Verhaal Cohen over winter 2020. Speculatief. Niet eens met modellen.

In zijn blog van deze week schetst Judah Cohen een beeld van de winterverwachtingen van de diverse modellen: zacht en positieve NAO/AO. Hij stelt daarnaast zijn eigen visie, daarin stelt hij dat op basis van zijn observaties. Hij wijst o.a. dat de kou in oktober in delen vh NH niet voorzien waren door de modellen (Sebastiaan -> negatieve NAO deze herfst hebben de modellen ook niet voorzien), lage zeeijsbedekking en uitgebreide sneeuwdek. 

Impacts

As we gear up for winter and what weather to expect, I did find one feature of the dynamical models very interesting.  With a lack of much of a tropical signal, I feel that more forecasters are looking to the Arctic for signs of winter.  Though the average or ensemble dynamical models tend to smooth out large anomalies or signals, one consistent signal that seems to have emerged is a fairly strong or robust positive AO/NAO signal.  In Figure ia, I show the sea level pressure (SLP) anomaly forecast for November, December 2019 and January 2020 from the C3S ensemble of seasonal forecast models which include the ECMWF, UK Met Office and Meteo France global climate models.  One feature that jumps out is low pressure across the entire Arctic region with high pressure across the subtropics and mid-latitudes of the North Atlantic basin.  This is a classical positive AO and NAO SLP anomaly pattern. A positive AO/NAO is also strongly related to relatively mild temperatures across the NH but especially the Eastern US and northern Eurasia.  The C3S surface temperature anomaly forecast is warm across the entire extratropical NH continents but especially eastern North America and northern Eurasia consistent with the positive AO/NAO forecast (Figure ib). Just as an aside, in contrast the models are predicting high pressure anomalies across all of Antarctica, likely related to the Southern Hemisphere sudden stratospheric warming observed in September. This forecast is looking good as the Antarctic Oscillation (AAO) is currently strongly negative and predicted to remain so into the foreseeable future.

Figure i.  a) predicted sealevel pressure and b) surafce temperature anomalies November, December 2019 and January from 2020 from the C3S model ensemble. Plots provided by https://climate.copernicus.eu/charts/c3s_seasonal/

Now if you are a regular reader of the blog and/or my research then you are already familiar with my arguments that Arctic amplification which not only includes amplified Arctic warming but low sea ice extent and high snow cover extent in the fall is favorable for high pressure in the Arctic and downstream troughing and cold temperatures at least regionally across the NH continents.

Figure ii.  Observed surface temperature anomalies October 2019.  Graphic found at https://twitter.com/RyanMaue/status/1188866944813932545

Global October surface temperature anomalies are shown in Figure ii (taken from Ryan Maue’s Twitter feed).  The Arctic exhibits amplified warming relative to the rest of the globe especially in regions where sea ice normally present but is currently absent (see below Figure 16).  In fact, Arctic sea ice is currently at record low extent beating the previous record low of 2012 pretty handily and sea ice extent is more typical for mid-September than late October as seen in Figure iii.  And though temperatures are relatively mild across the NH continents, cold pockets do exist downstream of the warm bullseyes in the Arctic basin.  Below normal temperatures are observed in western North America downstream of the warm maxima in the Chukchi and Beaufort Seas, in Scandinavia downstream of the warm maxima around Greenland and Mongolia downstream of the maxima in the Barents-Kara Seas.  The regional cold anomalies were not predicted by the subseasonal dynamical models (see NMME October temperature forecasts ).

Figure iii. Observed daily Arctic sea extent through 27 October 2019.  Also shown is the climtological values and the daily extent from 2012. Plot taken from https://nsidc.org/arcticseaicenews/

Snow cover is relatively extensive across the NH as seen in Figure iv.  Snow cover is above normal both across Eurasia and North America. Based on my observational analysis this also favors high pressure in the Arctic with downstream troughing and cold temperatures across the continents.  So, if you are considering Arctic predictors in your winter forecast, I would say that they are quite bullish for severe winter weather at least regionally and possibly on even larger scales if the stratospheric PV gets involved.  And in many ways this upcoming winter forecast is a good proxy or symbolic of the current debate whether accelerated Arctic warming is contributing to colder mid-latitude winters.  The dynamical models clearly say no.  I believe observational analysis says yes and it will be interesting to see if this winter is consistent with the modeling or observational/empirical analysis. Of course one winter alone does not prove a theory or analysis but rather is one data point in a series or collection of data points.

Figure iv. Observed snow cover extent anomalies through 17 October 2019. Plot taken from https://climate.rutgers.edu/snowcover/

As I have written many times previously, I do think Mother Nature likes to foreshadow the weather.  So, here are some recent anomalies that have caught my attention and some initial thoughts but are not a forecast:

1. Snow cover is more extensive across Siberia this October relative to last October. I was surprised that cold temperatures were not more extensive across Northern Asia last winter, but I expect this winter to be colder. More expansive cold across Siberia increases the risk for Arctic outbreaks to both East Asia and Europe.
2. Scandinavia has been surprisingly cold this fall, and snow cover extent has also been above normal. This could be an early sign that this winter will be colder across Scandinavia than last winter, which could expand to include more of Northern Europe.  This coupled with Greenland blocking since the summer could be suggestive of an overall negative winter NAO.
3. I am surprised that pressure/heights have been as low as they have been in the Barents-Kara Seas this October. This coupled with the cold in Scandinavia could be a sign blocking in the North Atlantic basin could be favored more towards Greenland and less towards Scandinavia/Urals/Barents Kara Seas. Though Greenland blocking is strongly correlated with colder winter temperatures across Northern Europe and the Eastern US it is also less favorable for a major disruption of the stratospheric PV.  I believe that widespread cold is difficult without a significant disruption of the stratospheric PV.
4. Snow cover has been above normal across Canada and the lower 48 of the US and below normal across Alaska. Could be an early sign that ridging and warm temperatures are favored in and around Alaska with downstream troughing and cold temperatures in parts of Canada and the US lower 48.
5. Sea surface temperatures are well above normal around Alaska and the eastern North Pacific. This favors, ridging over Alaska and the eastern North Pacific but I see mixed signals where the main axis of the ridging sets up.  If it is along the West Coast like 2013/14 and 2014/15 the cold is focused in the central and eastern North America but if it is offshore like 2017/18 then the cold is focused more in western and central North America.

Of course, the dynamical models don’t agree with me and say ignore all that is going on in the extratropics.  All that is important is that the Arctic is warm and if anything, that the warm anomalies will spread to the lower latitudes.https://www.aer.com/science-research/climate-weather/arctic-oscillation/