October 18, 2004

For many areas across the country, last winter brought its fare share of cold and snow to much of the same places that had seen it previously in the winter of 2002-03. New York City saw two back-to-back 40”+ seasonal snowfall totals. One would have to go all the way back to the winters of 1947-48 and 1948-49 to find two seasons with similar snowfall statistics. There were exceptions to the general rule however. Places such as Washington DC not only saw far less snowfall last winter than they did in 2002-03, they also recorded below average snowfall. Baltimore received only 18.3” of snow in comparison to their 2002-03 total of near 60”; which is a one year difference of 39 Inches!

So how about this winter?

Early indications are that 2004-05 may provide snowlovers with another great delight; especially those in the eastern and northeastern parts of the country. Our suite of global predictors--which we look at in order to determine how the season may evolve suggest that a cold and snowy winter may be on tap for many areas across the nation in 2004-05. We anticipate the onset of Weak El Nino conditions this fall, which combined with frequent high latitude blocking, can enhance the potential for cold and snow. Furthermore, there is also the potential for significant snow events--even in places which managed to evade them last year.

In this article, we will explore the factors shaping the winter of 2004-05, and offer insight as to what they could mean for us this winter. Unfortunately, as a result of extenuating circumstances, and major time constraints, I will only be able to issue one winter outlook for the 2004-05 season.

THE ANALOG METHOD AND FORECAST RATIONALE:

Over the past several years, the popularity of using the analog method approach to seasonal forecasting has proven very effective. Comparing past seasons showing significant hindcast skill with respect to similarities in global atmospheric circulations and patterns with those of the present allow us to better understand how they will shape the future, assuming the atmosphere continues behaves the same in the future as it has in the past.  Our selection of analog years will be based on analysis of various global predictors (climate patterns and oscillations)

The final component of this forecast is the output from numerical Climate Models.

A brief description of each factor will be offered in order to give everyone a general idea of what the feature is, and how it’s current or expected behavior will influence the 2004-05 winter.  

WINTER 2004-05 FORECAST:

Pattern Overview:

Temperatures:

Snowfall:

2004-05 Winter forecast:

Above normal temperatures are forecasted over the western Half of the nation where persistent ridging should be favored through much of the winter consistent with a mostly positive PNA pattern and Below average temperatures are anticipated across the southern plains ands and eastern United States. A mostly negative NAO should ensure the mean trough position will be located over this region.

El Nino-west QBO winters are notorious for above average snowfall across the eastern Portions of the United States especially in years where those conditions coincide with a neutral to negative NAO.  I anticipate an active sub-tropical jet stream which will result in above average precipitation over the southern and eastern portions of the nation. Further to the north, this should also translate to above normal snowfall.

The analog years suggest the core of the above average snowfall will be found over the interior portions of the northeastern half of the nation. A Significant ice storm threat also exists over the Carolinas. Below average snowfall is likely across the intermountain west and western high plains which is normal for El Nino-west QBO winters.

Below are the expected snowfall totals for the Major Cities of the Mid Atlantic and Northeast:

Washington, DC: 15-25”
Baltimore, MD: 25-30”

Philadelphia, PA: 30-35”

New York City, NY: 30-40”

Boston, MA: 55-65”

Above average snowfall is anticipated for the major Cities of the Northeast in 2004-05.

Blockbuster east coast Snowstorms: Research has shown that Major east coast snowstorms occur most frequently with Weak El Nino conditions in place in the equatorial pacific, and a tendency for a negative NAO. Such conditions are likely to be achieved this year, and I do anticipate at least one Major east coast snowstorm with widespread 10” or greater snowfall totals.

The NESIS scale (Northeast Snowstorm Impact Scale) developed by Paul Kocin, and Louis Uccellini attempts to rank east coast snowstorms according to various parameters. A presentation on this was given at the 19th Conference on weather Analysis and Forecasting/15th Conference on Numerical Weather Prediction. A link to the abstract can be found below:

http://ams.confex.com/ams/SLS_WAF_NWP/19WAF_15NWP/abstracts/47837.htm

A majority of the Major east coast snowstorms which occurred in our analog years have had NESIS categories of 2 to 4.

The link below provides a listing of those storms, NESIS rankings and their respective categories:

 

http://www.storm2k.org/phpbb2/viewtopic.php?t=49860

That said I anticipate one to two high-impact major east coast snow events during the coming winter.

