1ST  CALL   WINTER  PREVIEW   -- 22  OCT  2003  




1.  REVIEW  OF WINTER 2002-03 2.  USE of  ANALOGS
3.  Is there an El Nino Coming? 4.   Current  ENSO   conditions
7.   MEI 8.   QBO

review of 2002-2003  Winter   Forecast:   MOSTLY   SUCCESSFUL...

The  Wxrisk.com  2002-03 Winter forecast   was  a   major success and was easily  superior  to the    the  general or Consensus Forecasts (Hereafter  called CF). Although my January forecast for  2003   flat out  sucked...    the forecast for   December  2002  February 2003 and March 2003   was  nearly Perfect. Within the confines  of  the private sector  forecaster  the standard  for the  CF  which is set  by  CPC  or the climate  prediction Center.  After   potential  clients are looking for an edge or a better forecast and  if they get get a better  forecast for  free from  CPC  why pay for it?      The term "consensus forecast" (CF) is term that I use   to describe  the general view or a summary of what most PWSIP  (private weather  service  Information Services) /    NWS forecasters are saying.  The CF is very important in determining how the seasonal trade in Energy and Agriculture markets will unfold as well the operational side of the weather biz. For example… State DOTs will often significantly lower their Salt and Sand purchases if the CF is for a Mild winter in their areas. The CF often consists of  a) the CPC (Climate Prediction Center) …whose forecast carry a lot of weight...b) large well known Private Weather Service Information Provider (PWSIPs) and
c) some well known Energy/ Ag forecasters.

Last winter there was much  talk  of  the developing  El Nino in   the  equatorial  regions of the  Pacific  ocean.  Given the recent  History of  El Nino events  in the  70s  80s and   MUCH of the  90s...  it seemed  self  evident that ANY talk of an El Nino  automatically brings  about  certain  pre-conceived   notions  about  what the  Summer   and /or Winter   season will be like.  This can clearly see  by   CPC's terrible winter  2002 - 03 forecast.  This  map is a  standard CPC   "if it's an El Nino during the winter   this is what the winter  MUST be like...:"  But   as I   (and  some others)  pointed out in  OCT 2002   there were MANY  indicators   that  showed  the  developing autumn El Nino  was  NOT   going to bring about the  same  very  mild  winter  MOST   of the  USA  has seen in the El Nino  events 70s  80s and 90s   El Nino events.

winteroutlook121202b.jpg (240783 bytes)

Among those  clues    that   the  El Nino last  winter  was NOT going to bring about   "typical" El Nino  winter  conditions  in  the   CONUS were


REVIEW   of WINTER  2002-2003

For locations East of the Mississippi River   Winter  2002--03 ranged from  seasonally  cold to   VERY cold while  locations  WEST of the Mississippi river saw a  Warm  DRY winter  especially  over the  Upper Plains ands  western Great Lakes.   With respect to Snowfall... much of the  East Coast saw a winter that placed in the   top 10 to top 15  snowiest winter  ever.    January   2002  was   the coldest  month of the winter as the Jet stream surged  well to the south... which  forced the  southern track tracks off the Southeast coast.


OVERALL  DJF avgtemp.dfn_ltm_cont_200302.jpg (54710 bytes)
precip.dfn_200212.jpg (59958 bytes)  avgtemp.dfn.cont_200301.jpg (54903 bytes)  avgtemp.dfn.cont_200302.jpg (54393 bytes)
OVERALL  DJF precip.dfn.cont_200302.jpg (52297 bytes)

DEC  2002 precip.dfn.cont_200301.jpg (48858 bytes)   precip.dfn_ltm_cont_200302.jpg (53165 bytes)
12_02_2003_DvTempRank_pg.gif (80441 bytes) 12_02_2003_DvPrcpRank_pg.gif (78307 bytes)
01_01_2003_DvTempRank_pg.gif (77487 bytes) 02_02_2003_DvTempRank_pg.gif (77086 bytes) 03_03_2003_DvTempRank_pg.gif (77257 bytes) 01_01_2003_DvPrcpRank_pg.gif (77237 bytes)02_02_2003_DvPrcpRank_pg.gif (78337 bytes)03_03_2003_DvPrcpRank_pg.gif (78256 bytes)
12-02Statewidetrank_pg.gif (69792 bytes) 12-02Statewideprank_pg.gif (65892 bytes)
12Statewidetrank_pg.gif (68762 bytes)01Statewidetrank_pg.gif (67959 bytes)02Statewidetrank_pg.gif (66067 bytes) 12Statewideprank_pg.gif (68415 bytes)01Statewideprank_pg.gif (64470 bytes)02Statewideprank_pg.gif (66275 bytes)

I have seen  SEVERAL references   made within  various weather  web sites  discussions groups and some early winter forecasts that   OCT 2003  is  shaping up to  be   similar to OCTOBER 2002.  Such statements are usually  focused  on the  Sea Surface Temperature Anomalies  patterns in various locations around the   world .... OR    the  narrow  view  that  is premised on the idea  that   OCT 2003  at  " my house" is   running similar to  OCT  2002.     


OCTOBER   2002
10Statewidetrank_pg.gif (68631 bytes) 10Statewideprank_pg.gif (68518 bytes)


Such  assertions are of course  specious.   As the  2 maps  show  OCT 2002 saw a large areas of  temps   below   and Much below Normal   over the  Northern 1/3 of the Nation but especially over the  Plains    and  WET conditions over    the Lower Plains  into the  Southeast states and up the East coast.   OCTOBER   2003 has featured  days and days of   Record or near record   heat in the   central and Upper Plains   while the Lower Plains  have been very  dry.  ( again NOTHING   like   OCT of  2002)) . 

