REMOTE-SENSING OF BUSH FIRE DYNAMICS IN CENTRAL-AFRICA FROM 1984 TO 1988 - ANALYSIS IN RELATION TO REGIONAL VEGETATION AND PLUVIOMETRIC PATTERNS

Spatial and temporal subcontinental patterns of are distribution in Af rica, may vary strongly from year to year. These variations may be of climatic or anthropic origin. Global Area Coverage (GAG) data, from th e Advanced Very High Resolution Radiometer (AVHRR), are known both to provide a good description of fire calendar and on a continental scale , a good indicator of locations of intense fire activity. In fact, the GAC data, available on a daily basis back to July 1981, is the only s ource of information actually available to determine burning patterns both on a continental scale and over long periods of time. This study presents the spatio-temporal distribution of vegetation fire in Centra l Africa, as derived from GAC imagery, over the four 1984/85 to 1987/8 8 dry seasons of the Northern Hemisphere (November to February periods ). The analysis of the mean fire distribution patterns over the four d ry seasons, as defined by vegetation cover and pluviometric conditions , shows a north to south gradient of fire activity in the savannah dom ain: This gradient of fire activity has been previously related to the annual pluviometric calendar and the amount of rainfall. The strong i nterannual variability in fire activity observed for this four-years-p eriod could not, however, be explained by the monthly pluviometric pat terns. The observed fire gradient therefore seems to result from addit ional meteorological or anthropic parameters. The methodology used in this study demonstrates the potential of GAC data for fire monitoring at a subcontinental scale. Current work is extending the area of inter est to the entire African continent using the 10-yr-period of GAC imag ery.