The distributed recharge model developed by Scheidegger et al. (2015) is used to estimate the recharge values under extreme weather events. Synthetic extreme dry and wet rainfall and evaporation time series are produced by repeating a dry or a wet year within the historical rainfall and evaporation time series. The Standardised Precipitation Index (SPI) method is used to identify the most wet and most dry years. Heat maps showing the severity of drought or wet periods across the country are used. These maps show inconsistencies of the calculated indices across the country, with oddities observed in the north part of the country. Six scenarios are considered in which, the wet year is repeated once, twice, and three times and then the dry year is repeated in the same fashion. The estimated long term average recharge values are compared to the historical ones. On average, the groundwater system is expected to be in shortage of 9% of historical long term average recharge values calculated for the country when four successive years of drought years are considered. The groundwater system contains approximately 11 % more resources than that is calculated historically when four successive wet years are considered.AquiMod lumped groundwater model is used to estimate representative transmissivity and storage coefficient values for three catchments. Groundwater levels recorded at the boreholes in Chitipa, Endongolweni School, and Namwera are used for this purpose. The numerical model produces acceptable groundwater time series for the first two boreholes but fails to produce the groundwater level fluctuations at the Namwera borehole. It is believed that inconsistencies between the calculated recharge and the groundwater level time series are the reason for this failure. The optimised hydrogeological parameters lead to transmissivity values varied between 20 and 1500 m2/day. Storage coefficient (specific yield) on the other hand varied between 0.02 and 0.3.The AquiMod models were run using the synthetic meteorological extreme scenarios and the groundwater level fluctuations are compared to those produced using the historical recharge values. The uncertainties associated with the determination of extreme weather periods in the northern Malawi are propagated in this modelling exercise. Whereas the higher extreme weather signals in the south lead to the determination of clearly identifiable extreme weather events, the less clear signals in the north induce the production of incorrect synthetic wet scenarios for this region.
PDF
download
878987797
028-85220240
