Detection and monitoring of active landslides after the initial impoundment of the Baihetan Reservoir (China) using time series InSAR method

Abstract

The Baihetan hydropower station began impounding water on April 6, 2021, and its water level rose from 660 \(m\) to 817m by September 30, 2021. It is of great significance to catalog the landslides dynamically in the reservoir area and analyze the landslides deformation characteristics after impounding water. Stacking InSAR technology with ascending and descending Sentinel-1A data is applied to detect the potential active landslides on the upstream and downstream of Baihetan hydropower station from January 1, 2020 to April 6, 2022. Results show that a total of 76 active landslides were detected, and 21 landslides underwent significant deformation after impounding water. We analyze the relationship between slope stability and triggering factors, such as reservoir water level and rainfall to Mianshawan landslide, to which the displacement time series was obtained by small baseline subset (SBAS) InSAR technique. Results showed that the landslide was stable before impounding water, but increased to 15 cm/a after impounding water. Pearson correlation analysis revealed a 72-day lag between the displacement trend term of the landslide and the reservoir water level fluctuation. Wavelet analysis revealed a 60–90-day time lag between displacement trend term and precipitation, with common power on an interannual scale. The first impoundment of the Baihetan reservoir significantly enhanced the common power of precipitation and periodic term displacement on a quarterly scale. This research provides a reference for landslides detection and kinematics analysis in the reservoir regions.