Arctic Oscillation and Polar Vortex Analysis and Forecasts

January 22, 2018

Special blog on winter 2016/2017 retrospective can be found here - http://www.aer.com/winter2017

Special blog on winter 2015/2016 retrospective can be found here - http://www.aer.com/winter2016

Dr. Judah Cohen from Atmospheric and Environmental Research (AER) recently embarked on an experimental process of regular research, review, and analysis of the Arctic Oscillation (AO). This analysis is intended to provide researchers and practitioners real-time insights on one of North America’s and Europe’s leading drivers for extreme and persistent temperature patterns.

With the start of fall I will be transitioning to a fall/winter schedule, which is once every week starting sometime in October.   Precipitation forecasts will be replaced by snow accumulation forecasts.  Also there will be more emphasis on ice and snow boundary conditions (which are both now in their seasonal advance) and their influence on hemispheric weather.

Subscribe to our email list or follow me on Twitter (@judah47) for notification of updates.

Summary

  • The Arctic Oscillation (AO) is currently negative and is predicted to remain near neutral for the next two weeks.  But the AO has not been a good predictor of temperature anomalies or weather pattern reversals so far, this winter. Instead I have relied more heavily on the polar vortex (PV).  
  • The current negative AO is reflective of mostly positive pressure/geopotential height anomalies across the Arctic and mixed pressure/geopotential height anomalies across the mid-latitudes. The North Atlantic Oscillation (NAO) is currently positive with weak negative pressure/geopotential height anomalies across Greenland and Iceland but positive pressure/geopotential height anomalies across the mid-latitudes of the North Atlantic.
  • The North American circulation pattern, which has been dominated by ridging/positive geopotential height anomalies and warm temperatures in the west and troughing/negative geopotential height anomalies and cold temperatures in the east, will be mostly the opposite with troughing/negative geopotential height anomalies in the west and ridging/positive geopotential height anomalies and warmer temperatures in the east over the next two weeks.
  • Currently Eurasia is dominated by ridging/positive geopotential height anomalies across the north and troughing/negative geopotential height anomalies to the south. Northerly flow has allowed cold temperatures to become widespread across Siberia, Central Asia and Europe.  Over the next couple of weeks that cold air will continue to filter south.
  • Predicted persistent Ural ridging/blocking with downstream troughing near the Aleutians remains favorable for active energy transfer from the troposphere to the stratosphere.  Yet another pulse is predicted for this week in the troposphere and extending into the stratosphere by the end of the week. The pulse is likely related to relaxation of the pattern and mild temperatures across North America for the remainder of the month. 
  • The previous pulse that peaked last week resulted in a minor perturbation to the stratospheric PV including building of heights over Alaska and Eastern Siberia coincided with the rapid cooling over Siberia.   The next perturbation of the stratospheric PV could contribute to cross polar flow that is favorable to cold temperatures expanding across western North America at the very end of January into early February.
  • An important reversal is predicted over the next two weeks in the Gulf of Alaska from troughing/negative geopotential height anomalies to ridging/positive geopotential height anomalies.  This should lead to a pattern that is currently favorable to warm temperatures in the Eastern US to one that is favorable to cold temperatures in the Eastern US beginning in early February.

Impacts

The AO blog introduced a novel idea of ongoing analysis of seasonal variability at regular intervals and using recent weather model forecasts to update and revise seasonal forecasts similar to what is employed with synoptic meteorology. I was unsure whether it was appropriate to provide continuous “play by play” analysis of seasonal variability in real time similar to a sporting event.  Having done this for a few years now I have found value to the exercise of the blog, at least personally.  Hopefully those that are reading the blog are learning but I know for sure that I am learning from writing the blog and having to explain the ephemerality of the weather and the pattern changes while making sense of the chaos and anticipate the weather from days to weeks and even months.

One thing that has become clearer to me since I began writing the blog that it was overly simplistic to think of the polar vortex in a binary sense of weak and strong.  This influenced me to initiate the study that was published in a recent BAMS study.  This will be further expanded in a follow up study now in preparation.   This second paper I feel is very relevant for what transpired the first half of this winter.  We had a minor stratospheric PV disruption coupled with ridging near Alaska in late December that I believe triggered the intense cold air outbreak the following two weeks across eastern North America.  A second similar stratospheric PV disruption in late December coupled with ridging near Alaska took place last week that triggered deep and widespread cold across Siberia that is predicted to persist for at least the next two weeks.

