-band
rea of
y ‘of
same
OAL
1 coal
ound.
ution,
ment.
jobs;
viable
IAPRS & SIS, Vol.34, Part 7, “Resource and Environmental Monitoring”, Hyderabad, India, 2002
15.8°C to 19.2°C
1 19.2°C to 21.2°C
21.2°C to 256°C
25.6°C to 26.0°C
26.0°C to 27.3°C
27.3°C to 31.6°C
Figure 7. Estimated temperatures corresponding to thermal anomalies on Landsat
TM-6 image, Jharia Coal field, India (Courtesy: A. Prakash)
Remote sensing delineation of objects such as volcanic vents
and fires is based on thermal anomalies associated with them.
For temperature estimation, radiation intensity emitted from the
target (heat source) is measured by a remote sensor. Planck's
radiation equation is then used to convert the measured spectral
radiance to radiant temperature and then to kinetic temperature.
3.2 Methodology
The procedure of temperature estimation involves the following
main steps: determination of emitted radiation for each pixel;
subtraction of radiation from other sources, such as solar
reflected radiation, atmospheric scattering etc; conversion of
corrected DN values into emitted radiance; and finally
conversion of emitted spectral radiance into radiant temperature
values.
The sensors used for measuring thermally emitted radiance
include the aerial sensors operating in TIR and SWIR region,
and satellite sensors such as Landsat TM, ETM+, JERS-OPS
etc. For example, figure 6 shows the temperature sensitivity of
Landsat TM spectral bands.
The tools and methodology of temperature estimation depend
upon the range of temperature anomaly. In this context two
broad types can be distinguished: (a) buried hot features and (b)
surface hot features.
3.2.1 Buried hot features: Some of the thermal sources are at a
certain depth, e.g. subsurface fires, molten lava at depth etc.
The surface temperature of the ground above buried hot features
is generally quite low, owing to the low thermal conductivity of
rocks such as sandstone, shale, coal etc. The thermal IR band
data is best suited for sensing this order of temperature.
Figure 7 is an example of the measurement of temperatures
from thermal anomalies associated with subsurface fires in a
coal field.
3.2.2 Surface hot features: Volcanic vents with molten
magma, lava flows and surface coal fires are characterized by
high temperatures, reaching upto about 800°-1000°C. Features
with such high temperatures emit radiation also in SWIR region
(1.0-3.0u m), as indicated by Planck’s law. Therefore, although,
the SWIR region is generally regarded as suitable for studying
reflectance properties of vegetation, soils and rocks, it can also
be used for studying high-temperature surface features.
Further, the fire or volcanic vent need not occupy the whole of
the pixel; therefore, the temperature integrated over the entire
pixel, would be generally less than the fire or vent temperature.
As sensors in the SWIR bands (e.g. Landsat TM; JERS-OPS;
ASTER) have the capability to measure this range of
temperatures, their data can be used for studying surface hot
features.
For example, in a coal fire area, the pixel-integrated
temperatures for surface fires have been found to range between
217°C to 410?C. Figure 8a shows a Landsat TM image
exhibiting a thermal anomaly within a crater on Lascar, Chile,
with pixel-integrated temperatures shown in Figure 8b.
3.2.3 Sub-pixel temperature estimation: An active lava flow
will consist of hot, molten material in cracks, surrounded by
chilled crust. Similarly, in a coal-fire area, within a pixel only a
part is filled with surface fire. Therefore, thermally, the source
pixel will be made up of two distinct surface components: a) a
hot component (occupying fraction 'p' of the pixel) and b) a
cool component which will occupy the remaining (1-p) part of
the pixel. Using the dual-band method (Matson and Dozier
1981), the temperature and size of these two sub-pixel heat
sources can be calculated (Rothery et al., 1988; Oppenheimer,
1991; Prakash and Gupta, 1999)
Thus, remote sensing using SWIR bands can generate data for
understanding the volcanism cooling of lava flows and
473
