review of 2001-2002: WHAT WENT WRONG?
The speculation on what went "wrong" with the Winter forecasts of 2001 -2002 has been quite extensive within many of the WINTER 2002 -03 weather forecasts and discussions. In case you did not know... the general or consensus forecasts (CF) called for the winter of 2001-02 to be colder than normal over the Midwest and northeast US and snowier than normal. The term "consensus forecast" (CF) is a general view or summary of what most forecasts 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 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.
On the meteorological level the winter forecast for 2001-02 did not verify ( called "busted" in the weather biz) because of two main reasons. One the surprise jump in solar activity that occurred in the Autumn of 2001 and lasted into much of the winter. It has been argued that this increase in energy enhanced the High level Jet stream (200 MB and above) which made the flow zonal and resistant to buckling . which in turn lead to two developments
the very strong zonal Jet stream from OCT 2001 through March 2002 kept cold air bottled up in Canada and even there the icebox areas of Canada were not really that cold
prevented any sort of phasing between the Polar Jet ( PJ) and the Subtropical Jet (STJ) which greatly reduced the development any storminess hence the autumn and winter drought of 2001-02
The 2nd reason why the CF for the Winter of 2001-02 failed is that they
were all largely based on a FAULTY premise. The current El Nino has been forecasted by
various climate agencies both private and government since the Spring of 2000. In other
words the El Nino forecast /speculation began about 24 - 30 months ago! Last winter
CPC trotted out the old "weak El Nino is coming
bad winter" one more time
and we all assumed that THIS time the El Nino was going to develop. When I made my winter
2001-02 forecast I heavily emphasized that the call for a colder than normal and stormier
than Normal winter it was strongly contingent on a WEAK El Nino developing. By early
January when it failed to do so I acted.
Frankly I was amazed if not downright stunned by the number of forecasters -- both Govt. and PWSIPs that ignored crucial REAL time data by explaining it away with little more than a wave of the hand and favored Models that where NOT verifying.
LESSONS FROM LAST WINTER .
Use of long range climate Models data and analogies is all fine and good. Heck I use them as much as anyone. But after last winter there are important lessons to be learned and as I consider the various winter forecasts for 2002-03 it seems clear to me that some have NOT learned the lesson of last winter.
The advantage forecasters have with today's models research and technologies is that short terms and Medium Range (MR) forecasts can be adjusted to current or near real time data. For example suppose Weather models show that show a Low pressure area that may bring a winter storm to the East coast after it tracks across the Gulf Coast. If that scenario or SEQUENCE OF EVENTS (SOE) depicts a certain track yet the data shows the Low moving into say Northern Tennessee ( which implies or threatens more of a Rain scenario) then the storm track and overall forecast MUST be adjusted. That is how it is suppose to work: the use of various forms of data (including weather models) to come with a scenario or what I call a SOE. The advantage of using an SOE is that "surprises:" are less likely and good forecasters can make adjustments in the forecasts accordingly.
From my point of view a blown weather forecasts (" the bust") are part of the weather biz. If you are afraid or concerned with being wrong then sell shoes. Last winter my winter forecast was terrible. But I had the wherewithal to bail out of that forecast by the use of SOE method on January 4 2002 with approx. 75% of the winter remaining. Several PWSIPs meteorologists were surprised and stunned by my decision. "DT there is still 75% of the winter left Its too early to give up "
My bail out decision was based on several factors the lack of snow cover in Canada the failure of the much anticipated weak El Nino to develop and the failure of any blocking weather pattern to occur anywhere in the western Hemisphere. For example on or about DEC 15 2001 the Greenland Block had reached exceptional low negative values which should of flooded the eastern half of Canada and the US with cold air. Thus by January 4th when still NO cold outbreaks had developed I knew it was time to bail. The SOE that supported the forecast for a cold and stormy winter had not developed.
REVIEW OF THE CURRENT EL NINO
The current El Nino forecast and seasonal predictions since the spring of 2002 in the general sense have been pretty good. However with respect to the actual or what is referred to as "sensible weather" the seasonal forecasts have been nothing sort of stinking awful from various PWSIPs that are big names in Energy and Agriculture fields to CPC. There have been no aspects of the current El Nino sensible weather effects that has been forecasted correctly by the CF (consensus forecasts). It seems to be a widely held but poorly reason premise that all El Nino events always produce the same conditions. This faulty reasoning can be seen on a daily basis from various sources such as PWSIPs . NWS forecast offices around the nation as well as the Climate Prediction Center.
