Simulation of stream flows and climate trend detections using WEAP model in awash river basin

Abstract

The shrinking of water resources along with the frequent droughts plays an effective role in intensifying the water crisis. Climate change adversely affects the hydrological cycle at the basin level. These are often reflected in the water and food insecurity, frequent hydrologic extremes, frequent drought, flooding, and deteriorating ecosystem health. This study aims to calibrate and validate the stream flows in Awash River basin using a distributed hydrological Water Evaluation And Planning (WEAP) model. Mean monthly Precipitation; temperature and stream flow are used to run the model. The model was calibrated and validated using mean stream flow. The trends of climate over the study basin were also investigated using Mann Kendall trend tests. The results revealed that the value of R² ranged from 0.73 to 0.91 during calibration periods. On the other hand, the value of Nash—Sutcliffe efficiency (NSE) is ranged from 0.65 to 0.83. However, the values of coefficient of determination (R²) during validation periods ranged from 0.54 to 0.91 and the values of Nash—Sutcliffe efficiency (NSE), ranges from 0.50 to 0.84. Thus, The results at two important parts of the main river (upstream and downstream) gauges showed good agreement between the simulated and observed stream flow. The basin receives a significant amount of water during the rainy season (June to September). The mean annual rainfall of the basin ranges from 100 to 1700 mm with great spatiotemporal variation. The annual stream flow of the Awash River basin shows a sharp decreasing trend (Z = −0.120 during the study period. Thus, the findings of this study could provide insights for concerned bodies to implement effective water resources management techniques. The study also proposed that, to ensure the sustainability of water resources, better long-term management policies are required to be implemented in the basin and to meet future downstream water needs.

Keywords:

• Awash River Basin
• Water budget
• Stream flow
• Trends
• WEAP Model

PUBLIC INTEREST STATEMENTS

Climate change adversely affects the hydrological cycle at the basin level. The increasing demands of water resources in the world are the main problems for sustainable utilization of water resources. Over exploitation of water resources, population growth, pollution, and increasing of demand for economic development mainly cause water scarcity. Thus, this study aimed to calibrate and validate the stream flows in Awash River basin using a distributed hydrological Water Evaluation And Planning (WEAP) model. Mean monthly stream flow was used to run the model. The annual stream flow of the Awash River basin shows a sharp decreasing trend during the study period. Thus, the findings of this study could provide insights for concerned bodies to implement effective water resources management techniques. The study also proposed that, to ensure the sustainability of water resources, better long-term management policies are required to be implemented in the basin and to meet future downstream water needs.

1. Introduction

The sustainable utilization of water resources is a major global challenge due to the increasing water demand (Cai et al., 2017). Over-exploitation of water resources, population growth, pollution, and increasing demand for economic development mainly cause water scarcity (Divakar et al., 2011; Roozbahani et al., 2015). In river basins where water consumption is mostly for irrigation, sustainable management of water resources must consider two goals simultaneously: sustainable irrigated agriculture to ensure food security and protection of the environment. The Awash River Basin is the most utilized river basin for irrigation in Ethiopia, which covers a total area of 114,123 km². The river basin water is becoming scarce due to increasing demands and poor water resources management. The mean annual rainfall of the basin ranges from 100 to 1700 mm with great spatiotemporal variation.

Climate changes also adversely impacted the hydrological cycles of water resources in river basins across the globe (Pettinotti et al., 2018; Wang et al., 2015). The basin water becoming scarce due to increasing demand and poor water resources managements (Davijani & Banihabib, 2016; Hussen et al., 2018). Increasing water demand often results in unsustainable water consumption and leads to lack of sufficient amount of water for environmental protection. The effect is often reflected in the water and food insecurity, frequent hydrologic extremes, and deteriorating ecosystem health (IPCC, 2013). With the potential threats of climate change such as drought, pollution, flooding as a result of erratic rainfall and over consumptions, the basin’s water-related problems are expected to increase shortly (Gedefaw et al., 2019). The mean annual rainfall of the basin varies from 1600 mm to 160 mm in the northern part of the basin. The distribution of rainfall is bimodal. The Inter Tropical Convergence Zone (ITCZ) (Gedefaw et al., 2019) influences the distribution of rainfall in the basin. Thus, the basin climate highly influences regional precipitation and stream flow in the region. Hence, it is essential to quantify the water budget and the spatial-temporal variations of the basin for optimal allocation and effective management of the water resources. Various studies attempt to estimate the water budget of Awash River Basin so far. However, the estimations were limited due to lack of hydro-meteorological data in the basin.

