Ungauged catchment contributions to Lake Tana's water balance

Lake Tana's flow system is governed by four main components: the inflow from surrounding river catchments into the lake, the outflow at Bahir Dar through the Blue Nile, the direct rainfall on the lake and the direct evaporation from the lake. While recent studies applied simple pragmatic approaches to estimate runoff from ungauged catchments, here emphasis is placed on more advanced approaches based on regionalization and spatial proximity principles. In the regionalization approach, model parameters of the conceptual HBV (Hydrologiska Byrans Vattenbalansavdelning) rainfall-runoff modelling of gauged catchments are transferred to ungauged catchments to allow runoff simulation. Parameter transfer was attempted through regression, proximity procedures and catchment size. This yielded 42, 47 and 46%, respectively, of the total river inflow for the three procedures. Lake areal rainfall is estimated by interpolation of the rain gauges around the lake, open water evaporation is estimated by the Penman-combination equation while observed inflows and outflow data are directly used in the lake water balance. The water balance closure term was established by comparing the measured lake levels with the calculated levels. Results show that runoff from ungauged catchments is around 880 mm per year for the simulation period 1995-2001 with a water balance closure error of 5%. In addition, use is made of river and lake water chemistry to arrive at an estimate of the unknown inflow and outflow components through the mixing cell approach. The results obtained with this method also provide independent information with regard to the errors in the individual water balance components. Copyright (C) 2009 John Wiley & Sons, Ltd.