资源环境科技发展态势分析平台

Rain-triggered landslides in loosely packed granular deposits are highly hazardous because of their sudden failure, flow-like motion, and high mobility. Predicting their initiation and establishing monitoring and early-warning strategies based on the actual mechanisms of failure, rather than on statistical-empirical relations, can be challenging. Experiments on artificial granular slopes revealed a systematic instability-triggering process entailing the internal erosion and seepage-driven transport of small grains. This rearrangement of matter generates acoustic signals and microseismicity detectable prior to any pore water pressures spikes or macroscopic deformations, which are common proxies for slope stability monitoring. Acoustic emissions are more pronounced in loose assemblies, while they are much weaker in densely packed mixtures, which will not exhibit fluid-like motion. Acoustic sensors and accelerometers in support to monitoring networks for flow-like landslides can improve our early warning capabilities, as long as their good performance is confirmed in real-scale applications.

Plain Language Summary Some landslides initiated by rainfalls and evolving into very fast and destructive fluid-like movements are hardly predictable because they happen suddenly, with little warning. Traditional monitoring systems might fail to capture precursory signals, and thus, timely alarms cannot always be released. These systems mostly rely on known correlations between the amount and intensity of rainfall and the occurrence of landslides in the past or on the detection of water level and pressure changes by instruments installed within the slopes. However, these changes are often detected late, when the movement is already accelerating dramatically. In our experiments, we found that one of the processes that trigger flow-like landslides in granular soils is the internal erosion of small soil particles transported by water flowing within the slopes. This produces measurable vibrations (microseismicity) and sounds (acoustic emissions) well before water pressure changes and deformation trends can be identified. The performance of real-scale acoustic and seismic monitoring systems is still being verified, but we believe that complementing slopes monitoring with acoustic and seismic sensors can be the winning strategy to provide more timely alarms to evacuate people along the landslide paths and save more lives.