Kutia, M. and Li, J. and Sarkissian, A. and Pagella, T. (2023) Land cover classification and urbanization monitoring using Landsat data : A case study in Changsha city, Hunan province, China. Ukrainian Journal of Forest and Wood Science, 14 (1). pp. 72-91.
Full text not available from this repository.Abstract
The United Nations predicts that by 2050, 64.1% of the developing world and 85.9% of the developed world will be urbanized. This has resulted in a rapid change in land use and land cover types in the areas surrounding cities in all countries, particularly in China, which determines the relevance of this article. The aim of the study was to evaluate the dynamics of land cover change in Changsha City, Hunan Province, China, between 2005 and 2020, using Landsat time series satellite images and the Random Forest classification algorithm. The data acquisition, pre-processing, and analysis were conducted on the Google Earth Engine (GEE) publicly available online platform. Land cover thematic continuous raster maps were produced using ESRI ArcGIS 10.5.1 software. The overall classification accuracy was obtained by more than 83% for every produced map and the Kappa coefficient was 0.84 and higher, which approves the reliable classification results that are close to similar recent studies in terms of obtained accuracy. The study shows that from 2005 to 2020, the area of settlement in Changsha City, China, increased significantly, with an exponential increase in urban area from 3.23% to 15.95%. The proportion of forest cover gradually decreased from 2005 to 2015 but increased from 2015 to 2020. Cropland was the second most dominant land cover type, with a peak of almost 50% in 2010. Water bodies remained stable at around 3%. The proportion of open soil and bare land cover fluctuated between 180 and 400 km2 (1.5-3%). The study suggests that the offered monitoring approach provides reliable results, and the research findings can be used for sustainable urban planning and management, as well as conservation and development initiatives. The remote sensing data and advanced GIS technologies can provide decision-makers with the accurate data to ensure sustainable development in this area