Forest Fire Monitoring with SPOT-4 Satellite Imagery
3.4 Image Used
The image used in our study is the one shown in figure 1. The image was acquired on 7 July 2000 with instrument HRVIR2. The absolute calibration coefficient A is 5.49300 and the programmable gain G is 1.5. The digital value X of the big fire on the top left corner (refer to figure 1) is about 252. With this value, the corresponding radiance is computed to be 30.58 W/m2/sr/mm.
3.5 First Attempt in Temperature Computation
We first assume that the emitted radiance Le dominates the total radiance. This is motivated by the fact that most of the active fires can easily have temperature up to a few hundred degrees Celsius. Under this circumstance, the emitted thermal radiance is larger than the reflected radiance by a few order of magnitude.
Therefore, the emitted
radiance is taken to be 30.58 W/m2/sr/mm. Since
the image was acquired using HRVIR2, the effective
wavelength
l
e is 1.636 mm. With these values, the temperature is readily be computed from the Planck's equation
The temperature computed is 331.34 K that is equal to 58.34 °C.
3.6 Second Attempt in Temperature Computation
In our first attempt, the temperature computed is merely 58 °C which is far below a few hundred °C. This renders our assumption that the emitted radiance dominates the total radiance invalid. Therefore, we need to estimate the reflected radiance Lr of the active fire so that we can subtract it from the total radiance.
In figure 5, the shadow as indicated has a relatively low radiance than the surrounding non-shadow area. Therefore, the total radiance in this area is dominated by the emitted thermal radiance. Hence, we can use it as a reference to estimate the reflected radiance of a similar area.
The total radiances of the two areas A and B as shown in figure 5 can be written as
Here we have assumed that reflected radiance of area B is negligible as compared to the emitted radiance. Since the two areas have similar ground characteristics, we further assume that the emitted radiances of the two areas are the same. Therefore, we can compute the reflected radiance of these areas by taking the difference between the two total radiances.
We find that the reflected radiance Lr of area A is about 9.30 W/m2/sr/mm. Hence we estimate that the emitted radiance Le of the fire C is 21.28 W/m2/sr/mm which is obtained by subtracting 9.3 from the total radiance of 30.58. With this value, the temperature of the active fire is found to be 53.34 °C.
The relatively low value of temperature may be due to the fact that in the 20x20m pixel only a small part of the area was actively burning. Therefore, the surface temperature computed is just the average temperature of the entire 20x20m area.
3. Conclusions
The temperature sensitive SWIR band has rendered SPOT-4 to be more effective and efficient in forest fires monitoring. Even small fires, which do not have smoke plumes at all, can be detected in SPOT-4's 432 representation. Another advantage of SPOT-4 is that the SWIR band is relatively unaffected by atmospheric effects, therefore fires under thin cloud or haze can still be detected. One setback of SPOT-4's 432 representation is that the smoke plumes tend to be too thin for the purpose of determining the size of the plumes. A new display scheme in which SWIR, (NIR+GREEN)/2, RED bands display in RGB channels respectively, are found to be more suitable in forest fires monitoring. Under this scheme, both fire and smoke plume are clearly visible. The total radiance of the SWIR band consists of the reflected and the emitted radiances. We found that by comparing two areas, one with shadow and another without, the reflected radiance of the active can be estimated. With this, the temperature of the active fire is found to be about 53 °C.
References
- S. C. Liew, O. K. Lim,
L. K. Kwoh and H. Lim, "A study of the 1997
forest fires in South East Asia using SPOT
quicklook mosaics", Proc. 1998 Int. Geosci.
Remote Sensing Symp., Vol. 2, 879-881, 1998.
- S. C. Liew, L. K. Kwoh,
K. Padmanabhan, O. K. Lim and H. Lim,
"Delineating land/forest fire burnt scars with
ERS interferometric synthetic aperture radar",.
Geophysical Research Letters, 26(16), 2409-2412.
- V. Trichon, D. Ducrot and J. P. Gastellu-Etchegorry, "SPOT4 potential for the monitoring of tropical vegetation. A case study on Sumatra", Int. J. Remote Sensing, 20(14), 2761-2785, 1999