Recently is the wxrisk Daily Model Analysis newsletter I posted these Two links that talk about changes in HOW and WHERE El Nino events develop. Please take a look that these links and research if you get the chance
This link shows how the El Nino event of 1967 1965 and 1972 were Much different then the El Nino of 1982-83 86-87 91-92 and 97-98 http://www.soest.hawaii.edu/~bwang/bw/pubs/36.html
This link shows even more research on the vast differences between the El Nino events of the 50s 60s and early 70s versus the better known -- what I call CLASSIC El Nino from 1982-2001. This link is a GREAT paper !!!! http://www.soest.hawaii.edu/MET/Faculty/bwang/bw/paper/wang94.pdf
Consider the following:
CPC and various climate agencies have been forecasting this current El Nino event since the Spring of 2000 - about 24-30 months.
It did not show up last winter... like it was suppose to. When the current event did appear it was in the late spring.
· HURRICANE SEASON 2002
Many hurricanes forecasts --from Dr William Gray to Dr Frank Elsner to Tropical StormRisk.com used the developing ENSO as a KEY premise to make their hurricane forecast. In EVERY case over the last 30 years the development of the EN has resulted in reduced Tropical Cyclone (TC) activity in the Atlantic Basin. In every case. But the Hurricane season of 2002 featured EXCEPTIONAL TC activity for an El Nino year. The only El Nino year which had more tropical cyclone activity then the hurricane season of 2002 was 1969 - 1970. The activity in the month of September 2002 was also exceptional. Not only was it one of the most active September's ever... it was also the most active September for tropical cyclone activity in ANY El Nino year.
· SUMMER FORECAST 2002
Although you may not know it the Summer forecast for 2002 was another seasonal forecast bust of immense magnitude for most PWSIPs as well as CPC. The CF was for a mild and wet summer in the Plains and the Midwest. This forecast was based primarily on the El Nino since that is what occurs in a typical El Nino during the summer months. In fact it turned out to be the hottest and driest summer since 1995. It is not that I am complaining per se since I went for a particularly Hot & Dry summer in the Plains & Midwest and the CF busted so badly that I picked up by a lot of business and clients and made quite a bit of money. The point is that once again this current El Nino has not active in a way that has produced typical El Nino conditions.
· AUTUMN 2002
As we moved into the autumn of 2002 once again the classic El Nino weather pattern was brought out of the closet and the severe drought which had been ongoing for the last two years over the East Coast of United States was forecasted to continue "with no relief insight." September and October 2002 were exceptionally rainy and the drought has been greatly eased or diminished. Again the point is that forecasters that use the same old classic El Nino climate pattern have busted because this current El Nino is NOT of the same family of El Nino events we have seen for the past 30 years.
SNOW COVER IN N. AMERICA OVER THE LAST 30 YEARS ftp://ftp.ncep.noaa.gov/pub/cpc/wd52dg/snow/snw_cvr_area/NH_AREA/
As you can see from the Link above the mean snowfall cover for the month of October is 17.3 million square miles. Over the last several years snow cover in Canada has been either below normal by 30 October or just barely at the mean value. As the data link above shows however OCT 2002 saw the 2nd most snow on the ground since records began in 1973 ( the record is 1976 which was a Bitter cold extreme winter) . These numbers are simply STUNNING and once again exceed the typical classic El Nino normals.
Now lets be careful here . Deep early snow cover in Canada does NOT mean that the winter is going to be colder than Normal. What one can say is that late autumn and winter seasons that have:
A) deeper than normal and earlier than normal snow cover in Canada strongly favor at least normal winter temperatures or colder than normal winters. The data shows that there have been very few mild winters with in the US and Lower Canada when snowfall depth and coverage is significantly above the mean.
B) When arctic air masses form and drop south into the US they will run colder and last longer than normal. The deep October snow cover DRAMATICALLY affected the record cold air masses of October 2002 moved into the Plains and Midwest.