I do find it interesting that similar minor PV disruptions with warming and ridging near Alaska are associated with very different responses at the surface.  The core of the cold temperature anomalies associated with the first PV disruption were in eastern North America while with the second the focus of the cold is in Siberia.  The best explanation for the difference that I can think of is that the location of the maximum heating in the polar stratosphere is the dominating feature.  In the first PV disruption the maximum heating is on the North American side of the Arctic helping to focus the ridging in the troposphere near Alaska and Northwestern Canada while in the second PV disruption the maximum heating was on the Eurasian side of the Arctic helping to focus the ridging near the Chukchi Sea and Eastern Siberia (see Figure i). Another possibly important difference is the center of the ridging.  In late December the center of the ridging was firmly over northwest North America while in mid-January the center of the ridging was shifted to the west over the Bering sea and straddling between Alaska and Eastern Siberia.  The axis of the ridging in late December favored downstream troughing in eastern North America while the axis of the ridging in mid-January  favored downstream troughing in Siberia and East Asia.

Figure i. (a) Analyzed 10 mb geopotential heights (dam; contours) and temperature anomalies (°C; shading) across the Northern Hemisphere for 23 December 2017. (b) Same as (a) except averaged from 21 January 2018.

Also the global forecast system (GFS) forecast does predict for the warming to migrate from the Eurasian side of the Arctic to the North American side of the Arctic with subsequent predicted energy pulses.  This could eventually favor a return ridging closer to Alaska supportive of colder temperatures in eastern North America.  There is admittedly some added complexity of what came first the chicken or the egg in my explanation and these ideas need to be supported with more thorough analysis. Still I do feel that these ideas illustrate the nuances of PV disruptions and coupling with the troposphere and sensible weather.  

I think that I have been open about my ongoing challenges anticipating regional variability in the North Pacific that is so influential downstream on North America weather.  An obvious place to look for predictability is the Madden Julian Oscillation (MJO).  I regularly follow the MJO and comment on the MJO and its possible influence on North American weather in the SSTs section.  Though admittedly this is not my expertise, being able to attribute variability of North American weather to tropical forcing would be of great benefit to the field including myself who wants to provide the most accurate forecast and comprehensive analysis.

The transition of the circulation pattern from troughing in the Gulf of Alaska and ridging in eastern North America at the end of the month to ridging in the Gulf of Alaska and troughing in downstream across North America at the beginning of February is being attributed to the MJO transitioning from the warm phases to the cold phases (in relation to Eastern US temperatures).  Broadly this may be correct but a closer look makes attribution more ambiguous.  This past week the MJO was solidly in the warm phases (3 and 4) yet it was impressively warm in the Western US and cold in the Eastern US.  Also the MJO is not predicted to enter the first of the cold phases (7) until the beginning of February.  Also there is a lag between the MJO phase and the circulation pattern across North America of about 8 days.  Therefore the attribution of a pattern of ridging in the Gulf of Alaska with downstream troughing across North America to the MJO is from mid-February and on. Yet the transition to that pattern is taking place almost two weeks earlier at the very end of January or first few days of February complicating attribution wholly to the MJO.  This is not meant to argue that there is no influence from the MJO but attribution is complicated a common challenge in understanding climate variability.

Near Term Conditions

1-5 day

The AO is currently slightly negative (Figure 1), reflective of mostly positive geopotential height anomalies across the Arctic and mixed geopotential height anomalies across the mid-latitudes of the NH (Figure 2).  Geopotential height anomalies are slightly positive near Iceland and Greenland but more strongly positive across the mid-latitudes of the North Atlantic (Figure 2), and therefore the NAO is slightly positive.

Figure 1. (a) The predicted daily-mean AO at 10 hPa from the 00Z 22 Jnauary 2018 GFS ensemble. (b) The predicted daily-mean near-surface AO from the 00Z 22 January 2018 GFS ensemble. Gray lines indicate the AO index from each individual ensemble member, with the ensemble-mean AO index given by the red line with squares.

The omega pattern continues across the Eurasian continent with ridging/positive geopotential height anomalies centered over the Ural Mountains with closed lows/negative geopotential height anomalies upstream over Northern Europe and downstream over East Asia (Figure 2).  This blocked pattern across the continent is  resulting in cold temperatures for the interior of the continent across much of Northern and Central Asia, Northern and Eastern Europe including the United Kingdom (UK) with mild temperatures for the North Slope of Siberia closer to the ridging across the Arctic (Figure 3).  Ridging/positive geopotential height anomalies and southwesterly flow across Western Europe (Figure 2) is bringing mild temperatures to Western and Central Europe (Figure 3).

Figure 2. Observed 500 mb geopotential heights (dam; contours) and geopotential height anomalies (m; shading) for 00Z 122 January 2018.

