Monitoring Urban Growth Using Land Use Land Cover Classification

Abstract

The rise in population hastens changes in land use as well as land cover, increasing threat to natural resources. Monitoring and measuring land use land cover (LULC) changeover vast areas is crucial to study climate change patterns and in natural resource management. This paper examines changes in Vijayawada's urban growth over a period of six years (2016–2020) using the Google Earth Engine (GEE) cloud-based platform by classifying LULC changes. The NDVI and NDWI indices were used to accurately collect training points of ‘crop land’, ‘water bodies’ land cover classes, respectively. The random forest (RF) supervised classification technique was used to classify the LULC in the research area. Landsat-8 surface reflectance data was utilized for the LULC categorization between 2016 and 2020. ‘Water bodies’, ‘Crop Land’, ‘Built up’, ‘Bare Soil’, and ‘Shrubs’ were the five land use classes developed. For the years 2016 and 2020, the overall accuracy was 80.55% and 91.78%, and the kappa coefficient is 0.73 and 0.88, respectively.