Inverse analysis of the role of soil vertical heterogeneity in controlling surface soil state and energy partition

<jats:p>Surface soil moisture and temperature have been widely addressed in land surface processes modeling and satellite remote sensing because they play a key role in land surface energy and water budget. However, it is rather difficult for some land surface models to reproduce the surface soil state in areas with high soil vertical heterogeneity because these models use a single parameter set to characterize soil hydraulic and thermal processes. This study develops a single‐source land surface model to parameterize this heterogeneity. Its soil parameters are inversely estimated by minimizing a cost function that is objectively determined by the discrepancy between observed and model‐predicted values of soil moisture and temperature. The approach is then used to investigate how the soil vertical heterogeneity affects subsurface processes and thus controls soil surface state and surface energy budget. This approach is applied to a synthetic numerical experiment and a Tibet field experiment, where the horizontal heterogeneity can be neglected. We indicate that (1) vertical heterogeneous soils cannot be effectively approximated by vertically homogenous soils in a land surface model no matter how the soil parameters are adjusted; (2) soil vertical heterogeneity obviously affects soil subsurface processes and plays a very important role in controlling surface soil wetness and surface energy partition; and (3) in particular, the existence of dense vegetation roots in topsoils may significantly reduce thermal conductivity, increase soil water potential, and enhance surface evaporation. We therefore conclude that it is indispensable to take the soil vertical heterogeneity into account in land surface models, although some of them still assume vertically uniform soil parameters.</jats:p>