Abstract

In 1997 a drought commenced in Southeast Asia, this being directly related to the then ongoing El Niño‐Southern Oscillation (ENSO) event. Interaction between land clearance activities and this drought led to massive, uncontrolled vegetation fires that burned large areas of forest and agricultural land, most severely on the Indonesian island of Borneo. A similar situation in 1982–1983 led to the largest uncontrolled forest fire ever documented, damaging around 50,000 km2 of Borneo's forest. This paper investigates the extent to which nighttime advanced very high resolution radiometer (AVHRR) global area coverage (GAC) data can be used to detail the spatial and temporal evolution of these large‐scale Indonesian fire events, and particularly the 1997 activity. GAC data are a spatial subsample of the full resolution local area coverage (LAC) AVHRR data, but unlike LAC, they have been archived globally on a daily basis for almost 20 years. Despite the extreme subsampling involved in GAC data production, simulation modeling and analysis of real data indicates that surprisingly reliable fire statistics can be obtained for the 1997 Borneo fire event from GAC imagery. For three dates in October 1997 where coincident LAC and GAC data are available, once numerical adjustment for the GAC subsampling procedure is included, fire counts extracted from LAC and GAC data consistently agree to within 0.15–13%. Time series analysis of the 1997 GAC data indicates that fires began in July in south Kalimantan, principally at the interface between cleared, cultivated land and lowland rain forest. Fire activity moved generally southward and peaked in September 1997. GAC data show the equivalent of 9733 LAC pixels as containing one or more active fires on 2 September. Fire activity declined significantly in November owing to the onset of the monsoon rains, with all fires ceasing by December 1997. Analysis of the cumulative fire map distinguishes peat swamp forest as the most severely affected ecosystem on Borneo, with >20% of this land cover category being identified as directly impacted by active fires. If the performance of GAC data can be shown to extend to other El Niño years and also to less intense periods of burning, then the GAC archive offers a tool for analyzing long‐term changes in the spatiotemporal pattern of fire activity in the region. Any trend can then be investigated for its relation to variations in agriculture, El Niño‐related climate, and other phenomena relevant to regional change in Southeast Asia.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call