Space Apps Challenge 2017

Inspiration

Wildfires have global impact. They are hazardous to human health. The US spends almost $2 billion per year on the suppression of wildfires alone (i.e. not including loss from fire damage). Wildfire frequency has increased fourfold since the mid-1980s. Wildfires are a major source of atmospheric aerosols and carbon emissions

whAT IT DOES

Automatically detect fires in Landsat imagery. User can define any time and place on Earth for analysis. Enhanced spatial resolution allows fire location, size and shape to be quantified. Predicts fire propagation using local wind vectors. Synergy with existing fire detection systems that use MODIS-class sensors. Overlays precipitation data for contiguous US.

How we built it

This is built on Google Earth Engine via their Javascript API.

Challenges we ran into

Deciding which map technology to use. Technical issues such as discovering that there were some functions that could only be executed server side which forced us to code workarounds. Discovering that some satellites didn't provide the volume of data that we need. Figuring out how to show wind direction arrows.

Accomplishments that we're proud of

We provided a slick UI, created several map visual layers, created an algorithm to pick out likely fires while reducing false positives, created wind vectors for fires, and showcased a video showing fire patterns.

What we learned

Google Earth Engine API and satellite imaging.

What's next for HotMap

Make this a mobile app that fire monitoring teams can use in real time by teaming up with partners (e.g. http://live.farearth.com/) that are able to receive and process NASA data in real time as it is being received by ground station demodulators.

Try It

https://code.earthengine.google.com/c85bf3c19769d4... (Requires a Google Earth Engine developer account).

https://github.com/samsammurphy/nasa-space-apps (A clone of our repo due to issues with permissions on the Google Earth Engine repo).