Vahid Hatamzadeh

In recent years, preparing land use maps by digital classification of remote sensing data have been adverted as appropriate alternative for using this type of maps. Remote sensing is a modern and useful technique in land use ,updating maps and detecting new changes .using objective process technique , satellite images is one of the most updated methods in image processing that in addition to using spectral resolution , also used physical and geometrical properties like texture and form.in this research we use ArcGis pro for classification that is one of the most accurate and updated software for remote sensing’s process for detecting ۴ type of classes in Tehran city in IRAN. With standard kappa coefficient accuracy, and overall accuracy of data is proved. These results are beneficial decision-makers and officials for environment sustainability and effect of land changes. By considering ۴ essential classes in a major city and converted to maps of changes in linear regression concluded that build-up class have a significant slope increase ۳۴۲۲/۲۹ (hectar) fortunately plant class is improving during the study period as ۲۸۲۱/۷۱ (hectar) but these increment are inhomogeneous, water class has sharp drop as ۴۴۳.۵۲ (hectar), Furthermore the most of decrement is for the barren area which named soil class as ۵۸۰۰.۴۸ (hectar). Some part of accuracy in this research depends on severity of the numbers of test samples which given for classification that are more than۵۰۰۰ pixels to assessment reliable results. According to the standards of kappa coefficient that provided in USGS earth data site all off maps are acceptable.

remote sensing (RS), satellite images, land cover-land change, ArcGIS Pro, Classification, sentinel ۳