Appropriate modeling to quantify the water budget of the basin accurately is very essential. The hydrological models focus on understanding how water flows within a basin in response to hydrological events, while water resource planning models focus primarily on water allocation management. These helps to estimate water supply and demand, allocate water resources in water scarcity areas, and for proper sustainable management of water in river basins. Often-used hydrological models such as Soil and Water Assessment Tool (SWAT) (Arnold et al., 1998), a watershed modeling code that simulates the principal hydrologic fluxes at a daily time step; MODFLOW (Niswonger et al., 2011), a groundwater modelling code that simulates groundwater head and associated flow rates in a heterogeneous aquifer system have been used in numerous studies for water supply estimation and water resources management (Ehtiat et al., 2018; Guevara Ochoa et al., 2019; Taie Semiromi & Koch, 2020; Yifru et al., 2020). Simulation models address certain limitations of allocation models by solving physically based flow equations to offer spatially distributed water resources outputs for several parameters (Condon & Maxwell, 2013). However, allocation models are frequently used in the applied problem of water resources management. These models optimize water allocation from various resources to meet a range of demands and to find which design and operating policy would best meet the identified objectives under a set of priorities and constraints (Loucks, 2008). The Water Evaluation and Planning Model (WEAP), is a bridge between the basin hydrology and water management by combining the physical hydrological process in the water management framework (Yates et al., 2009).

Thus, this study chooses the WEAP model as it incorporates different hydrological components in data scarce areas and to evaluate water resources using a scenario-based system. The model also simulated the domestic, irrigation, and ecological water consumption in time and space as compared to other allocation models. The WEAP model resolves problems faced by water resources managers and planners using a scenario-based system by providing a set of objects and procedure, which can be done on reservoirs, river basins, and watersheds.

Different distributed hydrological models have been applied to Awash River basin to model the hydrological characteristics such as MODSIM model for water allocation (Berhe et al., 2013), SWAT model (Adeba, 2015; Arsano & Tamrat, 2005). However, this study chooses WEAP model for the successful calibration and validation of stream flow in Awash river basin. Furthermore, the trends of precipitation and stream flow were also detected. These enable to give insights on the spatial and temporal variations of water resources of the basin to concerned stakeholders.

2. Materials and methods

2.1. Description of the study area

Awash River basin located between latitudes of 7°53′ N and 12° N and longitudes of 37°57′ E and 43°25′ E (Berhe et al., 2013). The basin constitutes the central and northern parts of the Rift Valley and is bounded to the west, southeast, and south by the Blue Nile, the Rift Valley lakes, and the Wabeshebele basins, respectively (Gedefaw et al., 2019). It covers a total area of 110,000 km², with a length of 1200 km (Berhe et al., 2013). The basin is a home of about 15 million inhabitants (Gedefaw et al., 2019). The basin is the most highly utilized basin in Ethiopia for irrigation. The basin is divided into upper, middle, and lower valleys. The mean annual rainfall of the basin varies from 1600 mm northeast of Addis Ababa to 160 mm in the northern part of the basin (Figure 1). The distribution of rainfall is bimodal in the middle and lower parts of the basin and unimodal in the upper part (Berhe et al., 2013). The Inter Tropical Convergence Zone (ITCZ) influences the distribution of rainfall in the basin.

Figure 1. Location map of Awash River basin.