 Once again in this winter forecast that there will be a late-season update for any changes which I think should be made... and of course there will be a midwinter review which occurs sometime after the New Year. The science of seasonal weather forecasting  is half science and half arr. The mistake that forecasters make in my opinion regarding seasonal forecast is not so much attempting to make the accurate seasonal forecast and then having it bust.   The problem is in not following up on their forecast... seeing how the pattern is or is not developing and failing to issues any changes to  the forecast.

A classic case of this was  my Summer forecast for 2003.   My original Summer forecast ( issued Mid May ) was for a hot dry Summer across the Plains and the Midwest and  cool wet one for the East coast.  This idea was   based on a moderate La Nina developing  in the El Nino regions... and since there is a strong correlation between moderate La Nina and hot dry summers in the Plains and Midwest.... that forecast was fairly reasonable.   Moreover going into the month of MAY 2003 the data clearly showed a moderate La Nina event was building as the sea surface temperatures cooled significantly off of South America.

However the cold sea surface temperatures reversed itself and rapidly warmed during the month of June -- which of course is a month after I issued my summer forecast. The collapse of the developing La Nina effect is due the the - PDO which is a long-term weather phenomenon in the Pacific ocean that  enhances or inhibited the development of El Nino and La Nina events.  Thus on the basis of the   collapse  in the La Nina  in JUNE... I  updated   the Summer forecast right before the July 4th weekend--- which can be seen on the Web site.    In my opinion it is really quite pointless  to issue   the seasonal forecast  MONTHS before the season  under consideration actually begins    and  then issue NO updates.



Many  (but not all)  forecasters use "Analogs"  to help make a seasonal or monthly forecast. In weather forecasting the use of analog is an attempt to understand   the  seasonal forecast period  based (in past) upon similarities between a particular set of parameters.   The idea is by searching   for similarities  with  other  years or seasons    it gives a forecaster some clues as  to how the   monthly or seasonal forecast  may  develop. .

For example one may consider the fact that most  US winters   during   strong  El Nino years bring warm and dry conditions to the Upper Plains   and Midwest   and  rather wet and cool conditions to the Deep South and changeable conditions in the Northeast.   Or a forecaster may look at say the past 6 months... 12 months... or 18 months worth of temperature  and  precipitation data over a certain section of the country and use those similarities to the the current situation to assist them in making the monthly or seasonal forecast.

Done the correct way the Analog method has validity to it.   For example to simply assumed that all El Nino  events  produce   ONLY  one set of conditions in the winter months  is of course folly.    Some of the most severe winters in the last 100 years have been El Nino winters and likewise.... some of the most Mild winters in the last 100 years have also been El Nino winters.  Some forecasters  and   weather hobbyists   hold  the view   that analogs  are   not very helpful since  weather  records are incomplete or not sufficient enough to be considered as an indicator of what the upcoming seasons might hold.      For most locations temperature and precipitation data only extends back 100 or 130 years  and accurate Upper Air maps   only go back 70 years at best . That argument is   Valid  and ANY  seasonal forecast  using analogs  that is based on ONE   parameter is   very risky  and has a high degrees of failure  built within the forecast.

One way of getting around this is to use several factors parameters and cross matching the analogs to look for common ground.

Just   like   the  last several  seasonal    forecasts   I will use these   parameters and in this order of  importance

  1. ENSO

  2. SOI

  3. MEI  

  4. QBO  








The first parameter that I will consider will be the Sea Surface Temperature Anomalies (SSTA) in the ENSO regions. (Warm ENSO are called El Nino... cold ENSO are called La Nina). For the purpose of clarification I have presented a map of the four different ENSO regions. Officially  an El Nino or La Nina event is declared when the SSTA have reached a certain level for a 3 month time period.  Usually for El Nino events ENSO region 3.4 are  the critical regions.... while for La Nina events it is ENSO regions 1+2 .



Although there is little doubt that the El Nino events are somewhat is over hyped... the fact of the matter is that   El Nino and La Nina  events   are major players  and MUST be considered in  any attempt to make a seasonal forecast.  As I stated earlier the official CPC  (climate prediction center)  forecast  from   last winter was heavily based upon the idea that  a moderate to strong El Nino would develop and flood  the central  & eastern US  with mild air.   Instead the El Nino of  2002-03 collapsed in early January 2003.    Likewise in the Spring of 2003 the development of a La Nina led many forecasters (myself included)  into making a "Hotter than Normal"  and   "Drier than Normal"   Summer forecast for the central Plains and portions of the Midwest.    Instead the La Nina collapsed in June and the wet cool pattern held.

It should therefore be obvious and self-evident that a seasonal forecast which is premised on the idea that there will NOT be an El Nino this winter is likely to fail (and badly) in the event  that  an El Nino event were to develop.    Likewise  a seasonal forecast  which is based upon the idea of an El Nino   developing only to see  NO  development ( or a delayed  El Nino  later in the Spring of 2004)   is  also a Winter forecast that  just waiting  to Bust.     Indeed I have seen several forecasts from private weather services to the CPC to weather hobbyists   insist back in early September that there is NO   chance of a weak El Nino event developing.    In fact the data over the last 30 days  the data has turned decisively in my opinion towards the idea of a weak El Nino event developing.



