Ongoing climate warming and increased human activities have led to significant permafrost degradation on the Qinghai-Tibet Plateau (QTP). Mapping the distribution of active layer thickness (ALT) can provide essential information for understanding this degradation. Over the past decade, InSAR (Interferometric synthetic aperture radar) technology has been utilized to estimate ALT based on remotely-sensed surface deformation information. However, these methods are generally limited by their ability to accurate extract seasonal deformation and model subsurface water content of active layer. In this paper, an ALT inversion method considering both seasonal deformation from InSAR and smoothly multilayer soil moisture from ERA5 is proposed. Firstly, we introduce a ground seasonal deformation extraction model combining RobustSTL and InSAR, and the deformation extraction accuracy by considering the deformation characteristics of permafrost are evaluated, proving the effectiveness of RobustSTL in extracting seasonal deformation of permafrost. Then, using ERA5 soil moisture products, a smoothed multilayer soil moisture model for ALT inversion is established. Finally, integrating the seasonal deformation and multilayer soil moisture, the ALT can be estimated. The proposed model is applied to the Yellow River source region (YRSR) with Sentinel-1A images acquired from 2017 to 2021, and the ALT retrieval accuracy is validated with measured data. Experimental results show that the vertical deformation rate of the study area generally ranges from -30 mm/year to 20 mm/year, with seasonal deformation amplitude ranging from 2 mm to 30 mm. The RobustSTL method has the highest accuracy in extracting seasonal deformation of permafrost, with an RMSE (root mean square error) of 0.69 mm, and is capable of capturing the freeze-thaw characteristics of the active layer. The estimated ALT of the YRSR ranges from 49 cm to 450 cm, with an average value of 145 cm. Compared to the measured data, the proposed method has an average error of 37.5 cm, which represents a 21 % improvement in accuracy over existing methods.

受全球气候变暖影响，青藏高原冻土退化和地表失稳问题不断加剧，对基础设施的建设维护和区域社会经济发展造成阻碍。近年来，SBAS-InSAR技术已在冻土地表形变监测中得到广泛应用，但由于青藏高原部分地区存在较为严重的失相干现象，导致形变监测结果出现空间不连续性，进而无法获得全面且精细的监测结果。针对上述问题，本文提出了一种联合机器学习与SBAS-InSAR的冻土形变监测方法，选取西藏阿里门士乡为研究区域，使用2020年1月7日至2021年6月6日共43景Sentinel-1A降轨影像数据提取地表形变信息；综合多类环境因子数据生成训练集后，引入机器学习模型拟合SBAS-InSAR监测结果与环境因子之间的内在关系，从而获取研究区连续形变速率图。结果表明，联合随机森林模型与SBAS-InSAR的方法效果最优，通过该方法对冻土形变缺失区域进行插值能极大提高原有SBAS-InSAR方法的监测覆盖率，其插值结果平均误差和均方根误差分别为0.459和0.739 mm/a。

受全球气候变暖影响，青藏高原冻土退化和地表失稳问题不断加剧，对基础设施的建设维护和区域社会经济发展造成阻碍。近年来，SBAS-InSAR技术已在冻土地表形变监测中得到广泛应用，但由于青藏高原部分地区存在较为严重的失相干现象，导致形变监测结果出现空间不连续性，进而无法获得全面且精细的监测结果。针对上述问题，本文提出了一种联合机器学习与SBAS-InSAR的冻土形变监测方法，选取西藏阿里门士乡为研究区域，使用2020年1月7日至2021年6月6日共43景Sentinel-1A降轨影像数据提取地表形变信息；综合多类环境因子数据生成训练集后，引入机器学习模型拟合SBAS-InSAR监测结果与环境因子之间的内在关系，从而获取研究区连续形变速率图。结果表明，联合随机森林模型与SBAS-InSAR的方法效果最优，通过该方法对冻土形变缺失区域进行插值能极大提高原有SBAS-InSAR方法的监测覆盖率，其插值结果平均误差和均方根误差分别为0.459和0.739 mm/a。

