Monitoring for sinkholes using WT-SBAS-InSAR Analysis

Sinkhole events are becoming more frequent in urban areas, making reliable ground monitoring essential. Ground displacement patterns around four sinkhole sites in Seoul were examined using Sentinel-1 SAR (Synthetic Aperture Radar) data collected from 2018 to 2025. The SBAS-InSAR (Small Baseline Subset Interferometric Synthetic Aperture Radar) technique was used to detect long-term subsidence trends. At each site, displacement data were compared between points close to the sinkholes and points farther away. To improve the detection of unusual surface changes, a method combining WT (Wavelet Transform) and SBAS-InSAR was developed, referred to as WT-SBAS-InSAR. Wavelet transform was applied to the original InSAR time series to identify localized frequency changes. These changes appeared near the time of known sinkhole events. InSAR data from distant control points did not show similar frequency increases. The results suggest that satellite-based interferometric methods, especially when combined with time-frequency analysis such as wavelet transform, can help detect early signs of sinkhole formation. These findings also indicate potential for future use in predictive modeling to improve urban infrastructure safety.