Operational characterization of tropical forest change at high spatial resolution using a MODIS/Landsat data fusion approach in the Congo River Basin, Africa
Matthew
Hansen, GISCE, South Dakota State University, Matthew.Hansen@sdstate.edu
Erik
Lindquist, GISCE, South Dakota State University, Erik.Lindquist@sdstate.edu
(Presenting)
Systematic, unbiased characterization of tropical forest change is a goal of multiple international initiatives and is critical for modeling Earth’s biogeochemical cycles, monitoring habitat and biodiversity status and estimating anthropogenic effects on ecosystem services. The research proposed is a data fusion approach using moderate resolution MODIS imagery to normalize and train higher resolution, multi-temporal Landsat classifications across regional mosaics in the Congo Basin. The effects of sun-surface-sensor geometries on small field-of-view sensors will be addressed. Landsat SLC-off data will be included in the analysis to produce a mid-decadal estimate of tropical forest change. A suite of spatial statistics will be produced that provide a current baseline estimate of forest change characteristics and rates. The emphasis will be on minimal analyst input, building an automated and operational monitoring tool.