Spatio-temporal dynamics of urban sprawl in a rapidly urbanizing city using machine learning classification

Abstract

The world is urbanizing at an alarming pace, particularly in developing countries where exponential urban population growth leads to unplanned and uncontrolled urban expansion, resulting in urban sprawl. Mapping, monitoring, measuring urban sprawl, and identifying land cover transition contributors are of pivotal importance in formulating policies and management strategies for the sustainable growth of cities. The article presents an explorative study on the spatio-temporal dynamics of urban transition in Kochi Urban Agglomeration in Kerala, India, from 1988 to 2021, using the support vector machine (SVM) classification, Shannon’s entropy, and landscape metrics. From 1988 to 2021, the built-up share in the study area increased from 6.23% to 32.34%. The increased demand for land resulted in converting 186.94 km2 of cultivated land and 136.64 km2 of vegetation into built-up land. The level of land-cover changes points to the sprawl on the outskirts and densification in the inner city.