In Tigray-northern Ethiopia, lack of sediment yield data and appropriate methodologies for predicting sediment yield have contributed to poor planning resulting in rapid sedimentation in reservoirs and storage capacity loss. The objectives of this study were: (1) to assess the spatial variability of absolute sediment yield (SY) and area-specific sediment yield (SSY) and to identify their controlling factors for 11 representative catchments and (2) to develop models to predict sediment yield. We quantified sediment yield from reservoir sediment surveys and studied the role of bio-physical characteristics of the catchments and their interactions in controlling SY and SSY variability.
The average SSY for 11 reservoirs was 9.89 t ha− 1 y− 1 with a standard deviation (S.D.) of ± 4.46 t ha− 1 y− 1, which can be considered as a large spatial variation in SSY among the catchments. Total drainage length (TDL) and the proportion of the catchment area that is treated with soil and water conservation (SWC) practices are the strongest variables controlling the variability of SY and the SSY, respectively. Interactions between controlling factors were found: i.e. SWC practices and average catchment slope (Av_slope) (r = 0.80), SWC and proportion of cultivated land (CUL) (r = − 0.64) and CUL and Av_slope (r = − 0.81). SWC practices were found to be less implemented in catchments with a relatively high CUL-value and having moderately steep topography.
Best results were obtained with the SY regression model with a high model efficiency (ME = 0.88). The SSY model had a reasonable ME of 0.66. Therefore, the SY model allows a better prediction of SY in the planning phase of new reservoirs in Northern Ethiopia. However, such models need new calibration if they are to be used beyond the region where they were developed and they do not allow spatially-distributed input and output.
