1st issued  JUNE 28-29  --  posted  on web site   JULY  3


   While we are not quite at the midsummer point yet... over the last several days I have  made some major adjustments to my Summer 2003 forecast and released this information to the private clients that do business with WXRISK.COM.   Now as we are heading into the July 4th weekend I can make some announcements and adjustments to the SUMMER 2003 forecast here on the web site.

For those that are regular viewers of the web site... you will note that the Mid -Summer and Mid- Winter seasonal forecast assessment/ adjustment is a regular occurrence here.  Seasonal forecasting is half  science... Half  art...and even a little bit of guesswork as it still in its infancy even though there has been tremendous progress made in the last 10 years.  Many forecasters have their own particular method and use a number of different parameters. No two methods are the same and some of them are widely different. This being the case... one of my pet peeves is the failure or neglect of  those whom issue seasonal forecasts to make adjustments based upon what has happened recently and what the next several weeks look like. Certainly by July 4th one-third the summer will be over...   and we can see fairly accurately into the middle part of JULY which will give us about 50 percent of summer.

In the Winter of 2000 -01   my winter forecast was for a colder and snowier winter forecast than Normal  over the Midwest and Northeast.   (MOST forecasters-- but not all--  were calling for the same thing).  By   Xmas  2001  it was becoming clear to me that it was NOT going to happen.   On 4 January  2002  I made the  winter re-assessment  and changed the entire winter forecast... and released it on the  web site  JAN 10.

My Job here is NOT  to make myself look good...  or blow smoke  up your  ass and tell you the seasonal  forecast (in this case the Summer 2003 forecast)   is   going to turn out all right in the end.   If you want that  go to the  Climate Prediction center.   My  Goal is to make the best forecast I can and  when necessary   make the call... say I Busted ... and    get back ahead of the game.   I believe I  am going just that   here...  Much Like I did in January of  2001.

The key ingredient into my forecast for the summer of 2000 was the appearance of a weak to moderate LA NINA developing in the eastern equatorial Pacific off the coast of Peru . This feature developed in May BUT it has collapsed completely in the month of June.   Of course there is   still time  to see IF or How the La Nina event mat re-develop over the next few weeks.  The  SST anomalies on JUNE 23rd and JUNE 30  shows that the Cold water pool off the coast of Peru has weaken considerably . But even IF that were to happen   (La Nina  making a  "comeback")  it certainly would NOT affect  JULY 2003--perhaps  AUGUST  2003.   

My summer forecast did not feature the La Nina collapse or the somewhat unusual Jet stream pattern of May and June 2003. I placed the mean trough Position over the Pacific Northwest... which in turn would force the mean Ridge position to be over the Plains and the western portions of the Midwest. Such a position would favor the development of above normal temperatures and very dry conditions.

  Yet clearly that has NOT been the case.  If one were to look at where the warmest and coldest temperature anomalies have been in the months of May and June 2003, we would see that the warmest temperatures (relative to normal) have occurred over the Rockies West Coast and the Pacific Northwest what equal temperatures have occurred over the Upper Plains and the Northeast U.S.



  However unlike last year... the rainy pattern of SPRING 2003 did not end until the middle of the JUNE... whereas in the growing season of 2002 the rain period ended a few weeks earlier. This has affected the Soil Moisture Content / feedback which is a process that many forecasters as well as farmers are familiar with. During the times of drought very dry soil temperatures produce more Heat which ends up affecting the atmosphere and keeps the Heat Ridge reinforced and "locked" in place. Similarly excessively wet ground conditions also affect the overall pattern and this particular case the very wet winter and spring along the East Coast and the Southeast regions has established a trough over the Appalachian Mountains. (my forecast had the trough over the East Coast).

With the trough positioned further to the WEST the Heat in the Plains states has not been able to penetrate East of the Mississippi River for any sustainable period this summer.   Nor will it.   4 or 5 days of   temps  90 -95     followed by   decent  rains with cold front passages and 5 days or  normal or below  is  NOT going to hurt the  Corn of Beans   folks.   It will make for   GREAT  crops  possibly exceeding trend line yields...