受全球气候变暖影响，青藏高原冻土退化和地表失稳问题不断加剧，对基础设施的建设维护和区域社会经济发展造成阻碍。近年来，SBAS-InSAR技术已在冻土地表形变监测中得到广泛应用，但由于青藏高原部分地区存在较为严重的失相干现象，导致形变监测结果出现空间不连续性，进而无法获得全面且精细的监测结果。针对上述问题，本文提出了一种联合机器学习与SBAS-InSAR的冻土形变监测方法，选取西藏阿里门士乡为研究区域，使用2020年1月7日至2021年6月6日共43景Sentinel-1A降轨影像数据提取地表形变信息；综合多类环境因子数据生成训练集后，引入机器学习模型拟合SBAS-InSAR监测结果与环境因子之间的内在关系，从而获取研究区连续形变速率图。结果表明，联合随机森林模型与SBAS-InSAR的方法效果最优，通过该方法对冻土形变缺失区域进行插值能极大提高原有SBAS-InSAR方法的监测覆盖率，其插值结果平均误差和均方根误差分别为0.459和0.739 mm/a。

Introduction Surface deformation in the Three Gorges Reservoir area poses significant threats to infrastructure and safety due to complex geological and hydrological factors. Despite existing studies, systematic exploration of long-term deformation characteristics and their driving mechanisms remains limited. This study combines SBAS-InSAR technology and machine learning to analyze and predict surface deformation in Fengjie County, Chongqing, China, between 2020 and 2022, focusing on riverside urban ground, riverside road slopes, and ancient landslides in the reservoir area.Methods SBAS-InSAR technology was applied to 36 Sentinel-1A images to monitor surface deformation, complemented by hydrological and meteorological data. Machine learning models-Random Forest (RF), Extremely Randomized Trees (ERT), Gradient Boosting Decision Tree (GBDT), Support Vector Regression (SVR), and Long Short-Term Memory (LSTM)-were evaluated using six metrics, including RMSE, R2, and SMAPE, to assess their predictive performance across diverse geological settings.Results Deformation rates for riverside urban ground, road slopes, and ancient landslides were -3.48 +/- 2.91 mm/yr, -5.19 +/- 3.62 mm/yr, and -6.02 +/- 4.55 mm/yr, respectively, with ancient landslides exhibiting the most pronounced deformation. A negative correlation was observed between reservoir water level decline and subsidence, highlighting the influence of seasonal hydrological adjustments. Urbanization and infrastructure development further exacerbated deformation processes. Among the models, LSTM demonstrated superior predictive accuracy but showed overestimation trends in ancient landslide areas.Discussion Reservoir water level adjustments emerged as a critical driver of subsidence, with rapid water level declines leading to increased pore pressure and soil compression. Seasonal effects were particularly evident, with higher subsidence rates during and after the rainy season. Human activities, including urbanization and road construction, significantly intensified deformation, disrupting natural geological conditions. Progressive slope failure linked to road expansion underscored the long-term impacts of engineering activities. For ancient landslides, accelerated deformation patterns were linked to prolonged drought and reservoir-induced hydrological changes. While LSTM models showed high accuracy, their limitations in complex geological settings highlight the need for hybrid approaches combining machine learning with physical models. Future research should emphasize developing integrated frameworks for long-term risk assessment and mitigation strategies in reservoir environments.Conclusions This study provides new insights into the complex surface dynamics in the Three Gorges Reservoir area, emphasizing the interplay of hydrological, geological, and anthropogenic factors. The findings highlight the need for adaptive management strategies and improved predictive models to mitigate subsidence risks.

冻土的水-冰相变交替过程会造成水文环境与地表工程的破坏，从而导致路基塌陷、山体滑坡、洪水暴发以及冰川溃决等灾害，智能感知潜在风险对保护冻土区工程建筑具有重要意义。采用2017年01月—2023年04月194景Sentinel-1A 影像，利用SBAS-InSAR技术获取了黄河上游沱沱河盆地冻土区形变结果，冻土地表形变明显且空间分布不均匀，监测时间段内最大形变速率可达13 mm/年。冻土区青藏铁路路基形变呈现“冻胀融沉”的季节性变化，暖季匀速沉降，冷季缓慢抬升，在气候变暖背景下暖季逐渐长于冷季；分别将InSAR监测结果与近7年沱沱河盆地GNSS监测数据对比，两者趋势一致；引入降水和气温因素后发现冻土区形变具有显著聚集特征，在人类活动频繁地区存在较大形变。该研究对冻土防灾减灾、保障人民生命财产安全具有重要意义，为高纬度冻土工程建设提供一定借鉴。

