Determination of Urban Areas Using Google Earth Engine and Spectral Indices; Esenyurt Case Study

Identifying impervious surfaces for monitoring urban expansion is important the sustainable management of land resources and protection environment. Remote sensing provides an data source use/land cover mapping, these can be analyzed with various techniques different purposes. If aim to extract information easily rapidly, using spectral indices most appropriate solution, there are many created this purpose. The study carried out on Google Earth Engine (GEE) platform, Esenyurt, populous district Istanbul, was investigated Sentinel 2 MSI image, eight three vegetation indices. In addition, classification made, results were evaluated. As a result index applications, it has been seen that roofs more or less mixed bare soil areas, Normalized Difference Tillage Index (NDTI)gives best results. Accuracy assessment performed same points, due area density in application area, determined as 0.95% NDTI Vegetation (NDVI), 97% classification, respectively. GEE, high (-0.79) negative correlation observed between May mean values 2007-2022 population when NDVI time series applied entire within borders Landsat 5 8 images 1990-2022. rapidly increasing leads rapid urbanization, green areas disappearing at rate.