Evaluation of Philip's infiltration equation for cultivated upland terraces in Indonesia

Abstract

The effect of traditional soil management practices (ridge tillage) on the parameters and applicability of Philip's infiltration equation was evaluated during the rainy season (1987) in an upland area located in the Samin watershed, Indonesia, about 800 m above mean sea level. A double ring infiltrometer was used. Sortivity ( S) was determined by fitting field data, using the equation S = I t 1 2 at an elapsed time of 5 min, where I and t represent cumulative infiltration (cm) and time, respectively. Traditional tillage treatments significantly influenced the spatial variability of the sorptivity and soil surface saturated conductivity ( K s). The relationship between S and clay content, and initial soil moisture content, was weak. The non-tilled or older tilled areas exhibited mostly matrix K s and more recent tilled areas exhibited macroporosity. Equating a second parameter ( A) in the Philip's equation with the K s yielded a fair agreement of calculated cumulative infiltration with the experimental data. However, extra care was required in choosing the constant A in the case of newly cultivated areas. Averaging parameters S and A on the basis of tillage treatment gave a reasonable representation of the cumulative infiltration in the experimental area. The error varied from 17.7% to 33.2%. Excellent fit of calculated values of cumulative infiltration with the experimental data for individual measurement was obtained by evaluating parameters S and A using multiple linear regression analysis. Averaging the values of S and A resulted in a slightly more accurate predicted value than the previous method, with an error variation of 16.2–31.5%. Further, the sorptivity obtained from the multiple linear regression approach and the previous equation mentioned were highly correlated ( r = 0.96).