Assessing the potential of soil wetness data for landslide early warning
In mountainous areas, landslides triggered by heavy rain present a serious risk to people and infrastructure. Recent major events in central Switzerland have demonstrated the numerousness, abruptness and difficulty to predict shallow landslides. Consequently, a dedicated research effort has been initiated to advance fundamentals and develop tools for the early warning of landslides at the regional scale.
While most studies focus on the use of precipitation information to assess thresholds for the initiation of landslides (e.g. by intensities and accumulations), less work has been put into the utilization of soil wetness information to anticipate the imminent occurrence of landslides. In this respect, most attempts were made to estimate (spatial or local) soil saturation with numerical hydrological models to assess the criticality of the antecedent soil wetness in terms of slope stability. Such numerical models however, have limitations in regard to the representation of true soil conditions, and it is very challenging to run them in real-time.
The objective of this project is to assess the value of in-situ soil wetness measurements for its use in a landslide early warning system (LEWS). The study comprises following:
- A comprehensive analysis of available long-term soil wetness data from different research institutions and authorities at several sites all over Switzerland is conducted and the results are compared with a shallow landslide event data set. The aim is to identify statistical values of such time series that antecede observed landslide triggering events and that can separate landslide triggering events from non-triggering events.
- Different soil wetness measurement techniques are compared at a landslide-prone site in the Swiss Prealps (Napf area) to identify advantages or disadvantages, respectively, of certain soil-wetness measurement techniques and to assess the representativeness of flat meausurement sites for critical hydrological conditions in hillslopes.
- A state-of-the-art numerical model for the simulation of soil moisture dynamics will be applied to serve as a benchmark for the soil-wetness-measurement derived indicators.
The project pursues particular innovations in the analysis of soil-wetness time series by identifying characteristic patterns of soil moisture behavior (e.g. variability-mean relationships) that could be used as a precursor of landslides. The analysis of a large and complementary dataset of long-term soil wetness measurements in Switzerland is a unique opportunity to engage with the above research problems.
As an expected outcome of this project, decision-makers and experts responsible for the warning of natural hazards will receive a better knowledge base for the design of a national soil moisture observatory and to issue regional to national warnings regarding imminent landslide hazard.
Project details
Project duration
2018 - 2021
Project lead
Deputy
Adrian Wicki