Investigating the effects of land cover land use change on surface temperature using Landsat satellite images
Keywords:
Land use / Land cover, Urban heat islands, Random Forest, Google Earth Engine, Change detectionAbstract
Global and local population growth and the rapid increase in urbanization affect nature negatively by the destruction on forests and natural lands. An additional problem can be considered as the increment of the land surface temperatures due to the heat island phenomenon. Thus, long-term monitoring of rapidly developing cities is important. In this study, Izmir, which ranks 3rd among the big cities of the country, was chosen to monitor the long-term effects of urbanization. For this purpose, a 20-year period from 2020 to the past has been examined with Landsat images. As a first step, historical land cover – land use maps were produced from satellite images using the Random Forest algorithm in the Google Earth Engine platform. Secondly, the urban thermal field variance index (UTFVI) was calculated from thermal bands of Landsat images to examine the effect of urban heat islands and their relation to urbanization progress. Results of these analyses indicated that both cities faced urbanization at the expense of forest and semi-natural area loss in this 20-year period, which is well correlated with an increase in the UTFVI values. Moreover, the increase in UTFVI values on already urbanized regions proposed that the intensity of the urban areas also increased.