Factors affecting the spatial pattern of Infiltration capacity at the hillslope scale
2 Global Institute for Water Security, University of Saskatchewan
P 2.5 in …We collaborate with other disciplines and learn from each other.
Spatial organized patterns of infiltration capacity (Ic) have shown to be crucial in distributed hydrological modeling. Still the controls on the spatial organization of Ic in the landscape are poorly understood. The aim of this study was to investigate the characteristics and underlying principles of these patterns on a 5 ha, agricultural, low-angled field in Southern Saskatchewan prairies. For this purpose 62 infiltration measurements based on a systematic random sampling with a clustered component were performed using a Cornell Sprinkler Infiltrometer. Additional, the effect of factors on Ic, mainly easy accessible terrain indices, was assessed in random forest analysis and the resulting most important factors were further examined in Multiple Linear Regressions. The spatial correlation and patterns of Ic were analyzed using geostatistical methods involving the results of factors affecting Ic. A weak but significant spatial correlation of log-normal-transformed Ic with a correlation length in the order of about 210m was found. The predictive power of Ordinary kriging explaining 11% of the variance was low, but prediction accuracy was improved to 21% explained variance using the downslope index (DI) as surrogate in regression kriging. The low but significant negative linear impact of DI on Ic (R²= 0.23) indicates that overland flow related accumulation processes of finest material effect the distribution of Ic. How the provided patterns perform in rainfall-runoff models is focus of subsequent research.