Troughing/negative geopotential height anomalies over Alaska and the Gulf of Alaska is forcing ridging/positive geopotential height anomalies downstream over much of North America (Figure 2).  Low heights and northerly flow is predicted to result in below normal temperatures across Alaska and Northwest Canada while high heights and southwesterly flow are predicted to result in  above normal temperatures for the US and Southern Canada (Figure 3). Temperatures are predicted to be mild across the US despite a closed low and associated blizzard traversing the Central US (Figure 2) and the brief period of cold air in the wake of the storm.

Figure 3. Forecasted surface temperature anomalies (°C; shading) from 23 – 27 January 2018. The forecast is from the 00Z 22 January 2018 GFS ensemble.

Troughing and/or cold temperatures are predicted to yield new snowfall for Western Asia, Siberia, Canada, the North-central US (Figure 4).  New snowfall is even predicted unusually far south across Central and Southern China (Figure 4).  In contrast, southerly flow and warmer temperatures will result in snowmelt across Central and Eastern Europe, parts of the Rockies and the Northeastern US (Figure 4).

Figure 4.  Forecasted snowfall anomalies (mm/day; shading) from 23 – 27 January 2018. The forecasts are from the 00Z 22 January 2018 GFS ensemble.

Mid-Term

6-10 day

The AO is predicted to remain neutral to slightly positive next week (Figure 1) with weak geopotential height anomalies in the Arctic and mixed geopotential height anomalies across the NH mid-latitudes (Figure 5a). And with weak mixed geopotential height anomalies stretching from Greenland to Iceland, the NAO will likely be near neutral to slightly positive as well next week.

Figure 5. (a) Forecasted average 500 mb geopotential heights (dam; contours) and geopotential height anomalies (m; shading) across the Northern Hemisphere from 28 January – 1 February 2018. (b) Same as (a) except averaged from 2 – 6 February 2018. The forecasts are from the 22 January 2018 00z GFS ensemble. 

The ridging/positive geopotential height anomalies centered over the Urals is predicted to dissipate this period (this is often the tropospheric response to upward wave activity flux as predicted for next week). Instead ridging is predicted to strengthen near the UK and Eastern Siberia with troughing/negative geopotential height anomalies further south across Spain and East Asia this period (Figure 5a).  Troughing and northerly flow will bring colder temperatures across Southern Europe and Southern Asia while high heights and westerly flow favor above normal temperatures across much of Northern Europe including the UK and Northern Asia (Figure 6).

Figure 6. Forecasted surface temperature anomalies (°C; shading) from 28 January – 1 February 2018. The forecasts are from the 00Z 22 January 2018 GFS ensemble.

The pattern of troughing/negative geopotential height anomalies across Alaska and the Gulf of Alaska with downstream ridging/positive geopotential height anomalies across much of Canada and the US is predicted to persist (Figure 5a). This pattern will favor once again normal to below normal temperatures for Alaska and Northwestern Canada and normal to above normal temperatures for remainder of Canada and the US (Figure 6).

Figure 7. Forecasted snowfall anomalies (mm/day; shading) from 28 January – 1 February 2018. The forecasts are from the 00Z 22 January 2018 GFS ensemble. Note the new projection to provide finer resolution.

Troughing and/or cold temperatures are predicted to yield potentially new snowfall for Scandinavia, Northern Asia, much of Canada and possibly the Northeastern US (Figure 7).  Mild temperatures are predicted to result in snowmelt in parts of Europe and the Western US (Figure 7). 

11-15 day

With only weak negative geopotential height anomalies predicted in the Arctic (Figure 5b) the AO is predicted to remain near neutral (Figure 1). With weak pressure/geopotential height anomalies persisting across Greenland and Iceland, with positive pressure/geopotential height anomalies stretched across the North Atlantic, the NAO will likely remain near neutral as well (Figure 1). 

Predicted persistent ridging/positive geopotential height anomalies across northern Eurasia favor ongoing troughing to the south across Europe and East Asia (Figure 5b). Induced northerly flow by high pressure to the north and low pressure to the south favors widespread normal to below normal temperatures for much of Europe including the UK and the interior of Asia while westerly flow will result in normal to above normal temperatures across the North Slope of Asia (Figure 8).

Figure 8. Forecasted surface temperature anomalies (°C; shading) from 2 – 6 February 2018. The forecasts are from the 00Z 22 January 2018 GFS ensemble.

Ridging/positive geopotential height anomalies previously just west of the Aleutians are predicted to slide east across Alaska and into the Gulf of Alaska displacing troughing/negative geopotential height anomalies  downstream into Western Canada and the Western US (Figure 5b).  Meanwhile ridging/positive geopotential height anomalies will persist across Eastern Canada and the Eastern US (Figure 5b).  Cold air previously confined to Alaska and Northwestern Canada will spread southeastward and cover much of Canada and the Northcentral US while normal to above normal temperatures persist for much of the remainder of the US (Figure 8).