Forecasts of Teleconnection Indices (DEC-FEB average):

PNA: +0.1 to +0.5

NAO: -0.3 to -0.5

EPO: +0.1 to -0.2

TECHNICAL INFORMATION SECTION:

1.      2004 Hurricane Season-to-date

2.      El Nino returns in time for winter

3.      East Pacific Signal

4.      Northern hemispheric Snow Cover

5.      Summer 2004

6.      QBO

7.      SOI / MEI

8.      Climate Models

9.      Other Issues

10.  Crucial Assumptions

2004 Hurricane Season-to-date

The 2004 hurricane season has emerged as on of the most active seasons on record since 1950. As of the present time, a total of 14 Named storms, 8 Hurricanes, and 6 intense hurricanes of category three intensity or stronger have formed in the Atlantic basin.

http://www.weather.unisys.com/hurricane/atlantic/2004/

Using the link provided above (courtesy of UNISYS weather) we can see that the majority of the intense and noteworthy tropical cyclones which formed in the Atlantic basin have developed east of the Windward islands and taken long westward tracks before re-curving. These tropical cyclones include Charley, Frances, Ivan and Jeanne.

http://www.weather.unisys.com/hurricane/atlantic/2004/CHARLEY/track.gif

http://www.weather.unisys.com/hurricane/atlantic/2004/FRANCES/track.gif

http://www.weather.unisys.com/hurricane/atlantic/2004/IVAN/track.gif

http://www.weather.unisys.com/hurricane/atlantic/2004/JEANNE/track.gif

Three of These Systems—Charley, Frances, and Jeanne made landfall along the Florida coastline. Hurricane Ivan (which was at maximum intensity a category Five Hurricane) had the most Intense Hurricane days of any hurricane sine 1900; a total of 10 Intense hurricanes Days. Hurricane Frances, which Reached category four strength, had a total of 7.5 intense hurricane days. Frances however weakened near the Bahamas and eventually made landfall as a category two hurricane near Swells point, Florida on September 5 causing Almost Eight Billion dollars in damage to the state. The storm was also responsible for as many as 24 related fatalities.

2004 was an active “Cape Verde” season with eight of the Fourteen Named storms this year developing in this region of the tropical Atlantic between the west coast of Africa and the Windward Islands. This is typical for the Atlantic basin since a large number of the tropical cyclones which form in the August-September period develop in the Atlantic Inter-tropical Convergence Zone (ITCZ) where the Flow around the Sub-tropical ridges present in the northern and southern hemisphere provide a zone for enhanced convergence and corresponding increased low-level relative vorticity. What makes 2004 unique?

1.      The number of tropical cyclones taking long westward tracks toward the US Mainland and only a few systems re-curving out to sea east of 65 degrees west longitude.

2.      Excessive number of intense Hurricanes.

3.      Three total hurricanes (one category two, one category three, and one category four Hurricane) making landfall along the Florida Coastline.

The reason for the long westward tracking tropical cyclones and few re-curvatures this year in comparison to the last eight seasons (since 1995) centers around the positioning of the western Atlantic ridge between 30 and 50 degrees north Atlantic over the Northwest Atlantic and Enhanced Bermuda high which forced tropical cyclones such as Frances, and Ivan to take long westward tracks toward the united states mainland as a result of the enhanced easterly flow equatorward of the ridge.   

An excessive number of six intense hurricanes have formed in 2004 thus far. It is likely that the combined effect of well above normal SSTA over the Atlantic basin, intense low-level convergence and relative vorticity in the ITCZ, and Weak environmental vertical shear. It is also possible that the resultant effects of the west phase of the QBO (Quasi-Biennial Oscillation—a factor known for enhancing Atlantic basin tropical cyclone activity) may have played a key role in the above average activity. It should be noted that it is also very unusual to have such conditions simultaneously in place for an extended period of time during an El Nino season such as this.

So how does this effect the 2004-05 winter?

In choosing hurricane analogs, it is important to choose seasons which have similar upper air patterns, storm tracks, and landfall statistics. This year, the predominant upper air pattern across the western Atlantic and North America has been one where a persistent trough was observed over the western half of North America during the most active portion of the hurricane season, and a ridge over the northwest Atlantic. As far as storm tracks are concerned, they should conform to the observed upper level pattern. The aforementioned synoptic set-up would generally favor long—westward tracking tropical cyclones which affect the United States, potentially lending to an above average number of land falling hurricanes and intense hurricanes, as well as fewer re-curving tropical cyclones.

