The data-driven piece of my performance
The data-driven piece of my performance
Local artist Son Visual opened their Forest Fires gallery exhibition with an “Art & Activism Gathering” inviting presentations or performances related to themes of nature, fire, conservation, and grief. I took this as an opportunity to combine data-driven visuals and algorithmic music around a topic I’ve found fascinating: local fire ecology.
Each year is the same length of time, and has a burning and recovery section.
burning
Each prescribed burn is one note. The pitch and instrument of each note is decided by the fire’s DNR “Class” code: A–E, which roughly categorizes its size. The loudness of the note is taken more exactly from the fire’s size—from the acres burned in the fire.
The year’s fires are evenly spaced at the beginning of the year’s musical section (rather than relative to their timing in the year.) This means years with a lot of burns take longer to reach…
recovery
in which all the fire-notes echo, feedback, and fade, as the land heals. Highly burned years have louder echo material and less time before the year resets. During this time, I played echoing guitar riffs along with the resounding chord.
At the start of each year, the chord progresses.
(The underlying beat and my guitar playing are dataless, just musical.)
Bubble positions are taken from latitude-longitude of where the fire was started. This way, they gradually fill in the shape of Michigan! The size of the bubbles is taken from the size of the fire—though not to scale, to remain visible. Their color is taken from the fire’s time of year, evenly fading from greens-orange-teal for the spring-summer-fall-winter. Lots of green as most burns are conducted in the spring! In the recovery period, fires fade but their impressions remain.
The Michigan DNR offers an interactive map of fire events. Behind this is a database from which I was able to download their complete records of completed fires. (2007–”Present” refers to date of download, 9/26/2023.)
I used Google Sheets to filter these records down, eventually working with only each burn’s start date, size, and location.