Tuesday, 20 September 2016

Global Warming goes viral

NOAA's Climate Monotoring Chief Deke Arndt tweeted in May this year: "I look at this stuff every day and it still astonishes me" and showed this graph. Note, the locally typical units.

Honey, I broke the graph. Again.

How the temperature has risen the last years is amazing.

The prediction made this week for 2016 by Gavin Schmidt of NASA GISS does not look good. With a bit of British understatement he is 99% confident 2016 will be a record year.

Attention grabbers

That temperature jump is one reason people have woken from their slumber. Another reason is that people have started visualising the global temperature increase in interesting new ways. Let me try to explain why they work. It all started with the animate spiral of the global temperature of Ed Hawkins that went viral.

The spiral went viral because it was a new way to present the data. It was also very good timing because the spiral especially shows well how extraordinary the current temperature jump is.

The modern currency is attention.

So just after the Olympics I tried to top Ed Hawkins with this visualisation.

By my standards it went viral. It works because visual connects global warming to the famous Olympic photo of Usain Bold running so fast that he can afford to look sideways and smile at the camera.

I guess the virus did not spread beyond the few thousand people discussing climate change every day because without the axes you need to know the temperature signal to get it. Adding axes that would destroy the beauty.

In the last episode of Forecast Scott St George tells about his project to convert the climate signal into music. The different instruments are different climate regions. At the time this creative idea generated a lot of media attention. Works well on radio and TV than a static graph.

More regional detail can be seen in a so-called Hovmöller plot. The plot by Kevin Anchukaitis shows time on the horizontal axis and the colours indicate the average temperature over latitudinal bands. In the lower half of the figure you see the Southern Hemisphere, which warms less than the Northern Hemisphere at the top.

The additional energy that is available due to the stronger greenhouse effect can go into warming the air or evaporating water. The Northern Hemisphere has much more land, is drier. Thus evaporation increases less and warming more.

The front of the new State of the Climate also shows the observed temperature signal in red and brown.

Understanding climate change

Probably the most eye opening graph to understand the difference between short-term fluctuations and long-term trends of the temperature signal is this one. An important source of fluctuations is the El Nino in the Pacific ocean. By plotting years with El Nino, its counterpart La Nina and neutral conditions separately, you immediately see that they all have about the same long-term trend and that El Nino is mainly important for the short term fluctuations. No need for statistics skillz.

I realise that it is just a small tweak, but I like this graph by Karsten Haustein because it emphasises that the data in WWII is not very reliable. The next step would be to also give the decades around 1900 the colour of the orange menace. The data in this period also has some issues and it may well be warming.

This animation by John Kennedy shows the uncertainty in how much warming we had by displaying a large number of possible realisations. This is the uncertainty as estimated in the temperature dataset of the UK Hadley Centre (HadCRUT) due to instrumental changes; it does not include the uncertainty from not having measurements in some areas (such as the Arctic).

If you plot monthly temperatures rather than yearly averages, the warming graph becomes more noisy. Mitigation sceptics like to plot data that way; the trends naturally stay the same size, but the noise makes them seem smaller. This beautiful solution by John Kennedy plots every month separately and thus shows that all months are warming without distracting noise.

You can also show the seasonal cycle like in this NASA example, or animate it.

The sun is the source of nearly all energy on Earth and naturally important for the climate, but also for climate change? The sun was quite stable the last century and the last decades the sun may even have become less bright. By plotting the sun and temperature together Stefan Rahmstorf illustrates that the variations of the sun are too small to influence the climate much.

A longer perspective

Back in 2013 Jos Hagelaars combined the temperature reconstructions (from tree rings and other indirect information sources), the instrumentally measured temperatures and the temperature projection up to 2100 into one graph, also called The Wheelchair. It makes clear how large and fast the expected warming will be in a historical perspective.

The projected warming jumps up so fast in the graph of Hagelaars that you cannot see well how fast it is. Randall Munroe of the comic XKCD solved this by rotating the graph, so that it can be plotted a lot longer. To see how we are warming the planet, the graph had to become very long. Happy scrolling. See you below.

