Winter Outlook Part 3: Larry Cosgrove

I have a great deal of respect for Larry Cosgrove of Weather America. While we never met in person, we shared a weather center in Philadelphia. I worked at WPHL on a freelance basis in 2001, and again in 2003. Larry was there much longer, before he moved up to New England. Tony Pann and I had him on our radio show Weather Talk as a frequent guest, and I enjoy his in depth analysis of the atmosphere. Below is his disection of the winter of 2008-2009. It is long, but worth the read:
2008 - 2009 WINTER SEASON FORECAST

Transient +PNA and -AO with recurrent NAO make for a very wintry Midwest and Eastern Seaboard!

Overview

Using persistence, climatology, and comparisons with analogue years that showed certain similarities to trends seen in weather so far in 2008, a forecast for the upcoming winter season was compiled showing expectations for temperature and precipitation in North America, the European Union, and Asia. Special emphasis was placed on potential for extreme conditions (cold vs. warmth, heavy snow or ice risk, and excessive against minimal precipitation accumulation).

October Synoptic Pattern


There have been prevailing patterns around North America which, through persistence, may offer clues to the trends of synoptic systems that will occur during the winter months. If we count the two tropical storms (Hanna and Kyle), there have been no less than six storms paralleling the Atlantic coastline since early September. Deep tropical moisture connections are evident in all cases, so if such disturbances were to occur in winter, the prospects for snow and/or ice along the Eastern Seaboard will be higher than usual.

It should also be pointed out that, despite the variability of cold to warm conditions over the West, strong ridging routinely occurred at higher latitudes over the previous six weeks. You can see this signature in the 500MB longwave pattern graphic further down the page. Any ongoing tendency of high latitude ridging over the Northwest territories and Nunavut AR must be treated as a feature favoring drainage of Arctic air into the eastern two-thirds of the U.S. during the low sun period.

Possible Influence Of ENSO And PDO


For much of the summer and fall of 2008, the El Nino/Southern Oscillation signature has been in the neutral categories, between a warm (+ENSO or El Nino) episode and a cold (-ENSO or La Nina) phase. Most forecasts maintain this neutral position through the upcoming winter (although the most recent CFS outlook suggests a return to a weak La Nina), which seems a viable idea as most important changes in oceanic temperatures across the equatorial Pacific Basin will occur between spring and autumn. ONI measures are at zero, which, in the lack of any other forcing variable, should imply that simple climatology would have an edge in the seasonal forecast. One x-factor: we have seen two strong Kelvin waves progress out of the equatorial Indian Ocean since mid-September. These impulses could very well warm the neutral character of the waters between the Philippines and Hawaii, creating a window for a neutral/positive or even weak El Nino episode at some point late in the winter.

One possible factor in play for this winter is the Pacific Decadal Oscillation (PDO), based on the measure of SSTs in the Gulf of Alaska. If waters are colder than normal, a moderating or zonal influence is sometimes evident over North America. Should sea temperatures warm to above normal below the farthest north state, correlations exist that favor -EPO or +PNA ridge formation (often a warm condition in the western states versus colder air to the right of the Rocky Mountains. While a -PDO value has been in place for much of the past year, there has been some rightward flex from much warmer seas to the west and south, clearly evident on the October 20 chart. It should also be said that there are cases (such as January 1961) where a repeated negative height anomaly over the Gulf of Alaska was simultaneous with extreme cold and snow over the eastern half of the U.S. For this reason, I feel that the modifying influence of this signature may be muted during the coming months.

MJO And Kelvin Wave Signatures


Any discussion of possible contribution of the Madden-Julian Oscillation to winter weather over North America must be tempered by these facts: one, that the processes of the MJO and its effects are still not well understood (teleconnections). Secondly, that forecasts for the various phases of the oscillation are still quite poor. Thirdly, any estimation as to the position and strength of the MJO is virtually impossible beyond the traditional 15 day period.