The mean surface water resource of the Awash river basin is approximately 4.9 × 10⁸ m³ (Mersha, 2018). Irrigation used 44% from the surface water resources. More than 70% large scale irrigated agriculture in Ethiopia is found in this basin. The irrigation potential of the basin is estimated to be 206,000 ha as reported in ministry of water and energy office. The total mean annual evaporation is 1810 mm and 2348 mm in the upper and lower parts of the basin (Mersha, 2018). The estimated mean annual runoff within the basin is about 4.6 km³ (Taddese et al., 2004). Many rivers are functional only in rainy seasons (July to September) especially in lowland parts of the basin. However, Mojo, Akaki, Kessem, Kebena, and Mile rivers are functional throughout the year. Since the population is highly dependent on rainfed agriculture, this has made the population and the economy vulnerable to impacts of climate change and droughts (A. B. J. A. B. Khalil, 2015).

2.2. Methods

Mann- Kendall test for monotonic climate trends

The non-parametric Mann-Kendall (MK) test is commonly employed to detect monotonic trends in series of climate data or hydrological data which was first developed by Kendall (1975) and Mann (1945) (Pingale et al., 2014). It is used to detect the trends in time series data of precipitation and temperature. A positive or negative value of MK (Z) value indicates an upward or downward trend, respectively. At 95% significance level, the null hypothesis of no trend is rejected if |Z| > 1.96. Mann Kendall statistic (S) for a time series X₁, X₂, X₃ … , and X_n is calculated as (Gedefaw et al., 2018):

(1) $\mathit{S}=\sum _{\mathit{i}=1}^{\mathit{n}-1}.\sum _{\mathit{j}=\mathit{i}+1}^n\mathit{sgn}\left(X_j-X_i\right)$(1)

(2) $\mathit{sgn}\left(X_j-X_i\right)=+1\mathit{if}\left(X_j-X_i\right)00\mathit{if}\left(X_j-X_i\right)=0-1\mathit{if}\left(X_j-X_i\right)0$(2)

The probability associated with S and the sample size, n, is then computed to statistically quantify the significance of the trend using normalized test statistic Z as follows:

(3) $\text{Z}=\frac{\text{S}-1}{\mathit{\delta }}\text{ifS0 \hspace{0.17em}0ifS=0 \hspace{0.17em}}\frac{\text{S}+1}{\mathit{\delta }}\text{ifS0 }$(3)

The variance is computed as (Gedefaw et al., 2018):

(4) $\mathit{Var}\left(\mathit{S}\right)=\frac{1}{18}\left[\mathit{n}\left(\mathit{n}-1\right)\left(2\mathit{n}+5\right)-{\sum }_{p=1}^q\mathit{tp}\left(\mathit{tp}-1\right)\left(2\mathit{tp}+5\right)\right]$(4)

where n is the number of data points, q is the number of tied groups and tp is the number of data values in the p^th group.

2.2.2. WEAP Model simulation and its description

WEAP is a modeling platform that can provide integrated assessment of climate, hydrology, land use, irrigation facilities, water allocation, and water management priorities of the watershed (A. B. J. A. B. Khalil, 2015). The land use, irrigation facilities, water allocation, and water management are the priorities of the watershed. The WEAP model uses a standard linear programming model to solve water allocation problems time step and its target function is to maximize the percentage of supplying demand centers’ needs, at any time step and its target function is to maximize the percentage of supplying demand centers’ with regard to supply and demand priority, mass balance, and other constraints. All constraints are needs, with regard to supply and demand priority, mass balance, and other constraints.

This study developed WEAP model for calibration and validation of the stream flow in Akaki, Hombole, melka Kuntre, and Modjo gauging stations of Awash River Basin (Figure 2). The model was run on a monthly basis. The WEAP model provides the integrated assessment of climate, hydrology, water resources allocation, and watershed managements (Xiao-Jun et al., 2011). It also addresses various issues such as water resources, water demands analysis in different sectors, provides priorities in water allocation, reservoir operation, and managements. It solves the water allocation challenges at user-defined periods either monthly or yearly based on linear programming structures (Adgolign et al., 2016).

Figure 2. WEAP model development .