ECMWF   ENSO  3.4     Forecast   CMB '   FORECAST SCRIPPS 

SEPT   9

14oct3.4.gif (9429 bytes) figf4b.gif (7836 bytes) 3464a.gif (17981 bytes) 3464.gif (18303 bytes) figf5.gif (10852 bytes) figf11.gif (76149 bytes)

As you can see from some of these recent forecasts from various sources...  there has been a significant shift over the last 30 days regarding the development of a weak El Nino event during a second half of winter. Thus I  come to our first critical assessment of this forecast. Based upon what I am seeing right now... my forecast is going to be heavily based upon the idea of a weak El Nino developing in the second half the Winter into the Spring and therefore the analogs which I am be considering have to be based on that.

Officially Sea Surface Temperature Anomalies 9SSTA)  in all four regions are very close to neutral. However over the last 30 to 45 days there has been a slow increase in the warm water  in the ENSO region 3. 4.   The October reading of  +0.8C is the warmest water in that region since last winter. Here are the official sea surface temperature anomalies for regions 1.2 and 3.4  based upon  this  LINK   and    this LINK


There are several sources that one can use to figure out which historical period might be considered as they analog. Below I have reproduced the well known Table that shows El Nino / La Nina years from the Climate Prediction Center that gives a general overview of the last 50 years by 3 month intervals. ( w= warm SSTA     C= cold SSTA      N = Neutral SSTA)    http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ensoyears.html    As you can from the CPC table the ONLY periods that match   the CURRENT  ( December 2002 - October 2003)   are the  periods of 1958    1978 and to a lessor degree 1979. I have highlighted these periods in Blue and RED.

1950 C C C C
1951 C N N W-
1952 N N N N
1953 N W- W- N
1954 N N C- C
1955 C C- C- C+
1956 C C C C-
1957 N W- W W
1958 W+ W W- W-
1959 W- N N N


1961 N N N N
1962 N N N N
1963 N N W- W
1964 N N C- C
1965 C- N W W+
1966 W W- W- N
1967 N N N N
1968 N N N W-
1969 W W- W- W-
1970 W- N N C
1971 C C- C- C-
1972 N W- W W+
1973 W N C- C+
1974 C+ C C- C-
1975 C- C- C C+
1976 C N N W-
1977 N N N W-
1978 W- N N N
1979 N N N N
1980 W- N N N
1981 N N N N
1982 N W- W W+
1983 W+ W N C-
1984 C- C- N C-
1985 C- C- N N
1986 N N W- W
1987 W W W+ W
1988 W- N C- C+
1989 C+ C- N N
1990 N N W- W-
1991 W- W- W W
1992 W+ W+ W- W-
1993 W- W W W-
1994 N N W W
1995 W N N C-
1996 C- N N N
1997 N W W+ W+
1998 W+ W C- C
1999 C+ C C- C
2000 C C- N C-
2001 C- N N N
2002 N W- W W
2003 W- N N


However  a  much  comprehensive  MONTHLY  data base   of  all the  ENSO  regions --  both   SSTs   and  SSTAs -- can be accessed  at this  site http://www.cpc.noaa.gov/data/indices/sstoi.indices.  



The  numbers  are SSTs Anomalies  in degrees  Celsius  from JAN 203  to  OCT 2003  ;  
= Negative Values     BLUE = .  POSITIVE values  BLACK   near Neutral
ENSO 1+2 (off PERU coast)
JAN 2003  through OCT  2003
JAN  2003  through OCT  2003
-0.13   -0.22 -0.50 -1.04  -1.78  -1.80
-1.00 -1.00  -0.90 -0.50
1.24   0.80   0.66  0.13  -0.39   -0.20
+0.30    +0.20  +0.30   +0.80
SUMMARY;   In ENSO regions 1.2 Sea surface temperatures anomalies   were fairly cold  in the spring and summer 2003 (weak  La Nina)  then  warmed  in  SEPT & OCT.     In ENSO region 3.4 sea surface temperature anomalies were rather warm in January then dropped to neutral levels during the summer and are now warming again in the autumn.

This is the pattern that we have to look for to see what IF any period in time matches the last 10 months.


The data at  that link  http://www.cpc.noaa.gov/data/indices/sstoi.indices.   shows  that  the only analogs   that  are   close   are:
1968   1958    1960  1979 and 1969...    the  BIG  5 as I will   call them.




ENSO 1+2     (of .Americas)

1 1958 0.47   0.70    0.59   0.99    0.74
0.02   0.53   -0.26 0.25   0.05 
1.91  1.55   1.22   0.63   0.35
0.61   0.26 0.52 -0.51   0.11
POOR match ENSO 1.2 as  JAN -MAY 1958   was way too Positive  early 1958; ENSO 3.4  closest match
2 1979 0.28   -0.34  -0.23    -0.07   -0.21
0.27    0.33   0.39    0.83     0.69
-0.01   0.46    0.46   0.20 0.05  
0.27  -0.24    -0.02   0.86  0.27
3 1960 0.08    -0.33   -0.33 +0.74   -0.58
-0.87   -0.84   -0.45     0.07   -0.69
-0.03  -035 +0.24   0.15   -0.12
-0.14   0.00  0.19   -0.01  -0.32
CLOSE match in  ENSO 1.2 ...
CLOSE 3.4  (JAN  to  JUNE 1960  in ENSO  3.4  was Much cooler  than  JAN -JUNE 2003
4 1969 0.27   -0.50   0.40   0.96   1.63
1.28    0.39   0.19   0.21   0.88
1.18   1.12   0.84   0.49   0.81
0.60  0.20   0.63   0.63   0.85
JAN  to JUNE 1969 was to Positive
in ENSO 1.2  Much  better in   ENSO 3.4
5 1968 -1.31  -1.26   -1.31   -1.38     -1.80
-1.42 +0.72  -0.09    +0.39   0.25
-0.63   -0.92  -0.60  -0.41   -0.57
+0.10 0.40   0.44   0.04     0.32
Closest match  in ENSO  1.2    POOR  match ENSO 3.4 :  Jan to May 68 was way too Negative 

