0.00 18.6 70.2 0.00 
Duhe PK 11.1 43.0 0.00 21.6 81.2 0.00 
HLT 1.03 2.53 0.00 1.81 5.57 0.00 
Nanhe SLP 1.05 2.60 0.00 1.95 5.67 0.00 
Baihe XDP 0.39 0.94 0.00 0.65 2.17 0.00 
Tanghe GT 0.22 1.94 0.00 0.63 6.02 0.00 
Hanjiang Outlet 11.0 43.0 0.00 21.5 81.2 0.00 
Figure 11

Spatial distribution of (a) R1 and (b) R2.

Figure 11

Spatial distribution of (a) R1 and (b) R2.

From the multi-year average perspective, it is shown that the current rain–flood utilization potential (R1) and extreme rain–flood utilization potential (R2) in the whole basin are 11.0 billion m3 and 21.5 billion m3 respectively. The areas with high values of rain–flood utilization potential are mainly distributed in the BAK and the mainstream areas below HJG. For example, R1 and R2 at the node of AK are 3.46 billion m3 and 8.66 billion m3 respectively, accounting for 31.5% and 40.3% of that of the whole basin. Similarly, R1 and R2 in the mainstream areas below HJG are 4.27 billion m3 and 6.0 billion m3 respectively, accounting for 38.8% and 27.9% of that of the whole basin. In contrast, areas with low values are concentrated in the northern part of the HRB, especially the Baihe River. R1 and R2 in this area are 390 million m3 and 650 million m3 respectively, which only account for 3.55% and 3.02% of that of the entire basin.

It is well known that rain–flood utilization potential mainly comes from uncontrollable rain–flood, which is limited by the current regulation capacity and the amount of rain–flood resources. Figure 12 shows the correlation between rain–flood resources and rain–flood utilization potential for some typical nodes of the mainstream of the Hanjiang River.

Figure 12

Comparison of R1 and R2 vs WF by the scatter plots at the node of (a) HJG and (b) HZ.

Figure 12

Comparison of R1 and R2 vs WF by the scatter plots at the node of (a) HJG and (b) HZ.

It can easily be found that R1 has a clear threshold range. Taking the HJG node as an example, when the WF is less than 37.88 billion m3 (corresponding to precipitation of 410 mm, having a return period of five to eight years) and greater than 20 billion m3 (corresponding to precipitation of 207 mm, having a return period of two to three years), it is an ideal space for the utilization of current rain–flood resources. In other words, the main target for rain–flood utilization in the basins tends to the small- and medium-size floods with a return period of two to ten years. If the limitation of regulatory capacity is not considered, from the perspective of extreme utilization potential, as long as the rain–flood resources are greater than 15 billion m3, the larger the rain–flood resource amount, the greater its potential, which fully reflects the restrictive effect of rain–flood regulation and control capacity on tapping rain–flood utilization potential.

Analysis of rain–flood utilization

Further calculating rates (rx) to evaluate the level of rain–flood utilization in the basin: Figure 13(a) displays the spatial distribution of rx from a multi-year average perspective. It is easily found that the average rain–flood utilization rate of the whole basin is 49.4%, and that of the sub-basin is 29% ∼ 90.4%. The high values of rx are mainly distributed in the northern part of the basin, such as Danjiang and Baihe River, and the rain–flood utilization ratio is more than 50%, while in the basin of SQ and the Nahe River, the level of rain–flood utilization is relatively low, for which rx at the node of SQ and SLP is only 29% and 31.3% respectively. From the perspective of water resources in zoning, the level of rain–flood utilization of the UHRB is generally higher, for which rx is about 54.9% while for the others it is about 44.1% ∼ 54.4%.

Figure 13

(a) Spatial distribution of rx and (b) scatter plot of rx vs WF at the node of HZ.

Figure 13

(a) Spatial distribution of rx and (b) scatter plot of rx vs WF at the node of HZ.