There have been two major Kelvin wave ejections from the area between Sri Lanka and Sumatra since September 1. Both of these impulses seemed to have the effect of amplifying the polar westerlies upon emergence into the equatorial Pacific Ocean (a case that was seen many times during the previous winter). keep in mind that after passage of these waves, waters tend to warm. This is an argument against the formation of a La Nina signature (while supporting continued neutral character of the western and central ENSO sectors). On the idea that we may speculate on the same or slightly increased activity with the MJO over the December - March period, at least six large winter storms and cPk/cA advection events may occur.

Analogue Years

There are three criteria used here for a comparison to apparent weather during the upcoming winter. One is the ONI signature for autumn, which characterizes the ENSO configuration immediately precedent to the winter season. Next is similarity in synoptic pattern, a way of how weather during 2008 looks went placed against previous years. Lastly, similarity in tropical season: how did incidence of warm-core cyclogenesis rank against previous years?

This fall has seen an almost purely neutral ENSO, with ONI measure of zero and an increasingly homogenous look to all sectors of the equatorial Pacific Ocean. Previous autumns with this reading are 1960, 1966, 1980, and 2001. So the winters of 1960-61, 1966-67, 1980-81, and 2001-02 are added to the analogue list. An additional year, 1973 is also added as its Multivariate ENSO Index (MEI) values closely follow those seen in the past few months; therefore the 1973-74 low sun period is considered for this study.

The spring and fall of 2008 bore uncanny parallels to that of 1993: multiple cases of heavy rainfall and severe thunderstorms across the Corn Belt with heat and drought an issue for much of the Eastern Seaboard. With this reflection in mind, the 1993-1994 DJFM period is included.

Possible comparisons exist with years where tropical cyclone output was more or less equal (+/- 1) between the Atlantic Basin and Eastern Equatorial Pacific Ocean, where the number of storms exceeded seasonal climatology. This caveat is significant, because there is an indication of a higher rate of interaction between the tropics and higher latitudes. Therefore, the winter seasons of 1954-55, 1959-60, 1961-62, 1988-89 and 2002-03 were added to the analogue group.

Character Of The Stratosphere, 250MB Jet Stream, And 500MB Features

In recent years, efforts to forecast stratospheric temperature have been made as an aid to surface weather forecasting. Typically, a warmer 30MB reading will correspond to colder surface values which, in cases where advection mechanisms are present, will impact lower latitudes about a week to two weeks after the foundation of the high-level ridge complex. The strongest intrusions of cold air come when the stratosphere warms suddenly above a vast ice and snow field. This was the case last year, when the mean positive thermal anomaly formed in February and was accompanied by near-record cold in some parts of the Great Lakes region.

Forecasting the development of these warm pools in the highest elevations of the atmosphere is still not easily accomplished, so the use of past-year analogues may not be helpful here. The presence and size of the polar cold core, however, might be a parameter worth watching, as the warmer alignment of temperatures will rotate below this feature. If so, this could mean an extensive period of cold across much of the U.S. in December, when the warmer banding over northern Asia works eastward. Suggested reading for this phenomenon include The Stratospheric Sudden Warming Website and Monitoring of the Stratospheric Circulation.

There are two important synoptic features seen in late September and October which stand out in relation to the coming winter. One is the character of the jet stream, which has been split over and about North America. The semizonal velocity maximum at 250MB has remained over the western and central Pacific Ocean, while the resumption of unified flow aloft tends to occur just to the right of the Grand Banks. If the upper level wind fields were to be strong and un-amplified across the northern U.S., then at least the start of the December through March time frame would be mild and dry. But the broken appearance of the wind fields, and the recurrence of above normal 500MB heights in Arctic Canada (-AO) would seem to suggest that potential for cyclogenesis and cold air advection is fairly high in areas east of the Continental Divide. Add to this what appears in analogue years as a high rate of occurrence for a negative phase of the North Atlantic Oscillation and transient examples of a +PNA flow configuration, and chances for advection of cold air into lower latitudes seem to be higher than average.