SUMMARYTHE  BEST   3 MATCHES  ARE   1958   1960   and 1979... WHICH   EQUATES  TO THE WINTERS  OF   1958-59   1979 -80    and 1960-61  



HOWEVER a key indicator that helps forecasters understand the coming El Nino or La Nina and its intensity is called the SOI Index. This Link has a brief and basic review of the SOI and for ease of understanding please refer to this Link

The Southern Oscillation Index (SOI) is calculated from the monthly or seasonal fluctuations in the air pressure difference between Tahiti and Darwin. Sustained negative values of the SOI often indicate El Niņo episodes. These negative values are usually accompanied by sustained warming of the central and eastern tropical Pacific Ocean, a decrease in the strength of the Pacific Trade Winds, and a reduction in rainfall over eastern and northern Australia. The most recent strong El Niņo was in 1997/98.     Positive values of the SOI are associated with stronger Pacific trade winds and warmer sea temperatures to the north of Australia, popularly known as a La Niņa episode. Waters in the central and eastern tropical Pacific Ocean become cooler during this time. Together these give an increased probability that eastern and northern Australia will be wetter than normal. The most recent strong La Niņa was in 1988/89; a moderate La Niņa event occurred in 1998/99, which weakened back to neutral conditions before reforming for a shorter period in 1999/2000. This last event finished in Autumn 2000.

For  our purposes   we  shall  use  this LINK   as the   primary source to see  what  time  periods  IF ANY   closely  resemble  the  last 10 months of  2003.   Generally since  last  winter  the  SOI  has  been   Negative  but  for the most part  at a   "weak"    value  with a  few  brief  surges in the  SOI   into  -10 to -15 values. Note that  there has only been  1 month in the last  12  with  ANY  positive  values... a  feeble    +2.9  in  JULY.    As of OCT 20  the  SOI   value  for  OCT  is  -4.51   and  it may end up    close to  -5.00. 


2002 -6.0 -10.6 2003 -2.0 -7.4 -6.8 -5.5 -7.4 -12.0 2.9 -1.8 -2.2

Looking  through the  data of  SOI values back to 1900.... several    time periods  appear to  have  the  SIMILAR   VALUES AND SOI  PATTERN  that    we have seen in the SOI    over the last  11 Months  ( NOV  2002 to   OCT 2003).   The   periods  that  match  up  closest   to the current  SOI  values of the last 11 months  are  


1957 -11.9 -3.5 1958 -16.8 -6.9 -1.4 1.2 -8.2 0.2 2.2 7.8 -3.4    Leading to  winter 58-59  
1978 -2.0 -0.9 1979 -4.0 6.7 -3.0 -5.3 3.6 5.8 -8.2 -5.0 -1.9 Leading to  winter 79-80
1959  11.1 8.2 1960  0.3 -2.2 5.6 7.8 5.2 -2.3 4.8 6.6 6.9 Leading to  winter 60-61
1967 -4.0 -5.5 1968 4.1 9.6 -3.0 -3.0 14.7 12.3 7.4 0.1 -2.8    Leading to  winter 68 - 69
1968 -3.4 2.1  1969  -13.5 -6.9 1.8 -8.8 -6.6 -0.6 -6.9 -4.4 -10.6   Leading to  winter 69 -70
1994 -7.3 -11.6 1995 -4.0 -2.7 3.5 -16.2 -9.0 -1.5 4.2 0.8 3.2   Leading to  winter  95-96   BUT  since 
  SSTA   do not  Match at all   95-96 is

WHAT IS  THE   MEI?      ==========>>  LINK

Dr. Klaus  Wolter of  the  CDC  (Climate   Diagnostic Center)  has developed  a   index  called   the   MEI  or  Multivariate ENSO Index.  The MEI is   well known within the   Meteorological  Community.  The   problem with  using  just   SSTA  in  ENSO   regions  1.2 and 3.4   OR  Just the  SOI  is that   those values do not consider the   El Nino / La Nina as a  whole complex system.   MEI  measures  six main observed variables over the tropical Pacific. (sea-level pressure,  zonal and meridional  components of the surface wind, sea surface temperature,  surface air temperature   and total cloudiness fraction of the sky ) .

Dr  Wolter  argues that  the  MEI  is   better for monitoring ENSO than the SOI or various SST indices because  the MEI integrates more information than other indices, it reflects the nature of the coupled ocean-atmosphere system better than either component, and it is less vulnerable to occasional data glitches in the monthly update cycles. 

The  last  2 months   has seen a  BIG  jump   in the  MEI   for  the autumn months and he acknowledges that we are getting close to El Nino  threshold.  Using the  MEI data here ====>> http://www.cdc.noaa.gov/~kew/MEI/table.html



1.22 0.90 0.82 0.38 0.001 -0.003 0.22 0.27 0.46 0.59est.

Again   carefully   going  through the   table at the LINK  provided  above   -- and I urge  YOU   to  use  the  Links I  have given  above   to   go through  the data and make sure that I        A) have NOT  Overlooked  anything      B)  that I  am accurate in my assessment --  and  I can  only  find  3 years  that   match  the  CURRENT  MEI  over the last 10 months.