In my opinion/ perspective last week’s strong (JUNE 25-26)   cold front was a Back breaker. Up to that point it had been 10 -12 days since the Plains and WCB    (WESTERN CORN BELT)  areas had seen rain and IF the JUNE 25-26 cold front had fallen apart   (like they did June and July  2002)… the soil moisture feedback process might have become a bigger factor in holding the development of Heat Ridges and resisting trough passages.

In my daily  WEATHER TRADERS EDGE   newsletter  and  weekly 30 Day  forecasts... I warned  of   the CRITICAL importance    of this front.   In the 16 JUNE 30 day forecast I stated

Given the excessively rainy pattern we have experienced for most of the SPRING 2003... in order to get any sort of significant heat developing over the Plains and the Midwest we are going to have to dry out the soil moisture level . That is just a given and is based in the way that weather works on a very basic level.

So… If the current data is right  and the JUNE 26-27 cold front is the ONLY rainfall event  in the last week June   then  the period of JUNE 10 –12 to JULY 6  COULD be a pivotal  development  and would meet the criteria I stated above in RED. In short it would represent a significant change in the overall situation. So to say that the June 24th –26th cold front is a critical event for early in the summer would be a correct assessment.

  The wxrisk forecast  on the JUNE 25-27   Major cold front was for  drier than the CONSENSUS  forecasts   ( but not as dry as some others!).  The  Rainfall with that front   EXCEEDED  ALL forecasts  and Model Guidance.   For example  NO model in the short range had  2 and 3 inch rain amounts covering 75% of  MO with that event.    NONE.

This has been a seasonal trend   much like  the Summer 2002   where the actual rainfall amounts was often less than the Model guidance.

  Currently  there  is  another HOT / Dry  period is  developing over  the Plains and Midwest that will last into JULY  6-8.    IF the JUNE  25 -27  cold front   Had indeed fallen part   we would be looking at a much different situation  with this current  ( JULY 2-7)   five day hot period.

This is not to say that it will not turn hot and dry in August. Indeed there is quite a bit of data that suggests August will be very much unlike June and July. That being said Energy and AG interests still have to get through what I now sees as a normal  JULY   after a very good  JUNE. 



Last year when I made my Summer 2002 seasonal forecast there were  a lot of factors of coming into play to make the forecast one of the more important ones we have seen in the last 12 years. As hard as it is to believe there is little doubt that the SUMMER 2003 is going to be one of the more critical summers for many different sectors of economy that we have seen in the last 25 to 30 years. The number of factors coming together at this particular time along with the state of the U.S. economy in several key sectors all seem to be aligning into a particularly critical seasonal forecast.

ENERGY: In the post 911 economy with the completion of the highly successful are Iraq war... a particularly harsh SUMMER  could have a significant impact on the U.S. economic recovery. A hot Summer and recovering economy would allow for increases travel / Gasoline consumption which in turn would have long-term implications for the heating oil and natural gas usage / supply in the winter months.

TOURISM: the winter of 2002 -- 2003 was particularly severe in eastern portion of United States with numerous coastal storms and above-average snowfall. In addition this spring season has been particularly wet only eastern United States which has slowed the beach recovery projects noticeably and left many coastal communities exposed in event of an active hurricane season. Coastal communities that are dependent upon tourism and travel for large amounts of income really like to see a calm hot Summer to bring the tourists back to the beaches. As many are aware tourism within United States was down dramatically last summer.


GRAIN COMMODITIES: the hot dry Summer 2002  developed against most of the Consensus weather forecasts that were  heavily influenced  by the CPC  (climate prediction center)  Summer 2002 Forecast as well as private weather forecasters. The HOT and dry summer may have had a significant impact in terms of the amount of "carryout"  grains which could be put into storage.

In addition the last two winters have been for most of the Plains as well as the western portions of the Midwest the 2 driest consecutive winters on record. This has removed much of the deep soil moisture from these regions and despite the relatively rainy SPRING of 2003 (similar to the rainy SPRING of 2002) there is a very low threshold for crops to work with in the event a dry pattern develops for the summer months.

Indeed the recent rallies in CORN  and SOYBEANS in early in mid-May was rather unusual given the drought of the winter. Typically the underlying premise is that "Rain makes Grain" so therefore significant Rains which fell in late April and May should have been viewed as being helpful and friendly to crop production. Instead prices rose as farmers and traders viewed the excessively rainy pattern as increasing the risk for planting delays and delays in emergence. This should be considered a "shot across the bow"  that any sort of weather problems in the SUMMER 2003 will result in significant reaction in the grain markets.

