Urban lands Extraction from Sentinel 1 and 2 satellite imagery based on Google Earth Engine (GEE)

The use of appropriate methods and up-to-date satellite images in various studies, especially urban studies, can play a major role in the production of urban maps. One of these important data is the map of urban lands that can be extracted using various methods. The aim of this paper is to extract urban lands of a number of Iranian cities using Sentinel 1 SAR satellite images and Sentinel 2 based on Google Earth Engine GEE. For this purpose, Sentinel 1 SAR and Optical Sentinel 2 images were selected as time series from 2017.01.01 to 2020.01.01 for 20 cities in Iran. time series Images entered to the Google Earth engine environment, and then the mean and standard deviation of radar images were prepared and by applying the threshold, potential urban lands were extracted. NDVImax, NDWImean and slope and mountainous areas were also extracted using Sentinel 2 images and DEM, and mask images were created by applying thresholds. Finally, by applying these images to the map of potential urban lands, the target urban land map was created and by applying a 3 * 3 filter to remove individual and false pixels, the final map of urban lands was extracted. The results show that the average Kappa coefficient for 20 cities is 86.16%. Also, cities in arid and semi-arid regions are less accurate. The results of this study show that the GEE system is able to process large amounts of data in a very short time with high accuracy.