1958 1.47 1.45 1.29 0.87 0.75 0.88 0.68 0.38 0.08 0.10
1979 0.62 0.40 0.04 0.32 0.43 0.46 0.36 0.62 0.81 0.70
1969 0.66 0.80 0.39 0.59 0.73 0.80 0.42 0.28 0.22 0.51
1960  (NO match   at all) -0.29 -0.24 -0.09 -0.33 -0.35 -0.27 -0.31 -0.25 -0.51 -0.39





The quasi-biennial oscillation(QBO) is  a Band of    High level  zonal winds  found   over the Equatorial  at very high  altitudes ( even above the jet stream). It is  a very well known periodic oscillations in atmosphere.     The has been a lot of reseach done  over the last  10 years that shows the  QBO  has significant impact of overall climate patterns  as well as Hurricane  season and   winter weather pattern.    The period of the  QBO oscillation is about  (a little over) two years.    The   QBO  has two  "phases"   with  occur within the   Oscillation  cycle-- the  Easterly  or  NEGATIVE  phase   and the   WESTERLY or Positive  phase.   Within the forecast literature    several prominent  private  forecasters   have often  cited  the  QBO as a key ingredient  to figuring   out  what  the winter  pattern across  North  America might be. But  alot of the  discussion   about How and why  the   QBO is  important   is  rather  speculative  and    fails  to hold up to any scrutiny.  For  example  it has been asserted that Winter the Eastern  half of the  US is often  cold and   snowier than  Normal  when  the  QBO is in   Westerly   or Positive phase  while  other have asserted  that the   Easterly or Negative Phase  is better  for  colder and snowier   Eastern  US  winters.

This  LINK   has the   QBO  data  going back to  1948. Going though the data   it is easy to make  specious assumptions.  For example if one was to look at the  QBO  phases  for  colder  and snowier than Normal  eastern  US  winters...  one  might  Notice   that    in the severe  winter of  1977 - 78    the   QBO was  Positive or in the westerly  Phase...   as was   the  winter of 1978-78...   but   the   active and snowy  winter of  1962-63  had a  strongly   Negative / east QBO   values. 

I believe  that  I have detected a  subtle  but   very  important pattern in the  QBO  data that  has been previously overlooked by Most (if not ALL)   forecasters. 


1950-51 -3.31 -6.01 -5.32 1955-56 3.82 -1.01 -3.23
1951-52 -3.98 -6.80 -7.71 1964-65 0.04 -1.03 -2.26
1952-53 -1.65 -2.94 -2.68 1969-70 5.00 0.30 -1.41
1953-54 -2.18 -2.94 -2.68 1973-74 2.31 -0.91 -1.31
1957-58 7.35 5.25 4.10 1984-85 -8.16 -0.37 4.21
1961-62 6.25 2.84 3.68 1997-98 -9.86 -3.57 1.94
1963-64 5.48 3.94 5.26 1998-99 -3.96 3.09 5.84
1977-78 1.41 3.84 6.54 1960-61 -11.35 -5.40 0.62
1988-89 -2.42 -2.87 -3.56 1972-73 -16.78 -4.40 0.08
1993-94 -6.00 -7.34 -9.94
1995-96 -4.57 -5.79 -6.90
1999-00 6.43 4.86 4.20
2001-02 1.48 4.64 9.00
2002-03 -0.50 -1.39 -1.44

It appears that the timing of when the QBO changes from West   / Positive Phase  to East / Neg phase  and  the Intensity of the QBO can have significant impact on the winters across the Midwest and eastern US.    In this table  we can see that there are several prominent winters which occurred when the QBO phase was weakly (4.00 to 8.50) in the WEST / positive phase OR weakly in the EAST /Negative phase. The winters of the early 1950s saw a consistent easterly phase of the QBO which I think is suspect given the data that is available for researchers to look at  . Those winters were not especially cold or snowy but were generally moderate across the Midwest and eastern US.    However if we notice from the late 1950s (when the data became much more reliable)    into the middle of the '60s    there are several winters listed which had moderate QBO  in  the  Positive  /   Westerly phase and those Winters are generally viewed as  colder than normal and significantly snowier than normal  for the Midwest and Eastern US.     

Furthermore notice the gap between the winters    from  1963-64 and 1977 78 -- there were NO winters where that saw    weak  East or West QBO phases  that held through the winter.   NONE.   Except for the winter of 1969 1970 where the QBO phased changed in Mid winter ... all the winters from 1964-65  until 1976-77... had strong (9.00 or higher) East/Neg. QBO phases or strong West /Pos. QBO Phases.  In that period of time    MOST of those  winter were  MILD  in the Midwest  and eastern  US.

Thus it is my assertion that when the QBO  stays in the Weak Westerly   OR  Easterly phase for the ENTIRE winter... the winters in the eastern United States are often colder and snowier than normal.  (2002-03   weakly negative  1995-96  weakly Neg.    1977-78   weakly    positive... 1957-58   weakly positive) 

Indeed looking at the data from the link posted above almost all winters which featured a QBO that was STRONGLY East or West were either not eventful winters for the Midwest and Northeast US    OR   had a suppressed pattern which favored storm tracks over the central Plains and Tennessee Valley and the lower mid-Atlantic or Southeast states.

