Beyond and beneath displacement time series: towards InSAR-based early warnings and deformation analysis of the Achoma landslide, Peru

Abstract. Detecting precursors to slope destabilisation with sufficient lead time and accuracy remains a challenging and unresolved issue in landslide hazard assessment and prediction. This is key, as catastrophic landslides often go unnoticed until immediately before failure, limiting opportunities for intervention. While in situ methods offer high accuracy at point locations, they are costly and require prior knowledge of instability. Satellite-based synthetic aperture radar differential interferometry (InSAR) has shown promise in identifying unknown landslides over large areas and has been proposed as a potentially useful tool for failure prediction. Typically valued for retrieving displacement time series, InSAR time series reliability depends heavily on successful interferogram unwrapping, which often leads to severe underestimations over landslides. Here, we analyse the deformation process of the Achoma landslide in Peru and demonstrate that the InSAR signal contains precursors based on alternative markers, even without displacement time series. Interferometric coherence shows the formation of gravitational structures up to five years before failure, as well as a critical shift in landslide behaviour three months prior to failure. Additionally, a marker based on the wrapped phase reveals and quantifies alternating periods of quiescence and motion, the latter becoming more frequent in the two years before failure. Our findings highlight the potential to use alternative InSAR signal markers to observe the deformation process and progressive failure leading up to the event, and to detect landslide precursors across extensive areas, providing valuable lead time for intervention and disaster prevention.