CONCLUSION ON THE CURRENT EL NINO SO FAR
I have seen some forecasters attempt to use the hurricane season as a analog year to the subsequent winter season. Such an attempt is a misuse of the idea of a weather analog year (a analog year is a year or seasons that offers similar conditions to the current period on a certain set of parameters that may lead one to thank that the future forecast may be similar) . The significance of the near record active hurricane season of 2002...along with the wet and cold latter half September October over much of Eastern North America... the ending of the drought over the eastern United States... the record October snowfall in Canada.... should be telling forecasters that the classic El Nino climate pattern is not viable for this current El Nino. It should be setting off alarm bells and for the most part it is not. This is a failure to learn from last year's debacle because the CF is being enforced by the obsession with climate models over actual real-time data.
Assuming that my premise is correct... that these real-time facts and conditions are clues or at least red flags that forecasters should be concerned with... is there some way we can look at the actual El Nino data and compare this current El Nino to the previous El Nino events and the classic El Nino pattern?
Sure there is. For those of you that don't know the El Nino area in the eastern and central Pacific has been divided up into several area as shown here.
As you can see the El Nino region 1+2 is the area in the far EASTERN
Pacific off the Peru coast
and El Nino 3+4 is over the central Pacific located near
These two web sites contain El Nino data broken up by El Nino regions. This first link has weekly data going back to 1990. This second link has monthly data going back to 1950.
Here are Two DATA links. This one showed weekly ENSO breakdown since 1990
and MONTHLY enso data since 1950. http://www.cpc.noaa.gov/data/indices/sstoi.indices
Here are some graphical representations of other El Nino events summer which are rather well-known. http://www.cdc.noaa.gov/map/clim/sst_olr/old_sst/sst_anim_4panel.shtml
This link shows the more recent El Nino events such as the 1997 98 El
Nino. As you can see here the warmest SSTs anomalies occurred off the coast of Peru or El
Nino region 1 2. Note the similarities to the 1982 - 83 El Nino where once again we saw
the strongest warming / anomalies occurring in El Nino region 1-2.
THE CURRENT ENSO
||ENSO 1+2 (off
JUNE through MARCH
|ENSO 3+4 (EQ.PACIFIC)
JUNE through MARCH
|2002||-0.36 -0.81 -0.86 -0.60 +0.23||0.94 0.90 1.08 1.19 1.45|
|Numbers are SSTs anomalies; all values are assumed to be POSITIVE values. 0.50 = SST anomaly that is 0.50 above normal above. NEGATIVE values are colder than Normal SSTs|
And analysis of the current El Nino (2002) must be focused on two aspects.
First HOW the current El Nino developed. The 2002 El Nino began with colder than normal sea surface temperatures (SSTs) in region 1-2 during the summer months that showed slow warning in the autumn and reached neutral levels and October. In region 3 and 4 the current El Nino has featured moderately warm SSTs for the summer with steadily rising values in the autumn.
Second WHERE the current El Nino developed. With regard to the 2002 El Nino the critical aspect that many forecasters are overlooking is back this current El Nino did not develop off the coast of Peru or El Nino region 1-2. Instead is centered or developed first over the equatorial Pacific by the Dateline... in other words region 3+4.
This Link here from CDC shows 4 well known El Nino events (1982-83 1991-92 1994-95 and 1997-98) in fast moving looping animation. And as you can see all of these events featured the Core of warmest SSTs anomalies reached their peak off of South America.
Based on what I see right now there appears to be little chance that minor warming in El Nino region 1-2 will OVERTAKE the warm SSTs anomalies in El Nino 3-4 region this winter. The characteristic of the current El Nino with its imbalance in the warm SSTs anomalies has HAD a major impact on the summer and autumn and this may be why so many of the Consensus climate forecasts have been so bad. Assuming that this ASSESSMENT and forecast is correct These factors will result in an unusual winter when compared to the Classic winter El Nino forecast for the CONUS.
PREVIOUS EL NINO EVENTS.
