PIV analysis of flow-type landslides under suboptimal image conditions

The dynamics of flow-type landslides, including earthflows and debris flows, are still not fully understood, primarily due to the complexity of the physical processes that govern the flow and the challenges in acquiring direct field measurements. In modern monitoring stations, cameras represent cost-effective data sources, providing essential information for characterising documented reactivation events. Particle image velocimetry (PIV) algorithms have been extensively employed in the literature to reconstruct velocity fields and rheological behaviour of laboratory physical models under ideal conditions. However, the resolution of camera footage in the field typically falls short of being optimal due to lighting and weather conditions, as well as non-zenithal recording geometry, hindering a straightforward application of PIV. This study presents two primary sets of laboratory flume tests conducted to explore a broad range of recording conditions, bridging the gap between ideal laboratory settings and actual field acquisitions. The experiments enabled the evaluation of PIV performance for each image quality scenario, detailing and quantifying the main uncertainties as well as their impact on the resulting velocity fields while discussing underlying reasons and mitigation measures. The experimental results reveal that, with due adjustments, suboptimal-quality footage can be used to estimate the actual flow velocity field and infer the rheological behaviour of the flow. Furthermore, distortions related to non-zenithal perspectives can be reliably minimised through suitable orthorectification algorithms. These findings support the potential for broader application of the tested PIV-based methodological approach in field scenarios to investigate the dynamics of flow-type landslides.