The hydrological characteristics of the arid and semi-arid parts of Kenya

Abstract

The purpose of the study was to investigate the hydrological characteristics of the arid and semi-arid lands (ASAL) of Kenya with a view of establishing the water resources potential of ephemeral streams in providing supplementary water needs. Laikipia district which has undergone remarkable land use changes resulting in water use stress of perennial river abstractions and ground water exploitation in the recent years has been used as a model of the Kenyan ASAL. Long term availability of hydrological records was limited due to poor spatial distribution of the hydrological network over the Ewaso Ngiro basin. Further, the river flow data was only available on the perennial streams of the basin. Five sub-catchments areas in the Upper Ewaso Ngiro basin were chosen for experimental design. The period of experimental study extended from January 1989 to December 1991. The available hydrological records and the data obtained from the experimental study were subjected to spatial and temporal analyses. Spatial analyses included correlation and principal component analysis (peA) while temporal analysis involved the determination of diurnal, seasonal, cyclical and inter-annual characteristics including spectral and cross spectral techniques . simple transfer functions and complex Transfer Function Noise, TFN (r,b,s) models were used to quantify the rainfall-runoff and runoff-sediment relationships for the individual experimental sub-catchments. The univariate ARIMA p,d ,q) models were also used to quantify the annual long term hydrological records. The quality of all the records was first investigated before carrying out the above analyses. The study delineated the ASAL area into five hydrologically homogenous zones based on results from peA. The physical significance of the delineated zones were confirmed through inter-correlation and vector plot analyses together with the geophysical characteristics of the area. In each of these zones, a representative network was established, for which additional hydrological data is now available for-ASAL water resources planning in the light of increased settlements and population growth. Runoff episodes for individual sub-catchments indicated remarkable variations and contrasting responses among the sub-catchments. A 160 mmjh storm intensity caused a runoff peak rate of 1.456 m3jsjkm2 in Sirima while only 57.2 mmjh peak storm intensity caused a peak runoff rate of 6.052 m3jsjkm2 in Mukogodo. A minimum of 17 runoff events were observed in Matanya while a maximum of 67 events were obtained in Ngenia and the ephemeral channels were continuously dry for maximum periods ranging from 118 to 338 days during the experimental period. Although the ephemeral catchments have a surface water supply potential of at least 8,000 m3 per km2 per year, major threats were identified as sediment transport, high evaporation rates and intensified land use changes. The linear rainfall-runoff models for events showed initiation of runoff at threshold rainfall of between 8 mm and 18 mm . which characterised sub-catchment storage. Whereas the annual perennial streamflows were fitted by ARIMA (1,0,1) model, the annual rainfall indicated ARIMA (3,0,1) model to be appropriate due to the inherent complex rainfall-runoff relationships. The TFN (r,b,s) models based on 15 minutes rainfall-runoff event time series identified TFN (1,1,1) as the optimum model while the TFN (2,0,1) model fitted the corresponding runoffsediment series. Whereas runoff lagged the input rainfall by one time unit, the runoff and sediment occurrences were instantaneous. Ephemeral streams therefore have a high potential of water harvesting which could serve as supplementary water sources in the ASAL parts of Kenya. Its utilization however requires proper conservation measures which may enhance replenishment of the ground water systems.