With the falling U.S. dollar significantly increasing U.S. grain exports... the USDA 30 MARCH  projected at the end of the U.S. CORN harvest September 79.022 million acres will have been planted (same as last year)  But   the US CORN   Carryout will  be 1.004 billion bushels... FOURTH  lowest is  since 1975.    Likewise the USDA projected that the U.S. harvest in   September 2003 for SOYBEANS at 73.182 million acres... which is 576,000 LESS than last year!  The  "Carryout" is  projected at 145  million bushels, the second lowest in 27 years. The rainy spring of 2003 has allowed for significant improvements in the quality of both the winter and spring wheat in the U.S.  and the fact that the World wheat stocks next year will be the tightest since 88/89  have developed yet another early season weather market. It is now apparent that margin for yield adversity in ’03 in the  wheat is extremely narrow for both the US and world.

The USDA projections are based upon Near record trend line yields ( 140-142 Bushel /acre for Corn ) which just goes to emphasize the impact weather has had this spring and will have in the coming summer months. The bottom line is that for the SUMMER  2003 the grain commodities are going to be in a weather market that may  match  the 1988 one.   Not necessarily because the weather this Summer will be feature concerns of  drought   and severe heat but because of several factors including the  recent dry winters... the increasing U.S. exports and falling U.S. dollar... and the lack of carryout which leaves no margin for error... and the USDA projections for near record corn and soybean production.





Throughout this forecast discussion and in the daily Weather Traders Edge Newsletters I make constant reference to the CONSENSUS  FORECAST  (CF). Sometimes a forecaster will have a forecast which goes against the CF  and sometimes  the forecaster will be part of the CF. The Consensus Forecast is more than just a conglomeration of the various forecasts. It is also the perception of what the forecast is by the general public, businesses,  agricultural, and energy traders. For example last SUMMER 2002 in  mid  July the Consensus Forecast was for the very dry pattern which had been developing at critical stage for corn and soybean crop in the Midwest to come to end because of a strong cold front which was appearing on  some of the weather models. This front would usher in a pattern change with a fairly rainy interval for the next seven days. The Consensus Forecast called for 70 to 80 percent coverage across the Midwest with the cold front passage and 2 to as much as 6 inches of rain over the next seven days. My forecast called  for 30  percent coverage with the cold front over the weekend and no such change to a wetter cold pattern.

In fact  the Mid July  weekend cold front came   through essentially dry with less than 15 percent coverage across the Midwest. Was my forecast correct? In academic sense it was not but when compared to the Consensus Forecast-- forecasts issued by a well-known large Private Weather Servers Information providers (PWSIPs) as well as the Weather Channel...Local National Weather Service forecasts... NCEP... and the climate prediction center- my forecast for that weekend and the following week was clearly superior.

In the battle for issuing reasonably accurate and economically beneficial seasonal forecast to private industry and businesses what gets often overlooked is not the forecast accuracy per se but whether not your forecast matches or is in conflict with the consensus forecast. It is a combination of both factors -- forecast accuracy and your forecast compared to the consensus forecast-- that determines " victory" in the issuance of a seasonal forecast within the government and the private sector weather industry.

To illustrate this point a brief review of the Summer forecast for 2002 and the winter forecast for 2002 - 03 will be helpful. As many here know the Climate Prediction Center is   the branch of the National Weather Service that offers an amazing amount of free data that is frequently updated and almost always on time and easily accessible. They also are responsible for issuing the official seasonal forecasts (among their many other duties). As you can see their SUMMER 2002  forecast 200205month.gif (58308 bytes) was for a wet and cool July across portions the Plains and the Midwest. That forecast was based upon the early development of a El Nino event which did not take place. Instead July 2002 was very dry and in the top10 hottest on record.


After that seasonal debacle CPC   issued the AUTUMN 2002 forecast. falloutlook2002b.jpg (40876 bytes) At the   time the East Coast of United States was experiencing a severe and long standing drought that ranged from Georgia into New England. CPC's Autumn forecast called for a continuation of the severe drought along the East Coast. In fact it turned out to be one of the Wettest  Autumns on record for many locations in the Southeast and Middle Atlantic states.

