Mapping of Coal Fire in Jharia Coalfield, India: a Remote Sensing Based Approach
Total Page：16
File Type：pdf, Size：1020Kb
J Indian Soc Remote Sens DOI 10.1007/s12524-016-0590-5 SHORT NOTE Mapping of Coal Fire in Jharia Coalfield, India: a Remote Sensing Based Approach Anjali Singh1 & Ashwani Raju2 & Pitambar Pati3 & Narendra Kumar1 Received: 28 October 2015 /Accepted: 25 April 2016 # Indian Society of Remote Sensing 2016 Abstract In India, Jharia Coalfield (JCF) has one of the dens- Introduction est congregations of surface-subsurface coal fires known worldwide. Systematic investigation and quantification of ac- Coal has a significant contribution towards India’seconomic tual scenario of coal fires in JCF is always necessary to plan growth. It feeds major percentage of industrial demand for sustainable mining, industrial growth and environmental re- energy production. Due to the unavailability or inadequate mediation on a long term basis. The present approach involves supply of other energy sources, coal is primarily used as fuel evaluation and mapping of coal fire using ASTER (Advanced in the industries. More than 90 % of the coal available in India Spaceborne Thermal Emission and Reflection) data. Mapping is confined to the Gondwana Coalfields. Out of the total avail- reveals that the area located around western, eastern and able resources of coal in Gondwana Coalfields, 88 % of the south-eastern parts of JCF covering territories of Shatabdi coal is of non-coking type (GSI 2004). Rest of the 12 % coal is opencast, Barora; Sijua opencast; Godhar colliery; Kusunda; coking type which is exclusively contributed by ‘Jharia Bokapahari; Kujama and Lodna are under intense fire with Coalfield (JCF)’. Due to this reason, this coalfield has been cumulative coverage of 6.23 km2. The ASTER derived Land exploited by intense mining activities for more than a hundred Surface Temperature (LST) of the anomalous areas have been years. Intense mining exposed the coal seams to an open en- subsequently validated by the field observations, carried out in vironment where high oxygen influx causes them to self- JCF in February, 2010. The methodology adopted in the pres- ignite and undergo spontaneous combustion (Feng et al. ent study would provide precise evaluation and monitoring of 1973). coal fire in Jharia. JCF is rigorously affected by surface-subsurface coal fires (Michalski 2004). It has witnessed numerous severe accidents Keywords ASTER . Surface-subsurface coal fires . Jharia . and uncountable loss of valuable coal reserves due to uncon- Mapping trolled coal fires. Coal fire affected areas are often inaccessi- ble. Field surveys may not be plan frequently and considered unfeasible from economic point of view. Therefore, satellite remote sensing has widely contributed in analysing thermal phenomenon like coal fires (Rosema et al. 1999; Quattrochi et al. 2009; Prakash and Gens 2010). Since two decades, re- mote sensing techniques have been significantly used for * Ashwani Raju mapping and quantification of surface (Prakash and Gupta [email protected] 1999; Yao-Bao 2010;Rajuetal.2013) and subsurface coal fires in JCF (Cracknell and Mansor 1992; Prakash et al. 1997; 1 School for Environmental Sciences, Babasaheb Bhimrao Ambedkar Chatterjee 2006; Chatterjee et al. 2007;Mishraetal.2011; University, Lucknow 226025, India Roy et al. 2015a). 2 Geological Survey of India, Jabalpur, Madhya Pradesh 482003, India Coal fires in JCF are highly dynamic in nature and vary 3 Department of Earth Sciences, Indian Institute of Technology frequently in their spatial extent (Roy et al. 2015b). A single Roorkee, Roorkee, Uttarakhand 247667, India coal fire map may not be useful for long term basis. Hence, J Indian Soc Remote Sens precise planning and mitigation measurements require repeti- Surface fires are relatively high temperature phenomenon tive monitoring of coal fires. Therefore, in the present study, of relatively local extent and can be detectable in bands oper- suitable method for detection and mapping of coal fires has ating within shorter wavelength region (Prakash and Gupta been used and surface-subsurface coal fires have been system- 1999). Shortwave Infra-Red (SWIR) bands are sensitive to atically analyzed in JCF using Advance Space borne Thermal both solar reflected radiation and blackbody radiation emitted Emission and Reflection Radiometer (ASTER) data. by fire. Separate retrieval of emission and reflected compo- nent from SWIR bands is a difficult task and beyond the scope of present study. However, ASTER SWIR band 9 operates in Study Area: Jharia Coalfield, India 2.36–2.43 μm range at normal gain setting and has the capa- bility to detect pixel integrated temperature ranges between Jharia Coalfield (JCF) is located in Dhanbad district, 66 °C–222 °C. In the present study, the pixels with highest Jharkhand between N 23° 37’ 57.88^ - N 23° 49’ 55.47^ radiance response have been carefully analysed in SWIR band and E 86° 08’ 13.72^- E 86° 28’ 42.50’ and covers an