2.3. Model accuracy

The accuracy of the model simulation result is evaluated by root mean squared error (RMSE), Nash—Sutcliffe coefficient (NSE), relative error (R²), percentage bias (PBAIS), and observation standard deviation ratio (RSR) (Chea & Oeurng, 2017; Gumindoga et al., 2017; Li, 2010). The R² value is an indicator of the strength of the linear relationship between the observed and simulated values, while the (NSE) simulation coefficient indicates how well the plot of observed versus simulated values fits the 1:1 line (Gedefaw et al., 2019).

The RMSE, NSE, R², PBIAS, and RSR can be estimated as follows:

(5) $\mathit{RMSE}=1/\mathit{n}\sum \mathit{i}=1\mathit{n}{\text{ O}}_i-S_{i}2$(5)

(6) $\mathit{NSE}=1-\frac{{\sum }_{i=1}^n{\left({\mathrm{O}}_{\mathrm{i}}-S_i\right)}^2}{{\sum }_{i=1}^n\left[{\left({\mathrm{O}}_i-{\bar{\mathrm{O}}}_{\mathit{i}}\right)}^2\right]}$(6)

(7) $R^2=\frac{{\sum }_{i=1}^n{S}_i-{\sum }_{i=1}^n{\mathrm{O}}_i}{{\sum }_{i=1}^n{\mathrm{O}}_i}.100\mathrm{\%}$(7)

Where, n is the number of observations, ${\mathrm{O}}_{\mathrm{i}}$ is the observed value, is mean observed value, and $S_i$ is simulated value. The best value for NSE is one and the negative value indicates the model is not credible.

(8) $\mathrm{PBIAS}=\left(\frac{{\sum }_{i=1}^n\left({Y_i}^{\mathit{obs}}-{Y_i}^{\mathit{sim}}\right)}{{\sum }_{i=1}^n{Y_i}^{\mathit{obs}}}\mathit{x}100\mathrm{\%}\right)$(8)

The optimal value of PBIAS is 0.0, with low magnitude values indicating accurate model simulation. A positive value of PBIAS indicates model underestimation bias, and negative values overestimation bias.

(9) $\mathrm{RSR}=\frac{\sqrt{{\sum }_{i=1}^n{\left({Y_i}^{\mathit{obs}}-{Y_i}^{\mathit{sim}}\right)}^2}}{\sqrt{{\sum }_{i=1}^n{\left({Y_i}^{\mathit{obs}}-Y^{\mathit{mean}}\right)}^2}}$(9)

2.4. Data availability

Assessment of a basin water balance is vital to understand the processes of hydrologic cycle. Digital Elevation Model (DEM) data with a spatial resolution of 1 km was used. The digital elevation model (DEM) hydrological modeling requires data on topography, soil characteristics, and land is the most important input to provide topographical information in the WEAP system. The DEM is a simple input layer in the model configuration and was used to extract the slope and drainage network of the study area. The dataset was developed at the U.S. Geological Survey (USGS) and accessed through the USGS EarthExplorer platform (https://earthexplorer. usgs.gov/). The DEM data were used to derive basin characteristics and sub-basin boundaries, river networks, and slope. The hydrological and meteorological data from 1980 to 2016 were collected from the Ministry of Water, irrigation and Energy and National Meteorological Service Agency (NMSA) of the of Ethiopia, respectively. All climate datasets (daily precipitation, temperature, and stream flow data) were collected for this study.

There is no specific guideline on the appropriate number of decimal places to use data when presenting climate change. This would have an impact on the interpretation of results and conclusion. The temperature and precipitation readings are usually taken one decimal place and therefore, it is appropriate to consider only one significant decimal to present temperature and precipitation data. However, scientists used different decimal digits to present yearly temperature trends. Intergovernmental Panel for Climate Change (IPCC) uses one decimal place to show the change in temperature over a century and two decimal places to show the change over a decade (IPCC, 2013). Other recent studies also used two decimal digits. Therefore, one decimal digit is used in this study to show changes in precipitation. The following preconditions are applied to select the meteorological stations. (a) All months should contain complete data unless 5 or less days (b) complete yearly data (c) stations have complete data except upto five years missing data. Therefore, based on these parameters, the study stations were selected.

