Assessing the Predictive Accuracy of SOILWAT Model at Two Agro-Ecological Zones of Katsina State, Nigeria
Publication Date : 15/04/2021
The soil water characteristics of two agro-ecological zones of katsina were evaluated to determine the level of predictability of the SOILWAT model developed by Saxton and Rawls. The experiment covers two agro-ecological zones of Katsina State and soils of two different depths of Katsina State. 84 soil samples were collected at depths 0 to 30 and 30 to 60 cm for top and subsoil respectively at 7 locations in the zones and reference points were also recorded. The measured and predicted values of the SWC were evaluated using coefficient of variation (CV %), roots mean square error (RMSE), model efficiency (EF) and coefficient of residual mass (CRM). The RMSE of bulk density has values that range from 0.09 to 0.1, and available water from 0.03 to 0.1, while values of coefficient of variation were less than 15%. The field capacity, wilting point and moisture content at saturation were satisfactorily predicted, low RMSE were observed but the CV is less than 30% at saturation and field capacity but highly dispersed in wilting point. Also, it poorly predicted the saturated hydraulic conductivity of the two zones, with the RMSE ranging between 20 and 37 with a CV of 67% and 101%, the values of the model efficiency were negative indicating that the measured result is better than the predicted. Therefore, according to this study, it is recommended that SOILWAT model can be adopted for predicting the Soil Water Characteristics in the two ecological zones except the hydraulic conductivity, which is a limitation.
No. of Downloads :