WHICH ONES ARE SIMILAR TO THE EL NINO OF 2002 ?
|COMMENTS||ENSO 1+2 off PERU coast
JUNE through MARCH
|ENSO 3+4 EQ.PACIFIC
JUNE through MARCH
|1951 - 52||NO||strong El Nino||1.54 2.19
1.61 1.05 1.29 1.22 0.18
-0.01 0.38 0.03
0.04 0.41 0.74 0.38 0.80 0.80
0.72 0.37 0.34 -0.26
|1957-58||NO||Actual SSTs anomalies do NOT match 2002 El Nino at all||0.58 1.89 1.97 1.51 1.42 0.85 0.72 0.91 0.47 0.70||0.41 0.75 1.27 0.36 0.79 1.42 1.61 1.91 1.55|
|1958-59||NO||This is a classic weak El Nino that weakened in NDJ||0.02 0.53
-0.26 0.25 0.05 -0.04
|0.61 0.26 -0.51 0.11 0.20 0.54 0.55 0.52 0.33|
|1963-64||SO-SO||This is a classic weak El Nino that weakened in NDJ||-0.42 0.16
0.64 0.69 0.09 0.24
-0.01 -0.32 -1.06
0.92 0.78 0.99 0.94
1.11 0.91 0.45
|1965-66||NO||1+2 in summer 65 was already strongly Positive||
1.54 1.48 1.20 0.92 0.76 0.62
0.61 0.79 -0.13
|0.78 1.15 1.34 1.29 1.56 1.62 1.84 1.31 0.98|
|1968-69||CLOSE||1+2 was Neg. all summer to Oct 68. This is a classic weak El Nino that weakened in NDJ||-1.42
-0.72 -0.09 0.39 0.25 -0.33
0.16 0.27 -0.50
0.44 0.05 0.32 0.85
0.80 1.18 1.12 0.60
|1969-70||NO||1+2 in summer 69 was already strongly Positive||1.28 0.39 0.18 0.21 0.88 0.65 0.610.44 -0.42||0.61 0.20
0.63 0.63 0.85 0.73
1.15 0.88 0.41
|1972-73||NO!||raging HUGE El Nino!||1.84 2.22
2.45 1.65 1.65 1.72 1.85
|0.85 1.09 1.43 1.22 1.78 2.11 2.14 2.14 1.74 1.29|
|1976-77||NO!||Note how 1+2 weakened in NDJ||1.37 1.61 1.37 1.21 0.61 0.51 0.32 0.86 0.82 -0.07||-0.10 0.13
0.37 0.56 0.98 0.96
0.73 0.87 0.33
|1977-78||NO!||La Nina 1+2 weak El Nino 3+4||-0.20
-0.10 -0.54 -0.61 0.07 -0.03
-0.35 -0.15 -0.06
0.14 0.32 0.71 0.77
0.69 0.83 0.24
|1979-80||NO||This is a classic weak El Nino||0.27 0.33
0.39 0.83 0.83 0.69 0.52
0.41 0.29 -0.27
-0.02 0.86 0.27 0.37
0.36 0.62 0.49
|1982-83||NO!||raging HUGE El Nino!||-0.06 0.67
0.96 1.36 2.06 2.95
3.33 2.99 2.13
1.25 1.02 1.23 1.47 2.05 2.29
2.74 2.85 2.44
|1986-87||CLOSE||closest match of the group||-0.98
-0.68 0.18 0.00 0.18 0.38
0.21 0.88 1.17
0.45 0.69 0.97 1.22
1.23 1.39 1.33
|1987-88||NO!||strong El Nino||1.11 0.98
0.94 1.00 1.50 0.99
0.67 0.21 -0.23
1.53 1.73 1.88 1.75 1.48 1.47
1.13 0.81 0.53
|1990-91||NO!||This is a classic weak El Nino La Nina 1+2 ElNino 3+4||-0.27
-0.80 -0.54 -0.51 -0.61 -0.81
-0.34 -0.57 0.00
0.09 0.18 0.35 0.11 0.38 0.21
0.43 0.50 0.24
|1991-92||NO!||1+2 were positive from June on. 2002 was Negative June- early Oct||0.10 0.25
0.29 0.31 0.24 0.54
0.63 0.40 0.71
0.73 0.43 1.03 1.34
1.89 1.90 1.94
|1993-94||NO||This is a classic weak El Nino||0.80 0.26
0.27 0.39 0.14 -0.05
-0.08 -0.10 -0.18
0.14 0.28 0.33 0.39
0.29 0.08 -0.10
|1994-95||CLOSE||1+2 was Neg. all summer until OCT 94
enso 3+4 was weaker than current
-0.59 -1.09 -0.28 0.64 0.77
0.82 0.91 0.46
0.65 0.36 0.90 1.36
1.40 1.04 0.76
4.01 3.96 3.69 3.99
4.13 3.80 3.01
1.45 1.85 2.14 2.29 2.64 2.80
2.78 2.59 2.17
The table above above clearly shows the SSTs anomalies in both regions were Higher in the more recent El Nino events (since 1982) .