3. Results

3.1. Model performance

The results of this study showed that the value of coefficient of determination (R²) was ranged from 0.73 to 0.91 during calibration periods. On the other hand, the value of Nash—Sutcliffe efficiency (NSE) ranges from 0.65 to 0.83. However, the values of coefficient of determination (R²) during validation periods ranged from 0.54 to 0.91 and the values of Nash—Sutcliffe efficiency (NSE), ranges from 0.50 to 0.84. Hence, in calibration and validation periods, the values correspond a perfect match between the observed and the modeled stream flow values. Thus, the performance of the model is acceptable between the trends of observed and simulated stream flow in both calibration and validation of the study basin. This also helps to accurately project the prediction of future discharges based on future scenario set.

3.2. Precipitation distribution

The long-term (1980–2016) mean monthly precipitation distribution of Awash river basin is given in (Figure 3). The mean annual rainfall of the basin ranges from 100 to 1700 mm with great spatiotemporal variation. The temporal variation shares 71% and 29% of rainy season (June to October) and dry season (November–May) respectively. The mean annual rainfall of the basin varies from 1600 mm northeast of Addis Ababa to 160 mm in the northern part of the basin (Figure 1).

Figure 3. Mean monthly precipitation of upper, middle and lower sub basins of Awash Basin.

The basin receives maximum precipitation during the long rainy season from June to September and little rain from March to May. However, in the remaining months, the basin stays dry. Thus, the finding shows that there is high variability of mean monthly rainfall across the basin (Figure 3).

As far as the trends of precipitation in the basin concerned, increasing trend was detected in Gewane (Z = 0.80), Fiche (Z = 0.82) station and decreasing trend in Sekoru (Z = 0.45) and Bui (Z = 0.69) stations (Figure 4).

Figure 4. Statistical values of precipitation trends.

3.3. Analysis of streamflow

The annual stream flow of the Awash river basin shows a sharp decreasing trend from 1981 to 1986 and from 2000 to 2005 with a Z-value of−0.12, indicating a decreasing trend during the period of 1980 to 2016 (Figure 5 e). The curve line shows a statistically significant increasing trend in Berga gauging stations from 2010 to 2012 (Z = 4.00) (Figure 5 b) and a statistically significant decreasing trend in Lake Bishoftu from 1985 to 1995 (Z = 1.47) (Figure 5 d). However, the trends in Awash Hombole and Holeta gauging stations show decreasing and increasing trends, although not statistically significant, with Z-values of−0.18 and 1.66, respectively. The trends of the Awash river discharge generally exhibited a downward trend from 1980 to 2016.

Figure 5. Trends of annual stream flow across stations (UB = −UF).

3.4. Calibration and validation

The model was calibrated and validated with monthly stream flow for selected stations from (1980–2000) and (2001–2014) respectively. The results at two important parts of the main river (upstream and downstream) gauges are depicted in (Figure 6) showing good agreement between the simulated and observed stream flow. The statistical value of this study indicated that the performance of the model is acceptable between the trends of observed and simulated stream flow in both calibration and validation of the study basin.

Figure 6. Calibration and validation results of the model.