As any good climotologist or whether hobbyist can tell you during the period from the late 1960s into the middle 1970s there are a series of very mild winters with a much below normal snowfall across large portions of the Midwest and eastern United States. In fact it was during this time when the snowless winter of 1972-73 occurred.... In fact that winter had a very strong easterly QB0 phase in December which rapidly turned around into the neutral status by February.

In the second column notice that there were several seasons where the positive phase of the QBO switched into the negative phase during the middle and second half of the winter --- 1955 -56 1964- 65 and 1969 1970. Those winters were considered to be moderate winters which featured some areas or periods that experienced SOME temperatures below normal but cannot qualify as particularly severe winters in either in the Midwest or the Northeast US. The two standouts are 1997-98 which was the year of the massive El Nino and 1960-61 which was one of the historically great winters in the eastern US.


0.39 -1.44 -3.30    -8.57 -13.94  -18.01   -22.99   -24.64 -22.51 ?????


Trying to figure out what the QBO is going to be a very risky proposition and about as risky as trying to figure out what the phase of the NAO is going to predominate during the winter months. It is  clear that the current negative phase of the QBO has reached its peak in the month of August and values have started to move back towards zero.  However the question remains what the QBO   will do during the winter months.   One solution is to have the QBO  stay Negative    right into February / March 2004 .... And if this was the case  than Analog year 1979 1980 would become the leading candidate.

Another scenario would have the QBO moving into the NEUTRAL or weak positive phase by early or Mid DEC with the QBO staying in the Weekly positive Phase for the winter... and as I have already shown this will significantly favor a cold and stormy winter over the Midwest & Eastern US.

The THIRD scenario has the QBO switching phase at a slower Place... with the phase change developing in January or the second half the winter... which incidentally is when IMO the weak EL NINO event will start to Kick in.

The climatology of the QBO change in phases is helpful here. The odds are heavily stacked against the QBO reaching Positive values By DEC... since 1953 the vast majority of the time the QBO after reaching its MAX Easterly value took 5 to 6 months to break into Positive values.

1954 east QBO took 7 months to reach positive values from its  max Negative value in AUG  -14.53
1955-56 east QBO took 6 months  to reach Positive values   from its   max Negative value  in NOV  -16.96
1959 east QBO took 6 months to reach Positive values   from its  max Negative value  in June -20.06
1960-61 took 5 months to reach Positive values   from its  max Negative value  in SEPT  -16.01
1963 took 3 months  to reach Positive values   from its  max Negative value  in MAY  -24.18
1966 took 4 months to reach Positive values   from its  max Negative value   in  -21.90 JAN
1968-69 took 7 months to reach Positive values   from its  max Negative value  in -21.82 NOV
1970-71 took 4 months  to reach Positive values   from its  max Negative value  in-22.48 NOV
1972-3 took 5 months  to reach Positive values   from its  max Negative value  in-21.17 OCT
1974-75 took 6 months  to reach Positive values   from its  max Negative value  in -23.32 OCT
1977-78 took 6 months   to reach Positive values   from its  max Negative value  in-21.67 JUNE
1979-80 took 6 months  to reach Positive values   from its  max Negative value  in    -23.32 OCT
1982 took 6 months  to reach Positive values   from its  max Negative value  in  -16.67 MARCH
1984-85 took 6 months   to reach Positive values   from its  max Negative value  in -27.90 JULY
1987 took 3 months  to reach Positive values   from its  max Negative value  in-21.57  JUNE
1989-90 took 4 months to reach Positive values   from its  max Negative value  in  -21.30 MARCH
1992 took 5 months  to reach Positive values   from its  max Negative value  in-17.79 APRIL
1994 5 months  to reach Positive values   from its  max Negative value   in   -28.65 JULY
1996-97 took 5 months  to reach Positive values   from its  max Negative value  in  -26.02 AUG
1998-99 took 4 months  to reach Positive values   from its  max Negative value  in  -24.70 AUG
1999-00 took 5 months   to reach Positive values   from its  max Negative value  in -24.45 JULY

Thus with the current East QBO having reached its maximum value in August it seems that   IF  we assume an average  time of five or six months... than the phase shift would occur in January... NOT December .  Of course there have been SOME instances where 3 or 4 month time period has shifted the QBO into the West phase so there is SOME Support for a faster turnaround. But as of this particular time the situation remains unclear and I can only use the climatology of the QBO to estimate when the changes going to be.

Thus my second critical assumption here is that the QBO is going to change phase in January. If the changes to WEST occurs in December and remains in the Weak Westerly phase through the winter than the odds of a significantly colder and stormy winter would be much higher for the Midwest &Northeast. Likewise is also possible that the QBO could stay strongly negative right into January which what I can favor the 1979 -- 80 winter pattern.




The climate models going into the winter 2003-04 are not looking nearly as promising as last years Climate Models. Last autumn the atmospheric general climate models or AGCM accurately forecasted a large area of below normal temperatures across much of the Midwest and the Northeast United States. That model picked up on the cold pattern in July 2002 and as we move closer to the winter the cold increased so that the last run of the AGCM before the winter-- November 2002 -- showed a very cold winter setting up for the eastern half of the US and Canada. For portions of the Upper Plains and the Midwest the AGCM Busted as the model depicted too much cold air too far to the west ... but in terms of the overall pattern and the intensity of the cold air in the eastern United States the autumn2002 AGCM get a very good job.