But this is what I consider to be a KEY factor. The El Nino events in the 50s and 60s saw many SSTs in El Nino region 1+2 dropped suddenly in late November December and January... Take a look at 51-52 57-58 58-59 63-64 and 68- 69.
Yet since 1972 El Nino 1+2 SSTs anomalies did NOT for most part show a big or sudden drop in NDJ... 72-73 saw region 1+2 SSTs anomalies HOLD or rise in NDJ... as did the 1982-83 event 90-91 91-92 94-95 and 97-98....
In 1976-77... 1979-90 and 1986-87 the SSTs anomalies Held.
Finally note what happened in 1993-94 which was the coldest winter since 79-80 in the eastern Half of the US.... Region 1+2 saw BIG drops in SSTs anomalies in NDJ.
At this time I would categorize the CURRENT El Nino as a moderate event. If the current warming continues at the current pace through the end of the year, the current El Nino will reach the category of a "strong El Nino event".
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.
The SOI values of September and early October if unchecked would support the continued warming of the current El Nino. However over the past 25-30 days there has been signs of a Major changes in the SOI values and not surprisingly over the last week the warming has slowed considerably. This is a CRITICAL development. Many of the climate models through the winter show SSTs in all of the El Nino regions leveling off while the actual data is running slightly warmer than forecast. The next 21-30 days will be Critical.
ACTUAL SOI INDEX DATA
Date Tahiti Darwin Daily** 30 day 90 day
Av. SOI Av.SOI
10-Oct-2002 1012.01 1011.85 -17.50 -11.19 -11.52
11-Oct-2002 1012.20 1011.75 -15.70 -11.56 -11.74
12-Oct-2002 1012.51 1011.05 -9.20 -11.31 -11.83
13-Oct-2002 1013.70 1011.10 -1.80 -10.76 -11.63
The strong Negative values of the SOI 30 day average that developed from late September into the early October were quickly followed by the rapid warming in the SSTs in El Nino region 3 and 4.
14-Oct-2002 1013.91 1011.70 -4.30 -10.13 -11.40
15-Oct-2002 1014.64 1013.10 -8.70 -9.65 -11.16
16-Oct-2002 1013.30 1012.80 -15.40 -8.92 -11.07
17-Oct-2002 1013.11 1010.40 -1.10 -8.03 -10.86
18-Oct-2002 1013.63 1010.55 1.10 -7.79 -10.68
19-Oct-2002 1013.91 1011.05 -0.10 -7.86 -10.59
20-Oct-2002 1013.30 1010.75 -2.10 -7.97 -10.60
21-Oct-2002 1011.88 1010.60 -10.40 -8.34 -10.68
22-Oct-2002 1011.15 1009.65 -8.90 -8.84 -10.67
23-Oct-2002 1012.80 1008.65 8.10 -8.99 -10.57
24-Oct-2002 1014.03 1008.50 16.90 -8.46 -10.49
25-Oct-2002 1013.66 1009.45 8.50 -8.11 -10.37
By the last week of October the steady but slow weakening of the negative SOI
Began to accelerate. By NOV 1 the SOI was 40% weaker than it was in OCT 10.