Usually in El Nino seasons such as this, east pacific tropical cyclone activity is above normal, and Atlantic activity is suppressed. 2004 however, saw fewer East pacific tropical cyclones than the Atlantic, even in spite of the El Nino.

http://www.weather.unisys.com/hurricane/e_pacific/2004/

Other Similar seasons Include: 1963, 1965, and 1969.

http://www.weather.unisys.com/hurricane/e_pacific/1963/

http://www.weather.unisys.com/hurricane/e_pacific/1965/

http://www.weather.unisys.com/hurricane/e_pacific/1969/

This leads me to believe that the State of the large-scale circulations I the Atlantic and Pacific basins are similar to what they were during the decade of the 1960s, and the effects of the El Nino conditions in the Equatorial Pacific have been overshadowed. In all three of the aforementioned cases, the following winters featured frequent and persistent high latitude blocking over the Pole and North Atlantic in addition to toughing over the eastern half of the United States.

The abovementioned statistics pertaining to the current hurricane season would indicate a propensity for a frequently negative NAO in the December to February period, and persistent toughing over the eastern half of the United States during the same timeframe—which when combined with the Weak El Nino would suggest below normal temperatures and above average Snowfall along the eastern Seaboard.

El Nino Returns In Time for winter

Perhaps the most important component in any seasonal forecast is the predicted state of the ENSO. El Nino and La Nina events can be the primary factors which shape the dominating upper level pattern in all seasons. In winter, however, these climate features are especially important.

El Nino events are characterized by an anomalous warming of Sea Surface temperatures over the Equatorial pacific lasting on average 7 to 12 months. It should be noted though that some events will last longer or shorter than others depending on other large scale features such as the PDO cycle. Normally El Nino events in the PDO long-term warm phase will persist longer, and tend to be far stronger than their counterparts which occur in the PDO long-term cold phase. Some characteristics of El Nino events include:

·         A frequently positive PNA Pattern

·         The depth of the equatorial thermocline is closer to the ocean surface

·         Weak or Absent walker Circulation.

·         Negative SOI and Positive MEI values

·         Negative Eastern Pacific Oscillation

·         Warmer than average SSTA over the Equatorial pacific

·         Enhanced Convection over the tropical Pacific.

La Nina is characterized by the opposite, where an anomalous cooling of Sea Surface temperatures (SSTs) occurs over the equatorial Pacific. In order to make the analysis and forecasting of Warm and Cold ENSO episodes more efficient, the equatorial pacific is split up into specific regions. These regions are the ENSO 1+2 region closest to the South American coast (10S – EQ / 90W), NINO region 3.0 (5S – 5N / 90-150W), NINO region 4.0 (150W - ~160E), and NINO region 3.4 which combines the western portions of the 4.0 region and eastern portions of the 3.0 region. Click the link before for a graphical representation of these regions:

http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ninoareas_c.jpg

It is the NINO 3.4 region which is most important region to look at when anticipating El Nino formation. For La Nina episodes, the 1+2 region is most important. 

During the course of September 2004, ENSO region 3.4 and 4.0 SST anomalies continued to warm with region 4.0 anomalies reaching +1.12 degrees Celsius. This is the highest Monthly NINO 4.0 anomaly value observed since December 2002, and most certainly indicative of an intensifying Equatorial Pacific warm episode.

NINO 1+2

-0.43

In the table above, we Notice a distinct split in the SSTA over the equatorial pacific, with an anomalously cool east, and warm west. 

http://www.cdc.noaa.gov/map/images/sst/sst.anom.month.gif

Other years with similar statistics include 1977, 1986, 1990, 1994, 2002, and 2003.

0.00

-0.31

-0.28

-0.60   

-0.48

A majority of the winters above had the core of the warm SST anomaly associated with the El Nino centered across the central and western pacific in the NINO 3.4 and 4.0 regions. The figure below shows (December-February) 500mb Height in winters with a warm west and cold east.

In the majority of winters where a warmer than normal west and colder than normal eastern Equatorial pacific is observed, strong high latitude blocking over the northwest Atlantic (a negative NAO) and below normal heights are observed over the eastern US.

