Modelling Hydrological Processes in the Chania River System

Abstract

Kenya is classified as a chronically water-scarce country in terms of availability of water resources. This study was conducted in the Chania catchment located within the upper Tana catchment. Common water resources related problems experienced here are mainly related to water scarcity and deforestation. Hydrology plays an important role in providing information which will assist in overcoming the water shortage related problems in catchments. The main objective of this study was to model hydrological processes in the Chania catchment. The SWAT model was calibrated and validated using stream flow data for the years 2001-2004 and 2005-2008 respectively. Hydrological processes governing the temporal variability of stream flow were investigated using the Fourier Amplitude Sensitivity Test (FAST). The spatio-temporal variability of water balance components was determined from the model upon successful calibration. Effects of land use changes were investigated by using various land use scenarios. Results of calibration gave NSE of 0.67, R2 of 0.69 and Percent Bias (PBIAS) of 8.87 signifying a good performance. For validation NSE was 0.53, R2 was 0.59 and PBIAS was 12.24, signifying a satisfactory results. Results from FAST show that runoff processes were active mostly during precipitation events. Soil and ground water processes were most dominant in this catchment with a distinct presence during low flow seasons. The water balance components varied in both time and space. High quantities of ground water and water yield were noted in the north western side of the catchment. The average annual precipitation varied from 2376.57mm to 1499mm. Surface runoff ranged from 344.03 to 129.22mm with the highest value being recorded during the highest precipitation events. Ground water values range from 503.66mm to 765mm. Lateral flow values ranged from 291.40mm to 510.5mm while ET ranged from 670 mm to 850 mm. Investigations of effects of land use change on water balance show that decreasing forest led to a reduction in ground water, water yield and lateral flow. Urbanization reduced ground water, lateral flow and ET but lead to an increase in water yield and surface runoff. In conclusion, the SWAT model can be successfully used in xv Chania catchment to model hydrological processes. The variability of water balance components depended on the amount of precipitation, land cover and soil type. Since deforestation exacerbated water scarcity, it is therefore recommended that deforestation in the Chania catchment should be controlled while afforestation should be undertaken to restore the catchment