A data-driven snapshot algorithm for high-resolution soil moisture retrievals for the upcoming NISAR mission
Abstract
The NASA-ISRO Synthetic Aperture Radar (NISAR) is in the developmental stage and is planned to launch in Jan 2024 with two different microwave frequency bands L-band (~1.25 GHz) and S-band (~3.20 GHz). The NISAR mission aims to provide fine-scale SAR observations at resolutions of ~10 meters and near weekly global coverage with ascending and descending overpasses. The fine-scale NISAR L-band backscatter observation can be used to estimate the hydrological state variables, such as soil moisture at high resolution with a single overpass and change in soil moisture over time with repeat overpasses. We proposed a snapshot algorithm for retrieving field-scale soil moisture (~100 to 200 m) at a global extent with optimal uncertainty while taking advantage of fine-scale L-band SAR backscatter of NISAR mission. The proposed snapshot algorithm is formulated to retrieve soil moisture at ~200 m by blending the coarse resolution (~9 km) soil moisture products of model/reanalysis/satellite and fine-scale L-band co-and cross-pol SAR backscatter (~10m) observations. The major aim of the snapshot algorithm is to avoid any complex modeling approach, dependencies on NISAR observation time series and multiple ancillary datasets in soil moisture retrievals to reduce the degree of freedom. Further, the proposed algorithm uses a model/reanalysis based coarse resolution soil moisture products which have proven good accuracy i.e., European Centre for Medium-Range Weather Forecast (ECMWF) to reduce the dependency on another satellite-based (i.e., SMAP) soil moisture product. The use of model/reanalysis products also added an extra advantage that it will be able to select the input coarse resolution soil moisture value closest to SAR observations and minimizes the potential errors due to temporal mismatch of the coarse and fine-scale data. Since NISAR L-band SAR data is expected to be available by mid-2024 after in-orbit check. Therefore, similar L-band SAR datasets such as UAVSAR observations from SMAPVEX campaigns and ALOS PALSAR-2 have been used for retrieving high-resolution soil moisture as substitute to NISAR L-band observations. ECMWF ERA5-Land reanalysis soil moisture product (~9km) is used as input of coarse resolution soil moisture. The proposed snapshot algorithm was tested on different hydro-climatic regions of CONUS and showed that the approach has a great potential to retrieve soil moisture at a 200 m resolution with very low ubRMSE (±0.05 m3/m3), RMSE and Bias. The result confirms that the proposed algorithm can meets the accuracy of the NISAR mission goal ±0.06 m3/m3 (over areas with vegetation water content below 5 kg/m2) and provides new insight into generating high-resolution soil moisture with active L-band SAR observations.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFM.H42G1379L