Mapping and Scaling of In Situ Above Ground Biomass to Regional Extent With SAR in the Great Slave Region
Abstract Global forests are increasingly threatened by disturbance events such as wildfire. Spaceborne Synthetic Aperture Radar (SAR) missions at L‐ (or P‐) band, such as the upcoming NASA ISRO SAR (NISAR), have great potential to advance global mapping of above‐ground biomass (AGB). AGB mapping wit...
| Published in: | Earth and Space Science |
|---|---|
| Main Authors: | , , , , |
| Format: | Article in Journal/Newspaper |
| Language: | English |
| Published: |
American Geophysical Union (AGU)
2022
|
| Subjects: | |
| Online Access: | https://doi.org/10.1029/2022EA002431 https://doaj.org/article/421348a508cc489d9edc8451f9b313fe |
| Summary: | Abstract Global forests are increasingly threatened by disturbance events such as wildfire. Spaceborne Synthetic Aperture Radar (SAR) missions at L‐ (or P‐) band, such as the upcoming NASA ISRO SAR (NISAR), have great potential to advance global mapping of above‐ground biomass (AGB). AGB mapping with SAR is challenging due to lack of available L‐ or P‐ band data, and because SAR data are sensitive to confounding factors such as hydrology and terrain. This study uses recently collected AGB validation site data (AGBV) to produce a 1 ha biomass map about the Great Slave Lake in Canada using SAR data, and reports on NISAR's anticipated performance. This study addresses errors inherent to the representativeness of AGBVs with coarser grid/landscape scale processes by evaluating model performance as data are aggregated over increasingly larger areas (AOAs). Air and spaceborne SAR data were found to be interoperable after processing them according to analysis ready data specifications, improving data availability. Owing to poor model performance at two AGBVs, root‐mean‐square errors (RMSEs) were ∼60 Mg/ha, irrespective of AOA. When instead using NISAR's more lenient assessment criteria, RMSEs decreased to 32, 15, and 21 Mg/ha for the small (∼0.1 ha), medium (∼3.5 ha), and large (∼14 ha) AOA. Thus, AGB mapping in this region appears to benefit significantly from coarser data aggregations than to be used by NISAR's. This approach is practical for identifying a suitable scale of correspondence between the AGBV and SAR data and the landscape‐scale processes and can substantially improve AGB mapping accuracy. |
|---|