Classification and Change Detection of an Urban Built-Up Area: A Case Study of Gurugram

Abstract

Numerous built-up indices have gained wide attention for classifying urban areas using satellite imageries since the mid-1990s. In this study, three spectral indices performance to classify images and detect changes in the built-up area have been examined using Level 1C multispectral instrument Sentinel-2A imageries and compared. The following three mid-infrared-based indices are used: normalized difference built-up index (NDBI), index-based built-up index (IBI), and urban index (UI). These indices are used to classify and detect the temporal changes in urban areas in Gurugram using the image thresholding technique. For each image, 100 reference points from Google Earth were used to assess the classification. The built-up area’s evaluation revealed that most of the urbanization is concentrated in the eastern part and is expanding in the central part of the Gurugram city. The total geographical area of the city is 473 km2, out of which the built-up area for the year 2016 is estimated as 357, 350, and 349 km2 using IBI, UI, and NDBI, which increased to 373, 365, and 369 km2 in 2019. The increase in the built-up area from 2016 to 2019 is estimated as 4.48, 4.28, and 5.73% using IBI, UI, and 244NDBI, respectively. The findings show the significance of built-up indices in the rapid, spatial and temporal assessment of larger city areas. Such studies would enable the urban planners and policymakers in urban cool/heat island modeling, rainfall-runoff modeling, urban sprawl model parameterization, and climate-change mitigation.