[Updated: 17th September 2019]
The visualisation technique of ‘small multiples’ is often used to communicate a simple message. The above example shows maps of temperature change from 1850-2017 – the overall warming trend is obvious even though the details are fuzzy.
Technical details: The HadCRUT4.6 dataset is used with anomalies from a 1961-1990 baseline period. An annual average for a particular grid cell and year is only shown if 6 or more months have data, otherwise it is coloured white. The colour scale runs from around -2.5C to +2.5C.
Other examples: NASA have made a similar visualisation for melting Arctic sea ice.
26 thoughts on “Mapping global temperature change”
Very good, interesting and non-controversial.
Suggestion 1: Add a legend
Suggestion 2: Create a second version that uses only the same coverage area as 1850.
( to keep contrast changing coverage )
What I love about the 19th century Hadcrut4 data is that you can clearly see the main shipping routes to India, Africa and across to the US. If you look at animations of the monthly data then you can often see individual voyages.
A couple of comments on your nice montage:
1) The grey background you are using has the effect of making the older data ‘appear’ cooler when in fact it is simply because there are no data available. A white background won’t work as you use are already using white in your colour table. You could try using a black Map background with a Grey overall background.
2) The temperature scale being used for the anomaly values is already saturating in 2015 (Bright Red). Now suppose warming continues well above +2C. Your plots would then all saturate red. The only way round that would be to renormalise the scale every 10 years or so. The psychological effect of your (Met Office) choice of colour scheme is to ‘enhance’ the impression of warming. Everything above 0.2C appears warm to hot, yet the error on grid values is at least 0.5C. The worst offender here is GISS who use a non-linear colour interval for each colour. Yellow is 0.2-0.5 and deep red is 2.0 to 4.0
I think it should be a blended colour scheme without sharp step changes of colour :
1) It’s not a Met Office colour scheme – it is based on those on colorbrewer2.org
2) The alternative you suggest is here: http://www.met.reading.ac.uk/~ed/smmult_large_altcol.png
Hi Clive – I am trying to reproduce your chart (http://clivebest.com/blog/wp-content/uploads/2015/11/frame1989.png)
Is that using the data for 2015/10 ? I see that the anomaly range in that line of data is from -3.4 to 8 – can you share how you derived a color from that – trying to prove my system is working by trying to reproduce what you did.
Why do some people deny man made climate change so vigorously? I have tried to reason with such folks, show charts, etc., but it is like trying to argue about religion… no possibility for common ground. Any ideas??
there are many articles and reasons for this
The Dunning Kruger effect is one of the main ones
People are so stupid they imagine they are smarter than experts
that an a billion dollar industry set up to convince them they they know more than the experts
As well as Dunning Kruger effect don’t forget that a lot of those people tend to be hard-core conspiracy believers. See, for example, the newly elected Australian Senator, Malcolm Roberts, who is not a climatologist, degree in mining engineering I think, who does not believe in Global Warming because he “knows” that NASA has faked the climate data.
In the USA it also seems to have become a matter of political affiliation. For some strange reason if you are a Republican you have to deny Global Warming.
This conservative vs liberal difference does not seem to exist in other countries
“not a climatologist, degree in mining engineering”
Careful with that criteria.
Not strictly climatologist by education also includes policy advocates:
James Hansen ( math, astronomy, and physics )
Gavin Schmidt ( math )
Stephen Schnieder ( mechanical engineering )
Now, sometimes, being outside a field leads to new insights from a different perspective.
However, I think the lack of meteorological education among the three above, leads to the erroneous claims of predictability of aspects of climate.
Global average temperature is somewhat predictable because the top of the atmosphere forcing change is relatively uniform and relatively independent of other changes.
Specific weather, by contrast, occurs because of dynamic motion within the atmosphere which is unpredictable. Had the above been educated in meteorology, they might not have wandered in to making pronouncements of precipitation, droughts, storms, heatwaves, coldwaves, etc. which are not predictable either individually or in the statistical mean.
Why do some people deny man made climate change so vigorously?
