This paper aims to demonstrate the potential of space-borne datasets for forest height estimation over Indian tropical forests. Fully polarimetric space-borne SAR interferometry (PolInSAR) data is acquired at the L-, C-, and X-band frequencies. The X-band data are acquired in two modes—single pass and repeat pass. The datasets are compensated for decorrelation due to SNR and varying spatial baselines. The remaining major decorrelation is the volumetric decorrelation, which is modeled using random volume over ground model to estimate the PolInSAR height for all the three frequencies. The modified Three-stage inversion algorithm is utilized for forest stand height estimation. The effect of forest biomass on height inversion accuracy is assessed. Furthermore, a first-order estimate of the absolute temporal decorrelation is demonstrated for the repeat-pass space-borne PolInSAR datasets. Extensive field validation campaigns are carried out in the tropical forest ranges. The forest height is inverted using data at all the three frequencies and validated with field measured values. The zero-temporal baseline bistatic TerraSAR-X/TanDEM-X and the L-band ALOS-2/PALSAR-2 result in a good height inversion with ${\bm{r}^2}$ and RMSE of 0.77 and 1.86 m (X-band) and 0.75 and 1.94 m (L-band), respectively. Furthermore, a comparison of estimated height of ALOS-2/PALSAR-2, TerraSAR-X, and RadarSAT-2 has been done with respect to the estimated height of TerraSAR-X/TanDEM-X. It is observed that RMSE of height inversion with respect to TerraSAR-X/TanDEM-X height is found to be 5.4 m, 7.6 m, and 12.8 m for ALOS-2/PALSAR-2, RadarSAT-2, and TerraSAR-X, respectively.

Physics-based algorithms estimating large-scale forest above-ground biomass (AGB) from synthetic aperture radar (SAR) data generally use airborne laser scanning (ALS) or grid of national forest inventory (NFI) to reduce uncertainties in the model calibration. This study assesses the potential of multitemporal L-band ALOS-2/PALSAR-2 data to improve forest AGB estimation using the three-parameter water cloud model (WCM) trained with field data from relatively small (0.1 ha) plots. The major objective is to assess the impact of the high uncertainties in field inventory data due to relatively smaller plot size and temporal gap between acquisitions and ground truth on the AGB estimation. This study analyzes a time series of twenty-three ALOS-2 dual-polarized images spanning 5 years acquired under different weather and soil moisture conditions over a subtropical forest test site in India. The WCM model is trained and validated on individual acquisitions to retrieve forest AGB. The accuracy of the generated AGB products is quantified using the root mean square error (RMSE). Further, we use a multitemporal AGB retrieval approach to improve the accuracy of the estimated AGB. Changes in precipitation and soil moisture affect the AGB retrieval accuracy from individual acquisitions; however, using multitemporal data, these effects are mitigated. Using a multitemporal AGB retrieval strategy, the accuracy improves by 15% (55 Mg/ha RMSE) for all field plots and by 21% (39 Mg/ha RMSE) for forests with AGB less than 100 Mg/ha. The analysis shows that any ten multitemporal acquisitions spanning 5 years are sufficient for improving AGB retrieval accuracy over the considered test site. Furthermore, we use allometry from colocated field plots and Global Ecosystem Dynamics Investigation (GEDI) L2A height metrics to produce GEDI-derived AGB estimates. Despite the limited co-location of GEDI and field data over our study area, within the period of interest, the preliminary analysis shows the potential of jointly using the GEDI-derived AGB and multi-temporal ALOS-2 data for large-scale AGB retrieval.

Forest above-ground biomass (AGB) estimation from SAR backscatter is affected by varying imaging and environmental conditions. This paper quantifies and compares the performance of forest biomass estimation from L-band SAR backscatter measured selectively under dry and wet conditions during the 2019 AM-PM NASA airborne campaign. Seven Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) images acquired between June and October 2019 over a temperate deciduous forest in Southeastern United States with varying moisture and precipitation conditions are examined in conjunction with LIDAR and field measurements. Biomass is estimated by fitting a 3-parameter modified Water Cloud Model (WCM) to radiometric terrain corrected SAR backscatter. Our experiment is designed to quantify the biomass estimation errors when biomass models are calibrated and validated on varying acquisition conditions (dry or wet). Multi-temporal estimation strategies are also evaluated and compared with single-acquisition estimation approaches. As an outcome, the experiment shows that the WCM model calibrated and validated on single acquisitions adapts to different soil moisture conditions with RMSD up to 18.7 Mg/ha. The AGB estimation performance, however, decreases with RMSD upwards of 30 Mg/ha when the model is cross-validated on moisture and precipitation conditions different than the calibration conditions. Results confirm that calibrating the model over the multi-temporal data using averaged backscatter or weighted combinations of individual AGB estimates, improves the biomass estimation accuracy up to about 20% at L-band. This study helps design biomass cal/val procedures and biomass estimation algorithms for dense time-series to be collected by low-frequency radar missions such as NASA-ISRO SAR (NISAR) and BIOMASS.