Tuesday, March 9, 2010

Cold winter 2010

Geert Jan van Oldenburgh of KNMI has prepared an analysis "Klimaat 2010: a winter of extremes - Cold in Europa, Siberia and the US, but above normal globally".

His summary is: From Ireland to Siberia the winter of 2010 was colder than usual. The winter temperatures in the eastern half of the United States were also below normal. However, on Greenland and in north-eastern Canada it was much less cold than normal in winter and the Middle East was also milder. This pattern was caused by a well-know climate mode of variability: the Arctic Oscillation (AO), also referred to as the Northern Annular Mode. The Arctic Oscillation had its largest negative excursion since the beginning of observations in 1900. The net effect on the global mean temperature was close to zero: the cold air brought south to Siberia, Europe and the eastern US was compensated by warm air going north into other regions. The global mean temperature was in fact near record high, pushed up by the combination of the slow trend and the effects of the 2009/2010 El Niño.

Full text: http://www.knmi.nl/cms/content/78168/2010_a_winter_of_extremes


P Gosselin said...

"This pattern was caused by a well-know climate mode of variability: the Arctic Oscillation (AO), also referred to as the Northern Annular Mode. The Arctic Oscillation had its largest negative excursion sinc the beginning of observations in 1900."

So here we're talking about a phenomenom (pattern or cycle) that has been around for quite a long time.
I just think climate change would mean new, odd patterns emerging, and not old ones from the past simply repeating themselves.

Hans von Storch said...

No, P. Gosselin,
climate is the statistics (frequency distributions, cross-correlations, EOFs, spectra etc.) of weather (in a somewhat general sense), climate change is thus a change of these statistics. While it can not be excluded that some new phenomena may show up, first of all we keep "our weather", but the frequency (and possibly intensity) of events is changing.
Thus, there will be sunshine, rainfall, storms, dry days, frost and snow in future, ENSOs, positive and negative NAOs. More warm days, less cold days, less cold winters, more hot summers - but not: no cold winters, only hot summers.

Bayesian Empirimancer said...

A nice summary, though it a curious use of the word 'caused'...

_Flin_ said...

Concerning the AO: Daily Kos reported a connection of the Gulf Stream to the West Greenland Current happened on December 30th, due to the extreme AO. Did that really happen? I cannot find any other source for that.

And if it did happen, can this be one of the reasons that East Canada Sea Ice was so low?

Hard to keep track of causes and effects here for an amateur.

Anonymous said...

I have again a very stupid question:

Doesen't this special weather pattern lead to a higher albedo? This could be one of the negative forcings in climate?

Best regards and thanks again for the kind answers!!!!


P Gosselin said...

What do you mean by "events"? Weather? Cycles?
The terms mode, pattern, oscillations are used by van Oldenborgh. He makes no mention of "events".
I ask this because I'd like to know which "events" are changing in frequency and perhaps intensity. Are these changes in frequnecy in agreement with the models?

Hans von Storch said...

P. Gosselin - yes, I mean "weather", as we are speaking about a change in the statistics of weather. When we run a climate model, we get a sequence of weather (if you wish: events), sampled every 10 minutes or whatever the time step of the model is. From this realization of the weather(which will be different, if we rerun the model with very minor modifications, such as one day earlier) extending over many years and decades, even centuries and millennia, we derive by averaging the statistics of weather, i.e., the climate. Thus, what the models do, is what you say - simulating different weather.
Which weather elements are changing - I would say: all - but many only to an insignificant degree. But, check out our regionaler Klimaatlas für Deutschland - this web-page displays ranges of scenarios of changing weather statistics, such as number of summer days and the like.
The next question is - detection and attribution-type - is the projected change in frequency (and other stats) consistent with the ongoing change? This takes a bit longer, but should be discussed sometimes.

_Flin_ said...

@Eddy: To figure out whether this phenomenon leads to higher albedo you'd have to compare the increase of albedo in the cold regions due to more snow cover to the lower albedo in the warm regions. Over the last 28 years, however, snow extent (in the northern hemisphere) has a decreasing trend.

Concerning the "negative feedback", of the top of my head I doubt that this is a feedback at all, neither for warming nor for higher CO2. In the winter AO index there doesn't appear to exist a trend at all, although I did not calculate it. To me it looks more like a natural event. Which is just an opinion.

Rob said...

Attempt to another answer to Gosselin's remark: "I just think climate change would mean new, odd patterns emerging, and not old ones from the past simply repeating themselves."

Of course, the KNMI fellows spotlighted a more or less cyclic phenomenon. I think it was their intention to deliver an explanation for this cold winter, since many people are going to doubt about warming. You should consider that climate scientists are trying to figure out any single "cause" of climatic variations. it is like spectral analysis of light, noise etc. There are sunspot cycles, Arctic cycles, Ninos and Ninas, and last but not least Multi Decadal Oscillations (which - according to Latif - will contribute to relative cooling in the upcoming years). The more cyclic effects are identified, the more accurate figures of linear trends can be outlined (where the actual "average weather" is given by the sum of all these effects). It's up to you to accept or deny linear temperature increase "under the hood", as long as it is not unambiguously proved, but please be aware that we cannot afford to play with fire when linear tendencies can only be stopped or braked with huge, huge delays.

eduardo said...

