Forest Fire Monitoring and Research Activities at the Centre for Remote Imaging, Sensing and Processing (CRISP), National University of Singapore
(IFFN No. 19 – September 1998,p. 16-20)
CRISP’s Remote Sensing Ground Station
The Centre for Remote Imaging, Sensing and Processing (CRISP), National University of Singapore, operates a remote sensing ground station since September 1995 (Fig.1). It has been receiving data from the SPOT, ERS and RADARSAT satellites. The reception circle of CRISP’s ground station for these satellites is as shown in Figure 2. To date, about 155,000 scenes of SPOT, 13,000 scenes of ERS and 6,000 scenes of RADARSAT images have been archived. A World Wide Web catalogue-browse system (http://www.crisp.nus.edu.sg) allows users to search the complete archive, with location maps and quicklook images available online.
Fig.1. and 2. The antenna of the remote sensing ground station of CRISP (upper) receives satellite data within the geographic circle shown on the map (lower).
Research Activities at CRISP
CRISP conducts research in three areas: coastal and ocean studies, tropical vegetation studies, and remote sensing data processing techniques. CRISP maintains fruitful collaborative projects with international research organisations, e.g., tropical forest studies with CESBIO, internal wave and other studies with University of Hamburg, paddy field studies with IRRI and University of Can Tho, etc. CRISP is also the principal investigator for two ADEOS projects with partial funding from NASDA.
CRISP has also been actively using the remote sensing facilities for environmental monitoring. In 1996, it captured an ERS image of a ship discharging pollution into the ocean. With the help of this image, the culprits were convicted in court, and this image has become the first ever remote sensing image to have been accepted as court evidence. CRISP is now collaborating with the Maritime and Port Authority of Singapore in an project of oil slick monitoring with synthetic aperture radar remote sensing images. Since last year, CRISP has closely monitored the fires raging across the forests and plantations of the Southeast Asian region. With the high resolution SPOT images, CRISP was able to determine accurately the location of fires and to help determine whether the fires were associated with human activities. In collaboration with Singapore Ministry of the Environment, CRISP has implemented a daily fire monitoring operation. A careful analysis of the fires of 1997 is being carried out using imagery from SPOT, ERS and NOAA satellites. With the SPOT quicklook images, preliminary maps of fire-affected areas have been derived.
Forest Fire Monitoring and Research
1. Burnt area assessment using SPOT quicklook mosaics
The burnt areas in Sumatra, Kalimantan and Sulawesi have been mapped using mosaics of SPOT quicklook images before (January- June 1997) and during (August-December 1997) the 1997 fire episode, in order to estimate the total area burnt and the types of land cover on fire. Figure 3 shows the map for Sumatra with about 1.5 million hectares affected by the 1997 fires and Figure 4 shows the map for Kalimantan with about 3.0 million hectares affected. In Sumatra, the peat swamp areas stretching from the northern coast of South Sumatra to Jambi Province were observed to be affected by fires. In Kalimantan, the main fire activities occurred mainly on the Southeast and Southwest corners of the Borneo island. Most of the land affected by fires were associated with swamp forests and plantation/agricultural areas.
2. Fire monitoring operation
A daily fire monitoring operation is being implemented at CRISP, in collaboration with the Ministry of Environment, Singapore. SPOT images over hot-spot areas in Southeast Asia are analysed on reception and reports and annotated images of fires are transmitted to the Ministry with minimum delay. The high resolution SPOT images are visually inspected to detect smoke plumes and burnt scars associated with fires, to provide the exact locations of fires and to assess the extent of fires and types of land cover on fires.
Fig.3 and 4. Maps of Sumatra (upper) and Kalimantan (lower) showing the area burned by the 1997 fires of ca. 1.5 million ha and 3.0 million ha respectively.
Fig.5 and 6. Interferometric coherence image of an area north of Banjarmasin, South Kalimantan, derived from the ERS-1/2 tandem SAR data on 10-11 April 1996 (upper). The lower image shows the same area derived from the SAR data acquired on 22-23 October 1997. In these two images, vegetated areas (low coherence) have dark tones while the bright areas are land clearings. Areas marked A and B were vegetated in April 1996 but were cleared in October 1997 by fires.
3. Change detection using ERS interferometric synthetic aperture radar images
Interferometric coherence and radar backscattering amplitude maps of selected fires affected areas in Kalimantan have been processed from ERS synthetic aperture radar (SAR) data acquired during the ERS-1/2 tandem missions in April/May 1996 and October 1997. Burnt areas can be correlated to the areas with low coherence in 1996 but an increased coherence in 1997. Figure 5 is the interferometric coherence image of an area north of Banjarmasin, South Kalimantan, derived from the ERS-1/2 tandem SAR data on 10-11 April 1996. This area is known to be severely affected by fires from observation using AVHRR and SPOT imagery. Figure 6 is the coherence image of the same area derived from the SAR data acquired on 22-23 October 1997. In these two images, vegetated areas (low coherence) have dark tones while the bright areas are land clearings. Areas marked A and B were vegetated in April 1996 but were cleared in October 1997 by fires. The linear features visible in the 1997 image are indicative of plantation/agricultural land use. Area C has already been cleared in 1996 while vegetation in area D has still not been cleared in 1997. Unlike optical imagery, SAR imagery is not affected by clouds or haze. Hence, SAR data can be used to complement optical images in mapping fire affected areas. A project to systematically map out the burnt area using coherence change over the whole southern coast of Kalimantan is in progress.
Hock Lim and Soo Chin Liew
Centre for Remote Imaging, Sensing and Processing (CRISP)
National University of Singapore
Lower Kent Ridge Road