North Atlantic Oscillation phases, tree growth, and droughts

From the UNIVERSITY OF THE BASQUE COUNTRY, and the “climate phenomena not related to CO2” department.

Droughts and ecosystems are determined by the interaction of two climate phenomena

What is causing the droughts that the Iberian Peninsula regularly endures? Why are the winters sometimes mild and rainy and other times cold and dry or cold and damp? Is climate change of anthropogenic origin exerting an influence on these processes? How are these cycles affecting the productivity of terrestrial ecosystems? And finally, can these cycles be predicted and the economy thus adjusted to them? The work, published this week in Nature Communications, was led by the University of Alcal de Henares. It was conducted in collaboration with the UPV/EHU, the University of Geneva and the University of Castilla-La Mancha and offers important keys for answering some of these questions.

The North Atlantic Oscillation (NAO) is a large-scale fluctuation in the atmospheric mass located between the area of subtropical high pressures and the low polar pressure in the North Atlantic basin, and is largely responsible for the periods of drought on the European continent. Previous studies show that the NAO has a great potential effect on various aspects, from carbon fixing and tree growth to fruit production and forestry pest cycles. However, the connection between long-term forestry productivity and the NAO presented some inconsistencies, such as periods in which climate cycles did not correspond to what was expected in terms of the NAO value. In their work the researchers in fact show that these inconsistencies may be originated by periodical anomalies in the surface temperatures of the Atlantic Ocean, known as the Atlantic Multi-decadal Oscillation (AMO). They are ocean phenomena that appear in the North Atlantic and by which ocean temperatures follow a cycle that takes about 70 years in total. These temperature changes in the ocean affect the atmosphere not immediately but with some delay.

19th-century data analysed using modern tools

The work in which Asier Herrero, the post-doctoral researcher in the UPV/EHU’s FisioClimaCO2 research group, has participated in is the result of a painstaking line of research that was started over five years ago and which includes data from historical archives, climatology, statistical models and forestry ecology.

“It has been a fascinating piece of work, dusting off archives of the end of the 19th century to get accurate estimates about how the productivity of forests across the Peninsula evolved over the last century and analyse them using 21st-century tools to understand the causes of the climate cycles and their consequences for the productivity of ecosystems,” explained the researchers.

The research integrates data on pine forests in various localities in the Spanish regions of Castilla-La Mancha and in Castilla y Len. “These pine forests were the means of livelihood of many rural areas from the 19th century onwards, and that was why a detailed quantification of the available resources, timber, pastures, resin, etc. used to be carried out,” he pointed out. The problem is that many of the previous pieces of work were based on projections of models and, furthermore, did not take into consideration the interaction between the two climate modes, the NAO and the AMO. Thanks to the existence of this time series, the study shows for the first time that it is the interaction of both climate modes which largely controls the productivity of ecosystems.

So the results of the work show that the AMO+ NAO+ and AMO- NAO- phases exert a high degree of control on forestry productivity owing to the reduction in rainfall and wintertime temperatures. The NAO is like a key that opens up and closes off the entry of areas of low pressure. What is needed, however, is the control of the AMO (linked to the temperature of the Atlantic at extratropical latitudes and the formation of areas of low pressure), which eventually determines the temperature and humidity of the air that reaches the Peninsula.

Volume increment as function of NAO and AMO. Three-dimensional regression plot showing predicted wood volume increment (VI, as surrogate of forest productivity) responses to standardized North Atlantic Oscillation (NAO) and Atlantic Multidecadal Oscillation (AMO) interactive effects. The colored layout (i.e., from red to blue increasing VI values) represents the marginal response curve when the other variables in the best model are kept to mean values (i.e., mean tree size and the calendar year). Letters indicate the areas in the response curve related to each potential sign of NAO and AMO combination

“The monitoring of the climate modes analysed may help to predict periods of severe drought, although it would not be an easy task, thus encouraging the applying of measures to adapt the forests more effectively,”

said Asier Herrero. During a drought, such as the one thrashing the Mediterranean in recent times, these findings could be crucial for water, agricultural and forestry planning, and in particular for assessing the climate vulnerability of the ecosystems.


The study:

“Forest productivity in southwestern Europe is controlled by coupled North Atlantic and Atlantic Multidecadal oscillations”. Nature Communications, 2017. DOI: 10.1038/s41467-017-02319-0 

(open access)



newest oldest most voted
Notify of
Richard Greene

I was so excited by the possibility,
of having the first comment,
I forgot what I wanted to say.
heh heh

Tom in Florida

You wouldn’t be the Richard Greene who starred in the Adventures of Robin Hood back in the late 50’s would you?


That guy was a dab hand with the long-bow.
I would like our present correspondent to show a similar talent.


Pablo an ex Pat

Couldn’t they just homogenize and adjust the data to show a trend agreeing with AGW like everybody else does ? Sarc Off. Nice to see that there are some open minded scientists in the game still !

Richard Greene

Oh yeah, I just remembered:

I don’t trust data from models.
or data from before 1900,
worrying about missing data,
and accuracy.

