Wang, Zhaoxi, Cao, Chen, Yu, Qingbo, Xu, Xin, Wang, Qing, Niu, Cencen, Shen, Jiejie, Han, Mengxia, Fu, Huicheng, Zhu, Kuanxing, Liu, Jing, Xia, Weitong, and Sun, Xun
[Display omitted] • 1. The deformation and land surface temperature data were integrated for analysis. • 2. The canal slopes experienced temperature-derived spatially differential deformation. • 3. The mechanisms that the shady slope was prone to landslides were revealed. Hydraulic engineering has positive effects on agricultural production and soil desalination. However, freeze–thaw cycles can cause instability in canal slopes, emphasizing the need for effective deformation monitoring. Using Sentinel-1B data acquired from November 2017 to August 2019 and applying the small baseline subset (SBAS) InSAR method, two unstable canal segments with vertical deformation rates exceeding –40 mm/year were identified in Songyuan, Northeast China. The canal slopes uplifted in winter and experienced accelerated subsidence in spring, with the shady slope experiencing a deformation rate approximately 10 mm/year faster than the sunny slope. Further investigation, integrating Landsat–8 data collected from November 2017 to August 2019, revealed that the shady slope received less thermal radiation than the sunny slope, with an average annual temperature difference on the land surface of approximately 1 ℃. In winter, the saline–alkali soil particles formed a thick unfrozen water film, which, driven by temperature and soil–water gradients, migrated towards the freezing front, froze, and caused soil frost heave. The shady slope experienced longer freezing periods and deeper frozen depths, up to 240 cm, causing more unfrozen water migration and greater uplift deformation than the sunny slope. In spring, as temperature rose, the soil thawed from the surface downward. Notably, the shady slope thawed at a slower rate than the sunny slope, forming a frozen lens near a depth of 100 cm. Frozen lens acted as a barrier for meltwater infiltration, impeding the dissipation of pore water pressure, making the shady slope prone to surface landslides. The integration of multiple data sources is beneficial for identifying deformation risk in hydraulic engineering. [ABSTRACT FROM AUTHOR]