Physics-based Multi-resolution Radar-Radiometer Soil Moisture Estimation within the SMAP Mission Framework

Abstract:

To further develop our understanding of global carbon and water cycles and to support the NASA Soil Moisture Active-Passive (SMAP) mission efforts have been made to develop joint and combined radar and radiometer soil moisture estimation algorithms. Taking advantage of the complimentary sensitivities of radar backscatter and brightness temperature to soil moisture and vegetation has the potential to greatly improve global soil moisture estimates. With the advent of SMAP, not only combing radar and radiometer information is of interest, combing multi-resolution data becomes critical.

The work presented here will discuss methods to estimate soil moisture within the SMAP framework via a global optimization technique. Fine resolution radar backscatter measurements (3 km for SMAP) are combined with coarse resolution radiometer data (36 km for SMAP) in a joint cost function. Brightness temperature disaggregation and soil moisture estimation are then performed at the radar resolution. Furthermore, to capture the underlying physics of emission and scattering within the cost function, physics-based forward models which link emission and scattering from first principles are employed. The resulting effect is the ability to define a parameter kernel shared between emission and scattering models. Preliminary investigation yields improved soil moisture estimation when radar and radiometer information are used jointly. Furthermore, over a wide range of soil moisture (0.04 – 0.4 cm³/cm³) and vegetation (0- 5 kg/m²) physics based joint estimation yields the least retrieval errors.