You know about our series by now. We’re investigating the ‘Real Global Temperature Trend‘. That means, as we swim through a turbulent ocean of climate records, we want to maintain a clear view on the horizon – to see where exactly we’re going, and how fast.
To do so the below graph is very helpful. Very helpful for anyone who wants to know what’s going on with the climate. Because before you can establish what exactly should be the height and shape of the climatological trend line (spoiler: that one is higher than this one, do follow our series to understand why) you of course first need to be able to draw a proper statistical trend line.
Global temperature trend line from (all) observational data of 5 established temperature datasets between 1979 and February 2016 – with 3 statistical trend line definitions. Graph in Dutch, made by climate data journalist Stephan Okhuijsen. CC Sargasso.nl – for data collecting effort see dataset comparison at Datagraver.com. [For comparison, here's the same temperature graph for NASA GISS only, stretching back to 1880.] (Thick fat disclaimer: unfortunately different temperature datasets use different base references. So please just look at shape and peak/dip comparison mainly.)
That’s what climate data journalist Stephan Okhuijsen did once more. What you’re looking at is a number of statistical climate trend lines of all the world’s established temperature measurements between January 1979 and February 2016, an average of all the measurements of the datasets of NOAA NCDC, NASA GISS, RSS (remote sensing systems), the UK Met Office’s Hadcrut4 and the Japanese dataset of JMA.
That is a lot of measurements. When you lump all these together, and plot the global monthly averages you get the thin yellow line. The somewhat thicker orange line is the 36-month (or 3-year) average trend line (not a real trend of course, but a nice indication of annual developments), the red line is the 132-month (or 11-year) climate trend line and the blue line is the 360-month or 30-year climate trend line – the official standard definition for ‘trend’ in statistical climatology.
How do recent records compare to the (statistical) trend?
Now of course our big question (for follow-up articles) is establishing the height of the ‘real global temperature trend’ – that means the one you get when you eliminate all masking factors, stuff like global dimming and thermal inertia as good as you can. To get there we need to develop a better understanding of climate sensitivity. Or rather climate science – all of it. We’re ambitious.
But before we continue on that quest today we simply let statistics* answer the above question:
- February 2016 was 1.04 degrees Celsius warmer than climate average Februaries (1951-1980 baseline)
- February 2016 peaked 0.57 degrees Celsius above the 3-year temperature trend line
- February 2016 peaked 0.46 degrees Celsius above the 11-year temperature trend line
- February 2016 peaked 0.33 degrees Celsius above the 30-year temperature trend line
*) Again, see disclaimer below graph – the 5 temperature datasets use different baselines. You the above numbers for trend deviation are not exact temperature deviations, but just useful indications.
This style of temperature trend definition is close to evening out the natural cycles within the observed temperature rise. That means eliminating both El Niño peaks and La Niña dips. Indicative is the fact that the record temperature anomalies were smaller compared to the 30-year climate trend line (that evens out ENSO) and bigger compared to the 11-year temperature trend – which is still strongly influenced by the La Niña temperature plateau, therefore cooler than the longer-running trend.
As said above this is still statistics – not climatology. Stay tuned for more updates – and when we get there, a very properly defined ‘real temperature trend line’!
© Rolf Schuttenhelm | www.bitsofscience.org