World of Science | Review ISSW2018: Avalanche detection - industry and science

This time: Avalanche detection - Industry and research

When, where and what kind of avalanche has occurred is important information for many institutions in the mountain regions. Obviously, avalanche warning services use this information to make, validate and improve their forecasts - every avalanche that is reported helps everyone else. A slightly different use of avalanche detection systems is closely linked to artificial triggering devices. It is often difficult for the piste and road safety services to directly assess the success of an artificial trigger: Although you can hear the explosion, you cannot see the avalanche at night or in fog.

Detection also plays an important role in natural damaging avalanches. Not every avalanche path can be built with a gallery, a tunnel or retaining structures in the avalanche outcrop, but it may be possible to install alarm and warning systems. The difference between these two systems in the case of the alarm system is the direct activation of measures such as track closures and traffic lights when an avalanche is successfully detected. And a warning system is characterized by the fact that there is a message even before the avalanche starts.

Exactly such a warning system is presented in article P7.6. The Weissmies and its steep ice walls are first observed with a radar that reacts very sensitively to surface changes in the centimeter range (interferometric radar), later only with a high-resolution camera and "image correlation analysis". The authorities receive a warning of an ice avalanche when there is increased deformation or movement of the ice masses. This is what happened in September 2017, for example, when around 300,000m³ of ice accelerated. The authorities evacuated the affected residents of Saas Grund and less than 24 hours later, the ice avalanche released in several bursts so that it did not reach the village and did not cause any damage.

Researchers want more than just binary measurements

Research uses radar systems not only for detection (yes/no is a binary result), but above all, the characterization of the impact properties of different avalanche shapes is the typical scientific application. Radar research is prominent in the Swiss avalanche test site "Vallée de la Sionne" in Valais. Three articles provide an overview of the measurements taken in recent years: P7.1 about the Doppler radar measurements, P7.7 about a radar that records the position of an avalanche in high resolution, and O7.4 compares the high-resolution radar with seismic measurements recorded directly in the ground of the avalanche path.

The full information content in the seismic data is difficult to extract, as the individual impacts of the avalanche, which affect the ground and thus the geophones in the avalanche path, are too "chaotic". Nevertheless, the contribution O7.5 manages to extract important parameters about the type and size of the avalanche from the amplitude, the impact frequencies and, above all, from the signal of the incoming avalanche.

As mentioned above, one difficulty with seismic data is the many sources of interference, making it difficult to automate the reliable detection of avalanches in continuous data. Computer algorithms similar to speech recognition methods are used in the contribution O7.10, thus enabling operational avalanche detection with seismic sensors within a radius of up to 4 km in the future. An operational seismic application is presented in article O7.11: The sensors are installed in the outcrop areas of potential avalanche paths.

Continued on next page -->

Four articles deal with detection using infrasound sensors. Articles O7.6, O7.7 and O7.8 compare the detection rate based on manual avalanche observations and visual recordings with, for example, panoramic cameras. They come to the similar conclusion that avalanches with a length of more than 500m (size 3 and larger) can be detected relatively well within a radius of 3km around the station. Contribution O7.2 reports on low-cost infrasound sensors based on Arduino microcontrollers, which can be used easily and mobile.

Surprisingly, there is only one contribution in the session, P7.2, which extracts signatures of avalanches from satellite data. For this, they use the algorithms from the Norwegian researchers, who are leaders in the field and have presented many papers in the operational remote sensing session.

Conclusion

This session at ISSW2018 is a prime example of the motto "merging theory and practice", as there are roughly equal numbers of contributions from the detection industry and from research into detection methods. Radar, seismic and infrasound each have advantages and disadvantages for detection purposes, and until these are eradicated or circumvented by combining methods, manual observation and feedback will remain important for avalanche warnings and avalanche commissions. For us winter sports enthusiasts, this means: report avalanche activity to the local avalanche warning services, especially after storm periods or from remote areas.