VandalFire ML Framework: Validation Part 3
收藏资源简介：
Vandal Fire Intelligence (VFI) : Validation Part 3 (2025) This document is part of a two-part research package submitted to Zenodo. It is supported by the companion Jupyter Notebook 'vf-validation-part3.ipynb`, which contains all source code, data diagnostics, and visualizations referenced herein. The PDF version of that notebook is `vf-validation-part3.pdf`. Contact: Jeffrey Logan, Earth and Spatial Sciences, University of Idaho Email: jeffrey.logan@uidaho.edu ORCID: 0009-0001-9415-5809 VFI is a high‐performance, Dask powered platform developed to deliver dynamic, scalable flammability predictions to serve hyperlocal needs and global flammability analyses (Logan, Smith 2024). VFI informs fire managers, climate researchers, and NASA/NSF-aligned programs about evolving flammability risk patterns under changing environmental conditions. This builds on other documented efforts to explore single climate variables associated with wildfire activity in the Columbia River Basin Area of the United States from 1979-2025, such as Wind Speed Trends (https://doi.org/10.5281/zenodo.15485100) 1000 Hour Fuel Moisture Trends (https://doi.org/10.5281/zenodo.15446652) Vapor Pressure Deficit Trends (https://doi.org/10.5281/zenodo.15391290). Vandal Fire Intelligence (Columbia River Basin https://doi.org/10.5281/zenodo.15580220).
