Avalanche Forecasting with Geospatial Analysis in Southwestern Colorado
Open Access
Author:
Tyrrell, Emma
Graduate Program:
Spatial Data Science
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
November 01, 2023
Committee Members:
Patrick Joseph Kennelly, Thesis Advisor/Co-Advisor Mike Cooperstein, Committee Member Anthony Robinson, Program Head/Chair
Keywords:
Avalanche GIS Spatial Data Science High Resolution Rapid Refresh Model Colorado Avalanche Information Center ArcGIS Pro ArcGIS Insights Arcpy
Abstract:
Avalanches are natural disasters that can be forecast with spatial analysis. This paper performs multiple analyses on avalanches from 2018 to 2022 in January and February along highways in the San Juan Mountain region of Colorado. This project uses data from historical natural avalanches (NatHA) and non-avalanche days (NonAvD) to determine statistically significant variables that contribute to such events via a binary logistic regression (BLR) model. The output of this BLR model provides the basis for an original, weighted Forecast Value Index (FVI) that determines avalanche likelihood. This FVI was tested as a forecasting principle on data that was not included in the initial analysis. Based on days considered within the forecasting model, this approach resulted in a 64% to 72% success rate in predicting avalanches along highway corridor paths, varying on the days measured prior to the avalanche. An ArcGIS Insights web-based mapping application was used to display the data and is designed to be used by avalanche professionals.