@ Eddy,

it is an interesting comment. It would be a feedback if:
-CO2 forcing would gave rise to sustained positive or negative Arctic Oscillation
- and if a sustained positive or negative AO would be linked to a change in albedo

Concerning the second point I think it is reasonable to think there would be changes in albedo, but one had to consider the regions with increase and regions with decrease in snow cover.

Concerning the first point, climate simulations tend to indicate that greenhouse forcing causes a stronger AO, opposite to what is happening this winter.

So if a stronger AO leads to less snow cover over the Northern Hemisphere, this would amount to a reinforcement of greenhouse gas forcing, a positive feedback.

Geert Jan said...

Regarding the effect of the AO on the global mean temperature, the explanation given is a simplification to make the piece more readable. There are indeed snow feedbacks, with radiative cooling over snow amplifying cold signals. However, I did an easy calculation: take the global mean temperature and subtract the regression on the CO2 concentration. This removes most of the trend and leaves natural variability & other external forcings (solar, volcanoes, residual aerosol signal). Next make a scatterplot of this detrended Tglobal against the AO for winter. By eye there is no connection, the correlation coefficient is 0.06. This shows that in the end the AO does not affect Tglobal much. (You can repeat this using my climate analysis web site, http://climexp.knmi.nl/)

Anonymous said...

Hello Flin an Eduardo,

Again I have to thank you for the very kind answers!!!

I thought, that the snowcover in Canada and in Greenland might be about the same than during the last years. But in Europe etc. there is much more white than during the whole last decade.

This could even lead to a colder summer ...? Even if the summer has already begun in Israel and is VERY hot. ;-)

Could also be e very hot summer ... I'm not a prophet. ;-)

But it was just an idea. Thank you very much one more time. I'm really happy about this blog! :-)

Best regards

Neven said...

That was a very good question, Eddy. I've actually thought quite a bit about it when I discovered a website by a man called Tom Wysmuller who maintains that the melting of Arctic Sea Ice will actually cause cooling in the long run. This is based on a theory in the 1950's by two scientists called Ewing and Donn (you can read a bit more on that theory on the website of science historian Spencer Weart). Less Arctic ice means warmer sea water, means more evaporation, means more 'ocean snow effect' (similar to lake snow effect).

From Wysmuller's website:
as more Arctic Ocean becomes exposed for longer times before being sealed by seasonal pack ice formation, snows will fall earlier each year, in larger amounts, and linger on the ground longer into the spring. Albedo reflectivity from snow is negligible during the dark arctic winter, but both in early fall, and as spring arrives, the snow in higher latitudes remains on the ground longer, and reflectivity increases. Thus the central US and the East Coast will experience cooler springs and the transition into summer temperatures, although somewhat later, will be relatively rapid. The US West Coast will have warmer springs and summers, as will Western Europe. Warmer weather may may not make it through October’s end almost everywhere, and winters will be sharply colder. The massive pack-ice loss of 2007 set up ocean effect snows on the Arctic Shores that increased reflectivity. Colder Northern Hemisphere winters with sharp temperature contrasts are an inevitable result.

It's an interesting theory, but I'm not sure how it will play out.

gregor said...

the knmi article sets the temperature mean from 1971 to 2000 as "normal"; german meteo-services usually set 1961-1990 mean as "normal". this sounds so extremely arbitrary to me, as humans are altering the climate in diverse ways since they first invented agriculture millennia before today; the influence of the global human population on (local/regional/gobal) climate (= weather statistics) did surely not start in 1850. so how do climatologists set their "standards"? who tells us, what is the "normal climate". In my very personal opinion this is one of the biggest flaws in the current AGW hysteria.


Neven said...

Where do they state it is normal? I thought the function of a baseline is to be able to measure anomalies?

For me 'normal' climate is where the weather doesn't get so capricious so as to become structurally problematic for agriculture or cause structural damage to society.

An increase in droughts, flash floodings, hurricane intensity and/or frequency, is not what I would call normal (ie caused by natural variability). Emphasis on the word increase, before you think I'm saying these events cannot be caused by natural variability.

Tom said...

@ Gregor

"climate normal" is not the same as "normal climate". When calculating anomalies, useful for objectively identifying "high" and "low" periods, it makes sense to have a standard period for the mean from which you calculate the anomaly. The mean over 30 years is chosen as covering typical variations in climate, but which is not unduly influenced by long term trend. I don't know, but I guess the 1961-1990 period is most often used simply because it is the first 30 year period giving reliable measurements of most climate variables at most locations. There is no assumption that the data over the standard normal period represents the "typical" climate. And papers using anomalies always indicate the period of the standard normal used. Your suspicions are groundless.

gregor said...


i would insist that a majority of (non-science) people out there get the feeling that meteorological/climatological reference periods stand for "the typical climate" of the "good old times" although no serious scientist would argue this way. i guess this is mainly a personal impression that i get from media coverage of weather/climate issues. i would also insist, that the theory of ANTHROPOGENIC global warming - as represented by the IPCC - has quite a limited view of how the human race were and are altering local, regional and therefore global climate. i refer to articles of r pielke sr. or b ruddiman and the like. therefore the basic statement of the IPCC, that humans are altering the global climate in a significant (and dangerous?) way since (somewhere in between 1750 and) 1880 is imho misleading, as we started having major impacts on the climate thousands of years ago.