Figure 9.  Forecasted snowfall anomalies (mm/day; shading) from 2 – 6 February 2018. The forecasts are from the 00Z 22 January 2018 GFS ensemble. Note the new projection to provide finer resolution.

Troughing and/or cold temperatures are predicted to yield potentially new snowfall for much of Europe, East Asia, Alaska, Canada and the Northern US (Figure 9).  Return of mild temperatures are predicted to result in snowmelt in Eastern China and Northeastern Siberia. 

Longer Term

30–day

The latest plot of the polar cap geopotential heights (PCHs) shows warm/above normal PCHs in the troposphere but cold/below normal PCHs in the stratosphere (Figure 10).  Though not clear from my version of the PCH plot (see instead http://www.cpc.ncep.noaa.gov/products/precip/CWlink/daily_ao_index/hgt.s...) it does appear that the warm PCHs in the troposphere propagated down from the stratosphere to the troposphere last week initiating a turn to much colder temperatures across Eurasia.   Cold/below normal PCHs in the stratosphere are predicted to peak this week before weakening while weak warm/above normal PCHs persist in the troposphere.

Figure 10. Observed and predicted daily polar cap height (i.e, area-averaged geopotential heights poleward of 60°N) standardized anomalies. The forecasts are from the 00Z 22 January 2018 GFS ensemble.

The plot of Wave Activity Flux (WAFz) or poleward heat transport shows a relatively strong pulse last week (Figure 11) that initiated the latest minor stratospheric PV disruption.  A new pulse predicted for this week will cease cooling in the stratospheric PCHs (Figure 10), turning the positive stratospheric AO back to neutral (Figure 1) and maintain the stratospheric PV center displaced between Svalbard and Greenland (Figure 12).  It also reinforces ridging/positive geopotential height anomalies near Alaska in the stratosphere.  Warming temperatures in the polar stratosphere on the North American side of the Arctic favors a return of ridging/ positive geopotential height anomalies  to near Alaska at the very end of January and into early February in the troposphere.  Northerly flow between the stratospheric PV over Greenland and the ridging near Alaska favors cross polar flow and cold temperatures in Canada.  The orientation of the trough axis across Canada first favors cold temperatures in Western Canada (Figure 12a) but then a southeastward expansion of the cold air (Figure 12b).

Figure 11. Observed and predicted daily vertical component of the wave activity flux (WAFz) standardized anomalies, averaged poleward of 40-80°N. The forecasts are from the 00Z 22 January 2018 GFS ensemble.

In the troposphere, periods of more active WAFz/poleward heat transport often coincide with a return to milder conditions across the NH mid-latitudes with the possible exception of East Asia.  Therefore, I do think that the model forecast of a relaxation of the cold pattern in North America this week is related to the predicted WAFz pulse seen in the troposphere (Figure 11). I do believe that the pattern of pulsing of WAFz punctuated by quieter periods has created highly variable weather at least across North America.

Figure 12. (a) Forecasted 10 mb geopotential heights (dam; contours) and temperature anomalies (°C; shading) across the Northern Hemisphere for 28 January – 1 February 2018. (b) Same as (a) except averaged from 2 – 6 February 2018. The forecasts are from the 15 January 2018 00z GFS ensemble. 

The repetitive pulses have so far have not been of sufficient magnitude to cause a major disruption of the stratospheric PV. I still see no robust signs of a highly disrupted stratospheric PV but rather relatively minor PV perturbations. Even so, I still believe that a more significant stratospheric PV disruption is possible and as I mentioned last week our highly experimental PV model predicts that the most likely period of a significant weakening is the second week of February.  I have leaned towards a more disruptive stratospheric PV given the easterly quasi biennial oscillation (QBO) and if one does not happen this winter then I will consider QBO phase less next winter. Still the impacts on the troposphere even from minor PV disruptions can be significant especially if they are repetitive.  This was the situation in winter 2013/14.  Though I still believe that this winter has diverged from the winter of 2013/14 in some important aspects.

Figure 13. Forecasted average 500 mb geopotential heights (dam; contours) and geopotential height anomalies (m; shading) across the Northern Hemisphere for February 2018. The forecasts are from the 22 January 2018 CFS. 

