Arctic Ice and the AMO

By Paul Homewood

David Rose’s piece in the Mail on Sunday has already been picked by WUWT and Bishop Hill, amongst others. But I want to concentrate on one aspect, that of cycles.

The Mail report:

However, Dr Hawkins added that the decline seen in recent years was not caused only by global warming. It was, he said, intensified by ‘natural variability’ – shifts in factors such as the temperature of the oceans. This, he said, has happened before, such as in the 1920s and 1930s, when ‘there was likely some sea ice retreat’.

Dr Hawkins said: ‘There is undoubtedly some natural variability on top of the long-term downwards trend caused by the overall warming. This variability has probably contributed somewhat to the post-2000 steep declining trend, although the human-caused component still dominates.’

Like many scientists, Dr Hawkins said these natural processes may be cyclical. If and when they go into reverse, they will cool, not warm, the Arctic, in which case, he said, ‘a decade with no declining trend’ in ice cover would be ‘entirely plausible’.

Peer-reviewed research suggests that at least until 2005, natural variability was responsible for half the ice decline. But exactly how big its influence is remains an open question – and as both Dr Hawkins and Prof Curry agreed, establishing this is critical to making predictions about the Arctic’s future.

Prof Curry said: ‘I suspect that the portion of the decline in the sea ice attributable to natural variability could be even larger than half.

‘I think the natural variability component of Arctic sea ice extent is in the process of bottoming out, with a reversal to start within the next decade. And when it does, the reversal period could last for several decades.’

This led her to believe that the IPCC forecast, like Al Gore’s, was too pessimistic.

‘Ice-free in 2050 is a possible scenario, but I don’t think it is a likely scenario,’ she concluded.

The cycle they refer to is the Atlantic Multidecadal Oscillation, AMO. Below is the detrended AMO, and you can see it runs on about a 60-year cycle. It is currently around or just below its peak, having risen strongly since the mid 1970’s.

It may remain at the current level for a few more years yet, but it will then fall away for the next 30 years, just as it did from the 1940’s to the 1970’s.

http://www.esrl.noaa.gov/psd/cgi-bin/gcos_wgsp/tsanalysis.pl?tstype1=91&tstype2=0&year1=1895&year2=&itypea=0&axistype=0&anom=0&plotstyle=0&climo1=&climo2=&y1=&y2=&y21=&y22=&length=&lag=&iall=0&iseas=1&mon1=0&mon2=11&Submit=Calculate+Results

So, what effect will this have on Arctic ice? We can glean much from examining temperature trends around the part of the Arctic affected by the AMO.

First, Iceland. Below is a report from the Icelandic Met Office in 2008, “Past temperature conditions in Iceland from 1798 to 2007”, which uses the long running, high quality site of Stykkisholmur.

Temperature in Stykkishólmur (Western Iceland)

The temperature (figure 1) has in the long run been increasing during the last 200 years at the rate of +0.7°C per century. This is similar to the general temperature increase in the whole Northern hemisphere during the same period. The warming has been very uneven, dominated by three cold periods and two warm ones.

Annual temperature in Stykkishólmur 1798 to 2007

Figure 1. Annual temperature in Stykkishólmur 1798 to 2007. Note that the values prior to 1845 are interpolated from observations at other stations. The confidence is very low for the years before 1830 and the values are preliminary and should not be referenced. Work on quality improvement is ongoing. A few warm and cold years are highlighted.  

The time from 1925 onwards is dominated by a very large cycle that does not show an overall significant warming, although the temperature rise of the last 20 years is considerable.

There is also a large decedal variability before 1925. The year 1892 marked the end of a period dominated by a very large year-to-year variability and the end of a long run of very cold years. There was a relatively warm period during 1837 to 1858, and by overlooking the very cold year of 1835 and a few isolated cold months one can identify the interval 1813 to 1858 as a generally warm one.

The years 1807 to 1812 were very cold. Although the following warm period was considerably colder than the corresponding 20th century warm period it was noted as a generally favourable time for agriculture and the population of the country increased markedly.

The 20th century warm period that started in the 1920s ended very abruptly in 1965. It can be divided into three sub-periods, a very warm one to 1942, a colder interval during 1943 to 1952, but it was decisively warm during 1953 to 1964.

The cold period 1965 to 1995 also included a few sub-periods. The so called “sea ice years” 1965 to 1971, a slighly warmer period 1972 til 1978, a very cold interval during 1979 to 1986, but therafter it became gradually warmer, the last cold year in the sequence being 1995. Since then it has been warm, the warmth culminating in 2002 to 2003. Generally the decription above refers to the whole country, but there are slightly diverging details, depending on the source of the cold air.

Note:

1) The reference to a long term temperature increase since 1798.

2) No significant warming since 1925.

3) Reference to cold and warm periods. The most recent ones being the cold period from 1965 to 1986, followed by the recent warming. Note how these, and the warm period culminating in the 1940’s, correspond with the rise and fall of the AMO.

If we look further afield, we find patterns in Greenland (Godthab and Angmassalik), Norway (Vardo) and Russia (Murmansk and Salehard). The following graphs are from GISS, and use unadjusted data.

 http://data.giss.nasa.gov/cgi-bin/gistemp/show_station.cgi?id=431042500000&dt=1&ds=1

 http://data.giss.nasa.gov/cgi-bin/gistemp/show_station.cgi?id=431043600000&dt=1&ds=1

http://data.giss.nasa.gov/cgi-bin/gistemp/show_station.cgi?id=634010980003&dt=1&ds=1

http://data.giss.nasa.gov/cgi-bin/gistemp/show_station.cgi?id=638221130005&dt=1&ds=1

http://data.giss.nasa.gov/cgi-bin/gistemp/show_station.cgi?id=222233300002&dt=1&ds=1

All these stations, ranging from western Greenland to Siberia, show essentially the same pattern, a warm period around 1940, comparable to now, and a much colder interlude in the  1960’s and 70’s. And, of course, these all closely follow the ups and downs of the AMO.

There seems little doubt that the Arctic will be in for another cold period during the next 30 years or so, and that, as Judith Curry indicates, we will see a long term recovery of Arctic ice extent.

We fail to learn the lessons of history at our peril!