I hope the paper has a long,
discussion about data quality,
from the 1800s,
although I’d be surprised if it did.

I’ll read the study now,
and shut up.


I don’t trust the AMO data from whoever
or the AMO data from before 1950s
so I did some tests, and guess what, I trust them even less now (see my comment further down)


Basques never have bought into the global warming crap….


Sorry to break the news, but the study was leaded by the University of Alcalá de Henares in Madrid (first and last authors). The Basques in the study are co-authors, as the people from Albacete, Valladolid, and Geneva.

As regarding the position of Basques on AGW, as the rest of the Spaniards they are mainly bought into AGW. Public opinions in Spain against AGW, like mine, are very rare. But it is also an issue of least concern to Spaniards despite the constant propaganda by the media. When asked about the three main problems, climate change is not even an option, and is assumed to be a part of environmental problems. These environmental problems appear to be a main worry for just ~ 1% of the population.


Javier….I think you just agreed with me….1%


Fascinating! The interaction of the NAO and the AMO is decidedly non-linear, as if the feedback between them may be EITHER positive or negative, depending on the relative values. Since this is based on empirical data spanning centuries, it is a pretty strong indication that the assumptions if the computer models that individual feedbacks are either positive OR negative are not consistent with Reality.


l would have thought that rainy winters in Spain are linked to blocking over northern europe. Where the jet stream and the area’s of low pressures are forced track further south then normal. The cause of cold damp winters are more likely due to blocking over Greenland/northern Atlantic which can bring cold air from the north down across Spain. Where warm dry winters are likely to be due to a northern tracking jet which allows the Azores high to ridge across europe.


That is correct. Blocking conditions are more common during NAO– winters. If AMO is positive (according to the study) then Spain should get the most winter rain. However forest productivity is low in winter whether it rains or not. Spring rains are the most important ones for forest productivity, followed by late winter snow.

Despite being widely known, “the rain in Spain falls mainly on the plain,” is actually not correct. We get most of the rain at the mountains.


OMG, they got it in 1964!!!!


Here in England am pretty certain that recent warming of our winters is largely due to the lack of blocking during Jan/Feb. Because most of the winter warming has been happening during these months and its during these months when blocking is most likely to happen. Where during Dec there has been little change,which l think is due to the fact that during this month the most common weather is going to be “Atlantic weather” anyway.
My 40 year recording of the first winter’s snow also backs this idea up. As l have seen little in the way of delay in the timing of when the first snow over this period turns up..


But you also seem to “still” have snow. I though that that was supposed to be off … for good!


Yes we still get snow and cold spells, its just that there not turning up as often or lasting at long as they did during the 1960’s. lts this what l suspect is the reason for the recent warming rather the any delay to the start of the winter season.


From the 3D plot figure it is clear that when the NAO phase is fully positive, the AMO phase exert minimal effect on forest productivity (VI). And when the NAO phase is fully negative, AMO phase dominates VI.

The question I have is to what resolution of AMO phase does rhe author use? Annual AMO fluctuations or the long-term ~60-70 yr cycle phase (11 year running avg shown here below).

There is almost a religiously firm belief in the 60 years AMO quasi-periodic oscillation, which has filtered into N. Hemisphere temperature assemble (or vice-versa) and from there into global temperature.
However, there is some doubt about veracity of 60 year periodicity since neither the CET (England) or Armagh (N. Ireland) data show such periodicity, both the best temperature sets of data available and both at the edge of the N. Atlantic.
Since climates of Ireland and England are very similar, and are to a great extent influenced by the N. Atlantic it is possible that the AMO 60 year data periodicity is at best a crude integration of what is actually the case.


Nope, no cycle in Ireland data
comment image


comment image


There is a quasi-periodic multidecadal ‘oscillation’ but apparently not of the periodicity 60 or near years as found in the AMO data.


I happen to think that the Armagh data is probably better indicator than the CET in relation to what is going on in the N. Atlantic. Armagh is a good set of data from a single station, while the CET is amalgam of number of different stations with the influence of European continent’s climate to a greater extent than island of Ireland
It looks as the Armagh temperature is directly related to the Icelandic atmospheric pressure system – Reykjavik observatory.
Apparently the AMO has no much effect on the either of two, contrary to what I would expect.

And surface temps follow the moisture that is blown downwind.


I’m pretty certain that the rain in Spain falls mainly on the plain. Beyond tgat, I’m not so sure….

Phil R

Just a quick question to anyone who might have an answer. Not a climate scientist or whatever, but been following this for several years. I thought several years ago (2009-2010, maybe?), the prediction was AMO would be going negative around 2015 or so (and would roughly coincide with the PDO going negative). By the graphs above, looks like it will be a while longer. Did I misunderstand (most likely), or what ‘s the deal with the AMO?

Richard M

I think you might be confusing “going negative” with “reaching its peak and heading down”. The AMO appeared to reach its peak around 2010 so it will be awhile before it reaches negative territory.


“…accurate estimates…” – love it 🙂


or precise guess?


“…calculated guesstimates based on Finagle’s constant”?

Philip Mulholland