In my opinion, this essay is a must read because it clearly illustrates correlation between ocean cycles to; Arctic ice loss and gain, glacier advance and retreat, and land surface temperature rise and fall. As I said graphically in a previous post…
Guest post by Juraj Vanovcan
The following article shows, that decadal oscillation in North Atlantic sea surface temperature is the driving force behind observed variations in European climate during 20th century. Long-term North Atlantic SST trend is well correlated to European temperature station record, Alpine glacier retreat/advance and changes in Arctic ice extent as well.
Considering the problems with ground station record being contaminated by urbanization, land use changes and selective use, SST record offers an alternative metrics of changes in climate record, since it is free of at least some issues mentioned above. North Atlantic SST record is unique in this view, since it is quite reliable also in the early part of 20th century, when the ship measurement coverage of Atlantic between American continent and Europe had been much denser than in other parts of the globe .
Here is presented North Atlantic sea surface temperature record since 1850. While the pre-1880 data are rather noisy, probably because of sparse coverage, the 20th century record shows regular cyclical pattern of warming and cooling. The cycle length is 65 years, with cold minimums reached in 1910 and 1975 and warm maximums in 1940 and 2005.
Figure 1: North Atlantic SST record, expressed as monthly anomalies against 1971-2000 period (HadSST2 dataset)
Let’s now compare the North Atlantic SST record with the European ground stations within 40-70N and 10W-30E.
Figure 2: North Atlantic SST record compared to European ground stations
European station record is well correlated with the Atlantic SST changes, and lags the SST record by some 5 years. It is thus obvious, that it is the Atlantic decadal variability, which dictates the European climate. Some excessive surface warming to the end above the SST record (observed also in global surface and SST datasets) is either explained as a sign of quicker response of the surface to increasing radiative forcing, but critics consider it as a sign of urbanization and land use changes, plaguing the station record. This might be especially true for Europe, where population density and its growth have been considerable during the last 100 years. This dispute can be resolved by comparing the North Atlantic SST trend with long-term rural station record.
Armagh Observatory (Ireland) is one of the few rural stations with long historical record, located near small town of Armagh and its surrounding has been claimed to be basically intact since its start in 1796. Lomnicky peak Observatory (Slovakia) is located on the top of the Lomnicky Peak (2655), the highest mountain of Carpathian ridge and measurements are available since 1941.
Figure 3: North Atlantic SST record compared to rural ground stations
From the graph above, it is obvious that the North Atlantic SST record is extremely well correlated to selected UHI-free surface station records from both Western and Central Europe. Amplitude of warming and cooling cycles is slightly more pronounced in the station records.
There are several points worth of interest.
- The rate of warming in 1910-1940 period has been equal with the warming period 1975-2005.Even if one suggests that the anthropogenic forcing is superimposed on natural variations in the background, it is difficult to identify the alleged “increased anthropogenic forcing” in the record to the end of 20th century.
- There has been pronounced cooling period since 1940 until 1980, which completely erased the early century warming against the 19th century average. The 1982-centered decade in Armagh and CET records has been actually colder than end of 19th century and the decade centered around 1870, which again questions the concept of anthropogenic forcing, which should already manifest with the CO2 increase. Surprisingly enough, looking back at the whole length of the both records, 80ties in Europe were equally coldish as average of the Little Ice Age period.
- The overall warming trend since 1900 (0.6 deg C/century for SST and 0.9 deg C/century for the station record) is partially created by the fact, that beginning of the century starts with the cycle minimum and ends with the cycle maximum. By more proper procedure – comparing the differences between 1910/1975 minimums and 1940/2005 maximums – one gets constant warming trend of 0.3 deg C/century for SST record.
- Despite a string of cold years in early 1940s (much more pronounced in the Central/Eastern European record), individual years in 1940-1950 decade were comparably warm as during the last decade. But the fact is that the last decade as a whole has been warmest in record in both Armagh and Atlantic SST data.
Figure 4: 0-700m ocean heat content in North Atlantic, 1955-2010
In the monthly Atlantic SST record, we can observe that the recent warm phase peaked in 2005 and subsequent cooling of North Atlantic started, despite the recent AMO peak as a response to 2009/2010 El Nino. This climate shift is even better visualized in the 0-700m ocean heat content record for the Northern Atlantic. Based on previous records, we can expect the European climate to follow the SST record and to mimic the 1940-1975 cooling trend.
