Tag Archives: Rahmstorf_Stefan

Explaining the pause that wasn’t


We’ve only just seen that statistically there never was a recent ‘pause’ or ‘hiatus’ in warming when two articles show up explaining the pause that wasn’t. A key word here is statistically. No-one is suggesting that there may not be variations from time to time that do not breach the trend.

Also climatologists are always interested in understanding the physical mechanisms behind both short-term variability and longer term trends.

First, via Carbon Brief there has been a new study examining the impact of aerosols from volcanoes.

Virtually all research into the climate influence of volcanic aerosols has used satellite measurements of particulates in the upper atmosphere (the stratosphere). These satellite measurements only monitor the volcanic aerosol at heights of 15 km and above. The new paper by David Ridley and colleagues studied the amount of volcanic aerosols in portions of the stratosphere that lie below 15 km.

They found that 30 to 70% of aerosols from recent eruptions lodged below the 15 km mark. Further, these lower level aerosols had been responsible for significant cooling since 2000. Taken together with additional heat being stored in the deep oceans, the slowdown is “both fully accounted for and temporary.”

Estimated volcanic cooling from this source is not included in climate models.

A second study looks at the influence of Pacific winds on warming since the 1890s. They did this by analysing the chemical make-up of corals.

The coral record suggests, for example, that trade winds were weak between 1910 and 1940 when the Earth warmed by 0.4 degrees, and were strong from 1940 to 1970, during a period of relatively little increase in global temperatures.

Apparently the winds have been very strong since the turn of the century.

The winds in question are the trade winds that move tropical surface water from east to west. This two-dimensional image attempts to illustrate the complex pattern. Trade winds just north and south of the equator drive warm surface water westwards:


The water is replaced by cool water rising from the deep, which is known to affect global average surface temperatures. A simplified mechanism of how this works is given in the following:


Knowledge of the influence of the wind on temperature variations is not new. For example, Matthew England of the University of NSW and others published a paper early in 2014 on the subject. England says of the new paper:

“This is a really important study: it confirms the crucial role of the Pacific Ocean in driving decadal climate variability at a global scale.”

The strength of the trade winds is associated with a natural climate phenomenon called the Interdecadal Pacific Oscillation (IPO). The IPO has positive and negative phases, which switch every few decades. Lead author Professor Diane Thompson says:

“We do know that the typical lifetime of a phase of this cycle is less than 30 years, and the current one began about 15 years ago. So although the timing of that switch remains difficult to anticipate, it would be most likely to happen within the next one to two decades.”

We don’t know the future of volcanic activity, nor when the Pacific wind pattern is going to switch. What we do know is that both have been inhibiting warming in recent years.

Of course Pacific wind patterns are also associated with ENSO where El Niño years are warmer and La Niña years are cooler. A year ago Stefan Rahmstorf looked at The global temperature jigsaw where he identified a variety of influences at work. It included a graph from a 2012 paper which used a multivariate correlation analysis to take out the influence of ENSO, volcanoes and solar activity. The pink line shows observed data and the red line with those three factors removed:


The ‘pause’ effectively goes away.

The observed data (pink line) in the above graph is an average of five data sets. It is likely that two of them, HadCRUT and NOAA omitted the Arctic entirely. I’ll post again the recent HadCRUT4 hybrid data from the satellite era which Rahmstorf suggests is now the best:


Again a pause is pretty hard to find, especially if you ignore 1998 as a outlier year. Of course 1998 should not be ignored as it appears to have had a crucial role in transferring stored heat from the ocean to the surface. It’s worth remembering just how little of the planet’s energy is stored on the surface compared to the ocean:


Rahmstorf did predict that when we had another El Niño year a new record would be set. 2014-15 is not an El Niño, not yet, but is being heralded as the warmest ever. I’ll wait for it to actually happen before reporting on it further. Meanwhile it’s a fair bet that we will have strong warming in the next few decades.

The end of the warming pause that wasn’t

The bottom line is that statistically there never was a recent ‘pause’ or ‘hiatus’ in warming. If you ever thought there was you’d have plenty of company, but 2014 sees temperatures returning to the trend line.

At RealClimate Stefan Rahmstorf took a look at global surface temperatures since the satellite record in 1979. He used Kevin Cowtan’s interactive temperature plotting and trend calculation tool and HadCRUT4 hybrid data, which now has the most sophisticated method to fill data gaps in the Arctic with the help of satellites. This is what he found:


Rahmstorf then looks at some shorter intervals using more recent starting dates. He finds:

There simply has been no statistically significant slowdown, let alone a “pause”.

He then displays an analysis done for Realclimate by Niamh Cahill of the School of Mathematical Sciences, University College Dublin using a technique called change point analysis. This time GISTEMP data from NASA GISS is used from 1880 to the present. 2014 was represented by January to October data, the latest available.

The algorithm used sorts through the data looking for changes in the trend lines, which it finds automatically. This is how it came out:


Change points were found approximately at 1912, 1940 and 1970. The trend since 1970 is firmly in place.

Tamino at Open Mind took a look at the same data. People often cherry pick a starting date so using a different statistical technique he examined whether there had been any change in trend starting from any year between 1990 and 2008. Often people choose a data set that suits their argument so he ran the technique over all eight commonly available data sets.

He could find no trend change – not even close.

I repeat: not only is there a lack of valid evidence of a slowdown, it’s nowhere near even remotely being close. And that goes for each and every one of the 8 data sets tested.

He then plots the GISTEMP data from 1970:


It’s all happening between the trend lines. Of course there are fluctuations, but none of significance. Not yet.

For a time the Hadley data simply left out the Arctic area where there are no weather stations. Of course the Arctic is warming faster than lower latitudes and it came to a point where this mattered. Recently they’ve corrected this by using satellite data, not directly, as the satellites measure the troposphere rather than the surface. I gather the satellite data is used comparatively to infer the surface temperature across the Arctic.

Looking back over the archive, there was a study by Cowtan and Way (see RealClimate and Climate clippings 89, Item 4) which compared the HadCRUT4 data of that time with ‘filled in’ data. This is how it looks:


Notice how the heavier corrected lines make 1998 a little cooler and recent peaks a little warmer.

Even so, I recall that Ross Garnaut in preparing his report about five years ago asked two leading statisticians to examine the trends. They said the long term warming trend was still in place. Seems they were right.