79-87
Evaluation of Philip’s and Kostiakov’s Infiltration Models on Soils Derived from three Parent Materials in Akwa Ibom State, Nigeria
Authors: Utin U.E., Oguike P.C.
Number of views: 418
Knowledge of water infiltration into soils is required in determining how much water is stored in the root zone for crops and for the design of irrigation systems. Therefore, infiltration models are useful tools in predicting water infiltration into the soil. A study was conducted to evaluate two infiltration models on soils derived from coastal plain sands (CPs), sandstone (SSt) and river alluvium (ALv) in Akwa Ibom State, Nigeria. The models were Kostiakov’s (KOS), and Philip’s (PHI) models. Ten observation points were selected from each of the three parent materials totaling thirty (30), where infiltration studies were carried out. Topsoil (0-20 cm) samples were also collected for laboratory analysis of some soil physical properties. Data generated were summarized using mean, standard deviation and coefficient of variation. Coefficient of determination (R2), Nash-Sutcliffe model efficiency (NE) and root mean square error (RMSE) were used to determine the goodness of fit of the infiltration models with the field-measured data. Infiltration model with the highest R2 and NE as well as lowest RMSE was adjudged best in predicting water infiltration. Results showed that the highest means of infiltration model parameters were observed in SSt soil as follows: Kostiakov’s K (1.376 cm min-1) and α (0.893) and Philip’s S (0.667 cm min-1/2) and A (0.813 cm min-1). Alluvial (ALv) soil had the lowest mean values of the models’ parameters as follows: 0.067 cm min-1 and 0.800 for Kostiakov’s K and α, respectively as well as 0.068 cm min-1/2 and 0.018 cm min-1 for Philip’s S and A, respectively. In the CPs soil, Philip’s model with the highest R2, highest NE and lowest RMSE of 0.999, 0.999 and 0.013 predicted water infiltration better than Kostiakov’s model with R2, NE and RMSE of 0.998, 0.998 and 0.015, respectively. In the SSt soil, Kostiakov’s model, with R2, NE and RMSE of 0.999, 0.999 and 0.012, respectively gave better predictions of water infiltration than Philip’s model with R2 of 0.998, NE of 0.998 and RMSE of 0.018. In the ALv soil, KOS had R2, NE and RMSE of 0.998, 0.997 and 0.016 while the values for PHI were 0.997, 0.994 and 0.019, respectively, indicating that KOS predicted water infiltration better than PHI in this soil. Philip’s model gave the best prediction of water infiltration in CPs soil while the Kostiakov’s model was best for soils of sandstone and alluvial parent materials. The Philip’s model was therefore recommended for the prediction of water infiltration in soils derived from coastal plain sands while the Kostiakov’s model was recommended for sandstone and alluvial soils of Akwa Ibom State, Nigeria.