Urban Change Detection Analysis during 1978–2017 at Kolkata, India, using Multi-temporal Satellite Data

Abstract

This paper focused on urban change detection and growth pattern analysis for the period of 1978–2017 at Kolkata using remote sensing data and GIS. The supervised Maximum Likelihood Classification technique is used to classify the multi-temporal satellite data in five classes which are urban built-up, open land, vegetation, agricultural land and water body. Results revealed that urban built-up area has progressively increased by about 21.17% (239.097 km2) during study period due to the new construction of roads, flyovers, settlements, etc. Other geographical features such as open land, vegetation, agricultural land and water body have gradually declined. To assess the accuracy of classification, 88.27%, 92.42%, 91.62%, 90.18% and 89.34% overall accuracy and 0.851, 0.904, 0.893, 0.875 and 0.866 Kappa statistic have been achieved for the images of 1978, 1988, 2000, 2010 and 2017, respectively. The degree of magnitude of urban sprawl has measured using the Shannon entropy method which demonstrates that the overall entropy values are progressively increased throughout the entire region that means the urban built-up is gradually extended in different positions. Moreover, entropy outcomes revealed that urban development occurred more in the northern and southern regions as compared with the other regions. From this study, four urban growth patterns have been found which are low density continuous, continuous linear, noncontiguous linear, and leapfrog development. The important patterns are continuous linear and noncontiguous linear because most of the urban development happened along the sides of the major roads or highways. Future prediction has been obtained using CA–Markov chain model and estimates that the urban built-up may be increased by about 56.18% (509.82 km2) in the year of 2031.