The next thing we like to look for is how SSTs have trended over the past three eleven months. There have been a total of 12 seasons which saw similar conditions to the past eleven months with respect to the ENSO signal. For our purposes here, we will be using the three-month running mean ONI (oceanic NINO index) values since the November-December-January period of 2003-04 up to the present. 

The ONI uses three-month running mean of NOAA ERSST.v2 SSTA in the NINO 3.4 region based on 1971-2000 normals.

El Nino conditions per ONI values are achieved when the threshold of SSTA greater than +0.5 Degrees Celsius is met for at least five consecutive three-month periods.

The links below provide more information on the ONI

http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ensoyears.html

http://www.cdc.noaa.gov/Correlation/oni.data

Other similar eleven moth periods have occurred in the winter, spring and summer of

1951-52, 1952-53, 1962-63, 1969-70, 1976-77, 1978-79, 1979-80, 1981-82, 1989-90, 1990-91, 1993-94, 2001-02

1951-52 is excluded since no El Nino developed following the spring and summer.  the same can be said for 1952-53, which remained neutral then trended toward La Nina. 1962-63 can be considered a match. 1969-70 cannot be used since the following winter was a la Nina winter. 1979-80 is out since 1980-81 saw Weak cool conditions. 1981-82 would be a good match however the El Nino that developed was one of the two strongest events of the past 50 years. THIS EL NINO WILL NOT BE AS STRONG AS THE 1982-83 El NINO.

This leaves us winter winters of:

1962-63 ------> 1963-64

1976-77 ------> 1977-78

1978-79 ------> 1979-80

1989-90 ------> 1990-91

1990-91 ------> 1991-92

1993-94 ------> 1994-95

2001-02 ------> 2002-03 

Many of the seasons above saw El Nino conditions peak during the early winter or very late fall and decline rapidly thereafter. This could imply that the 2004-05 El Nino may decline rapidly as well after reaching its peak.

Above are 500mb heights in the aforesaid winters. Note blocking over western Canada and a strong Aleutian low.

The forecast will reflect a weak to borderline moderate warm episode which peaks during the early winter and steadily declines during the later half of the season. 

East Pacific Signal

El Nino events normally point toward a negative Eastern Pacific Oscillation in winter. The EPO reflects a tripole of height anomalies, with those of similar sign located over the east central Pacific and Eastern Canada (centered near Hudson Bay), and an area of dissimilar height anomalies centered over Northwestern Canada, Alaska, and the extreme northeast pacific.

http://www.cdc.noaa.gov/people/gary.bates/tele/epo.cmp.gif

The positive phase of the EPO reflects a dipole of higher than average heights over the East central pacific and Eastern Canada, with an area of below normal heights covering the Northeast pacific, Alaska and Northwestern Canada.

Equally important to determining the favored phase of the eastern pacific oscillation in winter is the evolution of SST anomalies over the north pacific, and the short-term variations in the PDO (pacific Decadal oscillation) index. In winters with a positive EPO, cool SSTA are located in the Gulf of Alaska with above average SSTA northwest of Hawaii. This leads to the development of a negative height anomaly near the west coast of North America and an intensification of the Pacific Jet. We will use October-December 2001 as an example:

SSTA

http://www.osdpd.noaa.gov/PSB/EPS/SST/data/anomnight.10.29.2001.gif

http://www.osdpd.noaa.gov/PSB/EPS/SST/data/anomnight.11.13.2001.gif

http://www.osdpd.noaa.gov/PSB/EPS/SST/data/anomnight.12.11.2001.gif

A negative EPO represents the opposite. Above average heights are found over Northwest North America with below average heights over Eastern Canada, and the east central Pacific. The flowing is associated with a negative EPO:

·         Suppressed pacific Jet

·         Split flow and rex-blocking set-up across the eastern pacific.

·         Strong Polar vortex near Hudson Bay

·         Below average heights and temperatures over eastern North America

The SSTA configuration associated with a positive EPO is reversed, where warmer than average SSTA are found in the Gulf of Alaska, and cold SSTA are found Northwest of Hawaii. This is most frequently achieved in El Nino winters during the PDO warm phase. An example of this would be 1986, 1993, and 2002.

http://www.longpaddock.qld.gov.au/SeasonalClimateOutlook/SeaSurfaceTemperature/1986/Oct.gif

http://www.longpaddock.qld.gov.au/SeasonalClimateOutlook/SeaSurfaceTemperature/1993/Oct.gif

http://www.longpaddock.qld.gov.au/SeasonalClimateOutlook/SeaSurfaceTemperature/2002/Oct.gif

Other factors such as above average geomagnetic activity during and immediately following the solar maximum may result in a positive EPO even in spite of otherwise favorable SSTA and ENSO conditions due to the propensity for a weaker than average Aleutian low during the cold season.