* * *
Multidecadal oscillation in European climate is also tied to European glacier growth/decline. We often hear about the recent Alpine glaciers retreat, but the fact is, that similar retreat occurred in early 20th century as well, and most of the observed glaciers advanced just three decades ago. Data from Swiss Glaciology Institute, covering more than 100 Swiss glaciers, show ratio of advancing, stationery and retreating glaciers during the 20th century, presented here against the AMO index.
Figure 5: Swiss glacier advance/retreat related to Atlantic Multidecadal Oscillation (older years are to the right)
Compared to North Atlantic SST record, the period with most glacier growth/retreat lags the ocean by 5 years, matching the lag in surface record. Extremely warm European summer in 2003 is clearly recognizable, when all observed glaciers retreated. But similar period occurred in 1945-1950, followed by years with prevailing growth in late 70ties/early 80ties. This glacier behavior is also discussed in recent study “100 year mass changes in the Swiss Alps linked to the Atlantic Multidecadal Oscillations” . Based on the AMO peak in 2005 and observed 5-year lag, rebound of Alpine glaciers in the near future is expected.
* * *
North Atlantic seems to have decisive effect on Arctic temperature and ice extent as well. This is understandable, since the Gulf Stream brings masses of warm Atlantic water into the Northern Ocean. Plotting the post-1979 satellite era ice extent against both North Atlantic SST anomalies and Ocean heat content shows reasonable correlation.
Figure 6: Arctic ice extent as a function of North Atlantic SST record, 1979-2009
Figure 7: Arctic ice extent as a function of North Atlantic 0-700m ocean heat content, 1979-2009
By extrapolation this correlation backwards, it is understandable, that the North West Passage has been open for shipping in both 1942-1944 and again in 2007-2009 period. Beyond this SST range, also other positive/negative amplifying effects may change the linear correlation suggested above. Starting rebound of Arctic ice extent since its 2007 minimum is well explainable in light of recent climate shift in the North Atlantic to the cooling mode.
In light of these facts, the alleged Arctic ice history often presented as a “proof” of “unprecedented” ice retreat in the 20th century is unsupported.
Juraj Vanovcan 26th September 2010
My thanks to Juraj for this excellent essay. The conclusion from this essay is that the oceans drive the temperature of the atmosphere, not the other way around. The polar ice responds to the AMO, and glaciers in Europe respond to the AMO. When the AMO and PDO coincide to both be negative, forecast to be sometime around 2015, there’s gonna be some ‘splaining to do.
As the New Scientist finally came to realize and publish on this week, the sun and the oceans play a bigger role than many give credit for. – Anthony
Here’s some additional information via appinsys:
PDO Plus AMO / US Temperatures
Joseph D’Aleo has conducted a correlation analysis between the PDO, AMO and temperatures [http://icecap.us/images/uploads/US_Temperatures_and_Climate_Factors_since_1895.pdf] and [http://intellicast.com/Community/Content.aspx?a=127]. The following figures are from D’Aleo’s analysis.
The following figure shows the 5-year means of PDO, AMO and PDO + AMO.
The next figure shows the US temperature anomalies as calculated by NASA’s James Hansen (2001). The periods when the temperature anomalies are positive correspond almost exactly to when the PDO+AMO changes between warm and cool phases.
The following figure compares the PDO+AMO with the US average annual temperatures. D’Aleo calculated an r-squared of 0.85 between the two – an extremely good correlation.
The next figure compares the same temperature data with atmospheric CO2. D’Aleo calculated an r-squared of 0.44 between the two – a fair correlation, but poor in comparison to the PDO+AMO correlation. Although correlation does not prove causation, lower correlation is evidence of lower probability of causation.
The following figure shows the combined effect of PDO and AMO on drought in the United States [http://oceanworld.tamu.edu/resources/oceanography-book/oceananddrought.html]. Further information on these drought relationships can be found at [http://www.pnas.org/content/101/12/4136.full]