Water Resources in Lake Tana Basin: Statistical Analysis of Hydrological and Meteorological Time Series

Abstract:

Nowadays, time series environmental flow analysis is becoming one of the most important tasks in ecohydrology in order to design process based system solutions. Thus, the purpose of this research paper was to understand temporal and spatial variability of stream flows, rainfall, and inflows and outflows to and from the Lake Tana basin. Autocorrelation and cross correlation tests were applied for the long years’ daily stream flows and rainfall using R languages. These methods were used to see how the stream flow or rainfall data were serially correlated and rainfall, stream flow and lake level time series data were cross correlated with each other. Autocorrelation tests of daily rainfall were carried out for many rainfall stations and the outputs indicate that there are no spikes showing significant seasonal signals. The annual rainfall map was produced for the whole catchment based on long years’ records at different stations inside the catchment using inverse distance weighted interpolation (IDW) method in the GIS environment. Based on this map there is high spatial variability of annual rainfall in the catchment. The average maximum, minimum and mean annual rainfall values are 1506.4, 798.7, and 1238.1 mm respectively. According to the cross correlation tests done for stream flow & rainfall, better correlations were observed after 15 to 30 days lag time due to late response of the catchment for runoff generation. The study also prevailed that the Lake Tana water level and Blue Nile discharge at Bahir Dar station have positive cross correlation with maximum value at time lag of zero. There is a dramatic drop in the lake level and stream flow volume at the same location since 2000 due to human induced local climate forcing. In general, this research indicates that there is high temporal and spatial variability in rainfall, Lake water level and stream flows.