I hope you did not miss the mouse-over text:
"[After setting your car on fire]
Listen, your car's temperature has changed before.

Some complain that the temperature reconstructions are smoother than the instrumental data. This although it is even explained in the comic. How much more geeky can you get?

They want to suggest that thus a peak like we see now could be hidden in the reconstructions. That is theoretical possible, but there is no evidence for that. More importantly: the current warming is not a peak, it is a jump, it will still get a lot warmer and it will stay warm for a very long long time. If anything we are doing to the climate now had happened in the past it would jump out in the reconstruction.

Chris Colose solves the problem a little more technically and puts the current and future warming in the context of the period during which our civilization developed in this animation.

Mitigation sceptics

HotWhopper gathered and visualised some predictions from mitigation sceptics. Pictures that stand in sharp contrast to scientific predictions and in a sane rational world would discredit their political movement forever.

David Archibalds' prediction from 2006.
Based on solar maxima of approximately 50 for solar cycles 24 and 25, a global temperature decline of 1.5°C is predicted to 2020, equating to the experience of the Dalton Minimum.

Pierre Gosselin of the German blog No Truth Zone in 2008:
-2.5°C by 2020!...My prediction is we’ve started a nasty cold period that will make the 1960s look balmy. We’re about to get caught with our pants down. And a few molecules of CO2 is not going to change it.

Don Easterbrook in 2001.

Christopher Monckton prediction in 2013.
A math geek with a track-record of getting stuff right tells me we are in for 0.5 Cº of global cooling. It could happen in two years, but is very likely by 2020. His prediction is based on the behavior of the most obvious culprit in temperature change here on Earth – the Sun.
Maybe the Lord's math geek got a minus sign wrong. Will be hard to get so much cooling in 2020, especially as promissed due to the sun.

Normal science

A normal scientific graph can be very effective. I loved it how the audience was cheering and laughing after having to endure the nonsense of Australian Senator [[Malcolm Roberts]], when the physicist Brian Cox replied with: "I brought the graph." A good sign that the public is fed up with the misinformation campaign of the mitigation sceptical movement.

The graph Cox showed was similar to this one of NASA-GISS.

You know what they say about laughing at geniuses. I hope.

Related information

Ed Hawkins made a list of Visualisation resources for a wide range of climatological topics

Q&A smackdown: Brian Cox brings graphs to grapple with Malcolm Roberts

Temperature observation problems in WWII and the early instrumental period

Early global warming

A similar idea as the orchestra playing global warming is this tune based on the flow of the river Warta in Poland. The red noise of nature sounds wonderful.

* Temperature curve of XKCD used under a Creative Commons Attribution-NonCommercial 2.5 License.


  1. The State of the Climate front is beautiful. It also shows a sea level plot in blue and CO2 emissions in grey. Not sure what the ice one is? Glaciers, ice sheets?

    I've been experimenting with creating generative audio from climate data. Very rudimentary so far, and not very successful - the basic translation of daily date into a soundwave just results in monotonous noise, predictably given the regularity of the annual cycle. Looking to advance to something more like triggering sound events by thresholds and multi-instruments.

  2. I guess to hear more than a hiss one would have to draw out the time series or average over the series so that every note can be heard individually. The homepage describing the river flow tune also mentions that they had to round the values to the nearest tone to make the music more harmonic.

  3. Thanks for the post, Victor. There seem to be a competition going on these days ... not to the detriment of the public I'd argue :)

    I think the plot of mine with GISS data starts to look more sensible pre-WWII if switched to HadCRUT4-CW. Here we go (for the sake of completeness):
    HadCRUT4-CW 1850-2016 seasonal temperature anomalies



Comments are welcome, but comments without arguments may be deleted. Please try to remain on topic. (See also moderation page.)

I read every comment before publishing it. Spam comments are useless.

This comment box can be stretched for more space.