As for this year, the ongoing El Nino conditions would imply a negative EPO, but the current SSTA configuration and trend since the summer is not as clearly defined as it was in the aforementioned years to sell me on the idea of a mostly negative EPO this winter. The years with similar September-October SSTA configurations across the North pacific were 1961, 1962, 1963, 1964, 1968, 1977, and 1990.

-2.69

0.6

-1.52

-0.6

-1.7

-0.69

-0.8

Given the information gathered, I expect the EPO to display quite a bit of variability, with strong positive and negative periods throughout the winter, however once all the numbers are in, I expect the EPO to in the end average neutral to slightly negative—in touch  with El Nino conditions, and the SSTA analog years. 

Northern hemispheric Snowcover

The spatial extent of northern hemispheric snowcover during the autumn is critical in the production of arctic air masses, and can also give us clues as to the state of the NAO and AO during the following winter.

Above average snowcover during this period would be a positive when looking for a predisposition toward colder than average conditions over a given portion of North America, whereas below average snowcover can signal that the development and sustainability of arctic air masses will be in question.

Furthermore, above average snowcover may signal a tendency for a negative Arctic oscillation (AO) during the winter, since the snowcover favors the intensification of the Siberian high, and may also correspondingly signal a negative NAO during the following winter.

There have been exceptions though—i.e. years where snowcover averaged well above normal during the fall and early winter, only to see warmer than normal conditions over the lower 48 states. Similarly, there have been several seasons with below average snowcover during this critical time period, however the following winters ended up seeing much below normal temperatures over portions of the United States. This is risky correlation, and more research needs to be conducted before additional forecast weight can be given to this factor as a viable predictor.

In any event, August 2004 snowcover averaged below normal—about 2.1 million square kilometers.  The average is 3.3 Million Square Kilometers. The text value for September is not out yet, however judging by the graphics, In my opinion, the September value is most likely four or five million square Kilometers—also slightly below the average.

ftp://ftp.ncep.noaa.gov/pub/cpc/wd52dg/snow/snw_cvr_area/NH_AREA

Based on this information, if October and November 2004 northern hemisphere snowcover extent is also below average, we might expect the following:

·         The buildup of arctic air masses will be slower and less cold air will be available.

·         Arctic Intrusions will be transient and less sustainable. 

·         Positive AO

·         Positive NAO

In the event things turn around during the next two months, then it is possible this factor will also favor cold and snow as several others we have already looked at do. I urge everyone to monitor daily changes in snowcover extent and be aware of trends in the data over the next two months.  

Summer 2004

Recent research has shown that the summer and fall patters leading up to a given winter can provide as a window into what that winter may be like.

The summer of 2004 was cooler than average in the central and eastern US and wetter than normal over the southern and eastern half of the nation which is pretty typical of an El Nino summer.

The wetter than average conditions along the east coast are indicative of a propensity for increased extratropical cyclogenesis along the east coast during the following winter if the pattern holds.

Through the month of September we saw several landfalling hurricanes which brought significant precipitation to the eastern half of the nation.

http://www.cpc.ncep.noaa.gov/CWI_images/mtd-precip.gif

This is a powerful signal for above average precipitation during the following winter along the east coast.

The past week has seen the development of the first major extratropical cyclone of the coming winter, with a major phasing event over the eastern part of the nation. Below is the 48hr ECMWF forecast from 12z October 13. Note the Split flow over the eastern pacific which is common during El Nino-west QBO winters, and the deep-high amplitude trough over the eastern US.

This may also serve as an indicator of things to come. Years where major deepening extra-tropical cyclones are seen during the fall, typically are analogous to similar events later on in the season.

The summer and early fall patterns clearly favor above normal precipitation and colder than normal temperatures over the eastern part of the nation during the 2004-05 winter, in addition to an above average potential for major extra-tropical cyclogenesis along the eastern seaboard.

