Determining the Accuracy of Water Infiltration Models for Different Land Uses in the Dry–Hot Valley Region of China

In the dry–hot valley region of Southwest China, water infiltration exhibits temporal variations due to the combined effects of land use type and the dramatic seasonal dry–wet cycle. To accurately compare and predict the infiltration characteristics, soil water infiltration processes and cumulative infiltration were quantified for five typical land uses—traditional corn (TC), plum orchard (PO), pine forest (PF), grassland (GL), and abandoned cropland (AC)—in a dry–hot valley region during both the rainy (July) and dry (November) seasons using a Mini Disk Infiltrometer (MDI). These data were then statistically analyzed using the Kostiakov, Philip, and Horton models. The results showed that the mean infiltration rate and cumulative infiltration during the rainy season were 1.34 times and 1.31 times higher than in the dry season, respectively. The water infiltration rate and cumulative infiltration for the five land uses generally followed the order of PF > GL/TC > PO/AC during both rainy and dry seasons. The model parameters related to the initial infiltration capability (Kostiakov parameter, a) and the steady infiltration capability (Philip parameter, A; and the Horton parameter, f_c) during the rainy season were all greater than those in the dry season. Compared to the Kostiakov and Horton models, the Philip model achieved the highest mean Nash–Sutcliffe efficiency (NSE) values in fitting soil water infiltration processes, the lowest mean relative error (MRE) values, and the highest determination coefficient values (R²) in predicting the cumulative infiltration, with relatively little difference between the two seasons. These results indicate that PF, GL, and TC exhibit superior soil water infiltration capabilities compared to other land uses during both the rainy and dry seasons. The Philip model is more suitable for estimating soil infiltration capacity in the dry–hot valley region during both seasons. Identification of the superior land use types and accuracy determination of the water infiltration model can help guide effective water conservation and vegetation restoration initiatives in the dry–hot valley region of Southwest China.