Automatic detection of avalanches using a combined array classification and localization

Abstract

Abstract.We use a seismic monitoring system to automatically determine the avalanche activity at a remote field site near Davos, Switzerland. By using a recently developed approach based on hidden Markov models (HMMs), a machine learning algorithm, we were able to automatically identify avalanches in continuous seismic data by providing as little as one single training event. Furthermore, we implemented an operational method to provide near real-time classification results. For the 2016-2017 5 winter period 117 events were automatically identified. False classified events such as airplanes and local earthquakes were filtered using a new approach containing two additional classification steps. In a first step, we implemented a second HMM based classifier at a second array 14 km away to automatically identify airplanes and earthquakes. By cross-checking the results of both arrays we reduced the amount of false classifications by about 50 %. In a second step, we used multiple signal classifications (MUSIC), an array processing technique to determine the direction of the source. Although avalanche events 10 have a moving source character only small changes of the source direction are common for snow avalanches whereas false classifications had large changes in the source direction and were therefore dismissed. From the 117 initially detected events during the 4 month period we were able to identify 90 false classifications based on these two additional steps. The obtained avalanche activity based on the remaining 27 avalanche events was in line with visual observations performed in the area of Davos.

Distributed Acoustic Sensing of Microseismic Sources and Wave Propagation in Glaciated Terrain

“…In microseismic studies, machine learning algorithms have proven useful for detecting near-surface seismic sources 40 and identifying noise time series suitable for interferometric studies 19 . These approaches could be applied to DAS records and enhanced with array techniques 41 such as matchedfield processing used here. Moreover, records from geophones or seismometers installed sparsely along the fiber-optic cable can help to efficiently scan DAS records for repeating glacier stick-slip events: template searches such as shown in Fig.…”

Effect of snowfall on changes in relative seismic velocity measured by ambient noise correlation

“…1b) with an eigenfrequency of 4.5 Hz, and data were recorded using a 24-bit acquisition system with a sampling rate of 500 Hz (van Herwijnen and Schweizer, 2011). To increase the signal-to-noise ratio, the sensors were buried 30 to 50 cm deep as suggested by Heck et al 2018. For this study, we used data from two sensors deployed at a distance of 35 m (yellow dots in Figure 1c).…”