Satellite monitoring of vegetation fires for EXPRESSO: Outline of activityand relative importance of the study area in the global picture of biomassburning
Jm. Gregoire et al., Satellite monitoring of vegetation fires for EXPRESSO: Outline of activityand relative importance of the study area in the global picture of biomassburning, J GEO RES-A, 104(D23), 1999, pp. 30691-30699
The satellite monitoring of vegetation fires for the International Global A
tmospheric Chemistry (IGAC) Experiment for Regional Sources and Sinks of Ox
idants (EXPRESSO) was designed to assist the assessment of the fluxes of tr
ace gases and aerosols emitted by savanna fires that occur during the dry s
eason in Central Africa. It is of particular interest that the study area c
overs the transition zone between the savanna and the forest domains. Satel
lite remote sensing is the only technology that allows consistent data coll
ection of the spatial and temporal distribution of fires at the scale requi
red by the EXPRESSO experiment. A portable monitoring system was developed
and operated in Central Africa for the in situ, real-time acquisition and p
rocessing of National Oceanic and Atmospheric Administration-advanced very
high resolution radiometer (AVHRR) imagery during the EXPRESSO dry season f
ield campaign, which lasted from October 1, 1996, to December 2, 1996. One
AVHRR scene per day was processed to locate active fires and burnt areas. T
hese results are compared with our daily global maps of vegetation fires de
rived from the International Geosphere-Biosphere Programme-Data and Informa
tion System Global AVHRR data set for the period April 1992 to December 199
3. This comparison shows the EXPRESSO fires in the context of the African c
ontinent and of the total global fire activity. The EXPRESSO study area is
part of one of the largest fire belts observed in Africa, and even over the
globe, spreading from Senegal to Ethiopia. This study also shows a big inc
rease in the number of fires in 1996 compared to 1993, from a total of 39,5
00 fire pixels detected in November 1993 to 124,500 in November 1996. We al
so show that the November period corresponds to a minimum of the global fir
e activity: only 3% of the fires detected during a complete year were detec
ted during this month. Moreover, the number of fires detected in this parti
cular month was 28% of those detected during the annual peak period observe
d in July-August. Current activity is focusing on the implementation of the
World-Fire-Web network: A system for globally mapping vegetation fires.