If we look at the last  few runs  of the AGCM for  2003   we see the exact opposite trend. The July and August AGCM showed large areas of cold temperatures across the Midwest and Northeast and Southeastern Canada but as we have move closer to the event the AGCM continues to show last last cold air. We can see this by comparing the September AGCM to the October AGCM.
SEPT  2003  AGCM....xstemp_us_current.gif (15115 bytes)xprecip_us_current.gif (15005 bytes)  OCTOBER  AGCM... xprecip_us_currentOCT.gif (15472 bytes)  xstemp_us_current.gif (15129 bytes)

Thus if one were to make a Winter forecast based simply on the AGCM -- which would be really a stupid idea--- than the winter forecast would feature Much less cold and one would have to forecast a rather normal winter across most of the continental United States.

My experience shows however that the new long-range GSM has been doing a much better job with climate forecasting than the AGCM. The last several runs  of the GSM shows a very cold pattern indeed for the central eastern US as well as eastern Canada during the winter months and a fairly strong ridge along West coast. At first glance it appears that this is a classic +PNA pattern... however the last few runs of the GSM continue to place the center of the cold air further and further south and deep in the trough over the eastern US to the point that it now poses a mean pattern which might take potential with the storms OFF the East coast of US and out to sea.


GZ5.200309.ano_PNA.gif (56004 bytes) GZ5.200310.ano_PNA.gif (53977 bytes) GZ5.200310A.gif (522294 bytes)




Recent research within the journal of climate has shown that the patterns which set up in October and November can often but not always the precursors to what the general winter pattern will be like in North America. Certainly this was true for the last few winters. For example last winter we saw a very active and cold October and November with several systems moving up East coast and that pattern continued right into the Winter of 2002-03. Likewise in the record warm winter of 2001 -- 02... there was a consistent zonal flow with no pattern amplification no cold outbreaks and very dry conditions throughout the month of October and November in the CONUS with no significant low-pressure areas developing anywhere in the Midwest or on East coast. Once again we saw that pattern continued into the winter months and we ended up with the record warm and dry winter. Another example of this idea generally working out he is the winter of 1995 1996 when again there were several systems that moved up East coast of United States during the autumn and that pattern continued to the winter.


In this regard is sometimes useful to look at the overall hurricane season and see what if any clues might be given by the pattern of the hurricane activity as well as what we are seeing so far halfway through autumn 2003. But one must be careful with this and NOT read too much into it.


The hurricane season of 2003 featured numerous Cape Verde systems - tropical ways and tropical depressions -- that failed to develop into significant Tropical Cyclones for the most part. There were two notable exceptions -- Hurricane Fabian and Hurricane Isabel. As the map shows most of the hurricane activity occurred in the Tropical and subtropical Atlantic  which  is why  the  Track of   Isabel is  such a standout.   We can clearly see that MOST of the TC activity recurved to   the East of the 65 degrees Longitude line.

In considering our analog years we can see that the season of 1960 and 1979 shows almost no similarity to the hurricane season that is now coming to a close. In the 1958 hurricane season we can see more of the recurvature that is similar to the 2003 hurricane season but by far away the Best match appears to be the hurricane season of 1969.

So far this season we have seen several attempts by the medium-range models to develop a large and powerful ridge on the West Coast and subsequently a rather deep and cold trough over the Eastern Half of the US. This has brought about comment from several meteorological sources as well as weather hobbyists that this AUTUMN (2003) is "just like last autumn" when of course that is utter nonsense.     Only in the Northeast US are temperatures resembling anything close to what we saw in October 2002.   The Plains states which in 2002 experience one of its coldest Octobers in the last 50 years has seen day after day of Much above temperatures in the 85 to 90 degree range. The snow cover is DRAMATICALLY LESS than OCT 2002..

In addition we have seen a very strong Pacific Jet stream develop that has bought in heavy rains and much-publicized flooding into far northwest Washington state as well as large portions of British Columbia. The development of the Pacific jet is going to be a problem for the first half the winter . Over the last 45 days we have seen system after system initially forecasted by the medium-range models to move to certain position or location.... Only to see that later model runs shift to the trough/ low or ridge further to east with subsequent model runs on a steady progression. For example in the last week of OCT 2003... there will be a major trough and surface Low developing over the Deep South and Midwest . Originally this is system was depicted by the medium-range models on the October 20-21 to form a closed off Low in the Upper levels of the atmosphere in MO or lower ILL. Instead we have seen a consistent shift with this major trough and surface Low so that it now looks like the system is going to form well to the East of the Mississippi River. In my opinion it is this shift towards the East which is going to be a problem with the Medium-range weather models all winter and is a common byproduct of having a active and strong Pacific jet stream.




Lastly there is the issue of Canadian snow cover. Sometimes the point about the Canadian snow cover gets overlooked... my experience has been that when Canadian snow cover is around the norm is not really a particular issue. However there are winter seasons where the Canadian snow cover has been significantly above or below normal and as a result the production and sustainability of the Arctic air masses becomes problematical.

To be sure there have been winter is where the snow cover into the Month of OCT has been below normal and still turn out to be colder than normal and snowier than normal winter and likewise there had been Months where the snow cover the month of October and Canada was Above normal and its turn out to be a WARM or mild Winter. Case to in point the winter of 2001-02 when the October snow cover was 17.9 which is very close to the Mean of 17.5 billion square KM. Or 1998 -99 when October snow cover was well above the mean at 18.9! However the general rule is that Autumn Months which are running Above normal in the snow cover in October and November generally have colder and snowier than normal winter's across the Midwest and Northeast.

