Guest post by Jos Hagelaars. Dutch version is here.
Discussions on the Internet regarding climate change are sometimes about scientific details, sometimes about the climate sensitivity regarding the equilibrium situation hundreds of years from now, but the most prevalent discussion topic is probably: the global average temperature. Will it get warmer or colder, is there a temporary slowdown or acceleration in the rise in temperature, are the models correct or not, will the eventual warming of our earth in the future be large or small? New numbers are released on a monthly basis and every month megabytes of text are generated about them. My forecast is that 2015 again will lead to a discussion-spike.
The graph above shows the evolution of the global surface temperature anomaly for three datasets, where the average of the period 1981-2010 is defined as 0. For the year 2015 only data are presented up to and including June. So far 2015 exceeds all other years and the evolving El Niño makes it likely that 2015 will set a new world record.
The black line in the graph is a so-called smooth function (Loess) over 30 years. 30 years is not an arbitrary choice, it is the number of years on which the climatic definition is based. Natural variability, e.g. caused by El Niño’s and La Niña’s or temporary dips by volcanic eruptions are averaged out on this time scale. This climatically relevant perspective of 30 years shows that there has been a steady increase in the surface temperature on Earth since the 1970s, as expected given the steady rise in greenhouse gas concentrations in the atmosphere.
A comparison between the observations and forecasts of climate models for various scenarios in terms of greenhouse gas emissions is frequently presented on Real Climate by Gavin Schmidt of NASA. Also this year, where he compared various datasets with CMIP3 model calculations (used for IPCC AR4) and CMIP5 model calculations (used for IPCC AR5). The last year that is visible in these comparisons made by Gavin Schmidt is the year 2014. The Real Climate graphs show that the observations clearly fall within the range of the model calculations. So Gavin Schmidt eventually concludes:
By now 2015 is halfway through and – driven by curiosity – I added the data of the first half of this year to the graphs made by Gavin Schmidt. See the green rectangle in the two graphs below. 2015 so far is rather close to the average model (IPCC AR4) for the A1B scenario as predicted in 2000 (figure 2).
The graph in figure 3 contains the CMIP5 calculations for IPCC AR5 and also an update of these calculations. Several years have passed since the CMIP5 calculations for AR5 were made and now there is new information available on e.g. solar activity, volcanic ash in the atmosphere and the concentration of greenhouse gases. An update of the CMIP5 calculations on the basis of this new information is shown as dashed lines in figure 3. The green rectangle representing 2015 is fairly close to the center dashed line.
It is important to note that the exact timing of the natural variation is not present in each individual run of a climate model. When you take an average of the different model runs, as shown by the black line in the middle of the gray areas in figures 2 and 3, the influence of natural variation is completely gone. For example, there are no El Niño/La Niña spikes visible in the two black lines. Climate models may thus give an idea of the ‘forced’ development of the climate under different scenarios, but not of the weather-induced or other ‘unforced’ variability.
Besides the misspecification of climate forcings there may be other factors that create the appearance of a divergence between modeled and observed temperatures: The former is based on air temperatures, whereas the latter is based on sea surface temperatures over the oceans.
I am not aware of any climate models developed by the so-called climate skeptics that are based on physics in a way that is similar to the CMIP3 and CMIP5 models. Maybe I missed something? However, I encountered some forecasts in the past and I compared some of these with the observations including May 2015 (HadCRUT4). Two more ’serious’ forecasts are presented in figures 4 and 5. Figure 4 is taken from the book “Die Kalte Sonne” (2012 – English: “The Neglected Sun) written by Fritz Vahrenholt and Sebastian Lüning and Figure 5 is from a paper by Nicola Scafetta (2012). The green and gray lines are HadCRUT4 monthly data that I added to both figures respectively. In both cases the 2015 temperature value is located outside the range of the forecast, just a few years after the forecast was issued.
The funniest forecast is made by Don Easterbrook, an emeritus professor of Geology. He has presented this at a meeting of the Heartland Institute in 2010 and Hans Labohm (a Dutch climate sceptic) used this very seriously in a discussion with Bart Strengers (PBL Netherlands Environmental Assessment Agency) on the website of the Dutch news station NOS in 2009. According to Easterbrook we would get significant global cooling. See figure 6, again the green line consists of the HadCRUT4 monthly data which I placed in the graph of Easterbrook. Even at the time of his presentation in 2010 this forecast could already be seen to be way off.
More information about this silly forecast of the Don can be found at SkepticalScience.
I may think that the forecast made by Easterbrook is funny, sadly there are still a lot of people who take these sorts of wishful thinking type forecasts seriously. Science, however, tells a different and thoroughly substantiated story. It will get warmer this century. How much warmer depends to a great extent on choices we make regarding future emissions.