area 9 and no saturated pixel (pixel with DN value of 255) was of about 450 km2. The altitude of the area is 77 m above mean identified. Hence, ASTER SWIR band 9 has been preferably sea level. JCF comprises sedimentary litho-package belonging chosen to map coal mine surface fires. to the Gondwana Supergroup of Permo-Carboniferous age. The package is composed of alternating sequence of sand- stone and shale with inter-bedded coal seams of fluvio- Methodology Overview and Data Analysis lacustrine origin, deposited within intra-cratonic Archean gneissic basement. The litho-units have regional strike of E- In the present study, the methodology involves acquisition, W to NW-SE with shallow dip of about 3°-8° toward south. processing, analysis and interpretation of the satellite datasets All coal seams are restricted to the Barakar and Raniganj for precise detection of coal fires in JCF (Fig. 1). Since April Formations of Gondwana Supergroup. Jharia coal belongs to 2008 ASTER SWIR data are not useable (https://lpdaac.usgs. medium to high volatile sub-bituminous to bituminous range gov). Hence, for detecting high intensity surface fire ASTER coal containing 0.13 % to 2.81 % of moisture, 12.0 % to SWIR 2008 data is used. ASTER L1 A SWIR and TIR images 26.63 % of ash, 6.93 % to 28.40 % of volatile matter and dated 29th March 2008 and 18th November 2009 (ID: AST_ >60 % of fixed carbon (Karmakar et al. 2013). In JCF, fire is L1A_00303192008045415_04172009132739 and AST_ mainly confined to coal seam X and XIV-XVIII (Chandra L1A_0031182009163246_11250009543938) respectively 1992) and occurs up to maximum depth of 110–130 m. At have been procured from USGS LP DAAC (Land Processes present, nearly 67 active fire sites are reported from 23 large Distributed Active Archive Centre) archive (https://lpdaac. underground and nine large open cast mines in JCF (BCCL usgs.gov/data_access) and used for the analysis. Both 2008). datasets have been co-registered and resampled to 30 m spatial resolution and then processed for the further analysis. ASTER L1 A data further is processed using radiometric Theoretical Background calibration coefficient to obtain scaled radiance. Calibrated ra- diance emitted from the ‘hot ground features’ in form of ther- Different bands have different temperature sensitivity. malanomalieshavebeeneffectivelymeasuredbytheonboard Wavelength distribution of emission response by a blackbody satellite sensors. At-sensor spectral radiance (Lλ) for ASTER is inversely proportional to its temperature. ASTER registers data can be obtained using Eq. 1 (Abrams et al. 1999): the spectral response in five Thermal Infra-Red (TIR) channel functioning in 8.125 μm to 11.65 μm ranges of the electro- Lλ ¼ ðÞÂDNλ − 1 UCCλ ð1Þ magnetic spectrum (Abrams et al. 1999). It provides moderate spatial resolution data at high radiometric and temporal reso- Where, DN is the digital number of a pixel and UCC is the lutions that holds great potential for effective mapping and Unit Conversion coefficient (W m−2 sr−1 μm−1). monitoring of coal fires occurring worldwide (Martha et al. 2010). ASTER TIR channel acquires data in 12 bits high ra- Processing of SWIR and TIR Data diometric quantization level. It has high detectable tempera- ture limit with an ability to measure the small differences in In coalfield area, intense mining activity and coal burning radiation emitted from low temperature phenomenon like sub- have significantly contributed aerosol and particulate matter surface coal fire. However, subsurface coal fires are greatly to the atmosphere. Spectral radiance received at the satellite varied in depth and spatial extent. It can be only detected by sensor is commonly affected by atmospheric transmittance (τ) the TIR sensor if its magnitude and intensity is large enough to and upwelling radiance (LλP, atmospheric path radiance) due enhance the per pixel radiant response than background. to atmospheric haze, aerosols etc. (Gupta 2003). To J Indian Soc Remote Sens Fig. 1 The methodology followed in the present study Table 1 Details of field observations carried out in JCF Sr. No. Site Latitude Longitude Surface Land cover type temperature range (°C) 1 Bokapahari 23° 45’ 9.2^ 86° 25’ 4.2^ 43.6–71.2 Area of intense shallow subsurface fire with sparse vegetation. Linear cracks have developed parallel to the strike of coal seams. 2 Baghadih 23° 43’ 50.2^ 86° 24’ 52.8^ 22.7–39.5 Subsidence area with sparse vegetation. 3 Lodna 23° 42’ 58.0^ 86° 25’ 17.5^ 42.6–131.0 Overburden dump site. Anomalies are due to dump and subsurface fire with surface temperature ranges from 165.2 °C – 332.4 °C at places. 4Sudamdih23°40’ 1.5^ 86° 25’ 19.8^ 32.2–42.2 Anomalies are due to subsurface fire. 5 Bhowrah 23° 41’ 2.3^ 86° 23’ 29.9^ 41.5–76.8 Overburden dump site. Anomalies are due to smouldering fire in dump. 6 Tetulmari, Sijua 23° 48’ 24.8^ 86° 19’ 57.8^ 39.8–85.4 Opencast mine. Anomalies are due to high intensity coal seam fire (217.0 °C − 254.5 °C) exposed due to intense mining activity.