There is a book entitled The Myth of the Rational Voter in which the author attempts to analyze political adherence by this observation: it is not the predicates of policy that move people, but rather, the net results ( benefits accrued to some, detriments incurred by others ). People conceive of a positive or negative outcome ( usually wrt themselves ) and form their beliefs according to ensuring or avoiding the imagined outcome. I think climate is similar – people believe either the climate or the economic impact will be significantly negative and tend to maximize evidence which supports their case ( while ignoring nuances or contradictory evidence ).
I have tried to reason with such folks, show charts, etc., but it is like trying to argue about religion…
Yes, religion, like politics above, involve an unmeasured and largely untestable future state ( heaven or hell, and for many religions, nothing in between ), but never the less proscribes a particular course of action.
no possibility for common ground. Any ideas??
So, I tend to fall into the category of those opposed to government intervention, largely because I believe that while global average temperature is rising, global average temperature is not correlated with many of the harms I believe the IPCC has erroneously alluded to. Also, secular factors are at work which are likely making the point rather moot. We are most likely within three decades of falling human population. We are already experiencing a rapidly aging population and economic development continues to improve the efficiency of energy usage. These factors appear to have already led to decreasing rates of radiative forcing from CO2 and now may already have led to falling CO2 emissions.
But as to common ground, lets back up some.
I think we’d agree that:
* humans emit CO2
* emitted CO2 is accumulating in the atmosphere
* increased CO2 reduces the efficiency with which heat radiates from earth
* the most likely response is increased atmospheric temperature
* this cannot continue indefinitely without harming humans/environement
* indefinitely, however, is a long time.
Other things that are demonstrably true:
* radiative forcing from CO2 is logarithmic
* a constant rate of increase in forcing requires exponentially increasing CO2
* 10yr RF peaked in 2007
* we may already have passed peak global warming from CO2
* temperatures have been naturally higher in the past ( Eemian & HCO )
* The HCO was a period of advancement for humans ( Cradle of Civilization ).
* there is contradictory evidence to the IPCC notion of extremes
* global average temperature is not a term in the equations of motion
* global average temperature is not correlated with weather
* Increased CO2 increases vegetative growth
* Increased CO2 increases crop yield
* the uptake rate of CO2 has increased some 500% since 1950
* human fertility implies falling population soon
* CO2 emissions will fall with falling global population
* CO2 emissions may already have peaked in 2013
* the gradient changes of RF from the HCO were likely much greater than the gradient changes from 2xCO2
There are many other things to think about, but do any of these occur in common ground between you an me?
Peak warming figures to occur at our about the time that peak CO2 atmospheric concentrations occur.
If emissions due to industry etc. fall from their current levels, but to new levels that remain sufficient to swamp natural uptake mechanisms such as rock weathering, then CO2 levels will continue to rise. Peak warming later is better for us than peak warming sooner, because we’ll have more time to deploy solutions that become more feasible as R&D advances. Wind, solar, conservation, all sorts of responses we can make figure to become easier to deploy at scale if they don’t have to be deployed too rapidly and using immature technology. But it’s a circular situation. What we deploy now, imperfect though it is, helps buy us time.
“Peak warming figures to occur at our about the time that peak CO2 atmospheric concentrations occur. “
Peak warmth figures to occur at or about peak CO2.
Peak warming figures to occur when rates of forcing decline, which appears to already have happened. Future warming should still occur, but at a decreasing, not increasing rate.
Current rates of uptake are roughly equivalent to 2.5ppm/per year. Provided that number does not continue to increase ( which it has been since 1960 ), then emissions of 2.5ppm per year ( about 50% of current emissions ) would yield constant concentrations.
Nice plot sumarising global warming. I love the use of the early ship log data- reminds me of Master and Commander 🙂
Presenting this as an example of great communication through visualisation at a small internal meeting here at CSIRO in Melb-Aus. Looking forward to the next innovative idea!
While higher temperatures and higher carbon dioxide concentrations may raise levels of carbon fixation into carbohydrates by photosynthesis, a rise in mean night time temperatures will cause plants to respire more at higher temperatures, in the course of which they will use up more of the carbohydrates they made, so there will be less available for storage in fruits and seeds. A calculation by scientists at the International Rice Research Institute at Los Banos, in the Philippines, has detected a ten percent decline in seed production for a one degree rise in mean night time temperatures.
But it doesn’t seem to have been significant so far:
Another excellent vis Ed. Even effective paired right back to just red and blue with no grey or white (I tried it!)