Long Range Winter Outlook

Analysis and Forecast by Steve Gregory...



COLDER THAN NORMAL GREAT LAKES - NEW ENGLAND AND GULF COAST

A colder than normal winter, especially during the first half appears likely for the Great Lakes, Upper Midwest and New England region, while below normal Temps are also likely across the entire Gulf coast region. However, while Temps will likely average below normal, it’s very unlikely that this winter will be even remotely close to last years’ frigid winter.

In contrast to the east, the western half of the nation should see another unusually mild winter – but should be significantly stormier than last year – with above normal Precip expected from northern California to Washington State as the mean storm track shifts further south compared to last year - especially by January. Near normal Precip appears likely for southern California, allowing for at least some recovery from the extreme drought of the last several years.


FORECAST MODEL GUIDANCE – AN ENSEMBLE OF TECHNIQUES

Based on the extended range GFS and EURO global ensembles, dynamical trends, along with forecasts and tools available from NOAA, Australia’s Bureau of Meteorology (BOM) and various University and Research group models (all of which call for at least some further strengthening of El Niño) – point towards a generally cold start to winter. However, unlike last year when we had long duration cold air events that included several cross-polar flow patterns – there should be a more progressive and stronger maritime influence as a more westerly flow of air moves into western North America due to an El-Niño like wind pattern. Using the CFSv2 and NMME Model Ensembles (that are made up many well known global and specialized climate models) an El Niño-like driven winter pattern should develop by late DEC or by mid JAN. This should lead to larger precipitation events compared to last year in the west and along the Gulf coast region, with a simultaneous lessening of the cold anomalies over the eastern half of the nation that will likely highlight the upcoming 6 weeks.

The colder pattern that is now developing across the US is partially the result of an enormous and very powerful West Pacific cyclone developing from the remnant of Typhoon Nuri (with the GFS forecasting a central pressure of 918mb by SAT morning). This super storm is building a strong ridge ahead of it – with an even stronger downstream TROF expected to form over North America next week. These type of strong tropical storm forcing mechanisms on the overall hemispheric wind pattern generally last 10-15 days – but rarely, for longer periods - IF there is a lack of pattern forcing from the Eastern Pacific (EPAC).

Another, albeit small, input to the Long Range Outlook is a fairly well studied phenomena related to the early development of a large AND unusually deep snow cover in Siberia during October. Absent a strong Pacific ENSO event influence, a deep Siberian snow cover in the Fall often leads to a cold North American winter as the large, low level source of cold air over Siberia ‘sets up’ a jet stream pattern that favors deep, arctic air mass intrusions into much of North America. Unfortunately, real-time snow depth reports from the Euro-Asian region (and especially Siberia!) is simply not available. Snow extent (derived from Satellite imagery) is, however, available on a daily basis, and current snow extent (and arctic basin ice coverage) is slightly greater than at the same time last year, and significantly more widespread than in OCT 2012. Although I’ve only given small weighting to these indicators, it was enough to forecast a colder winter than that given in the official NOAA outlook, but not nearly as cold as most every other private forecast services are calling for.


WHAT COULD DERAIL THIS FORECAST?

Numerous things could negate the expected impact from the warm Pacific waters – including the near impossible to forecast development of a strongly negative Arctic and/or North Atlantic Oscillation (AO and NAO respectively, even when we have a moderately strong El Niño event. Strongly negative NAO and AO indices typically results in a very cold winter in the eastern half of the nation. (Ironically, last year’s AO and NAO averaged near to above normal – and were a very big exception to the above ‘rule’.) However, there are other, lesser known ‘oscillations’ or Teleconnection patterns like the PNA, EP-NP, and PT. Most of these teleconnections do not become established until early DEC – but even then, these teleconnections can vary dramatically over the course of the winter.

In addition to the above uncertainties, ‘spectacular’ and Sudden Stratospheric Warming (SSW) tends to occur every few years that can bring about major hemispheric pattern changes – with very cold weather resulting in either Asia or North America. These events are even harder to predict than teleconnections which themselves can only be reliably forecast about 2 weeks in advance at best.

But at the moment, the most uncertain aspect of the below Winter Outlook has to do with Pacific Basin Temperature anomalies. This includes the slowly strengthening El Niño, but just as importantly, the very warm conditions off the west coast of North America that extends into the Gulf of Alaska, with another large warm water anomaly extending from the California coast westward to past the Date Line. A wedge of colder waters divides this mid-latitude warm water from the very high latitude warm water in the Gulf of Alaska (GOA) region. As explained in some of the below Figures, the Pacific Basin SST anomaly pattern has been continuously evolving over the past year, and is quite unlike last year at the same time, and will likely remain significantly different from last winter. This seemingly non-stop SSTA pattern 'evolution' has introduced even greater uncertainty in the current Long Range Winter Outlook. But one thing is clear – last year saw the development of a strong, Negative Phase of the PDO, whereas we are now in a Positive PDO phase pattern (which dominated the later 1980's and 1990’s). And whereas last winter featured a weak La Niña event, this years equatorial SST anomaly is clearly in the El Niño camp.

