High-resolution Surface Soil Moisture with High Temporal Frequency using Passive/Active MW Synergy
Escorihuela, Maria Jose1; Merlin, Olivier2; Zribi, Mehrez2

We have implemented and validated a disaggregation algorithm over the modern irrigated area of Segarra-Garrigues in Lleida (Catalonia) to provide high-resolution (100 m) NSSM data from SMOS data (Merlin et al. 2013). The algorithm uses MODIS data to provide NSSM at 1 km spatial resolution every 2/3 days, and LandSat or ASTER data to provide NSSM at the field scale (100 m) every 16 days. The SMAP (Soil Moisture Active Passive) satellite (to be launched in 2015) will provide continuity to SMOS data.

Thermal-disaggregated passive microwave-derived NSSM data provide very high accuracy and do not require calibration. However its temporal frequency is limited by the availability of high-resolution thermal sensors (Landsat and ASTER have a repeat cycle of about 16 days). Zribi et al. 2011 have developed an approach to use high resolution Synthetic Aperture Radar data to estimate soil moisture. This type of sensors can provide adequate temporal coverage (Sentinel-1 to be launched in 2013 will provide high resolution radar data every 3 days), however their accuracy is lower than passive microwave-derived NSSM and retrieval algorithms require site-specific calibration (using independent estimates of NSSM).

In order to address this issue, we propose here an innovative approach to take advantage of the complementarity between thermal-disaggregated SMOS/SMAP NSSM (no need for calibration) and radar derived NSSM (with a need for calibrating the radiative transfer model) at 100 m resolution. On the days when radar data are available at approximately the same time as the thermal data, we will calibrate the inversion algorithm of radar data from the NSSM data estimated at similar spatial resolution by the thermal-based disaggregation method of SMOS/SMAP NSSM. As a result, we will be able to provide high-resolution NSSM every 3 days.