Temperature Forecast

Note that snow cover across the Northern Hemisphere has been increasing, approaching the seasonal normal for late October. Extensive snow fields are important for generation of cold temperatures, and with acreage of snow and ice over the northern portions of the continent now filling in (especially into QB, Nunavut AR and the Northwest Territories), a case can be made that building of high pressure into the U.S. and southern Canada will be accompanied by meaningful displays of cAk (coldest Arctic) air masses.

On the whole, I expect colder temperatures to dominate much of the lower 48 states. Core negative anomalies will be over the lower Great Lakes and St. Lawrence Valley, as well as central Canada. A tendency for Alberta Clipper type storms and flexed +PNA signatures will probably allow for cA intrusions to reach into the Mid-South and Eastern Seaboard. Occasional digging of shortwaves into Colorado (precedent to Colorado/Trinidad storm development) may allow for a secondary region of colder values in the Salt Lake Valley and central Front Range.

The tendency for ridge formation over the Pacific Northwest and Continental Divide may allow for relatively mild conditions in much of the Intermountain Region and the High Plains. It should be noted that if the -PDO environment strengthens, then this area of warmth would be compressed or perhaps eliminated. The tendency for ridging to develop in the Southeast, as noted in many of the analogue years for the period after January 15, favors the highest temperature profiles from the Gulf Coast into Florida and Georgia.


Month - To - Month Breakdown
Here is the forecast for the upcoming winter temperature trend, month by month. Notice that January is, by area and temperature deviation, the coldest month (while December chill impacts the Eastern Seaboard). A key issue with winters with a -PDO (coldness of Gulf of Alaska waters) is a recurrent ridge over the Southeast. I believe that as spring approaches, a tight gradient may emerge from the Ohio Valley into the Mid-Atlantic states. Think of the transition as a battleground between the entrenched colder regime associated with the transient +PNA and -AO related ridging , in concurrence with -NAO aligned blocking near Baffin Island or Greenland. This boundary could be a conduit for storms, with late-season snow and ice events across the Great lakes and Northeast. Colorado cyclones which undercut transient +PNA ridging may stimulate colder temperatures in parts of the Intermountain Region through the winter, while the unstable positive height anomalies over the Pacific Northwest keep mPk frontal structures off of the Gulf of Alaska from making frequent passages.

DECEMBER

Repeated cold intrusions from the Upper Midwest into the Eastern Seaboard. Generally warmer across the Intermountain Region and Great Plains, with area from Utah and Colorado into Missouri and Arkansas a buffer zone between returning warmth along the western Gulf Coast and a mild regime parallel and west of the Continental Divide. Some potential for occasional occurrence of colder values in the heartland of the U.S. due to Colorado/Trinidad "A" storms with increasing snow. Greatest threat for major winter storm will be over the major cities along the Interstate 95 corridor.

JANUARY

Widespread cold except for the West Coast and the Southeast. Active storm track from western Gulf of Mexico up along the Atlantic shoreline will create ice and snow hazards over the eastern third of the U.S. Lake effect snows may be crippling in Cleveland OH and Buffalo NY metro areas, with occasional issues from snowfall in Chicago IL, Indianapolis IN, Cincinnati OH and Pittsburgh PA. Ice storm risk may be higher through the interior of Dixie and along the Eastern Seaboard.

FEBRUARY

While cold strengthens its grip over much of Canada and the Midwest, I suspect that ridging will be an increased presence over the Southeast (see the analogue mean 500MB composite anomaly above for insight into this possibility). Very active procession of both Alberta Clippers and Colorado/Trinidad "A" cyclones along gradient between FL-GA positive height anomaly and cAk intrusions that will reach as far south as Texas and Appalachia (some spillover into the Northeast as well).

MARCH

Very similar to February scenario except that flat subtropical high may make inroads into the entirety of the Deep South (creating some warmth from Texas into the Carolinas). Extended winter for the Midwest as a whole; some risk present for late season surprise ice or snow event in the Northeast with a cold air damming scenario (i.e. backdoor cold front and overrunning).

