Item: Statistical Avalanche Runout Models: How Well Can Computers Predict Beta?
-
-
Title: Statistical Avalanche Runout Models: How Well Can Computers Predict Beta?
Proceedings: Proceedings, 2012 International Snow Science Workshop, Anchorage, Alaska
Authors:
- Alexandra Sinickas [ Department of Civil Engineering, Un iversity of Calgary, Alberta ]
- Bruce Jamieson [ Department of Civil Engineering, Un iversity of Calgary, Alberta ] [ Dept. of Geoscience, University of Calgary, AB, Canada ]
Date: 2012
Abstract: We reviewed the performance of a 23 m and 30 m Digital Elevation Model (DEM) and Google Earth in predicting beta (β) points for statistical runout modeling. Our objective was to find the resolution of DEM that was comparable with field survey error. We compared predicted β against fielddetermined β for 30 paths, using a field survey error range approximated from several error and repeatability measures. We found that none of the digital methods predicted β entirely within the range; however the 23 m DEM and Google Earth performed best. The 23 m DEM was biased to more conservative (downslope) estimates. The data suggested that a resolution of less than 23 m was required to match the error of a field survey; however, higher accuracy data may justify using a coarser resolution digital model. Judgment involved with β placement was a major contributor to uncertainty in this study.
Object ID: issw-2012-716-722.pdf
Language of Article: English
Presenter(s): unknown
Keywords: extreme avalanche runout, statistical model, alpha-beta, runout ratio, repeatability, error
Page Number(s): 716-722
Subjects: avalanche runout data terrain model field data
-