Wildfire Tracker

Inspiration

We were inspired by the recent events and the continuous impacts of climate change on society and the environment today. This hits close to home, as we have heard stories and experience from the people we know.

What it does

This application is a wildfire detector that takes in data from multiple sources including NASA satellite information to identify factories that contribute to forest fires and predict where they will occur.

How we built it

We started by researching sources of information about wildfires, primarily related to what factors were observable before and during such events. Using the API from NASA's FIRM platform, alongside historical fire data, we were able to make a heatmap using leaflet that will display recent and current wildfires as well as pipe data into a Google Gemini model that will analyze risk factors and predict the occurrence of wildfires in an area.

Challenges we ran into

Some of the challenges we ran into were issues with API and Cloud interfacing, including API keys being changed and applications timing out, stopping us from retrieving data. Branch management within our project has also presented some issues. As well, rendering the html for our website has caused some issues, but we have learned how to deal with issues of disorganization in our commits.

Accomplishments that we're proud of

Identifying 200 current fires and analyzing the weather, at a 90% threshold for confidence, it can identify 75% of fires.

What we learned

We have learned how to improve our problem-solving skills and teamwork in the environment of development. We were able to work together, allocate tasks, and advise and teach each other.

What's next for Wildfire Tracker

Continue to train our AI model with more data, and work to make an efficient and reliable web application that could assist individuals and local governments around the world to prepare for potential wildfires.