26-Oct-2002 1013.31 1010.55 -0.80 -8.10 -10.33
27-Oct-2002 1012.64 1009.80 -0.30 -8.12 -10.33
28-Oct-2002 1012.43 1009.95 -2.60 -8.53 -10.34
29-Oct-2002 1013.54 1009.55 7.00 -8.49 -10.17
30-Oct-2002 1014.35 1009.85 10.30 -8.05 -9.93
31-Oct-2002 1014.48 1010.55 6.60 -7.59 -9.81
1-Nov-2002 1013.70 1011.10 -1.70 -7.15 -9.72
2-Nov-2002 1013.88 1011.05 -0.30 -6.32 -9.58
3-Nov-2002 1013.26 1010.65 -1.70 -5.46 -9.43
4-Nov-2002 1014.58 1010.45 7.90 -4.48 -9.09
5-Nov-2002 1014.63 1009.70 13.00 -3.25 -8.75
6-Nov-2002 1014.04 1009.85 8.30 -1.97 -8.50
7-Nov-2002 1014.79 1010.35 9.90 -0.76 -8.21
8-Nov-2002 1015.48 1010.95 10.40 0.17 -7.98
The rapid weakening of the negative SOI has now completely erased what was once a strong signal of a moderate to strong El Nino event. This COULD mean the critical east winds over the equatorial have changed direction and the rapid warming in the SSTs will end by the Middle of November. In fact if this trend holds into NOV 15 -20 or so the waters in El Nino region 1+2 may begin to COOL.
This link here shows the current SSTs anomalies in the Pacific. Along the
Equatorial Pacific the El Nino can be clearly see with water temps that are 2.50 to 3.00
degrees above normal. Also note the amazing Pool of warm water off the West coast of
Canada. This is a Critical Piece of information and the existence of such a large pool of
water-rather uncommon-will have a major role to play in the winter of 2002-03
However the peak of the CURRENT El Nino appears -- at least as of Today Mid November -- to have been reached. Compare the current SSTs map with this one from OCT 27 when the Max anomaly was 3.09 degrees C above normal. In the space of 16 days SSTs have cooled over 0.25C
The PDO is a long term Oscillation that has been recently discovered in the Northern Pacific Ocean. The PDO has been the subject of some controversy. A well known meteorologist in his recent commented that since weather features such as the PDO are useless since they are not predictable and therefore a pretty much waste of time. Unfortunately that Mindset set really misses the point. While it is true that the PDO does NOT have a direct or daily/ weekly influence of weather patterns in North America there is clear and overwhelming data that shows the PDO does affect the strengths duration and locations of El Nino and La Nina (colder than normal SSTs) events. When the PDO is in the warm or positive phase it favors warm water in the tropical Pacific and El Nino events are typically stronger and longer. Likewise when the PDO is in the cool or negative phase the El Nino is weaker and shorter.
It may turn out to be that the Primary reason why the El Nino
of the 50s 60s and 70s were colder and snowier is the PDO inhibited
the large SSTs anomalies from building up.
For More info on the PDO use this link. http://tao.atmos.washington.edu/pdo/
The current PDO earlier in the year and last year may have acted as a break on the current El Nino. It may be the main reason why Current El Nino took 24-30 months to develop is because of the cool phase PDO.
THE USE OF ANALOGS
Some forecasters - in fact nearly ALL-forecasters use analogs to help make a seasonal or monthly forecast. In weather forecasting the use of analog is an attempt to seek understanding of the current situation in part based upon similarities between a particular set of parameters ( in this case monthly or seasonal similarities. The forecast or then uses these similarities with other seasons or years to assist him in making the monthly or seasonal forecast for the future.
For example one may consider the fact that most winters in El Nino years bring warm and dry conditions to the Upper Plains and Midwest... 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 precipitation data over a certain section of the country and use these 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's produced one set of conditions in the winter months during United States is of course folly. Some of the most severe winters and 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.
For example one of the analog years that some forecasters have been considering as a possible match for the current El Nino situation is the summer and autumn El Nino of 1991. Of course selecting such a season is a rather subjective but there are ways to get around this dilemma. One of them is to use the SOE method. There are numerous web sites which can allow weather hobbyists as well as professional meteorologists to seek similarities to the current situation and find any possible analog matches.
In the case of 1991 if you the use the link above you'll see that October 1991 was somewhat below normal over the Upper Plains but the rest of Midwest and the Northeast was closed normal. October of 2002 was very cold across the Upper Plains Midwest and in many locations such a Chicago Minneapolis Omaha and Indianapolis Oct. 2002 place to the top 10 coldest October is on record. November of 1991 was very cold across most of the Plains and Midwest and the cold lasted most of the month. November of 2002 has started off cold but the pattern has quickly collapsed and the rest of the month will see rather mild temperature readings across the Plains and Midwest as well as the Northeast.
