Effect of vegetation characteristics on near soil moisture retrieval using microwave remote sensing technique

Passive microwave remote sensing has shown potential for monitoring near surface soil moisture. This dissertation presents a new approach to representing the effect of vegetation on microwave emission by extending an existing model (Wilheit, 1978) of the coherent propagation of electromagnetic radiation through a stratified medium. The resulting multi-layer microwave emission model is plausibly realistic in that it captures the behavior of the vegetation canopy by considering the dielectric permittivity of the mixture of air and vegetation matter in the canopy and recognizing the vertical distribution of dielectric permittivity through the canopy. The model parameters required to specify the dielectric profile within the canopy are not usually available from data taken in typical field experiments, particularly the parameters that quantify the way the dielectric permittivity of the vegetation and air mix together to give the dielectric permittivity of the canopy. Thus, the feasibility of specifying these parameters using an advanced single-criterion, multiple-parameter optimization technique was investigated. The resulting model was also applied to investigate the sensitivity of microwave emission to specific vegetation parameters. The study continued with an investigation of how the presence and nature of vegetation cover influences the values of geophysical variables retrieved from multi-angle microwave radiometer spectrometer observations, using the upcoming Soil Moisture Ocean Salinity (SMOS) mission as a case study. The extended version of the Wilheit (1978) model was used to calculate synthetic observations of microwave brightness temperature at the look-angles proposed for the SMOS mission for three different soil moisture states (wet, medium, and dry) and four different vegetation covers (grass, crop, shrub, and forest). It was shown that retrieved values are only accurate when the effective values of the opacity coefficient used in the Fresnel model are made to vary in a prescribed way with look-angle, soil moisture status, and vegetation. The errors in retrieved values that may be induced by poor specification of vegetation cover were investigated by imposing random errors in the values of vegetation-related parameters in the forward calculations of synthetic observations made with the extended Wilheit model. The results show that poorly specified vegetation can result in both random and systematic errors in the retrieved values of the geophysical variables. (Abstract shortened by UMI.