Abstarct: Abstract Today, the increase in population and extensive human needs for water resources cause unconventional exploitation of water resources It has led to underground waters and a drop in their level. One of the main measures against this natural phenomenon is to determine the areas at risk or in danger, which is determined by different methods according to the conditions of the area and the factor that caused the subsidence. Among the new methods that are used to monitor this natural disaster; Remote sensing methods and aquifer modeling using the PMWIN numerical model can be mentioned as efficient and accurate methods. In this study, the changes of the Garmsar plain have been investigated using remote sensing method, and the changes of the underground water level in the study area in the period of 2014-2022 have been investigated by modeling the Garmsar aquifer with the help of PMWIN.In this study, the amount of subsidence in the plain was calculated and determined using the radar interferometric technique in the period from 12/01/2014 to 06/22/2022. For this purpose, the data of the Sentinel-1 sensor, which is in the C band of the radar, has been used. The maximum amount of subsidence calculated was 14.89 cm between 12/17/2020 and 06/22/2022. The distribution of subsided areas also indicates maximum subsidence in the central and western part of the studied area. Also, in this research, the simulation of the underground water table of Garmsar plain in stable and unstable state has been done using PMWIN model and MODFLOW-2005 computer code by finite difference method. Further, to calibrate the model, the observational water level data of 35 piezometric wells in the period of 1401-1392 have been used, and the observational water level data of piezometric wells in 1402 have been used for validation.Hydraulic conductivity parameter has been used for steady state calibration and storage coefficient parameter has been used for unstable state calibration. Finally, it has been validated using the observational data of 1402 calibrated models. The obtained results have an acceptable error rate of RMSE, which indicates the acceptable accuracy of the model. Also, by comparing the final results of modeling and the final results of subsidence estimation by remote sensing method, it can be said that most of the areas that have the highest amount of groundwater level drop also have the highest subsidence rate.