Estimation of Forest Above-Ground Biomass with C-Band Scatterometer Backscatter Observations

摘要

C-band scatterometer data are among the longest time record of spaceborne observations, dating back to 1991. Their use in the context of terrestrial carbon cycle studies has not been investigated yet. In spite of a weak sensitivity of the C-band backscatter to forest above-ground biomass (AGB), repeated observations have been shown to support the retrieval of biomass. In this paper, we investigate the use of C-band backscatter data at 25 km spatial resolution from the ERS WindScat and MetOp ASCAT sensors arranged in daily mosaics to estimate forest AGB. A retrieval approach based on self-calibration and regression between SAR backscatter and canopy density was developed, producing daily estimates of AGB that were found to well reproduce the global spatial patterns of AGB. A weighted average of the daily AGB estimates was then applied to generate annual maps of AGB. The evaluation of the AGB dataset with recent global datasets of AGB for 2010 confirms the overall reliability of our estimates.