Rain–flood utilization is limited by its regulation ability, and related to the rain–flood resource amount. The scatter plot of rx vs WF at the node of HZ is shown in Figure 13(b). Due to control errors of the water conservancy project during the flood season it is not always operated in strict accordance with the dispatching regulations. Moreover, the construction of the water conservancy project is constantly being improved, and the reservoir dispatching regulations will be adjusted accordingly. Therefore, there must be differences in the extent of rain–flood utilization over the years. Under the combined influence of these factors, the relationship between WF and rx cannot be expressed by a single mathematical equation, but there is a clear negative correlation, that is, the greater the rain–flood resource amount, the lower the rain–flood utilization rate. This is mainly due to the limited control capacity of water conservancy hubs and other supervised projects, resulting in excessive and uncontrollable volumes of water.

Table 4 specifically counts the thresholds of current rain–flood utilization at the typical nodes. It needs to be pointed out that potential objects of flooding utilization in the HRB are mainly for a rainfall process of about two to ten years return period, and some sub-watersheds only have two to five years. When the magnitude of rainstorms in the flood season is greater than that of the five or ten years return period, by the influence of rain–flood regulation capacity, the excavation potential of rain–flood will be restricted.

Table 4

Thresholds of current rain–flood utilization at the typical nodes

RiverNodeCurrent rain–flood utilization
RiverNodeCurrent rain–flood utilization
Thresholds (billion m3)Precipitation (mm)Return period (year)Thresholds (billion m3)Precipitation (mm)Return period (year)
Hanjiang SQ 5.70 ∼ 9.96 250 ∼ 415 2 ∼ 5 Duhe PK 2.20 ∼ 6.89 236 ∼ 732 2 ∼ 10 
AK 11.7 ∼ 17.2 300 ∼ 430 2 ∼ 5 HLT 2.72 ∼ 7.16 225 ∼ 633 2 ∼ 8 
BH 16.2 ∼ 23.5 270 ∼ 394 2 ∼ 5 Nanhe SLP 1.13 ∼ 20.2 216 ∼ 331 2 ∼ 6 
HJG 20.0 ∼ 37.9 207 ∼ 410 2 ∼ 8 Baihe XDP 2.35 ∼ 3.79 214 ∼ 346 2 ∼ 8 
XY 20.0 ∼ 40.4 215 ∼ 415 2 ∼ 10 Tanghe GT 1.50 ∼ 2.97 220 ∼ 432 2 ∼ 5 
HZ 22.0 ∼ 63.2 230 ∼ 430 2 ∼ 10 Hanjiang Outlet 24.6 ∼ 68.8 250 ∼ 490 2 ∼ 10 
RiverNodeCurrent rain–flood utilization
RiverNodeCurrent rain–flood utilization
Thresholds (billion m3)Precipitation (mm)Return period (year)Thresholds (billion m3)Precipitation (mm)Return period (year)
Hanjiang SQ 5.70 ∼ 9.96 250 ∼ 415 2 ∼ 5 Duhe PK 2.20 ∼ 6.89 236 ∼ 732 2 ∼ 10 
AK 11.7 ∼ 17.2 300 ∼ 430 2 ∼ 5 HLT 2.72 ∼ 7.16 225 ∼ 633 2 ∼ 8 
BH 16.2 ∼ 23.5 270 ∼ 394 2 ∼ 5 Nanhe SLP 1.13 ∼ 20.2 216 ∼ 331 2 ∼ 6 
HJG 20.0 ∼ 37.9 207 ∼ 410 2 ∼ 8 Baihe XDP 2.35 ∼ 3.79 214 ∼ 346 2 ∼ 8 
XY 20.0 ∼ 40.4 215 ∼ 415 2 ∼ 10 Tanghe GT 1.50 ∼ 2.97 220 ∼ 432 2 ∼ 5 
HZ 22.0 ∼ 63.2 230 ∼ 430 2 ∼ 10 Hanjiang Outlet 24.6 ∼ 68.8 250 ∼ 490 2 ∼ 10 

CONCLUSIONS

In this paper, we firstly put forward the evaluation method of rain–flood resource utilization availability at the distributed watershed scale. Based on this method, some concepts and calculation process have been defined. Then taking the Hanjiang River Basin (HRB) as a demonstration area, the rain–flood utilization availability of the basin has been analyzed. The main research results are as follows.

