GLAS-based Estimates of Biomass and Carbon in Siberia and Quebec
Ross
F.
Nelson, NASA-GSFC, ross.f.nelson@nasa.gov
(Presenting)
The primary objective of these two studies, one in south-central Siberia (811,414 km2) and the second involving the Province of Québec (1.27 million km2), was to assess the utility of the GLAS waveform measurements for estimating regional forest biomass and carbon. Both studies employed GLAS as a sampling tool to attribute digital land cover maps based on optical satellite imagery. The studies differ in the techniques used to create the equations used to predict forest biomass based on GLAS waveform measurements of forest structure. The Siberian investigators directly measured 51 forest plots illuminated by GLAS and used a neural network to predict timber volume and biomass as a function of various GLAS waveform measures. The Québec researchers used an airborne profiling LiDAR to tie tree measurements made on Provincial ground plots to ICESat/GLAS metrics. In Québec, the airborne profiling LiDAR was flown over 300 ground plots previously measured by the Provincial forest service, and also flown along 5000 km of GLAS orbital ground tracks. Siberian GLAS estimates of merchantable volume agree with ground estimates within 1% when GLAS shots on slopes ≤ 10° are considered. Ground vs MODIS/GLAS estimates increase to +6.0% when GLAS shots on all slopes are considered. In Québec, GLAS estimates of dry biomass differ from ground-based estimates by 6 – 26% at the cover type level within three southern ecozones, i.e., the Northern Temperate, Mixedwood, and Southern Boreal ecozones. Across all three southern ecozones, ground-GLAS estimates differ by 7%. Considering the entire Province including the 3 southern ecozones plus the Northern Boreal, Taiga, and Treed Tundra ecozones, dry biomass stocks amount to 5.04 ± 0.40 Gt, carbon, 2.52 ± 0.20 Gt (CV = 7.9%).
NASA Carbon Cycle & Ecosystems Active Awards Represented by this Poster: