SNOWPACK

SNOWPACK is the Swiss snowpack evolution model. This model simulates snowpack layering and characteristics such as density, temperature, and crystal type throughout the season using a set of relatively simple meteorological inputs. This model was developed by scientists from the Swiss Federal Institute for Snow and Avalanche Research (Michael Lehning and Perry Bartelt) with help from other scientists, among them Bob Brown and Ed Adams from Montana State University. It is currently being used operationally in Switzerland.

Chris Lundy setting up the data collection site in Montana's Bridger Range for running SNOWPACK.

The Forest Service National Avalanche Center worked with SFISAR and MSU scientists, as well as avalanche workers from Snowbasin Ski Area (Tom Leonard and Mike Jenkins) and the Utah Avalanche Center (Ethan Greene), to use and test the model during the 1999/2000 and 2000/2001 seasons. Instrumentation to run the model was installed at two sites at Utah's Snowbasin Ski Area. A third site was installed near Bridger Bowl, Montana by Chris Lundy, a graduate student in the Civil Engineering department at Montana State.

 Chris Lundy downloading data to use for running SNOWPACK  Chris Lundy collecting snow crystal data for comparision with SNOWPACK model output   Chris Lundy and Bob Brown at the data collection site

 

Though some instrumentation problems prevented us from collecting complete data sets, we did collect two seasons of data from our Montana site and a season of data from one site at Snowbasin. Chris Lundy analyzed the first season of Montana data for part of his MS thesis, and presented his results at ISSW 2000. He also published an article on this work. His thesis focused on an comparison of the model output with field data he collected. A sampling of some of his model output is shown below.

Modeled snowpack densities during the 1999/2000 season

 

 

Modeled snowpack temperatures during the 1999/2000 season

Modeled snowpack grain type during the 1999/2000 season

 

Our analyses highlighted some of the shortcomings of SNOWPACK, but also demonstrated its promise for simulating the evolution of the seasonal snowcover. Though it clearly cannot replace field observations, it does have the potential to supplement those observations as a tool for the human forecaster. At the 2001 Forest Service National Avalanche Center meeting, we presented the U.S. avalanche centers with a proposal to run SNOWPACK from a centralized location for all of the avalanche centers. Unfortunately, the cost of the project (about $70,000/year) was too expensive for the cash-strapped avalanche centers.

There were sizable improvements in the model perfomance between our two seasons, and the model is undergoing continuing refinements. At an International Glaciological Society conference on snow and avalanches in Switzerland in June of 2003, a number of people were working with the model from many different countries. The National Avalanche Center will continue to keep the U.S. avalanche centers informed on the continuing improvements and refinements to the model.