  • (1)

    From the perspective of theory, we put forward the evaluation method of rain–flood resource utilization availability at the distributed watershed scale, and define the connotation of relevant evaluation indicators, such as the rain–flood resource amount (WF), the available rain–flood amount (Wkly), the rain–flood utilization potential (R) and so on. The steps and calculation principles of rain–flood availability assessment based on water nodes are proposed systematically to guide practical applications.

  • (2)

    Taking the HRB as a demonstration area to develop the applications, it is indicated that the average annual WF of the HRB is 48.9 billion m3, the outflow is 29.9 billion m3, and the actual utilization is about 19.0 billion m3. The rain–flood resource amount mainly depends on precipitation, indicating the trend of increasing first and then decreasing. The obvious mutation is in 1990.

  • (3)

    The Wkly and Wkly_max in the HRB are 27.7 billion m3 and 38.5 billion m3 respectively. The UHRB is the key area for rain–flood utilization, where the Wkly and Wkly_max account for 62.5% and 65.7% of the whole basin respectively. In the Baihe and Duhe sub-basins, there are abundant available rain–flood amounts, while in the interval section of AK to BK, due to lack of large- and medium-size water conservancy projects, there are few available rain–flood amounts.

  • (4)

    The current rain–flood utilization potential of the whole basin is 11.0 billion m3, and the rain–flood utilization rate is 49.4%. There is high rain–flood utilization potential in the basin of AK and the mainstream areas below HJG. Due to the high utilization rate in the northern region, especially the Baihe sub-basin, this area has lower potential to be tapped, for which R1 and R2 are only 390 million m3 and 650 million m3 respectively, accounting for 3.55% and 3.02% of the whole basin.

  • (5)

    Limited by regulation ability and the rain–flood resource amount, the current rain–flood utilization in the HRB has a clear threshold range. Potential objects of flooding utilization are mainly for a rainfall process of about two to ten years return period. When the magnitude of rainstorms is greater than that of the five or ten years return period, by the influence of rain–flood regulation capacity, the excavation potential of rain–flood will be restricted.

In the practice of rain–flood resource utilization, although the rain–flood development and utilization model is becoming increasingly mature, its corresponding availability evaluation theory is still in the exploratory stage. In this paper, the evaluation method of rain–flood resource utilization availability is proposed systematically. Taking the Hanjiang River Basin as an example, the systematic and effective measurement of rain–flood availability and potential indicators is further enriched in the theoretical system of water resource utilization. The work in this paper is the basis of the practice of using water resources in the river basin, and the premise for effectively guiding and formulating adaptive countermeasures for water resources management.

ACKNOWLEDGEMENTS

The authors appreciate the Bureau of Hydrology, Changjiang Water Resources Commission for the support of data. This research work is financially supported by The National Key R&D Program of China (Item No. 2016YFC0400901), Youth Funds of National Natural Science of China (Project No. 51609007), Natural Science Foundation of China (Project No. 51479118) and Open Research Fund of Changjiang Academy of Sciences (Project No. CKWV2019766/KY).

