Soil moisture forms the interface at which the partitioning of the energy, carbon and water budget for the land-surface occurs. Its variability impacts different fields of application at varying extent scales like agriculture at the field scale, meteorology at the regional scale and climate change assessment at the global scale. However, past literature has focused on understanding soil moisture dynamics at this diverse range of extent scales using soil moisture data at the Darcy support scale which cannot effectively cater to soil moisture dynamics for the current eco-hydrologic models that describe complex heterogeneous domains at remote sensing footprint scales. This dissertation serves to push the envelope of our understanding of soil moisture dynamics and its dependence on land-surface heterogeneity at the coarse remote sensing scales. The research questions answered in this dissertation include 1) determining the dominant land-surface controls of near-surface soil moisture dynamics at scales varying between the Darcy (of the order of a few centimeters) support and satellite footprint scale (25.6 km); 2) generating a framework for quantifying the relationships between antecedent wetness, land-surface heterogeneity and near-surface soil moisture at remote sensing scales and 3) evaluating variability in the root zone moisture dynamics as evaluated through evapo-transpiration estimates at different remote sensing footprint scales. The dominant land-surface factors controlling soil moisture distribution at different scales were determined by developing a new Shannon entropy based technique and non-decimated wavelet transforms. It was found that the land-surface controls on soil moisture vary with hydro-climate and antecedent wetness conditions. In general, the effect of soil was found to reduce with coarsening support scale while the effect of topography and vegetation increased. A novel Scale-Wetness-Heterogeneity (SWHET) cuboid was developed to coalesce the relationship between soil moisture redistribution and dominant physical controls at different land-surface heterogeneity and antecedent wetness conditions across remote sensing scales. The SWHET cuboid can potentially enable spatial transferability of the scaling relationships for near-surface soil moisture. It was found that results from the SWHET cuboid enabled spatial transferability of the scaling relationships between two similar hydro-climates (Iowa, U.S.A and Manitoba, Canada) under some wetness and land-surface heterogeneity conditions. Evapotranspiration estimates were computed at varying scales using airborne and satellite borne remotely sensed data. The results indicated that in a semi-arid row cropped orchard environment, a remote sensing support scale comparable to the row spacing and smaller or comparable to the canopy size of trees overestimates the land surface temperature and consequently, underestimates evapotranspiration.
