Bias-Corrected CHIRP Satellite Rainfall for Water Level Simulation, Lake Ziway, Ethiopia

Abstract

Applicability of satellite rainfall products must be explored since rain gauge networks have limitations to provide adequate spatial coverage. In this study, Climate Hazards InfraRed Precipitation (CHIRP) satellite-only product was evaluated for rainfall-runoff modeling whereas the simulated runoff served as input to simulate the water levels of Lake Ziway from 1986 to 2014. CHIRP dataset was bias-corrected using power transformation and used as input to Hydrologiska Byråns Vattenbalansavdelning (HBV) model to simulate streamflow of Meki and Katar catchments. Results showed that gauged catchments of Meki and Katar contributed 524 and 855 mm to the annual lake inflow, respectively. The estimated runoff from ungauged catchments is 182 mm that amounts to approximately 8.5% of the total lake inflow over the period 1986–2000. The results of lake level simulation show good agreement from 1986 to 2000, but deteriorating agreement after 2000, which is mainly attributed to errors in water balance terms and human-induced impacts. For the period 1986–2000, the water balance closure error for the lake was 67.5 mm per year, which accounts for 2.9% of the total lake inflow from rainfall and river inflow. This study shows bias correction increases the applicability of CHIRP satellite product for lake water balance studies.