冻土的水-冰相变交替过程会造成水文环境与地表工程的破坏，从而导致路基塌陷、山体滑坡、洪水暴发以及冰川溃决等灾害，智能感知潜在风险对保护冻土区工程建筑具有重要意义。采用2017年01月—2023年04月194景Sentinel-1A 影像，利用SBAS-InSAR技术获取了黄河上游沱沱河盆地冻土区形变结果，冻土地表形变明显且空间分布不均匀，监测时间段内最大形变速率可达13 mm/年。冻土区青藏铁路路基形变呈现“冻胀融沉”的季节性变化，暖季匀速沉降，冷季缓慢抬升，在气候变暖背景下暖季逐渐长于冷季；分别将InSAR监测结果与近7年沱沱河盆地GNSS监测数据对比，两者趋势一致；引入降水和气温因素后发现冻土区形变具有显著聚集特征，在人类活动频繁地区存在较大形变。该研究对冻土防灾减灾、保障人民生命财产安全具有重要意义，为高纬度冻土工程建设提供一定借鉴。

冻土的水-冰相变交替过程会造成水文环境与地表工程的破坏，从而导致路基塌陷、山体滑坡、洪水暴发以及冰川溃决等灾害，智能感知潜在风险对保护冻土区工程建筑具有重要意义。采用2017年01月—2023年04月194景Sentinel-1A 影像，利用SBAS-InSAR技术获取了黄河上游沱沱河盆地冻土区形变结果，冻土地表形变明显且空间分布不均匀，监测时间段内最大形变速率可达13 mm/年。冻土区青藏铁路路基形变呈现“冻胀融沉”的季节性变化，暖季匀速沉降，冷季缓慢抬升，在气候变暖背景下暖季逐渐长于冷季；分别将InSAR监测结果与近7年沱沱河盆地GNSS监测数据对比，两者趋势一致；引入降水和气温因素后发现冻土区形变具有显著聚集特征，在人类活动频繁地区存在较大形变。该研究对冻土防灾减灾、保障人民生命财产安全具有重要意义，为高纬度冻土工程建设提供一定借鉴。

冻土的水-冰相变交替过程会造成水文环境与地表工程的破坏，从而导致路基塌陷、山体滑坡、洪水暴发以及冰川溃决等灾害，智能感知潜在风险对保护冻土区工程建筑具有重要意义。采用2017年01月—2023年04月194景Sentinel-1A 影像，利用SBAS-InSAR技术获取了黄河上游沱沱河盆地冻土区形变结果，冻土地表形变明显且空间分布不均匀，监测时间段内最大形变速率可达13 mm/年。冻土区青藏铁路路基形变呈现“冻胀融沉”的季节性变化，暖季匀速沉降，冷季缓慢抬升，在气候变暖背景下暖季逐渐长于冷季；分别将InSAR监测结果与近7年沱沱河盆地GNSS监测数据对比，两者趋势一致；引入降水和气温因素后发现冻土区形变具有显著聚集特征，在人类活动频繁地区存在较大形变。该研究对冻土防灾减灾、保障人民生命财产安全具有重要意义，为高纬度冻土工程建设提供一定借鉴。

冻土的水-冰相变交替过程会造成水文环境与地表工程的破坏，从而导致路基塌陷、山体滑坡、洪水暴发以及冰川溃决等灾害，智能感知潜在风险对保护冻土区工程建筑具有重要意义。采用2017年01月—2023年04月194景Sentinel-1A 影像，利用SBAS-InSAR技术获取了黄河上游沱沱河盆地冻土区形变结果，冻土地表形变明显且空间分布不均匀，监测时间段内最大形变速率可达13 mm/年。冻土区青藏铁路路基形变呈现“冻胀融沉”的季节性变化，暖季匀速沉降，冷季缓慢抬升，在气候变暖背景下暖季逐渐长于冷季；分别将InSAR监测结果与近7年沱沱河盆地GNSS监测数据对比，两者趋势一致；引入降水和气温因素后发现冻土区形变具有显著聚集特征，在人类活动频繁地区存在较大形变。该研究对冻土防灾减灾、保障人民生命财产安全具有重要意义，为高纬度冻土工程建设提供一定借鉴。