Assessment of Forest Biomass Estimation from Dry and Wet SAR Acquisitions Collected during the 2019 UAVSAR AM-PM Campaign in Southeastern United States

Abstract

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.

The Role of Time-Series L-Band SAR and GEDI in Mapping Sub-Tropical Above-Ground Biomass

“…This is important, especially for tropical regions where cloud cover can limit optical remote sensing for up to 6 months in a year. SAR backscatter has been extensively utilized for forest AGB estimation using a wide variety of SAR data acquired at different frequencies (Le Toan et al, 1992Luckman et al, 1998;Santoro et al, 2011;Englhart et al, 2012;Kumar et al, 2012;Schlund et al, 2018;Khati et al, 2020). Depending on the wavelength, SAR signals interact with different components of the forest such as stem, branch, leaves, trunk, and ground (Dobson et al, 1992;Ningthoujam et al, 2016Ningthoujam et al, , 2018Singh and Yamaguchi, 2018;Singh et al, 2019Singh et al, , 2020.…”

“…Depending on the wavelength, SAR signals interact with different components of the forest such as stem, branch, leaves, trunk, and ground (Dobson et al, 1992;Ningthoujam et al, 2016Ningthoujam et al, , 2018Singh and Yamaguchi, 2018;Singh et al, 2019Singh et al, , 2020. The SAR backscatter signal strength increases with AGB up to a saturation level (Yu and Saatchi, 2016;Joshi et al, 2017;Schlund et al, 2019), which depends on the sensor properties such as wavelength and polarization, as well as site conditions including stand structure, ground conditions, and moisture (Dobson et al, 1992;Le Toan et al, 1992;Ghasemi et al, 2011;Huang et al, 2015;Ningthoujam et al, 2018;Khati et al, 2020). The longer wavelength (P-and L-band) SAR backscatter thus saturates at higher forest AGB and is more suitable for forest AGB mapping.…”

“…The SAR backscatter-AGB relationship can be calibrated using field inventory plots, airborne laser scanning-(ALS-) derived AGB maps, or a dense grid of national forest inventory (NFI) data. The accuracy of the developed SAR backscatter-AGB model depends on these field plots (Santoro et al, 2002), which might have inherent uncertainties due to allometry, geolocation, and inadequate representation of the spatial variability of forests (Chave et al, 2004;Khati et al, 2020). Furthermore, the field plots can vary in size from 0.05 ha (hectare) (e.g., (Luckman et al, 1997(Luckman et al, , 1998Kurvonen et al, 1999)) to 50 ha (e.g., the 50 ha Barro Colorado Island plot (Hubbell et al, 1999)) or more (Houghton et al, 2001), and are laid in different configurations (circular, rectangular, or square) depending on the field campaign strategy.…”

“…However, NFI or ALS data are not publicly available over most of the tropical forests, due to geographic or regulatory restrictions. Further, the SAR backscatter-AGB relationship is influenced by soil and canopy moisture variations, seasonal changes in canopy structure (leaf-fall), and weather (Harrell et al, 1997;Pulliainen et al, 1999;Rauste, 2005;Lucas et al, 2010;Khati et al, 2020). These variations translate into uncertainties in derived AGB estimates, with Bouvet et al (2018) reporting up to 300% error in AGB estimates when L-band backscatter was trained and validated on regions with nonhomogeneous environmental conditions.…”

“…These variations translate into uncertainties in derived AGB estimates, with Bouvet et al (2018) reporting up to 300% error in AGB estimates when L-band backscatter was trained and validated on regions with nonhomogeneous environmental conditions. These uncertainties in AGB estimation can be reduced by integrating multitemporal SAR measurements (Kurvonen et al, 1999;Pulliainen et al, 1999;Rauste, 2005;Englhart et al, 2012;Tanase et al, 2014;Cartus and Santoro, 2019;Khati et al, 2020). The contribution of this study is to elucidate the feasibility of using relatively small inventory plots (0.1 ha) for forest AGB estimation using L-band SAR data, in contrast to previous works which use ALS or NFI data for model calibration.…”