Application of Landsat-TM
Thermal Band and IRS-1A LISS II Imagery in Delineation of Coal Mine Fire
in Jharia Coal Field V. K.
Srivastava Abstract Remote Sensing Unit Dept. of Applied Geophysics Indian School of Mines, Dhanbad -826004 In this paper an attempt has been made to study and interpret the thermal band image data of Landsat-TM (day time) in conjunction with IRS-1A LISS II colour composite images of famous Jharia Coalfield (Dhanbad) India in delineation of high heat areas due to subsurface and surface coal mine fires. This high heat areas map obtained from remote sensing images have been finally compared with known fire areas map. The study has brounth out a fair correlation with know fire spots in the region, however, the affected area has since increased. Also in order to be more successful in delineation of coal mine fire through satellite remote sensing images one need to study visible and near infrared images in addition to thermal band imagery of Landsat-TM followed by high resolution aerial thermal survey. 1. Introduction The Jharia coalfield which is the only coking coal source of India is located in the district of Dhanbad (Bihar) about w60 km NW of Calcutta. The southern edge of this coalfield is marked by perennial river Damodar. Coal min in this region was started as early as in 1890 providing a long span of hundred years of exploitation. There are 28-major coal seams. 19-in Barakar formations (the older stratigraphic horizon) and 9-in the Ranijanj formation (the younger stratigraphic horizon). Fires in cola seams of Jharia coalfield have been originated basically from spontaneous combustion occurring either underground or along the outcrops, and are restricated in Barakar formations with shallow depth of less than 40m. Mainly top seams which are thick and therefore more prone to spontaneous heating fires have also been caused due to burning of bantulsi, dumping of not ash in goafed out areas, illicit distillation in aabondoned working etc. There are about 20 fires spots covering an areas of 17.35 sq.k. and this coal mine fires is causing colossal loss of the country's valuable cooking coal reserved and in addition passing serious Environmental hazard by way of degradation land, damage to settlement, soil and vegetation cover. In order to save this valuable loss of coal reserve and also to minimise resulting environmental hazard, the current status of cola mining fire and affected areas should be known and mapped. Conventional methods for locating such fires are not so effective due to involvement of time consuming processes and the phenomenon being the dynamic one but the modern technique of remote sensing which provides image with synoptic view of whole project area in real time and in multispectral modes including thermal infrared region has been proved to be successful in such study. Particularly with the advancement of sensor technology like optomechanical scanner system, it is possible to map thermal variation of low order over earth's surface. The interpretation of these data can range from direct visual examination of photographic recording of the measured signal based on photo-interpretation technique to sophisticated digital processing using modeling analysis and pattern recognition techniques. However these investigations are limited by the complexities of the problem both in terms of physical phenomena as well as the number of different factors that influence the result. Therefore, most of the interpretation in such studies are based on quite simple theoretical models involving very limited assumptions and farley ideal circumstances of meteorological and geological setup. The use of thermal infrared imagery to detect subsurface coal fires is outlined in papers by slavecki (1964), Knuth (1968) and green & Moxhani (1968). In a study of 22-coal mine fires in Pennylvania, Green (1969) found that fires less than 10m depth were easily detected on thermal infrared imagery, fires between 10 and 30m in depth were detected only when heat was carried to the surface by convection in open cracks or if the fire had burned long enough (several years or more) to permit heat to reach the surface by conduction. Success through aerial thermal sensor has also been achieved in delineating and monitoring world wide high heat areas of effusive volcanism such as in Hawai volcanism such as in Hawai Island and in Italy; and areas of steaming altered Ground and hot spring activity in the western U.S. and areas of high heat zone in coal mining area (Jharia, Dhanbad) by various workers. Bhatacharya (1992, 1995) have also used Landsat-TM band imagery in delineating coal mine fire in Jharia Coalfied (India) by integrating ground truths collected through GPS. Shillin, Gornyi and Ermolaev Master (1987) have successfully used thermal aerial video photo images in dileneating coal mine fire area associated with coal seams outcrop in Mukunda open cast Project of Jharia Coalfied. For this same area Mansoor, Cracknell of U.K. nad Shilin Gornyi of Russia (1994) jointly published their integrated study using NOAA-AVHRR data and Landsat-TM data in delineating coal mine fire affected area in the region. They have shown that Thematic Mapper band-5 and 7 covering SWIR have been useful in locating burning areas where as thermal infrared band 6 has been useful in separating thermal areas from back ground solar warming region. However, they are of the opinion that the coarse resolution of Landsat-TM provides picture in gross sense. In the present study mapping of coal mine fire of JCF has been taken through comparing information's of standard false colour composites images of IRS -A (1988) and Thermal band images data of Landsat-TM, of (May 1987). 