Pedotransfer functions for saturated hydraulic conductivity for surface runoff modeling
The study involved development of pedotransfer functions (PTFs) for determining saturated hydraulic conductivity (Ks) used in surface flow prediction. This preceded evaluation of existing PTFs for Ks in flow simulation. The pedotransfer functions were developed to predict parameters used in the determination of Ks using selected basic soil properties. The Soil Water Assessment Tool (SWAT) model was used in flow prediction in the Naro Moru river catchment of the Ewaso Ng’iro river basin, Kenya. The developed pedotransfer functions were then used in the simulation of surface runoff on the catchment and their performance in surface flow prediction compared with that of existing pedotransfer functions. Initial model runs during flow simulation yielded poor daily flow simulations compared to monthly simulations. This was attributed to differences in the timing of peak discharges for the observed and simulated hydrographs. The model was calibrated for a three year period followed by a three year validation period based on monthly flows. Calibration results yielded acceptable, but modest agreement between observed and simulated monthly stream flows. The modest model performance was associated with input data deficiencies and model limitations. The results indicated that the model could be adapted to the local conditions. Manual flow calibration was performed to improve simulation results initially based on average annual conditions followed by monthly calibration. There was significant improvement in the model performance based on monthly flow simulations. The model simulation of surface flow registered better performance compared to base flow and total flow indicating the model to be a better simulator of surface flow than baseflow. Observed and predicted surface runoff was compared to evaluate performance of existing PTFs. Model performance was similar for the existing PTFs selected. There was diversity v in performance of PTFs when used for surface runoff prediction. It was felt there is the need for continued development of PTFs for predicting Ks. The developed PTFs were evaluated for accuracy and reliability. The PTFs developed for saturated soil moisture content (θs) produced better performance in reliability compared to the remaining parameters in the van Genutchen moisture retention equation. The developed pedotransfer functions were then used in predicting Ks for surface flow simulation. The model performance in surface runoff simulation using developed PTFs was found acceptable. The study provides insight in developing equations for predicting Ks from basic soil properties being an input parameter in hydrological models. Hydrologic modeling plays a significant role in enabling policy makers, watershed planners and managers make appropriate decisions consistent with sustainable management of watershed resources.