Prediction of permafrost degradation deformation trend based on Prophet model

In the context of global warming, the degradation of permafrost tends to intensify, making it crucial to analyze and predict its deterioration process. This study focuses on the Beiluhe region in the hinterland of the Qinghai-Tibet Plateau as a research area. By utilizing SAR image data from Sentinel-1A satellite spanning from January 2018 to December 2021, we acquired time-series deformation data in permafrost regions with a 12-day cycle. Through an improved Prophet model , we analyzed the periodic characteristics of frost heave and thaw settlement deformation in permafrost regions, mapped out changes in these characteristics, and predicted large-scale continuous subsidence caused by permafrost degradation. The stability of permafrost in this study area was evaluated by considering topographic and local environmental factors. The accuracy evaluation indicator for our constructed multi-year frozen soil degradation trend prediction model is based on comparing average RMSE values between output data and original data, resulting in an average accuracy rate of 10.94%. Cumulative change data regarding deformation trends over the past three years were categorized into four groups: I (range [0,20]), II (range (20,40]), III (range (40,65]), and IV (range(65,115]). Area I accounts for 17.53% of total area coverage mainly consisting of bedrock mountainous sareas where rivers are primarily located within melting zones. Zone II covers 34.37%, while Zone III covers 33.18%, and Zone IV covers 14.19%. Zones III and IV are predominantly distributed across the Beilu River Basin which represents a high-temperature unstable permafrost region. Furthermore, predictions have been made regarding surface deformation trends in this study area for the next two years to provide valuable data support for future research on permafrost degradation within this region.