Oral Session - Forests Chair(s): Gustafson, Eric ¹, ¹ Landscape Ecology Unit, Rhinelander, WI
Friday, April 2, 2004 3:00 PM - 5:20 PM Apollo Room 7

Modeling Forest Cover in Central America From 1880-2000 Using GIS. Cornell, Joseph^*,1, ¹ State University of New York, College of Environmental Science and Forestry, Syracuse, NY, USA

ABSTRACT- Changes in the distribution of forest cover in Central America from 1880 to 2000 were simulated in ten year intervals using spatial and nonspatial data. Spatial data included historical maps of forest cover for 1920 for all of Central America and for 1965 and 1982 for some countries as well as remote sensing data for the entire region for 1994 and 2000. Nonspatial data on the amount of forest cover in each country was obtained from the FAO online database and from extrapolation, yielding estimates for the entire 120-year period. A spatially-explicit model of Land-Use Change (LUC) called GEOMOD was used to convert these non-spatial estimates of forest cover into a time-series of maps showing the distribution of forest cover for each ten year period. GEOMOD simulates LUC by converting gridcells in an intial map of forest cover from forest to non-forest. GEOMOD requires non-spatial data on the amount of change that will occur during a given time period and a spatial image of forest cover at the beginning of the time period being simulated. GEOMOD uses "driver" maps to select gridcells for conversion. The driver maps used were maps of Holdridge life zones and of Parks and Protected Areas. Results were validated using the available spatial data. Model agreement with the spatial data ranged between 74 to 91 percent. The result of this study is a new time series of maps of forest cover for Central America for the entire 120-year period. These maps now can be used to inform other studies such as the role of tropical forests in the global carbon cycle. Preliminary analysis shows that between 1880 and 2000, about 7 Gt of carbon was released due to LUC. This loss represents about 60 percent of the total carbon storage potential of vegetation and soils in Central America.

KEY WORDS: land-use change modeling, central america, global carbon cycle, tropical deforestation, geomod