| Modeling of surface energy fluxes and evapotranspiration (ET ) requires the understanding of the interaction between land and atmosphere as well as the appropriate representation of the associated spatial and temporal variability and heterogeneity. This dissertation provides new methodology showing how to rationally and properly incorporate surface features characteristics/properties, including the leaf area index, fraction of cover, vegetation height, and temperature, using different representations as well as identify the related effects on energy balance flux estimates including ET.;The main research objectives were addressed in Chapters 2 through 4 with each presented in a separate paper format with Chapter 1 presenting an introduction and Chapter 5 providing summary and recommendations. Chapter 2 discusses a new approach of incorporating temporal and spatial variability of surface features. We coupled a remote sensing-based energy balance model with a traditional water balance method to provide improved estimates of ET. This approach was tested over rainfed agricultural fields ∼ 10 km by 30 km in Ames, Iowa. Before coupling, we modified the water balance method by incorporating a remote sensing-based estimate for one of its parameters to ameliorate its performance on a spatial basis. Promising results were obtained with indications of improved estimates of ET and soil moisture in the root zone.;The effects of surface features heterogeneity on measurements of turbulence were investigated in Chapter 3. Scintillometer-based measurements/estimates of sensible heat flux (H) were obtained over the riparian zone of the Cibola National Wildlife Refuge (CNWR), California. Surface roughness including canopy height (hc), roughness length, and zero-plane displacement height were incorporated in different ways, to improve estimates of H. High resolution, 1-m maps of ground surface digital elevation model and canopy height, hc, were derived from airborne LiDAR sensor data to support the analysis.;The effects of using different pixel resolutions to account for surface feature variability on modeling energy fluxes, e.g., net radiation, soil, sensible, and latent heat, were studied in Chapter 4. Two different modeling approaches were applied to estimate energy fluxes and ET using high and low pixel resolution datasets obtained from airborne and Landsat sensors, respectively, provided over the riparian zone of the CNWR, California. Enhanced LiDAR-based hc maps were also used to support the modeling process. The related effects were described relative to leaf area index, fraction of cover, hc, soil moisture status at root zone, groundwater table level, and vegetation stress conditions. |
