Burned area in Kalimantan, Indonesia mapped with NOAA-AVHRR and Landsat TM imagery

D. O. Fuller, M. Fulk

Research output: Contribution to journalArticlepeer-review

42 Scopus citations


Fire scars in Kalimantan, Indonesia were mapped using optical and thermal data from the NOAA-AVHRR satellite sensor. Forty-four high-resolution picture transmission images mapped at 1 km spatial resolution were selected for a period immediately after intense biomass burning from October-December 1997. Composite images of the Normalized Difference Vegetation Index (NDVI), surface temperature (ST), and albedo were used to examine the spatial distribution of fire scars. Owing to clouds and smoke, 10-week composites of surface temperature, NDVI and albedo were needed to reduce atmospheric effects and a range of different thresholds was applied to the composites to form sets of decision rules involving NDVI, ST, and albedo. The results of each AVHRR fire product were checked against a fire-scar map derived from a supervised classification of a Landsat TM scene. An overall classification accuracy of approximately 80% (K = 0.51) was achieved with criteria STa > 305 and NDVIa < 0.35, where STa was the surface temperature and NDVIa was the NDVI at the time of minimum albedo. Application of this decision rule to the portion of Kalimantan within the AVHRR window revealed a maximum of approximately 13.18 million ha of Kalimantan, or about 25% of the study area, burned during the latter half of 1997. The results suggest that multiple thresholds of NDVI, ST, and albedo may be used to map burned area in tropical environments similar to those of Kalimantan.

Original languageEnglish (US)
Pages (from-to)691-697
Number of pages7
JournalInternational Journal of Remote Sensing
Issue number4
StatePublished - 2001
Externally publishedYes

ASJC Scopus subject areas

  • Earth and Planetary Sciences(all)


Dive into the research topics of 'Burned area in Kalimantan, Indonesia mapped with NOAA-AVHRR and Landsat TM imagery'. Together they form a unique fingerprint.

Cite this