Carbon dynamics in northern forests using SIR-C/X-SAR imagery

A key large-scale question in ecology is the amount of carbon stored in terrestrial vegetation. The need to determine this over large regions is most efficiently met by remote sensing programs and is currently being tested using SIR-C/X-SAR imagery at the NASA Michigan Forests supersite. There are three goals in the present study. The first is to estimate carbon stored in living vegetation at the site. The second is to measure changes in carbon balance related to forest disturbance (clearcutting and thinning). The third is to determine carbon gain from annual growth (NPP). Biomass has been mapped in the image domain, and by applying an allometric relationship, results show that 12.82/spl times/10/sup 9/ kg of carbon are stored in the portion of the test site represented in the imagery. Clearcuts which were made between SRL-1 and SRL-2 have been extracted from the classified imagery and this analysis shows an estimated 300 ha of forest or 6.02/spl times/10/sup 6/ kg of carbon were removed. Ground data have been analyzed for annual growth; SIR-C/X-SAR images are currently being processed to demonstrate the sensor's sensitivity to this parameter.