DATA AVAILABILITY STATEMENT

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

REFERENCES

Abbaspour
K. C.
Yang
J.
Maximov
I.
Siber
R.
Bogner
K.
Mieleitner
J.
Zobrist
J.
Srinivasan
R.
2007
Modelling hydrology and water quality in the pre-alpine/alpine Thur watershed using SWAT
.
J. Hydrol.
333
,
413
430
.
http://dx.doi.org/10.1016/j.jhydrol.2006.09.014
.
Arnold
J. G.
Srinivasan
R.
Muttiah
R. S.
Williams
J. R.
1998
Large area hydrologic modeling and assessment part I: model development
.
JAWRA J. Am. Water Resour. Assoc.
34
(
1
),
73
89
.
https://doi.org/10.1111/j.1752-1688.1998.tb05962.x
.
Croley II
T. E.
Raja Rao
K. N.
1979
Multiobjective risks in reservoir operation
.
Water Resour. Res.
15
(
4
),
807
814
.
https://doi.org/10.1029/WR015i004p 00807
.
Deng
P. X.
Zhang
M. Y.
Bing
J. P.
Jia
J. W.
Zhang
D. D.
2019
Evaluation of the GSMaP_Gauge products using rain gauge observations and SWAT model in the Upper Hanjiang River Basin
.
Atmos. Res.
219
,
153
165
.
https://doi.org/10.1016/j.atmosres.2018.12.032
.
Hu
Q. F.
Wang
Y. T.
Yang
D. W.
2010
Assessment approach for flood resources availability and utilization potentiality and its application
.
J. Hydroelectric Eng.
29
(
4
),
22
29
(in Chinese)
.
Karthe
D.
Abdullaev
I.
Boldgiv
B.
Borchardt
D.
Chalov
S.
Jarsjö
J.
Li
L. H.
Nittrouer
J. A.
2017
Water in Central Asia: an integrated assessment for science-based management
.
Environ. Earth Sci.
76
,
690
.
https://doi.org/10.1007/s1266 5-017-6994-x
.
Kim
J.
Furumai
H.
2012
Assessment of rainwater availability by building type and water use through GIS-based scenario analysis
.
Water Resour. Manage.
26
(
6
),
1499
1511
.
https://doi.org/10.1007/s11269-011-9969-9
.
Liu
Z.
Huang
W.-Z.
Huang
Q.
Zhang
Y.-Y.
2009
Floodwater resources utilization based on flood forecast information of reservoir operation chart
.
Adv. Water Sci.
20
(
4
),
578
583
.
https://doi.org/10.3321/j.issn:1001-6791.2009.04.021
.
Liu
X.
Guo
S.
Liu
P.
Chen
L.
Li
X.
2011
Deriving optimal refill rules for multipurpose reservoir operation
.
Water Resour. Manage.
25
,
431
448
.
https://doi.org/10.1007/s1126 9-010-9707-8
.
Lu
G.-H.
Xiao
H.
Wu
Z.-Y.
Zhang
S.-L.
2012
Assessing the impacts of future climate change on hydrology in Huang-Huai-Hai region in China using the PRECIS and VIC models
.
J. Hydrol. Eng.
18
(
9
),
1077
1087
.
https://doi.org/10.1061/(ASCE)HE.1943-5584.00006 32
.
Moriasi
D. N.
Arnold
J. G.
Van Liew
M. W.
Binger
R. L.
Harmel
R. D.
Veith
T. L.
2007
Model evaluation guidelines for systematic quantification of accuracy in watershed simulations
.
Trans. ASABE
50
,
885
900
.
http://dx.doi.org/10.13031/2013.23153
.
Nash
J. E.
Sutcliffe
J. V.
1970
River flow forecasting through conceptual models: part 1. A discussion of principles
.
J. Hydrol.
10
,
282
290
.
https://doi.org/10.1016/0022-1694(70)90255-6
.
Navarro-Ortega
A.
Acuña
V.
Batalla
R. J.
Blasco
J.
Conde
C.
Elorza
F. J.
Elosegi
A.
Francés
F.
La-Roca
F.
Muñoz
I.
Petrovic
M.
Picó
Y.
Sabater
S.
Sanchez-Vila
X.
Schuhmacher
M.
Barceló
D.
2012
