The development of soil moisture retrieval algorithms for the upcoming NASA/ISRO SAR mission motivates investigations into soil surface scattering effects and their influence on soil moisture retrievals. To better understand the conditions that impact retrieval performance, an investigation is performed into the second-order solution small slope approximation (SSA2) for rough surface backscattering. Scattering amplitudes are calculated for varying soil moisture and surface roughness conditions and compared to the first-order solution. The results show the ability of the second-order solution to model cross-pol scattering and help pinpoint physical conditions that may influence soil moisture retrievals. Overall, the results suggest that the first-order SSA solution should be applicable under the physical conditions where NISAR soil moisture retrievals are expected to be performed.

The use of a “time-series ratio” soil moisture retrieval approach is under consideration for the upcoming NISAR mission’s soil moisture product. As such, it is of interest to characterize the algorithm’s anticipated error budget as part of pre-launch activities. This paper develops an error model to estimate retrieval errors for the proposed algorithm. The model accounts for error contributions from speckle and thermal noise as well as uncertainties that arise as part of the retrieval process. Spatial and temporal behaviors of the retrieval errors are determined using a SMAP-based soil moisture climatology. The results show that soil moisture retrieval errors from the time-series ratio method meet the 0.06 m3/m3unbiased root mean square error (URMSE) performance metric established for the NISAR soil moisture product.

The National Aeronautics and Space Administration (NASA) - Indian Space Research Organization (ISRO) Synthetic Aperture Radar (NISAR) mission plan to launch a SAR operating at L- and S-band with a 12-day repeat frequency. A global soil moisture product at 200 m spatial resolution derived from 200 m NISAR radar measurements is currently under development. Although several retrieval algorithms are being investigated, this paper focuses on a “time series ratio” retrieval approach. In order to understand and assess the performance of this algorithm, an error model has been developed and is reported in this paper. The model is applied to examine errors as a function of the instrument characteristics and for a given location. Initial progress in including vegetation effects and in predicting errors as a function of spatial location is also described.

L- and S-band observations from NASA's Passive/Active L/S band (PALS) sensor from the SMEX02 campaign were used to estimate soil moisture. The retrieval process is based on the “alpha approximation” method. This method utilizes a time-series of normalized radar backscatter measurements as well as ancillary information to estimate soil moisture over the Walnut Creek watershed. The resulting retrieved soil moistures are compared to in-situ soil moisture measurements at multiple test sites within the watershed. The calculations show reasonable results for both L- and S-band and provide further insight into the use of L- and S-bands for the upcoming NASA/ISRO mission.

The NASA ISRO Synthetic Aperture Radar (NISAR) mission is currently under development and will provide global L-band radar observations that will be helpful for various soil moisture applications. The final NISAR soil moisture product will have 200m spatial resolution with 12-day exact revisit time. A time-series ratio algorithm was implemented using NISAR simulated UAVSAR data collected during the SMAPVEX12 field experiment. In this paper, the performance of the time series ratio algorithm was assessed using in situ observations. Performance of the soil moisture retrieval algorithm was also assessed for dual polarization and quad-polarization observations modes.

S-band observations from NASA's Passive/Active L/S Band (PALS) radar from the SMEX02 campaign were used to estimate soil moisture. The “alpha” method is applied for this process, in which a time series of ratios of normalized radar cross section values at successive measurements is used to infer the corresponding soil moisture time series given ancillary information on the minimum and maximum soil moisture values expected over the time series. Results are examined as a function of the polarization and crop type. The results show reasonable retrieval performance, indicating the potential of using S-band observations from the future NASA/ISRO SAR (NISAR) mission.

The NASA ISRO Synthetic Aperture Radar (NISAR) mission is currently under development and is scheduled for launch in 2022. The NISAR mission will provide global data sets of Earth land surface dynamics that are critical for multiple Earth Science disciplines including observations of ecosystem carbon and water cycles. Global L-band radar observations at high spatial resolution will be helpful for soil moisture applications. One of the goals of the NISAR mission is to provide a global soil moisture product at 200 m resolution with a global revisit frequency of 6 days. A time-series ratio algorithm was implemented using NISAR simulated SMAPVEX12 UAVSAR data, which is an L-band airborne radar backscatter measurement. For a NISAR-like configuration, backscatter at incidence angles from 30 to 50 degrees was considered in this study. The initial retrieval statistics following comparisons with in-situ ground truth show correlation coefficients (R) to be about 0.81, and the unbiased RMSE to be about 0.06 m3/m3. Results from dual co-polarization and/or cross-polarization modes were evaluated and considered for performance improvement.