Assessing the effect of water harvesting techniques on event-based hydrological responses and sediment yield at a catchment scale in northern Ethiopia using the Limburg Soil Erosion Model (LISEM)

Runoff and sediment yield in semi-arid catchments are highly influenced by infrequent but very heavy rains. These events occur over short temporal scales, so runoff and sediment transport can only be understood using an event-based analysis. We applied a hydrological and soil-erosion model, LISEM, to the Gule catchment (~ 12 km2) in northern Ethiopia. The objectives of the study were: (a) to evaluate the performance of LISEM in describing event-based hydrological processes and sediment yield in a catchment under the influence of different water harvesting techniques (WHTs), and (b) to study the effect of the WHTs on catchment-scale event-based runoff and sediment yield. The model performed satisfactorily (NSE > 0.5) for most of the events when discharge was calibrated at the main outlet (Gule) and at a sub-outlet (Misbar). Runoff coefficients for the Gule catchment and Misbar sub-catchment were expectedly low due to the implementation of WHTs, which can store runoff from the rains and increase infiltration into the soil. Simulated and measured sediment yields were of similar orders of magnitude. LISEM generally overestimated sediment yield compared to the measurements. The poor performance of LISEM in predicting sediment yield could be attributed to the uncertainty of several factors controlling soil erosion and the inadequacy of LISEM in describing soil erosion on steep slopes. Catchment-scale model simulations indicated that runoff and sediment yield could be effectively reduced by implementing WHTs. The model estimated 41 and 61% reductions in runoff and sediment yield at the Gule outlet, respectively. Similarly, runoff and sediment yield at the Misbar sub-outlet were reduced by 45 and 48%, respectively. LISEM can thus be used to simulate the effects of different existing or new WHTs on catchment hydrology and sediment yield. The results of scenario predictions could be useful for land-use planners who intend to implement different measures of catchment management.