Characterization and disaggregation of daily rainfall in the Upper Blue Nile Basin in Ethiopia

SummaryIn Ethiopia, available rainfall records are mainly limited to daily time steps. Though rainfall data at shorter time steps are important for various purposes like modeling of erosion processes and flood hydrographs, they are hardly available in Ethiopia. The objectives of this study were (i) to study the temporal characteristics of daily rains at two stations in the region of the Upper Blue Nile Basin (UBNB) and (ii) to calibrate and evaluate a daily rainfall disaggregation model. The analysis was based on rainfall data of Bahir Dar and Gonder Meteorological Stations. The disaggregation model used was the Modified Bartlett–Lewis Rectangular Pulse Model (MBLRPM). The mean daily rainfall intensity varied from about 4 mm in the dry season to 17 mm in the wet season with corresponding variation in raindays of 0.4–26 days. The observed maximum daily rainfall varied from 13 mm in the dry month to 200 mm in the wet month. The average wet/dry spell length varied from 1/21 days in the dry season to 6/1 days in the rainy season. Most of the rainfall occurs in the afternoon and evening periods of the day. Daily rainfall disaggregation using the MBLRPM alone resulted in poor match between the disaggregated and observed hourly rainfalls. Stochastic redistribution of the outputs of the model using Beta probability distribution function improved the agreement between observed and calculated hourly rain intensities. In areas where convective rainfall is dominant, the outputs of MBLRPM should be redistributed using relevant probability distributions to simulate the diurnal rainfall pattern.

Research highlights
► Characterize temporally daily rainfall in the Upper Blue Nile Basin in Ethiopia.
► Evaluate performance of the Bartlett–Lewis rainfall disaggregation model (MBLRPM).
► Direct use of the MBLRPM cannot reproduce the diurnal pattern of convective rainfall.
► Redistribute the MBLRPM outputs by Beta probability function for better results.
► Adjust stochastically the MBLRPM outputs if convective rainfall dominates.