Using Sample Theory to Develop Forest Biomass Estimates from ICESat/GLAS Data

Sean P Healey, USDA Forest Service, seanhealey@fs.fed.us (Presenter)
Paul Patterson, USDA Forest Service, plpatterson@fs.fed.us
Sassan Saatchi, CALTECH/JPL, sassan.saatchi@jpl.nasa.gov
Michael Lefsky, Colorado State, lefsky@cnr.colostate.edu
Andrew Lister, USDA Forest Service, alister@fs.fed.us

Forest canopy height measurements developed from ICESat/GLAS data have shown excellent correspondence to independent field measurements of biomass. However, since the data are not synoptic, they alone cannot be used to estimate biomass for a particular area. One option is to treat the GLAS “shots” available for a landscape, or a subset of them, as a sample of that landscape. A GLAS-based sample could form the basis of statistical inference about population-level biomass levels – much in the way that the national forest inventory (US Forest Service FIA Program) makes official estimates about the country’s forest resources. We have developed specialized statistical estimators which can account for both sample uncertainty as well as the measurement error associated with deriving biomass measurements from GLAS. We have also devised a flexible sample framework which allows us to maximize the number of GLAS shots considered (thereby increasing precision) while satisfying critical statistical assumptions about the sample. This approach is being tested in the state of California as part of the NASA Carbon Monitoring System Biomass Pilot, and once validated, it should be applicable globally.

Presentation Type:  Poster

Session:  Science in Support of Decision Making   (Wed 10:00 AM)

Associated Project(s): 

• Healey, Sean: Assessing Potential Impacts of Ground Sample Bias in Global CMS Biomass Estimates, Now and in the DESDynI Era ...details

Poster Location ID: 165