Spatiotemporal dynamics of water use efficiency and its drivers in the Gilgel Abay Watershed, Ethiopia

Abstract In Ethiopia's rainfed highlands, water use efficiency (WUE), defined as the ratio of net primary productivity (NPP) to evapotranspiration (ET), is essential for balancing carbon assimilation and water loss. However, no previous studies have quantified long-term WUE trends or examined their biophysical drivers at the sub-basin scale in the Gilgel Abay watershed. Moreover, earlier research in Ethiopia has largely focused on vegetation indices, without explicitly linking WUE to key biophysical variables at the watershed scale, despite 20–30% rainfall variability posing serious threats to hydropower generation and food security for nearly three million people in the Upper Blue Nile Basin. Although global WUE studies are extensive, watershed-specific WUE assessments in Ethiopia, particularly in the Gilgel Abay watershed, which contributes approximately 60% of Lake Tana’s inflow, remain scarce. This study explored the spatiotemporal variability of WUE from 2001 to 2024 and its biophysical drivers using MODIS and Sentinel-2 MSI imagery. Mann–Kendall trend tests and Sen’s slope estimators were used to detect trends, while pixel-wise Pearson correlations were applied to assess relationships between WUE and the normalized difference vegetation index (NDVI), land surface temperature (LST), soil moisture index (SMI), and vegetation condition index (VCI). Results showed significant increases in ET (5.45 kg/m 2 /year) and NPP (10.11 gC/m 2 /year), but a non-significant upward trend in WUE (0.0005 gC/mm/year). Spatially, higher WUE, ET, and NPP were concentrated in the southern and central regions. WUE exhibited positive correlations with NPP, NDVI, SMI, and VCI, but negative correlations with ET and LST across most (≈75%) of the watershed. Land use/land cover (LULC) classification, with an overall accuracy of 94.4% and a Kappa coefficient of 0.93, indicated that cultivated land was dominant (72.39%), followed by shrubland (10.39%) and forest (3.48%). Forest areas exhibited the highest annual ET (596.80 kg/m 2 ), NPP (1055.90 gC/m 2 ), and WUE (1.78 gC/mm), whereas water bodies had the lowest values (ET: 456.40 kg/m 2 ; NPP: 625.80 gC/m 2 ; WUE: 1.29 gC/mm). These findings show that healthy vegetation and sufficient soil moisture together improve WUE, help ecosystems cope with high temperatures, and support restoration efforts and sustainable management of water and vegetation, ultimately enhancing the resilience of the Nile Basin.