In recent years, the Varamin Plain has experienced land subsidence. In this study, the subsidence rate was obtained using LiCSBAS software and Sentinel-1 satellite images. The maximum subsidence rate of 955 mm over 10 years occurred in the southern part of the region. To predict and investigate the influence of geological factors on the subsidence rate, the Random Forest algorithm was employed. Bedrock depth, groundwater level decline, groundwater depth, clay percentage of the aquifer, extraction from exploitation wells, and distance from faults and rivers were selected as model inputs. The model's coefficient of determination of 0.95, its good learning performance, and the cross-validation method indicate the model's high validity. The prediction map showed a maximum subsidence of 754.6 mm, consistent with the calculated subsidence pattern for the next 10 years. The Gini coefficient index identified bedrock depth, with a 27.5% decrease in model impurity, as the most influential parameter. Groundwater depth and groundwater level decline were identified as the next important factors. Based on the findings, groundwater decline is a prerequisite for subsidence but does not necessarily lead to its occurrence. The role of other factors, such as bedrock depth and groundwater depth, is also significant. The results demonstrated the efficiency of LiCSBAS software and the Random Forest algorithm for calculating and predicting subsidence rates and identifying influencing factors.