Applying Terrestrial Laserscanning to Continuous Snow Cover Monitoring in an Alpine Environment – A Feasibility Study : Workshop on Laser Scanning Applications

Köln

University of Cologner

2015

Poster

MonographieSonderdruck

192297

Promptly available information on the spatial and temporal distribution of snow depth in alpine terrain is essential to a wide range of monitoring applications in both snow science and practise. While in-situ data collection of the snowpack is labour-intensive and potentially hazardous, close-range sensing techniques allow monitoring snow depth from a safe distance. One of the most powerful and frequently applied instruments in this context are Terrestrial Laser Scanners (TLS). However, wintry conditions in (high-)alpine terrain, i.e. poor visibility due to snowfall, blowing snow or fog, low temperatures or high natural radiation, can often interfere with TLS data collection or render it impossible.
In the frame of the presented work, a Riegl LPM98-2K TLS was mounted in an easily accessible weatherproof containment, located in a ski area in Western Austria at 2260 m a.s.l., from where it monitored an avalanche slope in 90 to 400 m distance. From 31 January to 5 June 2013 the TLS acquired 632 scans (five scans per day on average), with a mean point spacing of 0.4 - 1.1 m, at an accuracy of } 0.05 m (1 Đ), plus a distance dependent error of . 20 ppm. An automatic scanner control routine was developed and implemented, allowing continuous remote TLS operation during the entire campaign. The data was collected in three separate scan windows, covering a total area of approximately 34,000 m2. A wide range of ancillary infrastructure and data were also available at this location: i) seven fixed avalanche-blasting devices were installed on the ridge above the slope, thus (mostly small) avalanches were artificially released; ii) a prototype avalanche mitigation measure ( eSnowcatcher f) was located on the monitored slope and instrumented with load sensors, measuring static and dynamic forces on the structure; iii) a webcam provided images from the study area every 30 min; iv) several automated weather stations in the vicinity provided meteorological data. By pooling data from all the above-mentioned sources, a systematic evaluation of the feasibility of applying TLS to snow cover monitoring in a (high-) alpine environment could be performed.
The results show, that despite the sometimes adverse weather conditions (e.g. temperatures below -20 C, high wind speeds and heavy snowfall), the instrument was fully functional for almost the entire measurement campaign. Thus, approximately 80% of the TLS scans provided legible information on the snow depth distribution on the monitored slope and its development throughout the winter. The total snow depth and relative changes were checked against the available meteorological data. On several occasions (e.g. 2 February 2013, 4 June 2013) the scanner was also able to detect both artificially and naturally released small snow avalanches and provide data on their cubature and runout length. The measurements could partially be confirmed by visual interpretation of the webcam images and analysis of the forces measured in the Snowcatcher, in case the avalanche had impacted its structure. However, the discussion of the results also showed the limitations of this TLS setup: During periods of intense snow fall, information on the current snow height is usually most critical (e.g. decision support for timing of artificial avalanche release) . due to poor visibility and the relatively slow scanning rate of the employed TLS (max. 4 Hz), the scanner was typically not able to retrieve any data during this time. The results from the analysis of scans recorded before and after these periods were often ambiguous, due to overlapping effects of snowfall, small-scale avalanche activity and wind drift. This setup has since been further developed: It is currently being applied to monitoring a large-scale avalanche slope, endangering road infrastructure in Western Austria, using a TLS with 1000 Hz.