Storm Track Scenarios

The tendency for storms over the western Atlantic Ocean and adjacent coastline stands out as we head into the colder months. Since stronger features tend to regress and deepen during winter, the lows which have formed near or east of Bermuda may instead pass closer to the Eastern Seaboard. Likewise, some potential exists for Piedmont cyclones, originating in the Deep South and passing west of the Interstate 95 corridor. The most likely options for track scenarios are for Colorado/Trinidad "A" cyclones and Hatteras (Nor'easter) Lows, as examples of these systems have already occurred, and will be favored to develop again under a transient +/- PNA configuration. Alberta Clippers may take shape in middle and late winter, precedent to cAk intrusions; Piedmont storms and Galveston Bay Spin-Up cyclones are also a strong possibility, with ridging over the Sargasso Sea setting up against digging energy from the High Plains.
Precipitation, Ice, And Snowfall
A very clear pattern emerges with the averaging the 11 seasons used as analogues: the best risk of precipitation is over the Pacific Northwest and the Old South. If the overall synoptic pattern verifies, the presence of high moisture advection against incoming shallow cold air sends all kinds of alarm bells out for the possibility of major ice storm formation around Appalachia and the Piedmont (including the markets of Birmingham AL; Atlanta GA; Charlotte NC; Raleigh NC; and Richmond VA).

Rather than attempt a calculation of total snowfall or probability of snow for the season, I felt that the best option was to match the possible synoptic signature against the mean temperature pattern set for the analogue years. This comparison could help to determine possible situations for snow vs. ice/mix vs. rain, envisioning the storm track types mentioned above. The way to read the map is fairly simple: the dark blue would be the farthest south demarcation of where snow is likely to occur; the electric blue where accumulations of snowfall are possible. Snow events that could cause moderate impacts on travel and utilities are marked in light blue, whereas disruptive or heavy snow threats are most probable in the gray zone. Where marked white are locations where memorable snowstorms could occur. The Black Hills vicinity are almost always hit by blizzards or mammoth snows in late winter and spring. But the area from WV and VA into SE ON represents a synoptic system in the heart of winter; I feel that middle Appalachia and the lower Great Lakes will be the area to watch this winter for a critical snowmaking cyclone, with lake-related additions after passage.

One point which should be made: the area seeing the best overall potential for precipitation has been affected by long term drought, with a lack of consistent water accumulation during the past two years. There should be ample potential to erode the water deficit during the upcoming winter.

Summary

Confidence in this forecast is about average. Complications which could deter verification are stratospheric temperature anomalies (somewhat cooler over North America this fall when compared to last autumn); the possible effects of a -PDO measure; the repeated presence of mid-latitude cyclones across much of the western Atlantic Ocean (which, if these storms do not retrogress as expected, would lead to a colder score across the eastern third of the nation. And the evolution of the ENSO signature across the Pacific Basin, which as of this writing was solidly neutral.

In general, placement of polar and Arctic regimes will be somewhat further south and east than last year. The average of analogue years strongly supports development of a subtropical high near the southeastern U.S. during February and March, which could begin to offset cold advection and lead to much-above-normal average temperatures in about the state of Florida.

A strong case is made for abundant precipitation, much of it frozen, over the Old South, Appalachia, the Great Lakes and the Eastern Seaboard. Ice storms may prove to be an ongoing hazard in the Piedmont and lower Appalachian Mountains. Overall threats for major snow events are higher, with chief risk for a critically high snow and wind situation over the lower Great Lakes and upper Ohio Valley.

Prepared by Meteorologist LARRY COSGROVE on
Thursday, October 30, 2008 at 4:20 P.M. CT

Disclaimer:
The previous statements are my opinions only, and should not be construed as definitive fact. Links provided on this newsletter are not affiliated with WEATHERAmerica and the publisher is not responsible for content posted or associated with those sites.

Copyright 2008 by Larry Cosgrove