REMOTE SENSING BASED GEO-ECOLOGICAL STUDIES IN JHARIA COALFIELD, INDIA

Abstract

Coal is one of the most abundantly available fossil fuels for the Homo sapiens and provides a substantial part of their energy needs. Globally, India is regarded as the third largest coal producer after China and USA. Also, coal is the major fossil fuel used for power generation in India, in as much as about 70% of the mined coal in India is used for power generation. It is therefore imperative that to produce more coal, mining activities increase. However, coal mining has resulted into environmental changes such as changes in landform, land use/land cover and vegetation distribution as well as degradation in the quality of air, water, and soil. Besides these, other effects of coal mining include activation of coal fires both surface and subsurface, and subsidence of land surface. Therefore, it becomes important to assess the changes at regular intervals so as to understand their impact on the environment. Satellite remote sensing coupled with geographic information system (GIS) offers an extremely powerful tool for such monitoring applications. The research presented in this thesis focus on the evaluation of impacts of coal mining and mine fires on the environment as a whole and vegetation in particular in Jharia coalfield, India using field based measurements, expert opinions and remote sensing data analysis. The study area Jharia coalfield (JCF) (lying between latitude 23° 38’ N - 23° 50’ N and longitude 86° 07’ E - 86° 30’ E) is important in Indian context as it is the only source of prime coking coal that meets a large percentage of industrial demand in the country. At present, it consists of 23 large underground and 9 large opencast mines. The area is also infamous for surface and subsurface coal mine fires which have been burning for more than a century now. In order to understand the impact of these activities in JCF, a comprehensive review of relevant scientific literature on identifying the parameters (vegetation, agricultural land, coal fires, soil, etc.) affecting the environment has been carried out. In this research, an ecological and environmental impact assessment study been postulated based on integrated datasets consisting of remote sensing, field and ancillary data (e.g., toposheets, geological maps, offline and online survey, surface fire maps) ii considering JCF as the case study. Field data includes gathering information on land use/land cover (LULC) such as vegetation pattern, active/closed mining areas, overburden dumps, land subsidence, surface-subsurface coal fires etc., collection of ground control points, temperature information and photographs of ground conditions in JCF. Remote sensing images of varying spatial, spectral and temporal resolutions from sensors deployed on Landsat and Resourcesat satellites have been used to carry out vegetation and temperature based studies. The remote sensing data has been pre-processed for atmospheric and geometric distortions. GIS has been used for data management, processing and analysis. Digital image processing operations such as ratioing, orthogonal transformation, image classification and land surface temperature (LST) computations have been performed on standard image processing software. For detail processing, analysis and interpretation, two subsets have been extracted from the images covering the JCF area, and have been named as subset 1 (Shatabdi opencast mine, area: 4410 hectares) and subset 2 (mining area in and around the main Jharia town, area: 2963 hectares). In order to assess the cumulative impact of mining on the environment, data are collected from a number of sources and analysed collectively thereby making the process multi-dimensional or multiple criteria based. A multi-objective, multi-criteria decision-making approach, namely, Analytical Hierarchy Process (AHP) has been used to carry out comprehensive environmental impact assessment (EIA). AHP has been successfully implemented in structuring the decision problem (EIA) into a number of levels and providing weights to criteria through pair-wise comparisons. Six parameters namely vegetation, agricultural land, topography, air, water and soil have been identified for carrying out EIA which have been further categorized into 26 decision alternatives. First, the AHP analysis has led to identification of significant environmental parameters. Later, it has helped in identifying the most significant alternatives outlined for each factor, which cause the maximum impact on that particular factor. The results from AHP reflect that the local environment has been degraded to a large extent in JCF. The impact on air has been found out to be the highest because it is the most frequently affected parameter. The reason being; coal dust stirred up during the mining processes, as well as released during haulage/transport, can cause severe and potentially deadly respiratory problems. The next iii most affected criteria are water and soil as they are polluted due to pumping out of mine water, settling of coal dust leading to contamination etc. Opencast mining destroys landscapes, forest and wildlife habitat at the mining site when trees, plants, and topsoil are cleared from the mining area. This in turn leads to soil erosion and destruction of agricultural land. However, this process takes place less frequently as compared to air, water and soil pollution, which justifies the lower AHP ranking of vegetation, agricultural land and topography changes. Deterioration of the environment has an impact on and limits growth of vegetation, which is a key indicator of a healthy ecological system. Change in vegetation plays an important role in the environmental