Overall, confidence in my current forecast is considered average, or a ‘3’ on a 1-5. It would be a ‘4’ if not for the larger than normal uncertainty for the overall Pacific Basin SST anomalies going forward (both relating to the PDO and intensity of El Niño) . An updated (and final) Winter Forecast will be issued at the end of the month, with a mid-winter update in JAN.


CLICK IMAGE to open full size image in new window

Fig 1: The evolution of Total Heat Content anomalies in the Equatorial Pacific over the past year. Starting in DEC 2013, cooler than normal conditions developed across the Equatorial EPAC, leading to a weak La Niña episode. By FEB, however, the strongest SSTA reversal on record developed, with heat content in the Equatorial Pacific reaching El Niño like levels (and which is why NOAA and most everyone else at the time began calling for the strongest El Niño event by this winter since 1997/98). This reversal was due to 3 strong westerly wind bursts related to a strong MJO signal over the west central Pacific. However, the pattern suddenly reversed again, with heat content falling off rapidly by the start of Summer. It was during the summer that the full impact of the early spring warming began showing up in the Atmospheric pattern, and this began a resurgence in warm Equatorial waters with a generally weak but warm ENSO episode now in progress.




Fig 2: One of the ‘noisier’ indicators for an ENSO episode is the Southern Oscillation Index (SOI). Despite the noisy aspect of the SOI, a negative SOI has taken hold over the past several months – and 5 month running means of -8 or lower are indicative of an El Niño event.




Fig 3: The above SST anomaly analysis charts show the stark differences in SSTA’s (Sea Surface Temp Anomalies) this year compared to last year across the Pacific Basin (Charts A and E). By the middle of last winter (Chart B), we had a significant cold ENSO event, but also a pool of very warm waters in the Gulf of Alaska. This anomalous combination, along with a basin wide Positive PDO SSTA pattern was the primary driver to the non-stop frigid winter weather over the US. By June (Chart C) we saw the cold ENSO event reverse to a weak warm one, with a continual warming across much of the Pacific Basin throughout the summer. This warm anomaly pattern peaked in AUG, but by SEP (Chart D), the SSTA pattern began changing yet again. The very warm GOA/Bering Sea SSTA’s reversed with much smaller warm anomalies in the northern GOA and a spreading wedge of cooler than normal SST’s from off the Japan coast eastward into the southern GOA (Chart F). In addition, a slow warming has continued in most of the Equatorial Pacific (Chart G shows the very large CHANGE in SST’s during the month of OCT alone). Finally, Chart H shows the subsurface SSTA’s in the upper 300 meters of the Equatorial Pacific. Though I have not included a time based animation of this chart, the typical ‘trend’ is for SSTA’s at depth in the western Pacific to spread eastward and rise ever closer to the surface in the EPAC over the ensuing 6-8 week period. For this reason, I expect warmer than normal SST’s in the EPAC to continue through the winter season.



Fig 4: The above SSTA patterns depict the ‘idealized’ PDO pattern for the Positive and Negative Phases that tend to last 15-25 years before reversing. Using the above SSTA charts, you can see that we are now in a Positive PDO Phase with a weak El Niño, unlike last year’s Negative Phase PDO, which also featured a weak La Niña event superimposed on it.




Fig 5: The above charts show the evolution of Sea Surface HEIGHT (SSH) anomalies courtesy of NASA’s JASON Satellite derived datasets, and considered the ‘Gold Standard’ for interpreting thermal based circulation characteristics of the oceans. Above normal SSH’s indicate relatively deep, warm waters, while lower heights imply colder waters at depth. What makes this image so useful right now is to compare it to the standard SSTA charts. Overall SSH’s have risen sharply from the central Equatorial Pacific to off the west coast of California, and have increased again over the Gulf of Alaska and Bering Sea – both of which show near to even below normal SST’s right now. This only makes sense when you realize that with the onset of winter, cold air moving across the Bering Sea and Aleutians is chilling down the surface waters in those regions. But at depth, warm water is again developing. In addition, the very low (cold) ocean heights in the far western and to some degree central Pacific is displaced south of the axis of cold water shown on the SSTA charts. Because of these differences, an even stronger Positive Phase of the PDO appears likely to develop over the next few months, with a warm ENSO episode superimposed on it. This is totally opposite of last years Negative PDO Phase and cool ENSO waters. As far as I know, in depth correlations and seasonal impact composites of such patterns (especially their relative intensities) has not been compiled. We ‘know’ what happened last year (and I believe the PDO and ENSO patterns contributed to at least 70% of last Years US Winter pattern) – but exactly how the reversal of the pattern this year will impact our weather remains unclear – but I expect it to play a major role.




