Hydrological forecasting is one of the key research areas in hydrology. Innovative forecasting tools will reform water resources management systems, flood early warning mechanisms, and agricultural and hydropower management schemes. Hence, in this study, we compared Stacked Long Short-Term Memory (S-LSTM), Bidirectional Long Short-Term Memory (Bi-LSTM), and Gated Recurrent Unit (GRU) with the classical Multilayer Perceptron (MLP) network for one-step daily streamflow forecasting. The analysis used daily time series data collected from Borkena (in Awash river basin) and Gummera (in Abay river basin) streamflow stations. All data sets passed through rigorous quality control processes, and null values were filled using linear interpolation. A partial autocorrelation was also applied to select the appropriate time lag for input series generation. Then, the data is split into training and testing datasets using a ratio of 80 : 20, respectively. Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and coefficient of determination (R2) were used to evaluate the performance of the proposed models. Finally, the findings are summarized in model variability, lag time variability, and time series characteristic themes. As a result, time series characteristics (climatic variability) had a more significant impact on streamflow forecasting performance than input lagged time steps and deep learning model architecture variations. Thus, Borkena’s river catchment forecasting result is more accurate than Gummera’s catchment forecasting result, with RMSE, MAE, MAPE, and R2 values ranging between (0.81 to 1.53, 0.29 to 0.96, 0.16 to 1.72, 0.96 to 0.99) and (17.43 to 17.99, 7.76 to 10.54, 0.16 to 1.03, 0.89 to 0.90) for both catchments, respectively. Although the performance is dependent on lag time variations, MLP and GRU outperform S-LSTM and Bi-LSTM on a nearly equal basis.

Reliable and accurate streamflow simulation has a vital role in water resource development, mainly in agriculture, environment, domestic water supply, hydropower generation, flood control, and early warning systems. In this context, these days, deep learning algorithms have got enormous attention due to their high-performance simulation capacity. In this study, we compared multilayer perceptron (MLP), long short-term memory (LSTM), and gated recurrent unit (GRU) with the proposed new hybrid models, including CNN-LSTM and CNN-GRU. Hence, we can simulate one-step daily streamflow in different agroclimatic conditions, rolling time windows, and a range of variable input combinations. The analysis used daily multivariate and multisite time series data collected from Awash River Basin (Borkena watershed: Ethiopia) and Tiber River Basin (Upper Tiber River Basin: Italy) stations. The datasets were subjected to rigorous quality control processes. Consequently, it rolled to a different time lag to remove noise in the time series and further split into training and testing datasets using a ratio of 80 : 20, respectively. Finally, the results showed that integrating the GRU layer with the convolutional layer and using monthly rolled average daily input time series could substantially improve the simulation of streamflow time series.

Evaluating climate change impacts and the suitability of potential land resources is crucial for sustainable irrigated agricultural systems. This study applied a multi-criteria analysis supported by the Geographic Information System (GIS) application to produce irrigation suitability maps for the Anger River basin’s (Ethiopia) irrigation command area to optimize its irrigation system. Six irrigation suitability factors, such as distance to water sources (rivers), slope, land use/land cover, soil texture, drainage, and depth, including climate change impacts, were used. These factors were spatially analyzed using a comparison matrix and overlying the factors with 30 m resolutions to estimate the potential irrigable area. About 40% of the study area was classified as moderately to highly suitable for surface water irrigation systems. Moreover, we found that a large proportion of the study area is suitable for surface irrigation system, suggesting the relevance of implementing an enhanced irrigation system for improving the surface irrigation water productivity of the basin. However, future climate change is predicted to negatively affect the irrigation suitable area due to water scarcity. Therefore, this study provides useful information on the irrigation suitability and potential of the study area that could be used to facilitate the water resource development and food security plans.

No abstract

Assessing available water resources and their potential for irrigation water use is vital for sustainable agricultural development and planning. This is particularly of interest in developing countries like Ethiopia, where a small portion of largely accessible land for surface irrigation applications has been utilized, despite the majority of the population relying on agricultural productivity. This study utilized the Dhidhessa River Basin (Ethiopia) as a case study and analyzed the main challenges to balance the sustainable water resources utilization and enhance agricultural productivity of the basin. The study mainly focused on estimating the available water resources and their potential for surface irrigation water use in the basin. This was achieved by utilizing Geographic Information System (GIS)-based tools, a hydrological Soil and Water Assessment Tool (SWAT) model, and a Crop Water and Irrigation Requirements Program of FAO (CROPWAT) model. While the SWAT estimated the water availability in the basin, GIS-tools such as Model Builder were used to map the irrigation potential of the basin. For irrigation water potential assessment, we selected six crops (cabbage, maize, tomato, pepper, groundnut and sugarcane) and estimated their irrigation water requirements using the CROPWAT model. We developed the SWAT model for the period from 1986 to 2012 using the available hydro-meteorological and geo-spatial data. Due to many parameters used in the model, we first performed a parameter sensitivity analysis and identified the most essential/sensitivity parameters via Sequential Uncertainty Fitting-II (SUFI-2). The identified sensitive parameters were subsequently used for model calibration (1989–2000) and validation (2001–2012) procedures achieved via SUFI-2. SWAT was able to reproduce the observed monthly streamflow values with a coefficient of determination (R2) and Nash-Sutcliffe Coefficient (NSE) of 0.85 and 0.87 for the calibration period and 0.91 and 0.89 for the validation period, respectively. The findings generally indicated a “good” performance of the model in simulating the hydrology. The annual available water of the basin is 9.26 billion cubic meters (BCM) whereas the 70% and 80% dependable flow is 7.56 and 6.97 BCM, respectively. Based on the Model Builder of ArcGIS, the SWAT estimated available water can potentially irrigate an area of 259,028 ha for slope less than 8%, 643,162 ha for slopes less than 15% and 1,023,581 ha for slopes less than 30%. Moreover, the irrigation water requirements were calculated by the CROPWAT model for the six selected crops indicated that although the need for irrigation water varies depending on the season, the potential irrigation area of the Dhidhessa River Basin is greater than its irrigated land. Therefore, it is concluded that the basin’s surface irrigation systems need to be expanded to enhance the agricultural productivity and improve the livelihood of the basin’s communities and similar basins elsewhere.