A GUIDE TO HOT SPOTS AND FOREST FIRES IN SUMATRA
What is a hot spot? The NOAA satellites carry a thermal detector that measures the average surface temperature of each square kilometre in a strip below the satellite. This temperature information is transmitted to a ground receiving station (e.g. FFPCP, Palembang) as the data are being acquired by the satellite. The receiving station can then view by computer a thermal image of the area scanned by the satellite.
In the case of Sumatra, with an area of about 474,000 km2, this means that a similar number of temperature values can be extracted from the image for analysis.
The ambient temperature of the earth's surface is about 300°K (°K=°C+273.15). To detect fires each km2 is checked by computer processing to see if the recorded temperature is significantly above this value. A typical threshold value set for fire detection is 320°K and all the sampled 1 km2 areas with temperatures above this value are selected. The location of each area is then plotted to produce a 'hot spot map'.
Are all hot spots fires? Yes, after sensible data processing and interpretation. The NOAA sensor is sensitive to both thermal radiation from the ground and reflected sunlight. Hot spots caused by solar reflection (e.g. from clouds or water) are classified as false fires and edited out. Satellite data collected during the night give the most reliable information on vegetation fires that might be burning uncontrollably. These are usually called wildfires.
Can the number of fires burning be determined precisely from NOAA data? No. The NOAA sensor integrates all radiation emitted from each km2 sampled, and this may come from one or several fires to give an underestimate of fire numbers. But a small hot fire can produce more than one hot spot in a contiguous array (possibly owing to heated gases and suspended particles from the fire emitting additional radiation) to produce an overestimate. One hot spot does not therefore equal one fire but is normally a fair approximation.
Does NOAA detect all fires occurring at the time of the satellite pass? No. Some cloud cover over Sumatra is normal and thick cloud is opaque to thermal radiation. Fires are detected through thin cloud (or smoke haze) but energy absorption and scattering reduce the sensed temperature which may result in their being rejected as hot spots during processing. Ground fires where subsurface peat is burning and surface flames are lacking are unlikely to be detected. However, if sufficient smoke is generated from a ground fire, this is detected using data from one of NOAA's visible channels.
Does NOAA detect small fires? Fortunately, no! Flaming vegetation fires with a fire front exceeding about 50 metres can be detected. Extremely hot fires such as oil-field gas flares of about 20 metres dimension are detected: there are currently three of these regularly detected in Sumatra - one in Jambi, two in South Sumatra.
Are most hot spots forest fires? Definitely not. The natural forests of Sumatra are prone to burn only after damage by man. Of the millions of hectares of Indonesia damaged by fire in 1997-8, less than 3% was estimated to be primary forest and then only the edges. Less than 20% was in secondary forest/woodland. Most of the fires burn bushland/grassland and areas deliberately cut in preparation for burning. The great majority of fires were associated with agriculture and land clearing operations for tree crops. Since November 1997, there have been no forest fires in Sumatra. However, there is a continuous loss of forest in Sumatra and fire is normally used in this process, but only after the forest has been destroyed by land clearing (clear-felling) to make way, typically, for oil-palm.
For technical descriptions of the NOAA satellite fire detection system go to:
in Indonesia: operating procedures for the NOAA-GIS station in Palembang,