2. Study Area Jharia coalfield which is a famous coalfield of Dhanbad is about 40m in length in widths in exposed portion and in width in exposed portion and in width in exposed portion and stretches from west to east in the shape of sickle. The landscape of the area is characterised by undulating rocky and gritty surface with thin venner of insitu-soil supporting thin and sporadic vegetation. Undulating terrain with flat vallied rainfed ephemeral streams are common features. Seasonal crop is grown over valley filled alluvium/colluvium soil. Mining quarries, mine waste dump, subsidence of land surface, settlement for mining activities etc. are very common in the region. Lower Gondana sedimentary roacks surrounded by Archaean metamorphic and granitic rocks constitute the general geology of the area. Location is shown in fig. 1. 3. Data and methodology Landsat-TM image data band-6 (day pass of path-140 and row 044) of 1987 along with F.C.C. of IRS -1A LISS II of 1988 were analysed and interpreted with the hypothesis that the areas above subsurface coal mine fire gets heated by conduction which will be seen as bright areas in thermal band images and will show light yellow tone in F.C.C. the image data were processed for radiometric and geometric corrections using the standard digital processed for radimetric and geometric corrections using the standard digital processing techniques at R.R.S.S.C., Kharagpur. Wide range of digital enhancement techniques were employed but linear contrast enhancement of the image data with density sliced colour coded map were found to be of much use. After the study of false colour composite and Thermal band image data a heat zonation map of Jharia coalifield covering 23o43'N-33o51'N lat to 86o9E to 86o27'E long has been prepared and compared with known fire spots as shown by open circle in Fig.2. Further the affected area of high heat zone has been calculated using PLACOM digital planimeter. Ground checking and measurements of radiant temperature were carried out at places using Telatemp thermal radiometer as supplied by Space Application Centre, Ahmedabad (on loadn). Radiant temp(Trad) measured by the radimeter were converted into kinetic temperature (Tkin) using the formula as given below: Where 'E' emvissivity of the material and Tkin is the concern traction of the kinetic heat of a body of material which is measured with a thermometer placed in direct contact with the material. The temperature shown in the fig. 2. Is the average temperature measured by the radiometer. On high heat areas. 4. Results and discussions Grey level map mainly based on thermal band imagery of Landsat-TM and F.C.C. of IRS-1A LISS II covering Jharia coalfield is shown in Fig.2. Areas of high heat which are coraborating with known fire spots have come out clearly in this map. oN a single channel thermal infrared B & W image, surface temperature is xpressed by the grey tone and so cool areas have dark tones and warmer areas appear light. Thus black and white thermal IR image shows fire areas as bright white to greyish white depending on depth and intensity of the fire. Where as in false colour composite (2,3,5,4 for IRS -LISS II), high heat area correspondingly is shown by light yellow tone due to Parching of surface soil. Though this map show some anomalous heat zone where no known fire exist particular around Baghmara and Mahuda area. However, Mahuda, has previous history of fire but pressently no fire exists. High heat area in and around Baghmara may be in and around Baghmara may be due to heating of industrial installation in day time. Tundu lead smelter plant in Baghmara may be reflecting this heat anomaly. However, in order to separate out such anomolous heat zone due to coal mine fire and industrial area, the night time long wave length data is required to be studied and analysed and also a detail knowledge of depth of fire and mode of occurrence need to be studied for accurate determination of coal min fire areas. Fire area centered around Bhowra, Bhulan Bararee, Lodna, Kustore, Kusunda and Jogta have come out very nicely. Intensity and extent of fires in the western margin (Block II, Nadkharki) of the basin is much more sever than what is depicted on the existing available map. It was observed that the available fire position map as provided by the coal Mining co. does not show current status of fire. Surface or near surface fires are usually large in area extent and have irregular boundary shapes in bright white tone while may underground fires lead to the development of second dray cracks through which heat second type. The third type of coal fire i.e. burning dumps can be easily identified based on their typical speckle or intergrowth texture or curved boundary lines. Many such dumps are usually fire prone and tend to have temperatures higher than have temperatures higher than surrounding normal ground and can be identified and mapped. Study discussed in this paper indicates that that the fires burning at shallow to intermediate depths (less than 30 meters) can be detected more easily as compared to deep fires where surface main festations are few and also conductive heat transfer from the fire source may take many years before it can be noticeably detected on the surface by remote sensing. 5. Conclusions From the study it has been found that analyses of thermal band imagery of Landsat-TM LISS II data have been of much use in delineating high heat areas in coal minin effected region of Jaharia coalfied (Dhanbad). The Affected area appears to be more than the reported one. However, the coarse resolution of thermal band (120m x 120m) has given a regional thermal picture of the area but for detail study including depth estimation, high resolution aerial thermal survey is to be carried out. It is concluded that in order to delineate fire affected region more effectively colour composite images in visible & near infrared are also needed to be studied alongwith thermal band imagery of Landsat-TM. Acknowledgement The author is grateful to Professor D.K. Paul, Director of Indian School of Mines, Dhanbad (Bihar) for encouragement and Institutional support for completing the present work. References
|