Awards & Nominations
Sin Rocket no hay Paraíso has received the following awards and nominations. Way to go!
PREDICTFIRE
High-Level Project Summary
Wildfires have increased over the past years, with one of the main reasons being climate change. They can grow out of control even for experienced fire prevention units. We have developed a cell-based fire spread simulator that uses NASA's open data for geographic data and soil conditions. We chose to model fire using Rothermel's simple fire model, a well-established fire spread model able to quantify how quickly it moves through terrain. Our simulation is able to predict safe and dangerous zones, along with the time that it will take them to burn. In an effort to analyze how the terrain affects fire spread, we increase terrain dryness and see that it can have deadly consequences.
Link to Final Project
Link to Project "Demo"
Detailed Project Description
* PREDICTFIRE is a cell-based fire spread simulation algorithm that includes Rothermel's "simple" fire model
* It takes as input several variables about geographic data, soil properties and weather (fuel load, depth and moisture, wind speed, terrain slope), and outputs the rate of spread of the fire (m/s)
* We collected this information using NASA's open data (Earthdata Search, GIBS, CMR) for the Moncayo region in Aragón, Spain, and parsed it (common data format, resizing, map alignment, etc.)
* Our algorithm lets the user choose an initial ignition source, and shows how this fire will evolve on the grid. User is able to evaluate safe and dangerous zone, fastest spread zones, etc.
* Terrain properties can be changed (e. g. simulate drier terrain caused by climate change) to see how it affects fire spread, leaving other conditions unchanged (spoiler: it is bad, spreads faster and covers more area)
The algorithm and data parsing for PREDICTFIRE have been developed using Python, using matplotlib for user input and output.
Space Agency Data
* Earthdata Search (ASTER and VIIRS datasets)
* Global Imagery Browse Services (GIBS)
* Common Metadata Repository (CMR)
Hackathon Journey
We are six people from different backgrounds (computer science, industrial and electronics engineering, we even have some experience with agricultural engineering). For some of us it was our first hackathon. To be honest, we chose this project based on an even crazier idea of a videogame where you try to fight fire using different firefighting capabilities, with climate change being one of the main points, as it could make it much more difficult to combat the fires. Being engineers, we spent so much time on the fire simulation that we did not have time to build a game on top of it. It was a great journey and we learned a lot about new technologies, math, NASA's open data and, of course, fire spread.
References
Rothermel's simple fire model implementation: http://www.prairieprojectknowledgehub.org/books/fire/page/rothermels-simple-fire-model
Python libraries: matplotlib, numpy, tifffile, h5py, imageio
Tags
#wildfire, #climatechange