This type of graph is called a meteogram, and is actually a de facto format for visualizing weather station data. There are about as many types as there are weather agencies. You don't see them on consumer facing weather services so much (In the quest for minimalism, nothing gets ignored quite like a 16 color rectangle with 12 different data series smooshed together).
This particular graph from NOAA, though, is my favorite. I think what makes it work is how well it correlates disparate series: Each plot takes up the same amount of height and is scaled to look like roughly the same amplitude. Closely related data is grouped together, but not so closely to become overcrowded. Subtle touches indicate directionality, or totals, or probability-- But only when it applies. Most impressive is how it behaves at the extremes: During big storms, the data is just as readable as during a relative lull.
If you're curious about the data, I recommend messing around with the URL and seeing what comes up. The URL takes a hilarious number of parameters, most of which are pretty easy to figure out based on their key names. Most interesting is pcmd, which is a string of 58 ones and zeroes:
&pcmd=1101111111111000000000000000000000000000000000000000000000
Each bit turns an extra feature on and off.
A while back, I started hacking together some other ways of consuming NOAA's Meteograms. It's a little broken at the moment, but it'll at least let you see the six day forecast version!
No comments:
Post a Comment