Application of Cellular Automata (CA) for Predicting Urban Growth and Disappearance of Vegetation and Waterbodies

Abstract

For sustainable development and also urban planning, the data regarding the type and amount of urban growth, and disappearance of vegetation and waterbodies, because of urbanization, in the near/distant future, are absolutely necessary. This work endeavors to know, how effective (freely available) MOLUSCE (MOdules for Land USe Change Evaluation) plugin of QGIS is, for the prediction of the growth of Asansol city (in Eastern India), and the disappearance of vegetation and waterbodies, because of urbanization, in the area around the city of Rajarhat New Town (in Eastern India). The study considers the closeness to the Asansol Railway Station as the only factor governing the urban growth; and the closeness to the Sector-V Metro Station (adjacent to New Town) as the sole factor governing the disappearance of vegetation and waterbodies. MOLUSCE has used suitably classified Landsat images (Landsat images are freely available), and utilized Cellular Automata technique, with Logistic Regression method for transition potential modeling, to produce the prediction maps of the Asansol city, the vegetation around New Town, and the waterbodies around New Town. Classified Landsat images of the respective areas have been used to validate these prediction maps; the correctness is 75.44% for urban growth, 78.18% for disappearance of vegetation, and 85.73% for disappearance of waterbodies. However, there is not much similarity between the corresponding classified images and the prediction maps, because of the consideration of only two simple factors for prediction. Nevertheless, the result shows the potential of MOLUSCE for land use change prediction.