Fig 6: Besides the ONI metric used by NOAA to determine the strength of ENSO events, another, more ‘complete’ metric is the Multivariate ENSO Index MEI which takes into account both SST’s and several atmospheric indices (the Index is expressed in Standard Deviations). This particular chart shows the evolution of ENSO events since 1950 that started off well in Neutral territory during the prior year. There were 4 statistically significant analogs to this year’s values, and in 3 out of 4 cases, a weak to moderate El Niño followed.




Fig 7: The above chart shows the top 7 El Nino’s since 1950. The true standouts remain the 1982/83 and 1997/98 events, when the index reached 3 standard deviations from normal. Both of those events induced extremely warm winters in the US. The remaining events ranged from 1 to 2 Std. Deviations, which were considered Moderate to Strong events. The current ENSO state is around 0.6 Std. Dev., but is expected to approach 1.0 by JAN (a weak to Moderate El Niño). But just as important is the warm water off the west coast, which when combined with a weak-moderate El Niño, should lead to a more pronounced El Niño atmospheric pattern than the ONI or MEI metrics alone would imply.




Fig 8: In support of the above position for El Niño is the POAMA index forecast from the Australian BOM that calls for additional strengthening of El Niño to well within the El Niño event category which is shown lasting into next summer.




Fig 9: A more detailed break-out of specific model forecasts for the Niño 3.4 region courtesy of the AUSI BOM for JAN and March, 2015. The NASA, EURO and Canadian models show the strongest warming, but all models call for at least weak El Niño conditions.



Fig 10: The above charts the ONI Index used by NOAA, and the CFSv2 Ensemble member forecasts for this index through mid summer 2015. Although the Niño 3.4 region serves as the basis for the ONI (and MEI) metrics, this ‘one size fits all’ – especially when there are significantly different SSTA intensities within the other Niño regions (like we have now) has more times than not led to busted seasonal forecasts. The warmer SSTA’s outside the 3.4 region, plus the warm water off the West coast and the positive PDO phase, tends to support (IMHO) the development of a well developed El Niño like atmospheric circulation pattern this winter.



Fig 11: The above representation of an El Niño pattern shows the impact on equatorial convection with warmer than normal SST’s in the EPAC. Convection is enhanced due to the warm waters, which forms a bit of a feedback loop with anomalous westerly winds spilling out of the T-storms in the Equatorial EPAC, which further reduces the normal upwelling off the South American coast and eastern Pacific waters in general. The convection also impacts the development, mean location and intensity of the winter season sub-tropical jet stream.




Fig 12: In a ‘classic’ moderate, and especially strong, El Niño episode, a relatively warmer flow off the Pacific tends to dominate North America – with above normal Temps and drier than normal conditions over the northern tier of states, and cooler, wetter conditions in the south due to the stronger sub-tropical jet.



CLICK IMAGE to open full size image in new window

Fig 13: A look back at the NMME Ensemble Model suite forecast for the month of OCT that was issued in early SEP. Two of the NCAR models had the best overall ‘pattern’ forecast, easily beating out the NMME ensemble itself. The NCAR model caught the cool anomaly in the east, but did not have the correct anomaly magnitude in the west. OTH, the NCAR CCSM4 model caught the general PATTERN overall, including the Temp anomaly magnitudes in the west, but was too warm in the east (though significantly cooler than in the west). With this in mind, a little extra weight in my forecast was given to these models than would normally be the case.




Fig 14: The latest available Winter Season (DEC-JAN-FEB) forecasts from the NMME model suite. While the ensemble solutions shown (NMME and IMME) call for a generally mild to ‘warm’ winter in most areas, the NCAR model was considered the best overall model – though it’s now in conflict with the CCSM4 model. (The CFS model, not shown, shows above normal Temps nationwide this winter. )




Fig 15: My Month-By-Month Temp anomaly forecasts for DEC-FEB. The cold start shown for DEC (but again, not nearly as cold as last year) is expected to gradually give way to warmer anomalies by later in Jan and in FEB. Temps should average above normal in the west throughout the winter, as will cooler conditions across the Gulf states region.




Fig 16: The full winter Temp Outlook is simply the arithmetic mean of the individual monthly forecast charts.




Fig 17: The primary storm tracks should result in above normal Precip in the PAC NW, and closer to normal over the SW US. Significant Gulf of Mexico and Ohio valley storm tracks are also indicated – all of which are typical of an El Niño circulation pattern.




Fig 18: The latest available NOAA forecast for DEC-FEB reflects a full fledged El Niño winter.

Steve Gregory