As you can see here the CPC winter of 2002 -- 2003 forecast was another terrible forecast.  winteroutlook2002b.jpg (47859 bytes) CPC  predicted a typical El Nino winter with above normal temperatures over the northern portions of the central & eastern sections of the U.S and above normal precipitation over most of the Southern U.S.  SEVERAL other   PWSIPs  did the same and in EVERY case there was an over emphasis  on the   El Nino event in  certain regions of the El Nino.  In fact it was exactly the opposite with temperatures running much Below  normal over the central  & eastern U.S. in January and February... and it was quite Dry across most of lower Plains and the Northeast United States experienced a very snowy winter. 

Let's take a look at MY winter forecast for 2002 -- 2003. Overall my seasonal forecast was good except for the month of January 2003. But in order to judge my forecast as a good one it must be compared to the CF which was heavily based on the idea that "El Nino is coming so its 1982/83 or 97/98 all over again"..

dec2002.jpg (112657 bytes)  jan2003.jpg (166192 bytes)  FEB 2003.jpg (193714 bytes)

There were several reasons WHY   the weak El Nino event  last winter was NOT going to produce a classic El Nino that the CPC was predicting. Among them were :

Now why did the CPC forecasts fail so badly for the SUMMER   2002... the AUTUMN 2002... and WINTER  2002--03 ? Well one of the reasons is that the climate models   are heavily "influenced"  by well known and infamous strong  El Nino and La Nina events of the past 25 years. When the events are weaker the El Nino /La Nina events  do not have nearly same impact in the atmosphere and the climate models have much  less value. In addition there is often far too heavy reliance upon these still new and relatively primitive Climate Models that meteorologists are now using.  It seem that little if any weight is given to analog years by  CPC that featured similar seasonal patterns and that might give one a clue as what the overall forecast might be down the road for the (next several months).

Classic case in point... In the AUTUMN of 2002 there was the early season buildup of heavy and deep snow cover across most of Canada. In fact by mid-November the snow cover was the second deepest and largest snow cover in Canada in the last 30 years. This meant that the cold air masses in Canada would be larger and deeper and would have a tendency to push further south that we have seen over the last several years. For example in the very very mild WINTER 2001-02 the snow cover over central Canada was 50 to 75 percent BELOW  normal. Last year it was 100 to 125 percent ABOVE  normal and we saw numerous cold air outbreaks across the central and eastern United States. However the impact of the snow cover was pretty much ignored at CPC and the consensus forecast since there are no computer climate models which handle snow depth and the impact of it.

The  current CPC (climate prediction center ) SUMMER 2003   forecast... they continue the trend from the early spring in forecasting Above Normal temperatures over TX and the Southwest states... and normal temperatures across the the Plains Midwest North East.

Given the recent track record of CPC... with its last several seasonal forecast busting quite badly... I would not make any business or trading decisions based upon the latest climate prediction center forecast. IMO there appears to be overwhelming reliance on computer models as opposed to actual real-time data. For the last 30 to 45 days the CPC forecast has been based upon the El Nino moving into a NEUTRAL state and holding across the equatorial Pacific Ocean right into Autumn 2003. . As you can see some of the earlier climate models that CPC used showed a fairly Neutral SSTs into the Summer into the autumn of 2003 . Only in the last two weeks have the climate models come around to realize that the water is rapidly cooling in El Nino region one into an as a result most of these climate models forecasts have to be changed dramatically. Indeed the Aussies issues a La Nina likely statement in early MAY as did IRI and the Scripps Institute. Recently (may 20) CPC finally issued an update and did declare a La Nina event is coming on fast.


For those who are not familiar with what a El Nino or La  Nino event is... I have provided this link which will give you a thorough understanding. http://www.elnino.noaa.gov/lanina.html

Basically the El Nino and La  Nino events are two sides of the same coin. There are several places around the earth where on usual ocean currents appear from time to time and cause the jet stream patterns near the events  to alter in a significant way. In the equatorial Pacific the waters from the South American coast out to the International Dateline often go through cycles or phases. When the Sea Surface Temperatures (SSTs) are significantly warmer than normal-- the anomalies are in the positive range-- and  El Nino event is declared. When the SSTs  in the El Nino regions are significantly below normal a La  Nino event is declared.

