A physically based hydrological connectivity algorithm for describing spatial patterns of soil moisture in the unsaturated zone

Hydrologic connectivity has been proposed as an important concept for understanding local processes in the context of catchment hydrology. It can be useful for characterizing the soil moisture variability in complex heterogeneous landscapes. The current land surface models (e.g., Community Land Model, CLM) could not completely account for flow path continuity and connected patterns of subsurface properties in the unsaturated zone. In this study, we developed a physically based hydrologic connectivity algorithm based on dominant physical controls (e.g., topography, soil texture, and vegetation) to better understand the spatially distributed subsurface flow and improve the parameterization of soil hydraulic properties in hydrological modeling. We investigated the effects of mixed physical controls on soil moisture spatial variability and developed hydrologic connectivity using various thresholds. The connectivity was used for identifying the soil moisture variability and applied in a distributed land surface model (CLM) for calibrating soil hydraulic properties and improving model performance for estimating spatially distributed soil moisture. The proposed concept was tested in two watersheds (Little Washita in Oklahoma and Upper South Skunk in Iowa) comparing estimated soil moisture with the airborne remote sensing data (Electronically Scanning Thinned Array Radiometer and Polarimetric Scanning Radiometer). Our finding demonstrated that the spatial variations of soil moisture could be described well using physically based hydrologic connectivity, and the land surface model performance was improved by using the calibrated (distributed) soil hydraulic parameters. In addition, we found that the calibrated soil hydraulic parameters significantly affect model outputs not only on the water cycle but also on surface energy budgets.