Item: Afrench local tool for avalanche hazard forecasting: astral, current state and new developments
Title: Afrench local tool for avalanche hazard forecasting: astral, current state and new developments
Proceedings: 2002 International Snow Science Workshop, Penticton, British Columbia
Authors: Laurent Merindol , Gilbert Guyomarc'h and Gerald Giraud, Meteo-France, Centre d'Etudes de la Neige (Cen: Snow Study Centre)
Abstract: In several ski resorts, we have distributed a tool named Astral. Every days, this software calculates an Euclidean distance and correlation with past days, in order to select ten analogous days by a nearest neighbor approach (see Bois 75 and Buser 89). The input variables are : temperature, daily amount of fresh recent snow and water equivalent, sum of temperature at 13 hour above zero threshold, wmd velocity, .... A principal component analysis is done on these raw variables in order to have a data reduction and almost to obtain independent variables before the analogous step. The outputs are the raw variables concerning the ten selected analogous days and the corresponding observed avalanche activity. A notable insufficiency of this present tool is that no internal variable of snow pack is used to determine analogous days. This is due to the lack of daily observed snow pits. To avoid this default, a new improvement is to use results of the operational snow modeling chain Safran (weather analysis) Crocus (snow analysis) Mepra (stability analysis). We add new variables for a better characterization of the snow pack and its stability. Our aim is after wards to use the data from La Plague resort to validate the result of this approach.
Keywords: statistical method, avalanche forecasting, weather and snow pack parameters
Digital Abstract Not Available