Sessional Rainfall–Runoff Variability Analysis, Lake Tana Sub-Basin, Upper Blue Nile Basin, Ethiopia

Abstract

Lake Tana sub-basin of Abbay (Blue Nile) River Basin is located in the high land areas with unimodal rainy season with spatial and temporal variation of rainfall and runoff. Depending on available resources, there are many developmental plans and projects which seek the wise planning and management of water resources considering both low flows and floods. The seasonal streamflow variability analysis of the basin was performed with recorded meteorological and hydrological data. The four seasons of the year are considered for seasonality analysis. The rainfall variability is analysed using seasonality and variability measures of coefficient of variation, seasonal relative rainy days and seasonal rainfall intensity. The rainfall variability is more related with latitude and longitude. Spatial and temporal seasonal rainfall variation is analysed from daily rainfall data. Seasonal runoff and streamflow variations are also analysed using HEC-HMS hydrological model to generate runoffs at required and selected points to detect spatial variation. Runoff variation for catchments with gauged stations was analysed from recorded time series streamflow data. Runoff coefficient is taken as a variability index for both generated and recorded streamflows. The runoff coefficient ranges from 0 to 1. The range is high in the dry seasons and less in the wet seasons. The average runoff coefficient value of the basin is 0.28 ranging from 0.18 to 0.36. The average seasonal runoff coefficient value from generated runoffs is 0.45, 0.3 for dry and 0.6 for wet seasons. From the results, it is shown that runoff coefficient is more dependent on antecedent soil wetness condition, land use and land covers. Catchments were categorised spatially and temporally as vulnerable, moderately vulnerable and less vulnerable to runoff based on the analysis. From hydrological data variability tests, it is clearly observed that seasonal time series data are not homogeneous, stationary and independent. Minimum flows are more stationary and homogeneous than mean and maximum flows.