QBO (Quasi-Biennial Oscillation)

The QBO is a very well known periodic oscillation in the upper levels of the atmosphere over the equatorial regions. The QBO represents a band of zonally symmetric easterly and westerly winds which alternate regularly in periods of one to two years (20-30 months on average) with an amplitude of generally 40 to 50 m/s maximized near 20hPa (Note: hecto Pascals (hPa) and Millibars (mb) are equivalent units). Easterly QBO phases tend to be stronger and persist longer than westerly phases.    

Plumb 1984 developed a theory explaining the QBO in which Kelvin waves provide westerly momentum and increase the westerly shear zone resulting in the weakening of the Kelvin wave and allowing the mixed rossby gravity wave to propagate vertically upward more efficiently (associated with the onset of the westerly QBO phase). The mixed rossby wave then provides the easterly momentum upward increasing the easterly shear zone—which when strong—the mixed rossby gravity wave weakens and allows the Kelvin wave to propagate upward (inducing the onset of the easterly phase). The cycle starts over again.

This year, the QBO is currently in the westerly phase. The peak of this particular cycle occurred in May 2004, three months after reversing from east to west.

It is fairly likely that the QBO will remain positive though the remainder of 2004 and into January 2005 at a weak value. Below are years in which met this criteria for the October-December period: 1955   1963 1964 1969 1973 1977 1978 1994 1997 2001

Data for those years can be found by clicking the link below:

http://www.cdc.noaa.gov/correlation/qbo.data

Note that the data from the early and mid 1950s win which the

 QBO appeared to be engaged in a long-term weak easterly cycle 

should be viewed as unreliable due to the lack of dependable data. 

Research has also shown that years where the QBO is weak westerly or easterly through the majority of a given winter, that winter tends to be a significant one across the central and eastern US. Such years include:

1957-58, 1961-62, 1963-64, 1977-78, 1988-89, 1993-94, 1995-96, 1999-00, 2001-02, 2002-03.

There are a few exceptions such as 1988-89, 1999-00 (although that winter did have the “Carolina Crusher” nor’easter), and 2001-02. The others, especially 1977-78 and 1993-94 and 1995-96 tended to be among the most severe in the list.

The aforementioned years would serve as viable QBO analogs, however because I anticipate the QBO to transition to the easterly phase in January or February, and as the forecast is geared to reflect, we must look for seasons in which the QBO switches phases at some point during the DJF period.

Now, when we combine the effects of El Nino and the westerly phase of the QBO, the results are astounding. The years which qualify are 1957-58, 1963-64, 1977-78, and 1994-95. Below are composite 500mb heights in those years:

Strong high latitude blocking is seen over the North Atlantic which extended all the way back to the west into Canada with well below normal heights over the eastern half of the United States.

If the QBO once again increases westerly (i.e. a secondary peak) a majority of the aforementioned winters would once again come back onto the table with respect to being analogs for this particular feature. It may also signal a much increased chance for a severe winter in the I-95 corridor of the Mid Atlantic and Northeast.

El Nino-West QBO conditions also tend to favor rapidly-deepening extratropical cyclones, loaded with moisture—which given sustainable cold air can lead to significant snow events along the eastern seaboard.

Southern Oscillation Index (SOI) and MEI

The SOI represents a measure of large scale fluctuations in air pressure over the equatorial pacific based from the differences in air pressure between Tahiti and Darwin, Australia. The SOI operates in two primary modes which can be used by forecasters to determine the state of the ENSO and corresponding to observed SST anomalies over the region (for example is there an El Nino or La Nina event present).

The first primary mode of activity is the SOI negative phase. During periods where the SOI is negative, Sea Level Pressure (SLP) is below normal at Tahiti and above normal At Darwin, Australia. Long term negative periods correspond to above average SSTs over the Equatorial pacific and El Nino conditions. The positive phase represents the opposite of the aforesaid, and coincides with below average SSTs and La Nina conditions. Below is a Link to more background information on the SOI for those who wish to dig deeper:

http://www.bom.gov.au/climate/glossary/soi.shtml

Our goal here is to take a look at the SOI values over the past eleven months—from November 2003 through September 2004. This will allow us to isolate trends in the data and hopefully serve as a means to determine the intensity of the upcoming El Nino. The 2003-04 values are illustrated in the table below:

Going through the table, (http://www.bom.gov.au/climate/current/soihtm1.shtml) we can determine that the following years were at least a close match: 1951-52, 1958-59 1962-63, 1964-65, 1968-69, 1971-72, 1975-76, 1976-77, 1978-79, 1981-82, 1986-87, 1990-91, 1992-93, 1993-94, 1996-97. Remember that the potential analog winter is the winter following the years above.

