Soil moisture and temperature vertical profiles vary quickly during the day and may have a significant influence on the soil microwave emission. The objective of this work is to quantify such an influence and the consequences in soil moisture estimation from microwave radiometric information. The analysis is based on experimental data collected by the ground-based PORTOS radiometer at 1.4, 5.05, and 10.65 GHz and data simulated by a coherent model of microwave emission from layered media [Wilheit model (1978)]. In order to simulate diurnal variations of the brightness temperature (T-B), the Wilheit model is coupled to a mechanistic model of heat and water flows in the soil. The Wilheit model is validated on experimental data and its performances for estimating T-B are compared to those of a simpler approach based on a description of the soil media as a single layer (Fresnel model). When the depth of this single layer (hereafter referred to as the sampling depth) is determined to fit the experimental data, similar accuracy in T-B estimation is found with both the Wilheit and Fresnel models. The soil microwave emission is found to be strongly affected by the diurnal variations of soil moisture and temperature profiles. Consequently, the T-B sensitivity to soil moisture and temperature profiles has an influence. on the estimation, from microwave observations, of the surface soil moisture in a surface layer with a fixed depth (theta(s)): the accuracy of theta(s) retrievals and the optimal sampling depth depends both on the variation in soil moisture and temperature profile shape. [References: 21]
展开▼