This map here  shows the El Nino REGIONS in a geographical presentation that is easy to understand. Essentially the El Nino region ONE and TWO refers to El Nino region 1 and 2 which is located off the South American Coast. El Nino regions THREE and FOUR... refers to as El Nino 3 + 4 and is the region of the El Nino area which extends out towards the international date line.

ninoareas_c.jpeg (25879 bytes)

LA NINA  CLUES... or what we are looking for...

The idea is to find some similar seasons and use them as a guide or starting point as to what might happen in this upcoming SUMMER 2003 season. This process is known as analog forecasting and it does have some value.   What we had here is a rather unusual situation. A weak El Nino event that developed in mainly in region 3+4 ... that featured much cooler SST anomalies in El Nino 1+2... that fell apart in the early Spring and saw a rapid drop in SST anomalies in El Nino region 1+2 in the months of March April and May.

Over the last 60 years re have only been five such seasons which resemble  this pattern. El Nino events that featured very strong warming in the various regions of the owned Nino are not valid since we did not experience that with this particular event last winter.

This table shows the numerical values of the sea surface temperature anomalies in the regions out of El Nino1+2 and 3+4 since November 2002.




2002 NOV 0.60 1.75
2002 DEC 0.60 1.62
2003 JAN -0.13 1.24
2003 FEB -0.22 0.80
2003 MAR -0.50 0.66
2003 APR -1.04 0.13
2003 MAY ? ?

The five seasons that resemble the  Current SST Anomalies   are 1958 -- 59... 1969 -- 70.... 1977-78... 1987 -- 88.... and 1994 -- 95....

1958 -1959 EVENT 1969 -1970 EVENT 1977 -1978 EVENT 1987 -1988 EVENT 1994 -1995  EVENT
YR MONTH 1+2 3+4 YR MONTH 1+2 3+4 YR MONTH 1+2 3+4 YR MONTH 1+2 3+4 YR MONTH 1+2 3+4
1958 NOV -.04 0.20 1969 NOV .65 .73 1977 NOV -0.33 .77 1987 NOV .99 1.47 1994 NOV .77 1.36
1958 DEC -.36 .54 1969 DEC .61 1.15 1977 DEC -.35 .69 1987 DEC .67 1.13 1994 DEC .82 1.40
1959 JAN -.40 .55 1970 JAN .44 .88 1978 JAN -.15 .83 1988 JAN .21 .81 1995 JAN .91 1.04
1959 FEB -.37 .61 1970 FEB -.42 .41 1978 FEB -.06 .24 1988 FEB -.23 .53 1995 FEB .46 .76
1959 MAR .59 .10 1970 MAR -.91 .28 1978 MAR -.94 .22 1988 MAR -.58 .17 1995 MAR -.25 .48
1959 APR .42 .28 1970 APR -.98 .45 1978 APR -.44 -.46 1988 APR -.79 .-36 1995 APR -.87 .25
1959 MAY -.07 .16 1970 MAY -1.21 -.07 1978 MAY -.87 -.37 1988 MAY -.64 -1.28 1995 MAY -1.14 -.04


Which are the seasons are good matches to the current sea surface temperature in the El Nino regions?

1958 59 is not a  great match since in the El Nino   region 3 +4 for was much cooler than what we had in the Autumn of 2002 and El Nino region 1+2     went   positive in March and April of 1979  and  we   have Negative SST anomalies in March and April 2003.

1969 -- 70 is a closer match.
1977 78 is a Poor Match since  there were NO  positive values in El Nino region 1+2   in NOV and DEC 1977... while l Nino region 3 +4  stayed below 1.00.  This is s nowhere close to our current event.

The 1987 1988 is an extremely close match. The data speaks for itself as it matches the current event almost stride for stride.

The 1994 and 1995 event is also a very close match.

KEY POINT the current ongoing event which is switching from a weak El Nino last winter to a moderate La Nino is very similar to two events. 1987 88... which is followed by the very hot and severe SUMMER of 1988 . The second event that bears strong similarities to the current event 1994 1995 event  Which is followed by a very hot Summer and very active hurricane season in 1995.

We can use the analog forecasting technique and take a step further by comparing the last 12 or 13 well-known La Nina / El Nino events and a what if any sort out temperature anomalies or biases may be apparent in those particular Summer season's.

