Alaska Project Ideas, mentored by the researchers and collaborators of University of Alaska and supported by open-source entities and enthusiasts in Alaska.

Python for Raw Sentinel-2 data (PyRawS) is an open-source software providing utilities to open and process Sentinel 2 RAW data, which corresponds to a decompressed version of Level-0 data with additional metadata. The software is demonstrated on the first Sentinel-2 dataset containing raw data for warm temperature hotspots detection/classification.

Deep Learning Models for Wildfire Danger Forecasting

A deep learning approach for mapping and dating burned areas using temporal sequences of satellite images

http://ibm.biz/cfcsc-wildfires - predict the wildfire/bushfire area for 7 regions in Australia for each day in February 2021

Teleconnection-driven vision transformers for improved long-term forecasting

Download wildfire hotspots detected by NASA satellites and the Fire Information for Resource Management System (FIRMS)

Building models that can predict whether an area is at risk of a wildfire or not on satellite images.

Implementation of Dabrowski et. al., "Bayesian Physics Informed Neural Networks for Data Assimilation and Spatio-Temporal Modelling of Wildfires"

Convolutional neural network model based on the architecture of the Faster-RCNN for wildfire smoke detection.

A large dataset (500+ images) of past wildfire from Copernicus EMS using Sentinel-2 images in the period 2017- 2023

Download wildfires data from the National Interagency Fire Center

Download wildfires data from NOAA satellites

Download wildfire incidents data from InciWeb

Detection & monitoring platform of wildfires

Download watch, warning and advisory data from the National Weather Service

Agent-based modeling 2D wildfire suppression simulator tool built on the mesa framework in Python

Simulator based on the real physical phenomenons acting in a wildfire, deals with wind, heat capacities..an more.

SAVeTrEE is a script within Google Earth Engine for classifying areas of vegetation mortality. It prompts the user for a year, duration, and spectral index for which a mortality map should be produced, then fits a trend line to an imagery time sequence of vegetative spectral index values calculated from Landsat multispectral data. The slope of t…

Implementation of several state-of-the-art Deep Learning models for fire semantic segmentation.