Item: FROM POINT TO SLOPE: ASSESSING WEATHER STATION SNOW DEPTH REPRESENTATIVENESS USING LIDAR AND GEOSTATISTICS
-
-
Title: FROM POINT TO SLOPE: ASSESSING WEATHER STATION SNOW DEPTH REPRESENTATIVENESS USING LIDAR AND GEOSTATISTICS
Proceedings: Proceedings, International Snow Science Workshop, Whistler, BC, Canada, 2026
Authors:
- Katreen M. Wikstrom Jones [ Alaska Division of Geological & Geophysical Surveys, Anchorage/Fairbanks, Anchorage, AK, USA ]
- Gabriel J. Wolken [ Alaska Division of Geological & Geophysical Surveys, Anchorage/Fairbanks, Anchorage, AK, USA ] [ Alaska Climate Adaptation Science Center, University of Alaska, Fairbanks, AK, USA ]
- Michael Janes [ Alaska Electric Light & Power, Juneau, AK, USA ]
Date: 2026-09-28
Abstract: Weather station data are widely used by avalanche forecasters to validate meteorological forecasts and to assess snowpack characteristics at specific sites. However, when stations are located in terrain that is not representative, due to slope, aspect, or exposure, this extrapolation can introduce systematic over- or underestimation of snow depth at the slope scale. Such representativeness errors propagate uncertainty into avalanche hazard assessments that typically rely on sparse point observations. Alaska Electric Light and Power (AEL&P) operates a hydroelectric transmission line between Snettisham Hydroelectric Facility and Juneau in southeast Alaska, crossing multiple avalanche paths. Along Peel Arm, the line traverses steep fjord walls and avalanche exposure is particularly high. Following a destructive avalanche in 2008, AEL&P implemented an avalanche mitigation program that included the installation of weather stations to monitor snowpack conditions. One weather station is located on a mid-slope bench ("the Balcony") and the other is at an avalanche diverter structure further downslope ("Diverter 4-4"). Data from these stations are used to assess snowpack stability in nearby avalanche starting zones. In this study, we evaluate the spatial representativeness of weather stations using spatially distributed snow depth information. We conducted seven airborne lidar surveys in 2019-2023 and produced high-resolution snow depth datasets of the Peel Arm study area. Using geostatistical methods, we characterize spatial snow depth patterns, quantify elevational gradients, and generate difference maps to examine interannual and seasonal variability. We then compare station observations with lidar-derived snow depths using buffer-based sampling and correlation analysis to evaluate their ability to represent slopes-scale conditions and as predictors for absolute and evolutionary snow depth in nearby avalanche starting zones. Results show consistent interannual patterns in snow depth across elevations and persistent late-season snow retention throughout the study area. Our station-focused analysis shows high spatial snow depth variability at the weather station sites. Our distance-weighted correlation model shows that weather station snow depth is not a reliable proxy for snow conditions in selected avalanche starting zones at Peel Arm. Our project findings highlight the limitations of extrapolating point measurements to slope-scale problems while providing guidance for improved weather station placement.
Object ID: ISSW2026_P3.31.pdf
DOI: https://doi.org/10.15788/1790099157
Language of Article: English
Presenter(s): Katreen Wikstrom Jones
Keywords: Weather stations, Snow Depth, Spatial Representativeness, Snow Prediction, Lidar.
Page Number(s): 1359 - 1366
-