Since the data for October and November is not yet in old I have to wait until the late-season update at the end of November before I can make a final the determination with regard to the Canadian snow cover issue. And because there are so many exceptions which prove the rule here I give this factor the least importance


           J       F       M        A     M        J       J      A      S      O       N       D    MEAN
1973 45.0  45.8   39.2   30.9  19.9   11.4    5.2   4.0   6.2  17.3   36.5   43.3   25.4
1974 44.9  43.8   39.8   31.5  23.3  11.9    5.2   4.1   4.7  19.4   31.3    38.8   24.9
1975 43.9  43.7   40.2   31.5  20.4  10.6   5.1   3.6   5.2  17.0   31.4   41.3   24.5
1976 44.5  44.3   40.9   30.8  22.5  14.3    6.6   4.4   6.3  27.3  33.8     41.3   26.4
1977 48.1   43.3  38.6   29.8  19.7  11.5    5.8   4.1   7.8  19.4   30.5    43.4   25.2
1978 48.5   50.2  42.4   29.2  22.7  15.0    7.7   5.6   7.4  18.5   32.4    43.5   26.9
1979 48.8   46.7  42.9   34.5  22.3  11.1    6.5   4.6   5.7  14.2   27.9    36.9   25.2
1980 44.5   46.5  41.1   33.6  20.3  10.2    5.3   4.7   5.4  14.1   31.9    36.2    24.5
1981 41.3   42.9  42.2   33.2  21.6  13.3    6.1   4.8   4.5  16.2  33.9   43.0    25.3
1982 47.5   45.2  40.1   31.1  18.3   9.3     3.6   3.2   6.2  19.6   34.0    42.6   25.1
1983 46.0   46.1  40.1   30.2  21.3 10.8   4.6    3.8    5.5  17.9   31.7   43.5     25.1
1984 44.9   43.8  40.1   29.6  17.3   7.3    3.6   2.7   4.1  16.4    32.8    41.5    23.7
1985 48.1  46.5  42.4   31.8  21.8  12.4   4.9    3.2   4.5  18.0    39.1   44.6     26.4
1986 44.7  46.1  38.2   30.4  20.0  10.4   4.4    3.2   6.1  17.1   34.0    40.2     24.6
1987 46.3  44.8  42.7   29.7  17.8  12.2   5.5    3.1   4.8  13.5   32.5    41.7     24.6
1988 46.8  44.2  38.7   27.4  18.8   7.8     3.6   2.5   4.3  13.1   30.6     40.2   23.2
1989 44.8  43.2  37.8   28.8  18.1   9.2     4.2   2.6   6.2  16.5   33.0    43.9    24.0
1990 44.6  43.0  36.5   27.3  16.3   6.4     3.3   2.6   3.9   14.7   29.7    43.9    22.7
1991 45.1  44.6  38.1   28.4  18.7   9.9     4.1   3.4   4.5   16.7   34.6    41.7    24.1
1992 44.9  42.2  37.7   28.1  17.7  10.0   3.9    2.8   5.3   16.2   34.3   44.6     24.0
1993 45.6  44.2  39.3   27.7  18.0   8.2    3.7   2.8    6.0   18.8     37.7   42.7    24.6
1994 44.1  44.7  38.1   28.2  18.5   9.3    4.0    3.1   4.9   14.5   32.2    42.1    23.6
1995 45.2  41.8  37.4   32.2  19.3   9.5    4.1    4.0   6.5   17.2   33.3   43.5    24.5
1996 46.0  43.9  40.5   32.8  20.3  10.5  4.4    3.0    6.7   19.7   35.5   40.6     25.3
1997 45.3  44.1  37.8   31.1  19.7   9.5    4.8   3.9    6.7   17.4  34.9    43.1    24.9
1998 44.7  43.5  40.7   30.0  17.8   9.4    4.3   3.5   5.1   18.9   34.1    42.3    24.5
1999 43.3  42.6  38.9   29.1  18.3   8.2    3.1   2.3   5.1   18.0   33.2    42.1    23.7
2000 46.4  43.8  38.0   27.7  18.6   9.1  3.0     2.5   7.0   17.8    34.7    43.1    24.3
2001 45.3  43.8  39.0   28.3  16.4   7.4    3.0   2.3   5.7   17.9   31.1    44.0    23.7
2002 44.9  40.9   35.5   29.4 18.2   8.1   2.9   2.5   5.4   23.0   35.5    44.5    24.2
2003 46.3  46.5   39.7  30.7  19.4   9.3   2.8   2.2    4.7   .0      . 0        . 0    

MEANS 45.5  44.4   39.5  30.2  19.5  10.1    4.5    3.4   5.6   17.5   33.3    42.1
STD DEV 1.6   1.8    1.8    1.9     1.8     2.0    1.2      .8     1.0    2.8     2.3     2.0

WINTER 2003 - 04


1. Weak  El Nino  develops   by  DEC   that    begins to affect the overall pattern  in  Mid January. 
2.  QBO    will continue to rise  and reach Neutral levels   by Jan 2004  and Positive levels  in  MARCH
3.  Snowcover will    reach   slightly above normal Levels  in     NOV in Canada.

4.  Pacific Jet will be  active and strong  all  winter  but especially  first half of  Winter.

5.  JAN  15- MARCH 20 will be colder and stormier   East of  The Mississippi River  then DEC 1-JAN 14
6.  Middle  DEC    significant snowstorm over  central Plains  Lower Midwest/ Tenn.  Valley  VA  NC Lower MD

This forecast will be  update   NOV  25-28

dec.jpg (114621 bytes)

jan.jpg (134023 bytes)

feb.jpg (136185 bytes)