Permafrost Deformation Monitoring and Interpretation Using InSAR Technique in Northeastern China and Aosta Valley Region, Italy

InSAR algorithms have been widely applied in monitoring and mapping for the purposes of engineering geology research and the InSAR derived products have been extensively used in geological risk management by administrative entities and Emergency Management authorities. In this PhD thesis, the research was focusing on the application of InSAR technique, including conventional DInSAR and advanced SqueeSARTM in cold regions, (1) to locate the deformation hotspot caused by permafrost degradation, then to improve the regional permafrost mapping; (2) to analyze the deformation characteristics caused by the degradation of permafrost temporally and spatially. The research was developed on two sites of cold regions of different geomorphologies. The first case study was focusing on the detection of permafrost degradation phenomena using DInSAR and SqueeSARTM in the low-land permafrost distributed hilly regions in northern Heilongjiang Province, northeastern China. The study area is located at the southern margin of Siberia permafrost region, which is the largest area of the permafrost distribution in the northern Hemisphere. Studies have been revealing that the increasing mean annual air temperature in this study area has been causing the degradation of permafrost in decades and have been bringing geological risks to man-made infrastructures. Initially, the SqueeSARTM analyses using C-band Sentinel-1 and L-band ALOS PALSAR were conducted to reveal the overall displacements time series in the extracted permanent and distributed scatterers of the study area. Then the DInSAR analyses using Sentinel-1 and ALOS PALSAR data were completed to map the deformation hotspot of the study area. Lastly, by combination of the results acquired by SqueeSARTM and DInSAR analyses, the possible spatial distribution of permafrost deformation hotspot was mapped. PS and DS in the permafrost deformation hotspot are selected and analyzed to reveal the characteristics of permafrost degradation in the study area. The results indicate that in the permafrost distributed areas, the deformation velocity has been reduced in recent years from 2015 to 2019 than 2007 to 2010, roughly a decade ago. It could be related to the final degradation phase of the permafrost in recent years. In the second case study, the feasibility of DInSAR and SqueeSARTM in the study of active rock glacier in Alpine environments of mountainous geomorphologies of Aosta Valley Region, Italy was demonstrated, and the deformation characteristics corresponding to the seasonality of active rock glacier were discovered. The seasonal dynamic feature of the active rock glaciers was then analyzed using the regional monitoring results of Aosta Valley Region using Sentinel-1 SqueeSARTM technique and the regional active rock glacier investigation dataset. Interestingly, according to the result of the analysis, the displacement time series of active rock glacier in Aosta Valley Region has shown a half-year lapse compared with the change of regional annual air temperature that is contradictive to the conventional understanding of permafrost deformation dynamics. The driving factor of such phenomena was left unfound and open to the future analysis. Second, combined with high-resolution optical remote sensing imagery, the conventional DInSAR analysis using ALOS PALSAR data collected in summer season of 2007 has detected more active rock glaciers than the regional active rock glacier investigation dataset. The results have provided the possibility of further analysis of improving active rock glacier mapping using InSAR method in the future. In conclusion, this research highlights the value of using spaceborne DInSAR and SqueeSARTM methods in mapping and monitoring active periglacial landforms in cold regions at regional scale. Thanks to its short revisiting time and medium to high spatial resolution, Sentinel-1 data can be used for systematic and continuous monitoring of ground deformation, especially slow and very slow periglacial processes due to the changing climate in cold regions but the effectiveness of the usage of Sentinel-1 in dealing with winter snow cover and dense vegetation should be enhanced. On the other hand, L-band ALOS PALSAR data has solved the problem of vegetation coverage, which extensively exists in the cold region that has limited the capability of InSAR monitoring. The methods mentioned in this thesis are intended to be implemented in the regional or local geological hazard management in both study areas.