HR: 0830h
AN: B41C-0890    [PDF]
TI: Land-Cover Transitions as Predictors of Fire Activity in Amazonia
AU: * Cardoso, M
EM: manoel.cardoso@unh.edu
AF: Complex Systems Research Center, University of New Hampshire, Durham, NH 03824 United States
AU: Hurtt, G
AF: Complex Systems Research Center, University of New Hampshire, Durham, NH 03824 United States
AU: Salas, W
AF: Applied Geosolutions, 26 Mill Road, Durham, NH 03824 United States
AU: Hagen, S
AF: Complex Systems Research Center, University of New Hampshire, Durham, NH 03824 United States
AU: Fearon, M
AF: Complex Systems Research Center, University of New Hampshire, Durham, NH 03824 United States
AU: Keller, M
AF: Complex Systems Research Center, University of New Hampshire, Durham, NH 03824 United States
AU: Moore, B
AF: Complex Systems Research Center, University of New Hampshire, Durham, NH 03824 United States
AU: Nobre, C
AF: Brazilian Center for Weather Forecasts and Climate Studies, CPTEC-INPE, Cachoeira Paulista, SP 12630-000 Brazil
AB: Fires are of special interest in environmental studies. Because of their short time scale and strong links to biogeochemical cycles, fires can significantly affect many environmental characteristics, such as carbon stocks and fluxes, air composition, and land productivity. Fires happen both naturally and as a result of anthropogenic activities. In Amazonia, previous studies have identified climate, land-use and land-cover as major factors for explaining large-scale fire patterns throughout the basin. In this study, high-resolution land-cover transitions rates are analyzed together with information from satellite fire products for a study region in Rondonia. Land-cover transition rates were based on an analysis of sequential Landsat images, and included spatially resolved estimates for transitions between three main vegetation classes: forest, secondary, and clear. Satellite fire data were based on NOAA12/14, and GOES-8 as daily fire pixel maps. The results for 1995-96 show that the various transitions types differed in their utility as predictors for fire activity. In particular, correlations between fire activity and land-cover dynamics were highest for transitions from secondary to clear, forest to clear, and clear to clear. Transitions from forest to clear were related to more intense fire activity. While data from GOES-8 reported more fires and correlated better with changes in land-cover, fire products reported similar trends in fire activity. These results suggest that spatial/temporal data on land-cover transition rates may provide valuable information for improving fire models in the region.
DE: 0400 Biogeosciences
DE: 1615 Biogeochemical processes (4805)
DE: 1640 Remote sensing
SC: Biogeosciences [B]
MN: 2003 Fall Meeting