Predicting the future landscape of Dhanbad District: an analysis of land-use change and urban sprawl through cloud computing and neural networks

The present study attempted to analyse the land-use change and its dynamics in a coal mining region of India. The land-use change from the year 2000 to 2010 was analysed in the Google Earth Engine using Landsat data archive to develop a Neural Network based Markov chain model. The developed model was applied to predict the land-use change for the year 2020 and was validated with the actual land-use data of the year 2020. The validated model was used to predict six different land-use classes for the year 2030 in the present study area. The findings of the study indicated that surface mining areas are gradually decreasing due to the transition from open cast mining to underground mining and restoration of the degraded mining areas. Furthermore, it was observed that the agricultural area in the study area declined by 10.27% between 2000 and 2010 and by 6.24% between 2010 and 2020. Vegetation cover increased significantly by 54.25% during 2000 to 2010 but, after that it remained relatively constant till 2020. The predicted land-use showed that the built-up area will continue to expand away from the existing urban centres and mining areas. Additionally, the study identified that no new urban growth centres are expected to emerge by 2030, but higher growth rate is expected in and around the two new urban centres that emerged in the year 2020. The study revealed valuable insights into the underlying drivers of land-use change that will help the policy makers to device suitable urban planning.