Every month, the Goddard Institute for Space Studies publishes an update on global surface temperatures. It includes maps showing how temperatures throughout the globe have varied from the long-term mean over the course of a month or multiple months; and graphs depicting the evolution of the globe’s mean surface temperature over time.
Yesterday, the image to the left, reproduced from the NASA GISS graphs page, caught my attention. Each panel represents 30 years of climate information, with the rows (running horizontally) representing months, and the columns (running vertically) showing years.
So each colored square depicts how the Earth’s mean surface temperature departed from the long-term mean during a specific month in a particular year. The warmer the color, the warmer the temperature. (See the scale at the bottom for specifics.) The graphic also shows minima and maxima in the Sun’s activity cycle (“m” and “M” respectively), as well as El Niño and La Niña episodes (“E” and “L” respectively, with upper case or lower case indicating strength).
I found the three panels quite compelling, but I also realized that if I stitched them together in one continuous, horizontal graphic, the climatic warming trend would be clearer to see — and that the resulting image would also be quite beautiful, if unsettling. The result was the long, horizontal image that follows.
Many thanks to the people who participated in my little contest. As the comments section indicates, three people answered the question correctly.
Oh, and one last thing: the little gray squares at lower right in the 2010 column represent the months of November and December. The January through October period has already been the warmest such period on record, according to GISS. Check back here for the final verdict on 2010, which will probably be out in early January of 2011.
Here’s the original post from yesterday, along with the graphic I created using the three panels of climate data:

No, this is not a long-lost Piet Mondrian canvas.
A hint: I’ve taken three charts of climate data and stitched them together to form a single panel. I haven’t done it simply to create a puzzle. I think this way of depicting the data shows a climate trend more clearly than the original three charts viewed separately.
But where did the originals come from? What do each of those little colored rectangles represent? And what specific trend is depicted here? If you think you know, share your answer in the comments section below. I don’t have any prizes to give away. But if you get it right, I’ll share your accomplishment with CEJournal readers, along with additional details.
One last thing: I don’t mean to make light of what this graphic depicts. But when I viewed the original images this morning, I was struck by the beautiful way they depicted scientific — and unsettling — information.
– Tom Yulsman

This thing has 18 Comments
Average monthly temperature deviation over the last 90-ish years?
Hey Jerry, let’s just wait and see if any one else posts a comment on this. But in the meantime, what does each box represent?
Decadally-averaged anomalies by 60S-60N lat/long over several recent decades.
Oops. Let me amend that to say each column has to be a single year, but still with 60S-60N. I tricked myself because I’ve seen the other thing done.
Rows are months, columns are years, colour is global atmospheric temperature anomaly.
Nice image!
Byron has it. 1998 is obvious now that I look more closely.
Byron is right! Hooray Byron!
But a consolation prize goes to Jerry, who weighed in almost immediately — and who was correct. He just didn’t offer enough detail to get the Kewpie doll.
So now, can anyone actually find the source? And what are the two gray rectangles at bottom right?
source – GISS temperature record?
two gray rectangles – blanks for Nov and Dec 2010?
BINGO! Steve is exactly correct. The source is the Goddard Institute for Space Studies Surface Temperature Analysis: http://data.giss.nasa.gov/gistemp/graphs/ And those two gray rectangles at the lower right are awaiting the data for November and December. In just a bit, I’ll publish some additional details about this, including the original charts on which this was based.
Thank you all for participating!
Nice challenge. Counting the number of rows was what gave it away for me.
Thanks so much Byron! (I just wish I had a prize to give away.)
So what was happening in the early forties to the planet? And no, i don’t think war and atmospheric nuclear testing are genuine answers.
Brilliant image! Makes me wonder what a diagram of Arctic Sea Ice or northern Hemisphere snow cove extent would look like.
Mauri: Thanks for contributing. And btw: You have a brilliant blog!
For other readers, it’s called “From a Glacier’s Perspective.” Here’s the link: http://glacierchange.wordpress.com/
Back to the image, I was struck by it because the use of color in that particular way communicated to me more viscerally than an ordinary line graph of global average temperature over time. But separating the three panels as NASA GISS does (to fit it on the page I suppose) weakened it. That’s why I stitched them together.
I see no reason why one couldn’t depict sea ice or other data like this. It would just be a very time consuming task! (Maybe worth it though.)
Tom take a look at the North American snow cover extent graph I made today using the same philosophy as you did, on the glacier blog. It needs work still but since my blog focuses on glaciers you can use this as a compliment. Tell me what should be changed. The key is warming leads to summer extent changes, since that is when temperature is the dominant control of snowcover.
Tom @ 8:18am and Mauri @ 3:52am
There are already graphs and maps just like that for Arctic and Antarctic Sea Ice. They’re available at James Hansen’s Columbia University web page on climate science:
http://www.columbia.edu/~mhs119/
(Just click on “Sea Ice Area” and then “more figures”)
Sea Ice Area: Graphs
http://www.columbia.edu/~mhs119/SeaIce/SI_moreFigs/
From those pages, there are plenty more maps/graphs for global temperature, sea level, CO2, etc…
Many thanks Michael!
I find this form of visual communication of scientific data particularly compelling. It is visually arresting and portrays the scientific content almost instantly — and viscerally.
Speaking of visual communication, this:
http://www.youtube.com/watch?v=JRayIgKublg
lots of good stuff in there.
One Trackback
[...] what story? Image stitched together by Tom Yulsman at CE Journal, question answered there. As he says, looks like Mondrian, but this isn’t about aesthetics it’s about effective [...]