These next two maps show the impact of the El Nino / La Nina   during the June July and August period and the July August and September period. Both of these maps are set up the same way... the LEFT -hand side are Temperature Anomalies... right hand side is Precipitation Anomalies. The top row is the El Nino cases... the middle row is a La Nino...  and the bottom row are Neutral events.

 plot.enso.t.p.noaa.jja.jpg (82280 bytes)    plot.enso.t.p.noaa.jas.jpg (81927 bytes)

In both diagrams we can see that the La Nina events had a strong correlation to Above and Much About normal temperatures over a large portion of the Plains and  Midwest both in the June July and August (mainly  July and August) and July August and September.  Even more impressive is the strong tendency to have Below and Much Below normal rainfall over the heart of the Plains &  Midwest region in JJA   and JAS.

Next  we can compare the REAL  time data -- not models and theories-- to one of the two close analog years I mentioned... the weak  winter El Nino of 1994/95 that turned into a Moderate La Nina by MAY of 1995.

The Map on the LEFT column  shows the   WORLD view of SST Anomalies in JUNE 1995.  NOTE the warm SSTs anomalies in the central and tropical Atlantic  Ocean.... the Cool  SSTs anomalies  off  the   South America coast... the LARGE pool of COLD water in the central Pacific... the cool waters off the East coast of the US.

The  Second Map is an CURRENT  SST anomalies ...   The THIRD MAP is the last 30 day SST Anomalies.
NOTE the very similar patterns.


SST Anomalies
JUNE 1995
MID MAY 2003
SST Anomalies
SST Anomalies
plaincolorwithscale1.gif (42791 bytes) sst_anom.gif (112932 bytes) sst.anom.month.gif (86655 bytes)





In their  MAY 15  Summer forecast  which was based on NEUTRAL   SSTs which showed  NO heat or dryness of any kind whatsoever in the Plains and Midwest.

off01_temp.gif (26841 bytes) off01_prcp.gif (23449 bytes)


Just this past week  CPC issued a updated  statement on the  La Nina event... the significance of this is that in last APRIL  CPC denied that a La Nina was building.  


Warm episode (El Niņo) conditions rapidly dissipated in the tropical Pacific during March and April 2003, as sea-surface temperature anomalies continued to decrease across the central and eastern equatorial Pacific and drier-than-average conditions developed over the central equatorial Pacific.   Significant decreases in SST anomalies occurred in all of the Niņo regions during April and early May  . By mid-May equatorial SSTs were near or below normal between 165°W and the South American coast, with only a small area of residual positive SST anomalies west of the date line between 155°E and 175E° .

Consistent with the cooling trend in SSTs, the equatorial easterlies have been stronger than average over the central and west-central equatorial Pacific since late February, and the equatorial SOI has switched from negative to positive . In addition, in recent months the depth of the oceanic thermocline has steadily decreased across the central and eastern equatorial Pacific, and negative subsurface temperature departures have developed and intensified in the upper ocean of this region . By late-April subsurface temperatures at thermocline depth were below average throughout the eastern Pacific, with negative anomalies ranging between -1°C and -3°C  . These observed trends in oceanic and atmospheric variables indicate that a transition to La Niņa is underway and that La Nina conditions are likely to develop over the next few months.

The latest statistical and coupled model forecasts show a large spread in the forecasts for the next several months. While some indicate the possibility that La Niņa will develop during the second half of 2003, others indicate a resurgence of El Niņo conditions by the end of the year. However, based on current conditions and recent observed trends, it appears likely that cold episode (La Niņa) conditions will develop in the tropical Pacific during the next few months.


As one can see I am strongly leaning towards an La Nina event developing early in the Summer and and lasting into the Autumn of 2003. Obviously the two-week question then becomes is there support for a leading event developing early in the summer? Indeed there is.  First  there is the natural  cycle of   the LaNina itself -- once it begins it continues through  until it "burns itself out".    

Second  the various experimental climate models which had have been  showing with each run a deeper and more significant LaNina event building this SUMMER 2003 .

This next figures represents the Climate Models that are used to forecast the SSTs anomalies in the El Nino region 3 and 3+4. The first map is from March 18th cmb.SSTfcst_nino3.gif (10554 bytes)and that you can see most of these colorful lines are ABOVE the ZERO line which indicates a weakening of the El Nino and SSTs anomalies would stay very close to 0. On this basis many forecasters believed that we would have neutral conditions for the Summer of 2003.