I include in this week’s blog the monthly 500 hPa geopotential heights (Figure 13) and the surface temperatures (Figure 14) forecast for February from the Climate Forecast System (CFS; the plots represent the mean of most recent four ensemble members).  The forecast for the troposphere is ridging for Alaska, Western Canada, the Western US, Greenland and much of the Arctic (Figure 13) with troughs in Eastern North America, Europe and East Asia (Figure 13).   This pattern projects onto a negative AO and would favor cold temperatures in the Eastern US, Europe and East Asia with mild temperatures in western North America, Scandinavia and Western Asia (Figure 14). As I tweeted yesterday, the CFS has been consistent with this forecast for the past week.  This forecast is consistent with the short range weather model forecasts and the ongoing disruption of the stratospheric PV.  I am often skeptical of the CFS forecasts but I believe that more confidence in this forecast is warranted.

Figure 14. Forecasted average surface temperature anomalies (°C; shading) across the Northern Hemisphere for February 2018. The forecasts are from the 22 January 2018 CFS.

Persistence does get stronger at the end of the winter and I do believe that the predicted pattern of cold in eastern North America can last for much of February and even into March.  Though as long as the WAFz remains active a period of milder weather is always possible.

Surface Boundary Conditions

Arctic Sea Ice

Ice grew very little this past week on the North Pacific side of the Arctic basin, therefore negative ice anomalies perist in the Bering Sea (Figure 15).  Sea ice is below normal in the Barents-Kara Seas as well and with time the greatest negative sea ice anomalies will likely reside in this region (Figure 15).  As I have been writing for much of the fall I do believe that the record low sea ice anomalies in the Chukchi and Bering Seas has focused ridging/blocking in this region.   The ongoing low sea ice extent could be supporting model forecasts of the return of blocking in the region.

Figure 15. Observed Arctic sea ice extent on 21 January 2018 (white). Orange line shows climatological extent of sea ice based on the years 1981-2010. Image courtesy of National Snow and Ice Data Center (NSIDC). Snow and Ice Data Center (NSIDC).

Recent research has shown that regional anomalies are important and the sea ice region most highly correlated with the winter AO is the Barents-Kara Seas region where low Arctic sea ice favors a negative winter AO and a cold Eurasia.  Continued below normal sea ice in this region may be contributing to more active WAFz/poleward heat transport predicted in the models that eventually could result in a stratospheric PV disruption.

SSTs/El Niño/Southern Oscillation

Equatorial Pacific sea surface temperatures (SSTs) remain below normal in much of the tropical and subtropical Pacific (Figure 16). El Niño/Southern Oscillation (ENSO) forecasts for this winter are that a La Niña is likely this winter. Observed SSTs across the NH remain well above normal though below normal SSTs exist regionally. Warm SSTs in both the North Pacific and North Atlantic could favor ridging in the ocean basins.

Figure 16. The latest weekly-mean global SST anomalies (ending 19 January 2018). Data from NOAA OI High-Resolution dataset.

Currently the Madden Julian Oscillation (MJO) is in four three and the forecast is for the MJO to enter phases five, six and then  phase seven (Figure 17).  Phases four through six favor ridging in the central North Pacific, troughing over the Western US and ridging over the Eastern US.  The predicted pattern of the MJO could be contributing to a milder  pattern across eastern North America the next two weeks. Phase seven favors troughing in the central North Pacific, ridgiing in teh Gulf of Alaska and troughing over the Eastern US.  The predicted pattern of the MJO could be contributing to a predicted colder pattern across eastern North America tin February.

Figure 17. I could not find an updated MJO forecast plot. Past and forecast values of the MJO index. Forecast values from the 00Z 22 January 2018 ECMWF model. Yellow lines indicate individual ensemble-member forecasts, with the green line showing the ensemble-mean. A measure of the model “spread” is denoted by the gray shading. Sector numbers indicate the phase of the MJO, with geographical labels indicating where anomalous convection occurs during that phase. Image source:  http://www.atmos.albany.edu/facstaff/roundy/waves/phasediags.html

Northern Hemisphere Snow Cover

North American snow cover (not shown) has remained steady and remains near decadal means.  Snow cover extent could expand further across North America.

Eurasian snow cover advanced this week but remains near decadal lows.  However with predicted cold temperatures, snow cover extent across Eurasia could expand further.

October Eurasian snow cover extent was above normal relative to long-term averages. Above normal Eurasia snow cover favors a strengthened Siberian high and a weakened polar vortex/negative AO this upcoming winter with cold temperatures across the continents of the NH.  The most likely timing of a weakened stratospheric AO followed by a negative AO is January.  Though despite the above normal snow cover extent, the rate of the snow cover advance for October was slower than the average rate of advance relative to the past two decades.  I do think that the slower rate of advance of snow cover favors a relatively strong stratospheric PV in the near term.