processes and is considered as a sensitive indicator for environmental and global changes. Thus, the study later focuses on vegetation analysis using LST, Normalized Difference Vegetation Index (NDVI) and Tasseled Cap Transformation (TCT). Coal fire areas are also marked by a sudden rise in ground temperature relative to the background. This rise is very conspicuous when the fire is exposed on the surface, but it is less pronounced in the case of subsurface fires. The increase in high temperature areas imply that the areal extent of vegetation and agricultural lands may decline further. This happens because the soil loses moisture and gets baked due to the fire making it less potent for plant growth. All this is likely to adversely affect the ecology of the area. Therefore, the fire areas have been delineated using images acquired in thermal infrared (TIR) band of Landsat TM sensor. The radiance images have been converted to radiant temperature and then into LST for subsets 1 and 2. The LST values have been suitably adjusted by emissivity values of different LULC classes to account for the class-wise variation in surface temperature. Emissivity of each LULC has been derived using the NDVI values. The LST images show the progression of fire areas from 1991 to 2013 in some portions of the study area. To account for the general seasonal variation in the area, the entire dataset has been divided into four groups based on the acquisition time namely, Early November, Late November to early December, Mid-December and Mid-January. The analysis of both the subsets shows that over the years, the temperature has increased in each group, which indicates general warming and the increase in fire. iv Out of the six criteria used in EIA, vegetation has been found to be the most significant as in mining areas, it plays a significant role in modification of water balance, erosion control and landscape rehabilitation. Vegetation helps in retaining water in top soil and preventing excess outflow of nutrients and thus is essential for maintaining a healthy ecosystem. However, in the course of mining, the vegetation cover and the draining regime are disturbed over enormous areas. Therefore, one of the objectives of this research is to study multi-temporal data to assess the change in vegetation cover, health and density over the years as a result of coal mining in JCF. This has been done through the use of NDVI and tasselled cap analysis of remote sensing images. Based on the time duration, two types of vegetation analysis have been done using NDVI images. Firstly, a long duration (acquired during 1972-2013) medium spatial resolution (30 m) analysis using multi-sensor Landsat data for subsets 1 and 2 has been carried out. Secondly, a short duration (acquired in 2007 and 2012) high spatial resolution (5.8 m) analysis using Resourcesat LISS IV data has been carried out. As the spatial resolution of LISS IV data is large, two new subsets namely subset 3 (Bhowrah colliery and surrounding areas, area: 1891 hectares) and subset 4 (Angarpathar, area: 988 hectares) have been extracted from this data. NDVI images have been generated and density sliced to show distribution of three classes namely, a) non-vegetated area, b) sparse vegetation, and c) moderate to dense vegetation. The areal extent of each class has been calculated. The vegetation analysis indicates that the dense vegetation has drastically decreased from 1972-2013, as the mining proceeded. The reason for this is the direct loss of vegetation when the surface is cleared for initiation of mining activity in an area or covered under overburden material or there is an increase in settlements around the mines. Further, the effect of high temperature areas associated with mine fires (as identified from the LST analysis) on vegetation has also been studied. A negative correlation between NDVI and LST has been found for all the years in both subset 1 and 2, which should be the case. In order to evaluate the spatial distribution of vegetation condition for different years, a tasseled cap analysis has also been carried out. The first two components namely, Brightness and Greenness, which are equivalent to the soil brightness index and the green vegetation index respectively, have been used for the analysis. The results show that the Greenness is high in some regions which has later reduced over the time pointing to the v fact that vegetation has decreased in these regions over the years. This has been shown in the form of scatter plots of Greenness vs. Brightness. For each year, the plot illustrates a negative correlation of Greenness and Brightness implying that as mining has increased, soil brightness has correspondingly increased. Further, Brightness and Greenness images of 1992 and 2013 have been subtracted to generate difference images. A colour composite of difference images shows changes in Brightness and Greenness that have occurred in both the subsets over 21 years. Finally, inter-correlations have been carried out between NDVI and TCT. A high positive correlation has been found between NDVI and Greenness component of TCT. The results of Greenness component of TCT supplement those of the NDVI and as TCT takes into account Brightness component too, so additional results help in better interpretation of changes in vegetation due to mining. The findings of this research clearly show the effect of mining on the environment of JCF as a whole and vegetation in particular. The study demonstrates that conjunctive application and analysis of field based measurements, expert opinions and remote sensing analysis hold considerable promise for assessing the impact of mining on the environment over the years and its monitoring