Assessing and forecasting the impacts of global change on Mediterranean rivers. The SCARCE Consolider project on Iberian basins
.
Environ. Sci. Pollut. Res.
19
,
918
933
.
https://doi.org/10.1007/s11356-011-0566-5
.
Rani
D.
Moreira
M. M.
2010
Simulation–optimization modeling: a survey and potential application in reservoir systems operation
.
Water Resour. Manage.
24
,
1107
1138
.
https://doi.org/10.1007/s1126 9-009-9488-0
.
Ren
K.
Huang
S. Z.
Huang
Q.
Wang
H.
Leng
G. Y.
Wu
Y. C.
2019
Defining the robust operating rule for multi-purpose water reservoirs under deep uncertainties
.
J. Hydrology.
578
,
124134
.
https://doi.org/10.1016/j.jhydrol.2019.124134
.
Tennant
D. L.
1976
Instream flow regimens for fish, wildlife, recreation and related environmental resources
.
Fisheries
1
(
4
),
6
10
.
https://doi.org/10.1577/1548-8446(1976)001<0006:IFRFFW > 2.0.CO;2
.
Wang
F.
Wang
L.
Zhou
H.
Valeriano
O. C. S.
Koike
T.
Li
W.
2012a
Ensemble hydrological prediction-based real-time optimization of a multiobjective reservoir during flood season in a semiarid basin with global numerical weather predictions
.
Water Resour. Res.
48
(
7
), W07520.
https://doi.org/10.1029/2011 W R0113 66
.
Wang
Q.
Xiao
Q.
Liu
C.
Wang
K.
Ye
M.
Lei
A.
Song
X.
Kohata
K.
2012b
Effect of reforestation on nitrogen and phosphorus dynamics in the catchment ecosystems of subtropical China: the example of the Hanjiang River basin
.
J. Sci. Food Agric.
92
(
5
),
1119
1129
.
http://onlinelibrary.wiley.com/enhanced/exportCitation/doi/10.1002/jsfa.4607.
Wasimi
S. A.
Kitanidis
P. K.
1983
Real-time forecasting and daily operation of a multireservoir system during floods by linear quadratic Gaussian control
.
Water Resour. Res.
19
(
6
),
1511
1522
.
https://doi.org/10.1029/WR019i006p01511
.
Xu
W.
Dong
Z. C.
Luo
X. L.
Fu
X. H.
Tan
J. J.
2016
Analysis of ecological flow in Luanhe River based on improved 7Q10 method
.
J. Hohai University (Natural Sciences)
44
(
5
),
454
457
.
doi:10.3876/j.issn.1000-1980.2016.05.013 (in Chinese)
.
Ye
A.
Wang
Z.
Zhang
L.
Wang
L.
Wang
K.
2019
Assessment approach to the floodwater utilization potential of a basin and an empirical analysis from China
.
Environ. Earth Sci.
78
(
2
),
52
.
https://doi.org/10.1007/s12665-019-8050-5
.
Zeng
X. T.
Zhang
S. J.
Feng
J.
Huang
G. H.
Li
Y. P.
Zhang
P.
Chen
J. P.
Li
K. L.
2017
A multi-reservoir based water-hydroenergy management model for identifying the risk horizon of regional resources-energy policy under uncertainties
.
Energy Conversion Manage.
143
,
66
84
.
https://doi.org/10.1016/j.enconman.2017.02.020
.
Zheng
X.
Wang
C.
Cai
W.
Kummu
M.
Varis
O.
2016
The vulnerability of thermoelectric power generation to water scarcity in China: current status and future scenarios for power planning and climate change
.
Appl. Energy
171
,
444
455
.
https://doi.org/10.1016/j.apenergy.2016.03.040
.
Zou
Q.
Lu
J.
Yu
S.
Ding
Y.
2015
Assessment of flood resources utilization in the Yangtze River Basin
.
J. Water Resour. Res.
4
(
5
),
432
442
.
doi:10.12677/JWRR.2015.45053 (in Chinese)
.