Some years will have to be eliminated since the following season does not match the current SSTA profiles. Those years include 1951-52, 1958-59, 1971-72, 1981-82, 1992-93, and 1996-97. This leaves us with:

 

1962-63 leading to 1963-64

1964-65 leading to 1965-66

1968-69 leading to 1969-70

1975-76 leading to 1976-77

1976-77 leading to 1977-78

1978-79 leading to 1979-80

1986-87 leading to 1987-88

1990-91 leading to 1991-92

1993-94 leading to 1994-95

The MEI: The multi-relative ENSO index was developed By Dr. Klaus Wolter of NOAA’s Climate Diagnostics Center for the purpose of better monitoring the progress of the ENSO. The MEI combines six observed variables pertinent to warm and cold Episodes in the Equatorial Pacific, these include:

·         Sea Level pressure (SLP)

·         Zonal and Meridional (U, V) components of the surface wind

·         Sea Surface temperature (SST)

·         Surface temperature

·         Total cloudiness fraction of the sky.   

See the link below for more information on the MEI:

http://www.cdc.noaa.gov/people/klaus.wolter/MEI/mei.html#ref_wt1’

As with the SOI, we are looking for trends in the MEI and comparing the past 11 months with other similar 11 month periods in the MEI in order to gather potential analogs.

Other similar years were 1951-52, 1976-77, 1978-79, 1985-86, 1989-90, 1990-91, 1993-94, and 2001-02. To view the data from these years, consult the link below:

http://www.cdc.noaa.gov/correlation/mei.data

1952-53 is out since no El Nino developed during the following winter, leaving us with:

1976-77 ----> 1977-78

1978-79 ----> 1979-80

1985-86 ----> 1986-87

1989-90 ----> 1990-91

1990-91 ----> 1991-92

1993-94 ----> 1994-95

2001-02 ----> 2002-03

Climate Models

Since August 2004, the old AGCM climate model from NCEP has been replaced by the New NCEP CFS (Coupled Forecast System) model. The new page for the CFS can be found at

http://www.cpc.ncep.noaa.gov/products/people/wwang/cfs_fcst/

It is important for everyone to understand that while the incorporation of such model output can be a valuable tool in seasonal forecasting—creating a seasonal forecast based completely on long-range climate model data is a VERY RISKY PROPOSITION and ill-advised.

The old AGCM had quite a bit of success prior to the winters of 2000-01 and 2002-03, however the model predicted widespread below average temperatures across eastern North America, with the core of which centered near the great lakes. That of course did not pan out due to the unexpected surge in solar activity and other less prevalent factors. 

The latest run of the new CFS from October 2004 strongly supports both our analog years, and similar winters with EL Nino-west QBO conditions. Below are CFS 200mb heights in overlapping three monthly increments for the period NDJ 2004-05 through AMJ 2005.

http://www.cpc.ncep.noaa.gov/products/people/wwang/cfs_fcst/images/glbz200Sea.gif

There is a signal for a frequently positive PNA pattern as evidenced by the positive height anomaly over western North America, and stronger than average Aleutian low. The aforementioned teleconnection supports the formation of another downstream negative anomaly over the Eastern United States.

As far as SSTs are concerned, the model is forecasting the formation of a weak to moderate El Nino centered over the central Equatorial pacific, which supports our analogs.

http://www.cpc.ncep.noaa.gov/products/people/wwang/cfs_fcst/images/nino34SSTSea.gif

http://www.cpc.ncep.noaa.gov/products/people/wwang/cfs_fcst/images/glbSSTSea.gif

The only discrepancy lies in exactly when the event peaks. The forecast will reflect an El Nino event which peaks in the December-January period. If the CFS is to be correct the event would peak somewhat later, perhaps in the January-March Period, then decline. Overall however, the upper level pattern and SSTA forecasts from the CFS seem reasonable and are in support of our analog seasons.