This next figure is the SST anomalies forecast based upon the data from MAY 6. figf4a.gif (9086 bytes)Notice how most of these colorful lines (each line represents a individual climate model) are now dropping below the 0 or neutral line and showing SST anomalies moving to the cold side.

The important thing to note is that by early May almost all of these computer models were under doing the amount of Cold water developing in the El Nino regions. Lastly we have the May 20th or most recent climate model forecasts and as you can see the colorful lines are dramatically further below the zero line... which indicate a La Nina event.

cmb.SSTfcst_nino34.gif (11603 bytes)

  This next diagram represents about 16 various common models all forecasting EC surface temperatures over the next year in the El Nino region 3+4. This was as though the middle of April and began as it to be plainly seen... the actual anomalies are much lower or colder than what the April forecast was calling for in the month of MAY. 

figf13.gif (11929 bytes) 



The AGCM a well-known and much talk about Climate model that has been developed by the climate Modeling branch which is part of NCEP located at Suitland Maryland. Sometimes this model is very good sometimes this tree that promotes the time it has to be read or looked upon in just the right perspective.

The AGCM stands for atmospheric global climate model and it is an ensemble of 20 computer runs of the American model ( GFS). As I discuss many times on the website and in the newsletters the American model has a well-documented and severe cold bias which constantly asserts itself and causes the model to come some wacky solutions. Research and studies have shown that because the AGCM is run off the American model with its cold bias ... it is almost impossible for the AGCM to show any sort of sustainable heat or warm pattern over the United States during the Summer months no matter what the actual pattern might be. The cold Bias becomes so severe in the ensemble on a seasonal level that the AGCM simply destroys any warm anomalies that might appear. As result the AGCM typically does not show any sort of large area of above normal temperatures.

In the winter of 2000 -- 2001 the AGCM predicted a very cold first half the winter across the Plains the Great Lakes and the Northeast whereas the CF were still calling for mild winter. In fact November and December turned out to be the two coldest November Decembers 60 day periods in the last 100 years over the Eastern 3rd of United States. In other words the AGCM was essentially perfect for climate model.

Last winter, however it was a much different story as AGCM in the winter of 2001-02 showed another pool of very large cold air or the Eastern 3rd United States ... yet the winter was one of dryness lack of snow and near record warmth.

        Taking a look at the AGCM from SUMMER  2002 .. as you can see the model just 1month before the verification time completely missed the Very dry conditions that developed over portions of the upper Plains and the Western corn Belt. It was not even close.

200206_US_Precip.gif (13020 bytes)

The second map shows the July temperature forecast based upon the APRIL 2002 data. The Upper map shows the actual temperature anomalies for MAY JUNE JULY 2002. Note the large area of above normal temperatures over the Rockies and Southwest states and the fairly small pockets of cooler than normal temperatures over central and northern Canada.

200206_US_sfc_air_temp.gif (11338 bytes)

The bottom map shows the 1 month forecast. Note how it had no sustainable heat anywhere over the Plains or the Eastern United States but show a large pool of generally cooler than normal temperatures over the Rockies and most of western central Canada.

200207_US_sfc_air_temp.gif (11697 bytes)

The next map again shows the actual temperature anomalies from last year's very hot Summer in the period of June July and August. However on the bottom panel again we can see that model refuses to develop any sustainable heat in the continental United States. And this is Just 1 month before the period cam up for consideration.

I raise this point to show that the AGCM has a severe cold bias and that one should not just look at the model and assumed that because it has no Heat or Above normal temperature anomalies anywhere in the central or Eastern United States that it is going to be a quite or calm season. As I have shown the model has done this on several times completely missing extremely warm temperatures temperature and and below normal precipitation.


In the May 6 issue of the AGCM there are some existing trends to be considered. If we take a look at the precipitation Max first ... this will take us into the heart of the Autumn 2003. According the the Model We will see areas in central Canada and the WCB running above normal precipitation in July August and September... While the East Coast United States is in a serious drought. The drought continues into September October along East Coast and the Southeast.