If we consider the 1991-92 El Nino on Table 1 will see that it does not match well at with El Nino SSTs anomalies in region of 1-2 or 3-4. Moreover the hurricane season that year was one of below normal activity nor was late September and October over the eastern United States particularly rainy. At the very least if you are using 1991 as analog and you call for a significantly milder than normal October and a significantly colder than normal November . and just the exact opposite happens then given all this is probably safe to say that the 1991 92 in Nino is not going to be a good match for this upcoming winter since there appears to be no similarities and vast differences.
As shown on Table #1 I have highlighted the El Nino regions sea surface temperature data and as I show the best analogs are the years of 1963 - 64 1986 - 87 1994 - 95 and some similarities to 1968 - 69. The closest match in my opinion is 1986 87 but this is based on the OCT 2002 data so the whole story is not yet known.
If we use the SOI history as a analog we can see no similarities with other years. The current SOI ratings are AUG -14.29 SEPT -7.35 OCT -7.58 http://www.longpaddock.qld.gov.au/SeasonalClimateOutlook/SouthernOscillationIndex/SOIDataFiles/LatestSOI1887-1989Base.txt
Here is a Table of all the SOI indexes that go back to 1876
As you can see the current El Nino has featured strong Negative value for AUG followed by a 50% rise in SEPT and a hold at the value of near -7.5 in OCT is a Singular event . I can find NO similar matches since 1876.
WINTER 2002 - 03
Assuming that the warming in the El Nino regions of 3 and 4 does NOT continue and that the change in SOI means that the SSTs warming is about to come to a halt or even undergo some minor cooling .. then winter are 2002 and 2003 will be a memorable winter for most of the United States.
There will be several underlying themes this winter with respect to the upcoming moderate El Nino Winter 2002-03.
1. The New improved GFS (American Model) appears to still have MANY of the old model problems and Biases. The extreme cold bias may be some what less but judging by what forecasters saw in October the GFS still has a tendency to crush Low pressure systems and short waves by over emphasizing the Polar Jet (PJ). This results in the model crushing or suppressing surface Low development over the southeast states and slide them off the coast.
2. The Canadian Model with its reputation for being a progressive model will be a more useful weather model this winter since it handles weather patterns that feature strong Jet streams fairly well.
3. The ECMWF continues to handle STJ the best.
4. Expect a Number of fantasy BIG storms or false signals. This will be due to the GFS over phasing or over developing Low pressure areas into Massive systems when other weather models are not. This false alert has already occurred a few times in October 2002.
5. The stronger than Normal STJ is going to make it difficult to get Energy in the PJ and the STJ get meet or PHASE in a big way this winter. That does not mean that there wont be Major events or wet systems ones.... nor does that mean there wont be Noreasters along the East Coast. But it terms of the classic Big Low "Bomb} idea .... the enhanced STJ may be an inhibiting factor.... unless there is Strong blocking (like we saw in October) .
6. There will be an update in the first week of DECEMBER BUT only with regard to the SSTs and El Nino regions. January 2-4 I will make a final and Only Update
December I see as a changeable month; I see one or two intrusions of cold air into the Eastern Half of the US but for the most part it looks to be fairly zonal. The Pacific NW will get battered by the Pacific Jet but these system will not make it across the Rockies intact... so that areas as well as the Plains look to be dry and mild. For the snow crowd out there I dont see a lot of snow chance but there may be one or two. IF the 1957-58 analog does come to fruition then DEC could be EXTREMELY warm.... 1957-58 is NOT one of my analog years but Dr Dew Point does mention it.
January 2003 may see the central Pacific or Gulf of Alaska Low to drop Southeast as the El Nino continues to weaken . This could bring a series of heavy storms to the CALIF. coast.I see the mean trough setting up juts off the west coast which favors some sort of FLAT SE ridge and some Blocking over SE Canada. Much of Canada will get darn cold and the problem will result in the cold air UNDERCUTTING the zonal jet stream flow and setting the stage for a series of overrunning heavy snow and ice event from I-70 / I-64 northward.
Eastern Canada could turn very cold here as the pattern resembles the OCT 2002 split flow blocking pattern that develops as the El Nino was Building up.... In this case the El Nino is breaking down.