CFS temperature and precipitation:

http://www.cpc.ncep.noaa.gov/products/people/wwang/cfs_fcst/images/usT2mMon.gif

http://www.cpc.ncep.noaa.gov/products/people/wwang/cfs_fcst/images/usPrecMon.gif

The ECPC’s GSM (Global Spectral Model) has outperformed the AGCM in previous years. Below is a Link to the ECPC website and GSM model

http://ecpc.ucsd.edu/

http://ecpc.ucsd.edu/projects/GSM_seasons.html

The October run of the GSM supports the CFS ideas and the analog years with the development of a strongly Positive PNA pattern and Negative NAO as evidenced by the blocking (positive Height anomaly) along the west coast of North America and over the far North Atlantic. The resultant effect produces a strong negative height anomaly over eastern North America. 

http://ecpc.ucsd.edu/projects/GSM_home_data/200410/GZ5.200410.ano_PNA.gif

http://ecpc.ucsd.edu/projects/GSM_home_data/200410/GZ5.200410.ano_global.gif

This sort of set-up with a Positive PNA pattern and negative NAO suggests a persistent eastern us trough and western US ridge, which at times may reach extreme amplitude during the cold season, which is common for El Nino west QBO winters with a Negative NAO in the means.

Temperature and precipitation:

 

http://ecpc.ucsd.edu/projects/GSM_home_data/200410/TMP2.200410.ano_PNA.gif

http://ecpc.ucsd.edu/projects/GSM_home_data/200410/PRATE.200410.ano_PNA.gif

Another interesting factor to note is that the model is forecasting above average Latent Heat Flux along the east coast during the heat of the winter. Latent heat flux refers to the global movement of latent heat energy through oceanic and atmospheric circulations. An example of this would be tropical cyclones, which transport latent heat from the tropics poleward. It can also play a critical role in explosive extratropical cyclogenesis (bombogenesis) along the eastern seaboard during the winter season.

http://ecpc.ucsd.edu/projects/GSM_home_data/200410/LTNT.200410.ano_PNA.gif

This would support an increased probably for such extreme cyclogenesis during the coming winter.

For more information on the GSM click here:

 

http://ecpc.ucsd.edu/projects/gsm_model/index.html

Other Issues

There are a few other minor factors which were not covered in the sections about, but are still noteworthy.

·         Solar Activity: 10.7cm radio flux values during the months of May and June dropped below 1000, however since then, an upsurge in solar flux was noted in July. Since then radio flux values continue to steadily decline as we trend toward the solar minimum. Other years with similar ups and downs in 10.7cm flux values include 1994, 1993, 1984, 1983, 1973, 1962, 1961, 1950. These swings appear to be most common two to three years following the solar maximum. As many are aware, high 10.7cm solar flux can lead to stronger than average ridging over the mid latitudes, whereas the corresponding geomagnetic activity can lead to a weakening of the Aleutian low, and intensification of the Icelandic low during the cold season resulting in a Positive AO/NAO and negative PNA pattern. Values are low enough this year that solar flux should not be a major issue.

·         ATC cycle: Since 1995, the Atlantic thermohaline circulation has reverses phases back to it’s strong cycle which favors above average hurricane activity in the Atlantic basin, more La Nina events, warm SSTA globally and a frequently Negative NAO in winter. Although short term shifts in the ATC cycle from the long-term phase do occur—based on 2004 hurricane activity, and Global SSTA, I have no reason to believe a short term shift has occurred. Thus, it’s likely that this signal will favor a Negative NAO during the 2004-05 winter.

 Crucial Assumptions

·         The 2004-05 El Nino will peak at a weak to borderline Moderate Intensity during the early winter, with a steady decline thereafter.

·         The EPO will average Neutral to negative

·         SOI averages solidly negative consistent with El Nino 

·         The PNA pattern will average at least weak Positive.

·         Solar Activity will not be an issue.

·         The QBO will reverse in the January-February period to easterly, however remain at near neutral values (no greater than +/-5.00).

·         The AO and NAO will average neutral to negative given the long term ATC strong cycle, and Atlantic hurricane activity. Perhaps solidly negative if Northern hemispheric Snowcover increases rapidly over the next month and a half.

·         Major December Nor’easter

·         Major Midwest and/or Severe eastern US winter storm/blizzard in January.

·         Severe march Nor’easter

·         1-2 High-impact Major east coast snow events.