interesting findings for me so far:

  • Overall, more acres burn on weekends than weekdays. Lightning burns more acres on Tuesdays, warming fires burn more acres on Sundays, and cooking fires burn more acres on Saturdays.
  • The vast difference between human-caused acres burned (193,554) and naturally-caused acres burned  (7,559). The bubble graph makes this very obvious.
  • Looking at the trend lines in the bottom graph, we see that human-caused fires have increased over time. This is to be expected as the population has also increased over time. More humans = more human-caused fires. What I would like to do next is add the population data into the dashboard and see if the *rates* are increasing proportionally. In other words, is the number of human-caused fires increasing proportionally with the population increase? If so, perhaps our fire awareness and mitigation campaigns could use a boost. But if human-caused fires are not increasing as fast as the population then perhaps our awareness and mitigation campaigns are doing some good.
  • This one is less obvious. You have to use the timeline editor to see it. It's in the top left graph, Fire Season by Month. When all the data is considered at once, the numbers of both human-caused and naturally-caused fires peak in July. But if you go to the timeline editor, grab the left handlebar slider and slide it right and stop around 2010, you'll see something interesting take place. This drops pre-2010 data from the analysis and just shows 2010-2014 and reveals something interesting: in terms of numbers of fires, human-caused fires in the graph shift to the left, peaking in June, while the naturally-caused fires still peak in July.  It appears that in the last five years, human-caused fires are peaking (in terms of numbers) a month earlier than they used to.
  • Even though human-caused fires have burned exponentially more acres (193,554) than naturally-caused fires (7,559), lightning remains the most frequent cause of all fires. Warming fires, cooking fires, and smoking come in 2nd, 3rd and 4th. 
  • Fires caused by smoking had their annual peak at 13 in 1986, and have declined ever since. Since 2006 smoking has at most caused only one fire per year.
  • Between 1981-2007, pyromania was the cause of less than 5 fires per year. Between 2008-2011 there were none. But in 2012 there were 19! Did the pyromaniacs discover new confidence in 2012? Or was the definition changed? What's going on here?

The purpose behind these visualization is to highlight aspects of fires in a way that educates the public. These visualizations should answer simple questions like:

  • When is the fire season?
  • Where have fires occurred?
  • How many fires have there been?
  • How big were the fires?
  • How were the fires caused?
  • Were there more human-caused fires or natural fires?