Monitoring deformation in high undulating mountainous environments is critical for surface process research and disaster prevention studies. Although observations based on interferometric Synthetic Aperture Radar(InSA...Monitoring deformation in high undulating mountainous environments is critical for surface process research and disaster prevention studies. Although observations based on interferometric Synthetic Aperture Radar(InSAR) are an excellent tool for monitoring deformation, the shadow phenomena can limit its application. Based on a series of geomorphic parameters and limited InSAR observation data, surface deformations were reconstructed in areas with missing observations by constructing a random forest model to compensate for the shadow phenomenon at the grid-scale. The findings suggest that this method can be used to rebuild landscape variation characteristics in places where observation data is lacking. The dominant slope direction in the observation area corresponded to a more significant correlation between the reconstructed topography deformation characteristics and the observation. In addition, when building this model, consideration was given to the geomorphic parameter selection, elevation variation, hypsometric integral value, slope form, lithology, slope variation,and aspect variation;these parameters can significantly affect the surface deformation, which is closely related to their spatial autocorrelation. These findings are significant for eliminating the shadow phenomenon, which often occurs in In SAR observations taken over alpine canyon regions. The terrain and lithology of the underlying surface should be considered when reconstructing the surface deformation characteristics of the shadow region by using satellite observation data.展开更多
基金financially supported by the National Natural Science Foundation of China (42107218)China Geology Survey Project (DD20221738)+1 种基金China Three Gorges Corporation (YMJ(XLD) (19) 110)the National Key Research and Development Program of China (2018YFC1505002)。
文摘Monitoring deformation in high undulating mountainous environments is critical for surface process research and disaster prevention studies. Although observations based on interferometric Synthetic Aperture Radar(InSAR) are an excellent tool for monitoring deformation, the shadow phenomena can limit its application. Based on a series of geomorphic parameters and limited InSAR observation data, surface deformations were reconstructed in areas with missing observations by constructing a random forest model to compensate for the shadow phenomenon at the grid-scale. The findings suggest that this method can be used to rebuild landscape variation characteristics in places where observation data is lacking. The dominant slope direction in the observation area corresponded to a more significant correlation between the reconstructed topography deformation characteristics and the observation. In addition, when building this model, consideration was given to the geomorphic parameter selection, elevation variation, hypsometric integral value, slope form, lithology, slope variation,and aspect variation;these parameters can significantly affect the surface deformation, which is closely related to their spatial autocorrelation. These findings are significant for eliminating the shadow phenomenon, which often occurs in In SAR observations taken over alpine canyon regions. The terrain and lithology of the underlying surface should be considered when reconstructing the surface deformation characteristics of the shadow region by using satellite observation data.