Improving Forest Above-Ground Biomass Retrieval Using Multi-Sensor L- and C- Band SAR Data and Multi-Temporal Spaceborne LiDAR Data

Abstract

Due to the great structural and species diversity of tropical forests and limitations of the methods used to estimate aboveground biomass, there is uncertainty in quantifying its carbon sequestration potential. Measuring carbon sequestered in the terrestrial ecosystem and monitoring its dynamics is one of the key objectives in sustainable development goals. Synthetic Aperture Radar (SAR) has evolved as a key satellite technology in measuring and monitoring terrestrial carbon sink stored as biomass in plants. This study attempts to model forest above-ground biomass (AGB) using a random forest machine-learning approach where the predictor variables are from C-band (Radarsat-2), L-band (ALOS-2/PALSAR-2), and multi-temporal spaceborne LiDAR data from the GEDI platform. Training and validation data for the machine learning approach are obtained from the field measured inventory campaigns to evaluate the modeled forest biomass accuracies. The results show that variables from L-band (HH, HV), C-band (HV), and canopy height from the GEDI LiDAR platform performed effectively to model forest AGB with the coefficient of determination (R2) of 0.81 and root mean squared error (rmse) of 19.35 Mg/ha (%rmse – 17.17). In the case of single frequency SAR data, the analysis shows that the model derived from the L-band SAR data and LiDAR performed comparably better than the combination of C-band SAR and LiDAR data with an R2 of 0.78 and rmse of 21.36 Mg/ha (%rmse – 18.94). The results, thus, demonstrate the potential of SAR data (both single frequency and multiple frequencies) in combination with GEDI LiDAR data in effectively modeling AGB over highly biodiverse tropical forest regions.

Assessment of above ground biomass and soil organic carbon in the forests of Nepal under climate change scenario

“…our model for the estimation of AGTB is found strong, i.e., 0.74, which is higher than or similar to the other previous studies that used different predictor variables to predict AGTB using RFM (Powell et al, 2010;López-Serrano et al, 2020;Nguyen and Kappas, 2020;Li Z. et al, 2022). Similarly, the RMSE percent of the AGTB model in our study is slightly higher than the results reported by Musthafa and Singh (2022), Wai et al (2022) and slightly lower than result of Zhu et al (2020). These studies completely used other predictors (Image pixel value, age, crown density etc.)…”

“…The RFM used in this study helps understand AGTB as functions of predictors such as altitude and climatic variables. Previous studies also used RFM to estimate AGTB, but were confined to a few predictor variables such as image pixel value, canopy height, topography, vegetation indices, and texture feature (Li Z. et al, 2022;Musthafa and Singh, 2022;Wai et al, 2022).…”

Seasonal flooding provides limitations and opportunities for ecosystem carbon accounting from space

“…Major limitations in our ability to scale these data have been the lack of consistent data collection methods between surveys and an inability to sample many locations. Light detection and ranging (lidar) instruments are recognized as among the most reliable technologies for mapping the three-dimensional structure of ecosystem topography and composition [11,12] and while they have excelled at regional, site and footprint scales they have not been able to provide the geographic coverage needed to derive global scale information with low uncertainty [13]. However, we are now entering an era where consistently collected lidar data is readily available.…”

Radar and multispectral remote sensing data accurately estimate vegetation vertical structure diversity as a fire resilience indicator

“…In the tundra biome, which features typically low biomass values and where the role of wildfires in shaping ecosystem structure and dynamics has been underestimated Bartsch et al (2020) and Jones et al (2013) evidenced the ability of a Sentinel-1 and Sentinel-2 data fusion approach to retrieve tundra vegetation structural characteristics. Conversely, it is expected that backscatter saturation at short SAR wavelengths in biomes with high canopy closure and biomass, such as tropical forests (e.g., Englhart et al, 2011;Huang et al, 2018;Musthafa & Singh, 2022), do not allow to evaluate resilience to fire in terms of vegetation structural complexity through the proposed approach. Future research should consider the use of SAR sensors with longer wavelengths such as PALSAR-2 L-band SAR on-board ALOS-2 satellite (JAXA, 2022) or the P-band SAR instrument of the future Biomass mission (ESA, 2022e), which would penetrate the canopy to a higher extent in both unburned and postfire scenarios with strong vegetation responses (Kasischke et al, 2007;Tanase et al, 2010).…”

“…Thus, the fusion of optical and SAR data would provide complementary information on the vegetation vertical structure (Lu et al, 2016) and reduce the soil background influence on the retrieval of structural parameters compared with the individual use of SAR images (Wang et al, 2019). This approach has been shown to improve the estimation of structural parameters such as leaf area index or aboveground biomass worldwide (e.g., Huang et al, 2016;Montesano et al, 2013;Naidoo et al, 2019), but, as with SAR data alone, there are no studies up to date that exploit this approach for mapping vertical structural diversity and vegetation resilience in burned areas.…”