Climate is commonly defined as the weather averaged over 30 years. The temperature trend in the 30 years until mid 1997 is rising by 0.14 ° C per decade, and the rise is statistically significant with very good margins. Most people agree with this. The question is whether the steady warming continued after 1997 or whether it levelled off.
The black points in Figure 1 show the monthly global surface temperatures in these 30 years, and the black line shows the rising trend 0.14°C/decade in the period. Each temperature is a temperature anomaly, which is the deviation from the mean temperature in that month.
The red line in Figure 1 is an extension of the trend line up to 1997. In the Continued warming scenario we expect that about half of the monthly temperatures in the next 18 years will be warmer than the red line, and that about half of them will be colder. The two dashed red lines embrace the expected range in the monthly temperatures.
The blue horizontal line in Figure 1 starts where the trend line up to 1997 ends. In the Pause scenario we expect that about half of the monthly temperatures in the next 18 years will be warmer than the blue line, and that about half of them will be colder. The two dashed blue lines embrace the expected range in the monthly temperatures.
Now in July 2015 we know the monthly temperatures in the 18 years from 1997 up to now. It makes sense to draw them into Figure 1 without making any other changes to the figure. Then we will see if they fit with the red line representing the Continued warming scenario or with the blue line representing the Pause scenario.
In Figure 2 the monthly temperatures in the last 18 years are plotted as green points. They have had no impact on neither the black trend line nor on its red extension.
More than three quarters of the monthly temperatures after 1997 are warmer than the red line representing the Continued warming scenario. Some monthly temperatures are warmer than the expected range shown by the red dotted lines, and none are colder. The temperatures have been warmer than expected in the Continued warming scenario. It shows that the warming in the last 18 years has been at least as strong as it was until 1997. Nothing suggests that this is about to change; the temperatures in the last ten months are all warmer than the red line.
The fact that more than three quarters of the monthly temperatures in the last 18 years are warmer than expected in the Continued warming scenario might indicate that the warming is accelerating. But there is not enough data to claim that. A period of 18 years is too short to draw such a conclusion. But the temperatures provide no basis whatsoever for claiming that there has been a pause in the warming. Virtually every monthly temperature after 1997 is warmer than expected in the Pause scenario, and none are colder.
Why choose 'the last 18 years' ?
The contrarians cherry pick mid 1997 as the start date when they claim that there has been no global warming in the last 18 years. The reason is obvious. Global temperatures rose sharply in the last months of 1997 and in the first half of 1998 due to an unusually strong El Niño in the Pacific. By choosing mid 1997 as the start time when calculating the temperature trend up to now, the starting point of the trend line is tilted a little upwards, and the trend line thereby shows less warming. Thereafter they concentrate on the trend line in the last 18 years without seeing it in connection with the temperatures in the preceding years. In this blog post I compare the temperatures in the last 18 years with the trend line calculated over the preceding 30 years. A trend line calculated over 30 year is not so sensible of specific start and end time. I can replace 'the last 18 years' with 'the last 17 years' and still get the same conclusion; there is no pause in the global warming.
What about the other temperature series ?
76% of the monthly temperatures in the last 18 years in Figure 2 are warmer than expected in the Continued warming scenario. The exact percentage varies a little depending upon the temperature series that is used in the analysis. The results when using GISTEMP and five other temperature series are shown below. The series are updated with temperatures up to May 2015, except the Berkeley BEST series that is updated with temperatures up to December 2014.
Temperature series Months warmer than continued warming scenario
NASA GISTEMP 76 %
Met Office HadCRUT4 81 %
NOAA NCDC 63 %
Berkely BEST 81 %
UAH LT satellite 76 %
RSS LT satellite 84 %
The satellite series for the lower troposphere (LT) starts in December 1978 (UAH) and in January 1979 (RSS). The trend in the years preceding 1997 therefore cover approximately 18 and a half year for these series.
The plots and the analysis in this blog post are inspired by the blog post 'Global Temperature: the Post-1998 Surprise' that Tamino wrote one and a half year ago. Tamino's plots were indeed a surprise to me. Why do so many people speak about a pause in the global warming when there is no pause? Due to my surprise I wrote programs that analyse the temperatures and plot the results in a similar way that Tamino did. My results are the same as the ones that Tamino showed in his blog post. I have described the mathematics used in the analysis in an earlier blog post, which is number three in a series of five posts describing statistical analysis of temperature series.
Tamino showed the results with the yearly temperatures up to 2013. I used monthly temperatures in this blog post to disprove the claim 'No warming in the last 18 years'. That claim was presented by the Norwegian Nobel Laureate Ivar Giæver on Norwegian television in June 2015, and due to Ivar Giæver's Nobel status the claim was repeated many times in other media. I therefore wrote a blog post in Norwegian, similar to the one you are reading now, and published it on a web site managed by one of the main Norwegian newspapers. Many contrarians commented the blog post. None of them questioned the analysis. They either focused on other topics, as trolls often do, or they focused on the the temperatures in the last 18 years without seeing them in connection with the temperatures in the preceding years
It is easier to read the plots with yearly temperatures, as Tamono showed in his blog post, than to read them with monthly temperatures. I therefore add Figure 3, which shows the results of the analysis when it is done with the HadCRUT4 yearly temperatures up to 2014. 12 of the 17 years after 1998 are warmer than the red line representing the continued warming scenario. All the 17 years are warmer than the blue line representing the pause scenario, and none are colder.