4. Discussion

The Water Evaluation and Planning (WEAP) software, developed by the Stockholm Environment Institute (SEI), is a practical semi-theoretical, semi-distributed, and deterministic water resource planning tool that incorporates both water supply and water demand issues, in addition to water quality and the conservation of ecosystems, as required by an integrated approach to basin management (Abrishamchi et al., 2007; Hamlat et al., 2012). The WEAP model was used to simulate both the hydrological processes and the anthropogenic activities of water resources to analyze the availability of water in the basin. The WEAP model has been implemented at national and international levels because it offers a flexible and comprehensive policy analysis framework, in addition to a method for managing water supply and demand. In this study, the observed stream flow data were used for calibration and validation for the period of (1980–2000) and (2001–2014) respectively. The results showed that, in calibration and validation periods, the values correspond a perfect match between the observed and the modeled discharge values. This shows, the performance of the model is acceptable between the trends of observed and simulated one. This also helps to accurately project the prediction of future discharges based on future scenario set. Decreasing of stream flow at the main outlet of Awash River basin could cause high pressure on available water resources. A similar study conducted on Didessa sub-basin of West Ethiopia found that, the existing water demand was 74 Mm³ (Adgolign et al., 2016), which will help to solve water management issues in the basin. Another study conducted in the Mae Kong Basin in Thailand estimated an average unmet demands for agriculture of 62 and 17 Mm³ under differing climate change scenarios (A. Khalil et al., 2018). In the current study, the WEAP model was also used to analyze external drivers (population growth, agricultural expansion, deforestation, industrialization, and climate change), which place significant stress on the existing water supply system. These demands intensify water scarcity in the catchment, particularly downstream, and encourage water resource managers to implement water management policies in the Awash River Basin. This study was supported with the findings of (Abbas et al., 2022; Ajami et al., 2016; Gedefaw et al., 2019; Tong & Guo, 2013; Touseef et al., 2021).

As far as the trends of climate data concerned, it was confirmed that precipitation is mainly caused by seasonal variations across the basin. The trends of precipitation in Gewane and Fiche stations were increasing while the trends were decreasing in Sekoru and Bui stations during the study period. Similar findings were reported by (Berhe et al., 2013; Gedefaw et al., 2018, 2019; Hailu et al., 2018). If the trend continues in the future, it could impact the sustainability of water resources recharge. Increasing of temperature also increases transpiration, which increases the chance of rainfall and may interfere groundwater recharge triggered by summer season reduction. Further study will be required to analyze the future water demand beyond this study period and for downstream dwellers. Therefore, an efficient and effective integrated water resources management approach must be implemented based on the results of this study. This study shows that the Awash River Basin’s high water management pressure stems predominantly from demand in irrigation. This rapid rise in demand is not only due to anthropogenic activities but also to changes in precipitation and seasonal variations. Thus, all the basin sections are impacted by the risks and cause conflicts between upstream and downstream dwellers.

5. Conclusions

This study developed a distributed hydrological WEAP model for the simulation of water supply and demand in Awash River basin and to evaluate the impacts of climate changes. The trends of climate change were also investigated using MK test. The model results were verified through comparisons of simulated values with the observed values of stream flows. The results showed that the Awash River basin is faced with water scarcity to meet the needs of the competing users. Studies so far also showed a water shortage of 1.27 BCM/year in 2011 and 2.82 BCM/year in 2012. Addressing the challenges of water scarcity will require both selective development and exploitation of new water supplies and comprehensive policy reform that encourages more efficient use of existing water supplies. Exploitation of new water supply like harvesting rainwater and storing the excess flood during rainy period can alleviate water scarcity in the basin to a great extent. So the immediate future task of the decision makers should be to protect the ecology of the basin from further deterioration of the resources and exploit different sources of water to mitigate the current water scarcity. Therefore, we propose that better long-term management policies should be implemented in the watershed to ensure the sustainability of water resources in the river basin. Based on the findings of the analysis of the climate changes trends and the model results, this study has high significance for sustainable water management of Awash River Basin. Thus, the policymakers and concerned stakeholders should take immediate measures to solve the aforementioned problems.

Authorship statement

I submitted a manuscript entitled: “Simulation of Stream flows and Climate Trend Detections Using WEAP Model in Awash River Basin” which will be published in Cogent Engineering journal. The authors certify that they have participated sufficiently in the work to take public responsibility for the content, including participation in the concept, design, analysis, writing, or revision of the manuscript. The final manuscript before submission was checked and approved by the authors.

Acknowledgments

The authors thank the NMSA for providing data inputs for this research work. The authors also would thank the editor and the two reviewers for their constructive comments and suggestions to improve the quality this article.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The Authors confirm that the data supporting the findings of this study are available within this article. The raw data that supports the findings of this study are available upon the request from the corresponding author.