Can big data and random forests improve avalanche runout estimation compared to simple linear regression?

被引:5
作者
Toft, Havard B. [1 ,2 ]
Mueller, Karsten [1 ,2 ]
Hendrikx, Jordy [3 ,5 ]
Jaedicke, Christian [4 ,5 ]
Buhler, Yves [6 ,7 ]
机构
[1] Norwegian Water Resources & Energy Directorate, Oslo, Norway
[2] Univ Oslo, Dept Geosci, Oslo, Norway
[3] Antarctica New Zealand, Christchurch, New Zealand
[4] Norwegian Geotech Inst, Oslo, Norway
[5] UiT Arctic Univ Norway, Tromso, Norway
[6] WSL Inst Snow & Avalanche Res SLF, Davos, Switzerland
[7] Climate Change Extremes & Nat Hazards Alpine Reg R, Davos, Switzerland
关键词
Snow; Avalanche; Empirical-runout; Random forest; alpha-beta model; PERIOD;
D O I
10.1016/j.coldregions.2023.103844
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Accurate prediction of snow avalanche runout-distances in a deterministic sense remains a challenge due to the complexity of all the physical properties involved. Therefore, in many locations including Norway, it has been common practice to define the runout distance using the angle from the starting point to the end of the runout zone (a-angle). We use a large dataset of avalanche events from Switzerland (N = 18,737) acquired using optical satellites to calculate the a-angle for each avalanche. The a-angles in our dataset are normally distributed with a mean of 33 degrees and a standard deviation of 6.1 degrees, which provides additional understanding and insights into alpha-angle distribution. Using a feature importance module in the Random Forest framework, we found the most important topographic parameter for predicting a-angles to be the average gradient from the release area to the beta-point. Despite the large dataset and a modern machine learning (ML) method, we found the simple linear regression model to yield a higher performance than our ML attempts. This means that it is better to use a simple linear regression in an operational context.
引用
收藏
页数:17
相关论文
共 51 条
[1]  
Adjel G., 1995, HOUILLE BLANCHE, V7, P100
[2]  
American Avalanche Association, 2016, Standard for Tissue Banking, P1
[3]  
Bakkehoi S., 1983, Annals of Glaciology, V4, P24
[4]  
Bartelt P., 2013, RAMMS RAPID MASS MOV
[5]  
Breiman L., 1984, CLASSIFICATION REGRE, DOI DOI 10.1201/9781315139470
[6]   Automated identification of potential snow avalanche release areas based on digital elevation models [J].
Buehler, Y. ;
Kumar, S. ;
Veitinger, J. ;
Christen, M. ;
Stoffel, A. ;
Snehmani .
NATURAL HAZARDS AND EARTH SYSTEM SCIENCES, 2013, 13 (05) :1321-1335
[7]   Automated snow avalanche release area delineation - validation of existing algorithms and proposition of a new object-based approach for large-scale hazard indication mapping [J].
Buehler, Yves ;
von Rickenbach, Daniel ;
Stoffel, Andreas ;
Margreth, Stefan ;
Stoffel, Lukas ;
Christen, Marc .
NATURAL HAZARDS AND EARTH SYSTEM SCIENCES, 2018, 18 (12) :3235-3251
[8]   Automated avalanche hazard indication mapping on a statewide scale [J].
Buhler, Yves ;
Bebi, Peter ;
Christen, Marc ;
Margreth, Stefan ;
Stoffel, Lukas ;
Stoffel, Andreas ;
Marty, Christoph ;
Schmucki, Gregor ;
Caviezel, Andrin ;
Kuhne, Roderick ;
Wohlwend, Stephan ;
Bartelt, Perry .
NATURAL HAZARDS AND EARTH SYSTEM SCIENCES, 2022, 22 (06) :1825-1843
[9]   Where are the avalanches? Rapid SPOT6 satellite data acquisition to map an extreme avalanche period over the Swiss Alps [J].
Buhler, Yves ;
Hafner, Elisabeth D. ;
Zweifel, Benjamin ;
Zesiger, Mathias ;
Heisig, Holger .
CRYOSPHERE, 2019, 13 (12) :3225-3238
[10]  
Canadian Avalanche Association, 2016, OBS GUID REC STAND W, P1