Retrieval of Surface and Subsurface Moisture of Bare Soil Using Simulated Annealing

Soil moisture is of fundamental importance to many hydrological and biological processes. Soil moisture information is vital to understanding the cycling of water, energy, and carbon in the Earth system. Knowledge of soil moisture is critical to agencies concerned with weather and climate, runoff potential and flood control, soil erosion, reservoir management, water quality, agricultural productivity, drought monitoring, and human health. The need to monitor the soil moisture on a global scale has motivated missions such as Soil Moisture Active and Passive (SMAP) [1]. Rough surface scattering models and remote sensing retrieval algorithms are essential in study of the soil moisture, because soil can be represented as a rough surface structure. Effects of soil moisture on the backscattered field have been studied since the 1960s, but soil moisture estimation remains a challenging problem and there is still a need for more accurate and more efficient inversion algorithms. It has been shown that the simulated annealing method is a powerful tool for inversion of the model parameters of rough surface structures [2]. The sensitivity of this method to measurement noise has also been investigated assuming a two-layer structure characterized by the layers dielectric constants, layer thickness, and statistical properties of the rough interfaces [2]. However, since the moisture profile varies with depth, it is sometimes necessary to model the rough surface as a layered structure with a rough interface on top and a stratified structure below where each layer is assumed to have a constant volumetric moisture content. In this work, we discretize the soil structure into several layers of constant moisture content to examine the effect of subsurface profile on the backscattering coefficient. We will show that while the moisture profile could vary in deeper layers, these layers do not affect the scattered electromagnetic field significantly. Therefore, we can use just a few layers to represent the soil. We will then apply the simulated annealing method to retrieve the moisture content of the layers. We will also examine the sensitivity of the inversion algorithm to measurement noise. This work is directly intended for SMAP L-band observations. Nonetheless, our representation of soil is more general than what is currently being used for SMAP and would benefit not only SMAP retrievals but also future lower frequency missions. References [1] "Soil Moisture Active and Passive (SMAP)," Jet Propulsion Laboratory. [Online]. Available: http://smap.jpl.nasa.gov/. [2] A. Tabatabaeenejad and M. Moghaddam, "Inversion of subsurface properties of layered dielectric structures with random slightly-rough interfaces using the method of simulated annealing," IEEE Trans. Geosci. Remote Sens., vol. 47, no. 7, pp. 2035--2046, Jul. 2009.