Geology Reference
In-Depth Information
review by Rathmore and Wright (1993) suggests that the TIR sensor works best below an altitude of 3 km (~9000 ft)
above average ground level. Rathmore and Wright (1993) also suggest that a spatial resolution of less than 3m works
best for detecting coal fires. However, part of the problem with these early surveys is that the TIR scanners used in
many of the studies did not have the resolution and temperature sensitivity needed to detect small or subtle anomalies.
It is also possible that many of the early TIR surveys were not conducted at the right time of day or during the right
season to reduce the effects of solar heating and other surface factors that can create false anomalies.
Later Use
O
ne of the characteristics of the later remote sensing studies of coal fires has been the use of several different
remote sensing methods in combination. This has given the researcher(s) different ways to look at the coal fires and
thus provided more information. One of the first studies to combine remote sensing methods was by Yong-fang
et al. (1990) to study coal fires in areas along the
In this area of Western China, some coal fires had
apparently been burning for more than 1000 years (Yong-fang et al. 1990). These researchers used a Daedalus 1230
model infrared scanner, which would have had an IFOVof 1milliradian in combination with CIR aerial photos to
locate and determine the direction of movement of coal fires (Yong-fang et al., 1990).
“
Silk Road.
”
Another early study to use TIR to go beyond the location and delineation of coal fires was a remote sensing study
by Shallenberger (1993) at the Centralia coal fire in northeastern Pennsylvania. He looked at imagery from
Landsat, Skylab, U2 aircraft, C-130 aircraft, and other aircraft mounted thermal infrared imagery. Shallenberger
(1993) used TIR imagery from surveys conducted in 1971, 1980, and 1982 to determine the progression of coal fire
at Centralia and found that the repeated use of airborne TIR surveys were the most useful remote sensing method
for determining the advance of the coal fire.
Shallenberger (1993) found geobotanical anomalies present in CIR imagery at Centralia (i.e., areas where the
vegetation did not show the normal healthy red color). He used National Aerial Photographic Program CIR photos
(1:24 000 scale) that had been taken in August of 1987 for his Centralia study. He noted that the distinctive thermal
and geobotanical anomalies were evident only where the burning coal bed was at relatively shallow depths. This
observation was supported by a thermal model, which suggested that no detectable surface anomalies would occur
below a depth of ~30.5 m (100 ft) (Shallenberger, 1993).
Hong et al. (1996) briefly described using airborne TIR and Landsat TM band 6 to detect coal fires. They noted that
the time of day the data were acquired, the type of TIR sensor, the surface morphology of area, and the extent and
temperature of any surface anomalies could all affect the ability to detect coal fires. The authors noted that predawn
TIR surveys were best for detecting coal fires.
Peng et al. (1997) used both airborne TIR imagery and color infrared photographs to depict the fire front and
outcrop of coal seams. Then the location of the fire front and its direction of movement were combined with
topographic data using geographic information systems (GIS) to estimate the depth of the coal fire. Using these
techniques, they estimated the depth of several coal fires to range between 54 and 125m. Note that this depth is
greater than the 30 m depth that had been considered the lower limit for detection.
Zhang et al. (1999) used a combination of satellite and airborne remote sensing methods including TIR to detect coal
fires in northern China. These authors used high-resolution optical satellite imagery, satellite TIR data, and satellite-
based microwave data with airborne TIR imagery and aerial photography using computer programs to obtain better
results than could be obtained from one system alone. In addition, they were able to obtain a wider coverage using the
satellite imagery than would have been possible using airborne imagery alone. This was important because they were
working in a large area of northern China that was often remote and lacked ground access.
Kick et al. (2004) describe an experiment combining a TIR scanner and a Fourier transform spectrometer to detect
and analyze coal fires. As part of their approach, they fuse (i.e., combine the data from the two instruments
together).
Gielisch (2007) described the effort by the People
s Republic of China to control coal fires and noted that infrared
photography has been very important in that effort. He indicated that TIR surveys were necessary for locating and
defining the limits of coal fires.
'
