Exactly what I was thinking. There are 3 different off-white colors that indicate vastly different temperature variants. This map is incredibly ambiguous. Which white zones are -8° below average, roughly the same as average, or 13° above average?
The neighbouring colors are different though, so since it’s not going to jump from +10 to -8 you can unequivocally tell. Honestly I think I hate this less than having 8 shades of slightly different red
I might be the only one who doesn’t mind the color scheme. For the temperature range on the map, the only colors that don’t fit into the typical blue-to-red gradient are the green for really cold, which makes the coldest anomalies really pop, and the white for really hot. I get the complaints about white being used for both normal average temperature and for extreme heat, but it’s easy to tell which is which by the neighboring colors, and it almost looks as if France got so hot that it got a third degree burn, which chars the flesh, nerves included, and no longer feels like anything - thus insensate white.
So MAYBE, and I’m pulling this out of my ass with no background here, but expectation is temperature doesn’t jump, but flows as a gradient.
Using France as the start, we’ve white fading to dark greys then reds, which is the hottest of the three white possibilities. As going hotter then that gets pink again, the top end is white in France. We then decrease down the scale until we get those. green pockets. Light green/white touching green would signify the lower of the three white temps.
Not a great map, but perhaps it’s an understood practice with the field.
After all, how do you convert a quantitative scale into qualitative data.
You can’t really just number everything (people have shit attention spans and they’ll gloss over immediately. Anybody who’s delivered technical data to management can attest to that lol)
Color works well for that, but has a limited useful spectrum.
Getting too specific in a single spectrum muddies the graphic (what’s the exact color over Lisbon here? This kinda salmony guy? So 8? Or closer to 9?)
Basically, their use of brightness within hue bins makes this a non-function. Notice how its gets whiter closer to 0 and closer to 13. If two values of Y can get plugged into a function and they both return 0 for the expected X term, thats a non-function.
Temperature, or rather, difference in temperature from expected, which is whats being plotted here, is about divergence from normal.
To fix this map, pick a divergent color scheme, center 0, then create bins at either a specific interval or at quantile intervals.
While I do appreciate the information on better infographic creation, the example map has such a small range comparably. There’s over 30 values, not to mention the shades in-between values. I think a two color gradient would end up being very smooth at this scale. Sorta looks to generally drop in temperature as you go east here, nice red to blue fade.
Expanding the color palette does give more room for distinction, but that’s seemingly how they got where they did.
To be fair, from my friends who’ve actually had color theory and graphic design classes, STEM folks tend to do a poor job of communicating well.
So eh. Maybe it’s pointless for me to argue against it.
To be fair, from my friends who’ve actually had color theory and graphic design classes, STEM folks tend to do a poor job of communicating well.
it’s a way of thinking about your audience, i think. when i worked in accounting and had to present to and teach accountants: simple formatted tables, stuff that looks like excel, it didn’t need to be bright or colorful. it just needed to have colors that were unobtrusive. i’m working in music now and if my seminars don’t have showmanship (and if i use powerpoint at all) i’m hosed. different audiences have completely different styles that they are used to communicating with, and if you adapt to that you’re going to have more success
so like, i’m sure the STEM folk are fine at communicating within their field. outside? well, a few of my uncles were engineers. does it show?
What a horrendous color map. The true anarchy in this post is the cartographers use of hue.
Exactly what I was thinking. There are 3 different off-white colors that indicate vastly different temperature variants. This map is incredibly ambiguous. Which white zones are -8° below average, roughly the same as average, or 13° above average?
The neighbouring colors are different though, so since it’s not going to jump from +10 to -8 you can unequivocally tell. Honestly I think I hate this less than having 8 shades of slightly different red
Like I don’t have time to get into it cus I’m at work, but their use of hue is a non-function the color space.
It’s clear enough to me.
Could you try instead telling me the general message you get from the map?
Context clues
I get it,it is still interpreatable. but still a poor colorbar choice.
Do you honestly believe that someone like Joe Rogan could actually understand a graph with this much ambiguity?
If comprehension by Joe Rogan is the standard I think we might as well just give up now.
Just a map with “hot as fuck bro!” printed across it.
Maybe some flames at the hottest spot.
Also, fuck that moron.
I might be the only one who doesn’t mind the color scheme. For the temperature range on the map, the only colors that don’t fit into the typical blue-to-red gradient are the green for really cold, which makes the coldest anomalies really pop, and the white for really hot. I get the complaints about white being used for both normal average temperature and for extreme heat, but it’s easy to tell which is which by the neighboring colors, and it almost looks as if France got so hot that it got a third degree burn, which chars the flesh, nerves included, and no longer feels like anything - thus insensate white.
So MAYBE, and I’m pulling this out of my ass with no background here, but expectation is temperature doesn’t jump, but flows as a gradient. Using France as the start, we’ve white fading to dark greys then reds, which is the hottest of the three white possibilities. As going hotter then that gets pink again, the top end is white in France. We then decrease down the scale until we get those. green pockets. Light green/white touching green would signify the lower of the three white temps. Not a great map, but perhaps it’s an understood practice with the field. After all, how do you convert a quantitative scale into qualitative data. You can’t really just number everything (people have shit attention spans and they’ll gloss over immediately. Anybody who’s delivered technical data to management can attest to that lol) Color works well for that, but has a limited useful spectrum. Getting too specific in a single spectrum muddies the graphic (what’s the exact color over Lisbon here? This kinda salmony guy? So 8? Or closer to 9?)
So there is a lot of work on this in color theory, and you can go deep into this… Here are a few chapters on the matter: https://slcc.pressbooks.pub/maps/chapter/9-2/ , https://courses.ems.psu.edu/geog486/node/876 , and this part in particular https://courses.ems.psu.edu/geog486/node/877
Basically, their use of brightness within hue bins makes this a non-function. Notice how its gets whiter closer to 0 and closer to 13. If two values of Y can get plugged into a function and they both return 0 for the expected X term, thats a non-function.
Temperature, or rather, difference in temperature from expected, which is whats being plotted here, is about divergence from normal.
To fix this map, pick a divergent color scheme, center 0, then create bins at either a specific interval or at quantile intervals.
Something like this: https://colorbrewer2.org/#type=diverging&scheme=RdYlBu&n=11
While I do appreciate the information on better infographic creation, the example map has such a small range comparably. There’s over 30 values, not to mention the shades in-between values. I think a two color gradient would end up being very smooth at this scale. Sorta looks to generally drop in temperature as you go east here, nice red to blue fade.
Expanding the color palette does give more room for distinction, but that’s seemingly how they got where they did.
To be fair, from my friends who’ve actually had color theory and graphic design classes, STEM folks tend to do a poor job of communicating well.
So eh. Maybe it’s pointless for me to argue against it.
it’s a way of thinking about your audience, i think. when i worked in accounting and had to present to and teach accountants: simple formatted tables, stuff that looks like excel, it didn’t need to be bright or colorful. it just needed to have colors that were unobtrusive. i’m working in music now and if my seminars don’t have showmanship (and if i use powerpoint at all) i’m hosed. different audiences have completely different styles that they are used to communicating with, and if you adapt to that you’re going to have more success
so like, i’m sure the STEM folk are fine at communicating within their field. outside? well, a few of my uncles were engineers. does it show?