Looking at the temperature forecasts from the AGCM we see again a very similar type of setup with the SUMMER forecast 2002 Models. As I have already mentioned this Climate model problem is fact that it is heavily baseball upon the American model which has a severe cold BIAS. Since the time scale in these forecasts is over a month and sometimes as much as 6 months in advance the cold bias that the AGCM begins to overwhelm everything.This results during the Summer months the model is incapable of showing anything close to above normal temperatures in the CONUS. Looking at the May edition of the temperatures from the AG CM we can see the Model actually does Bay area of above normal temperatures are across the central Plains and the WCB during June July August. While this little blob of the yellow hardly seems to be that impressive.... Recall how last year this common model never forecasted anything close to the above normal temperatures and very dry conditions we saw. The summer forecast 2001 from this models was also just as bad.

xprecip_us_current.gif (16963 bytes) xstemp_us_current.gif (15070 bytes) xprecip_us_current.gif (17125 bytes) xstemp_us_current.gif (15091 bytes)


Here are the latest runs from the RSM which is a weather model upon from the Scripps Institute in California. This is an ensemble run of 20 members of the American model run on a weekly basis and smoothed out to present a weekly mean or average of what the jet stream is going to look like. I have presented the May 10th in the May 17th on both valid for the same time frame and as you can see the month of June a very strong which develops over the central portions of United States. This would be a fundamental change from the current pattern.

Of course this could change but to the point is that the RSM appears to be showing the development of a La Nina Like pattern by the middle and end of June.

THE MAY 10  RUN f_g50usa2003051000.week_04.gif (24298 bytes) f_g50usa2003051000.week_05.gif (24355 bytes) f_g50usa2003051000.week_07.gif (23369 bytes) f_g50usa2003051000.week_09.gif (22724 bytes)
THE  MAY 17 RUN f_g50usa2003051700.week_03.gif (24657 bytes) f_g50usa2003051700.week_04.gif (22336 bytes) f_g50usa2003051700.week_06.gif (23124 bytes) f_g50usa2003051700.week_08.gif (23194 bytes)



IF assume that the five selected Summer La Nina seasons are the Correct or Valid ones... and that is a BIG IF... then we can consider the hurricane seasons of the five selected seasons and to seeks some analysis / trends as to see what if any tendencies might be picked up in the upcoming hurricane 2003 season.

Other five seasons selected the 1995 hurricane season was most active... with 19 been named storms. 1988   had 12... 1978 at 12... 1970 at 10 and the 1959 hurricane season had 11 named storms . The main difference that I can see is that the 1995 hurricane season had the QBO in the WEST phase which is rather favorable for active hurricane development.  ( The  "QBO" is a high level  wind that blows across the tropics and can enhance or supress Hurricane activity). The other four years had the QBO in the easterly Phase  (much like this coming season).

The 1995 hurricane season... while memorable for its overall activity had no East Coast landfall hurricanes   (Most of East coast Big Hurricanes are  Cape Verde hurricanes  as they form way out there by the Cape Verde Islands off the African coast). The 1995  season had all the Cape Verde  storms recurve out to out to sea well EAST of 70 degrees longitude. There was however significant activity in the Gulf of Mexico with three named tropical storms and 4 named hurricanes.

The 1988 hurricane season featured a lot of Cape Verde hurricanes but some of the systems passed south of the Caribbean islands and into the Gulf of Mexico... with only one named tropical storm hitting East Coast of United States . There were to Cape Verde systems that rapidly turned out the sea... well East of 50 degrees longitude.

The 1978 hurricane season had 12 named storms but 4 of them were weak tropical storms that dissipated. There  were NO East Coast landfall hurricanes (again)  but there was several Cape Verde systems that turned out to sea.

The 1970 hurricane season had 10 named storms but with no activity on East Coast... and quite a bit of activity in the Gulf of Mexico.

The 1959 hurricane season again featured no Cape Verde hurricanes for the East Coast but with numerous systems forming off the south East Coast and in the Gulf of Mexico.

In the table below I have summarized some of the overall hurricane activity. If you notice in all the 5 years selected:

#  TC
#Trop Storms #
#   Intense  Hurricane Seasonal Breakdown
. . . . . Trop storm
G of M
western Carb
East Coast THREAT
W of 70
60W -70W
East of  60W
1959 11 4 7 2 1 2H 4 TS 2 H 3 1
1970 10 5 5 2 1 3 H 4 TS 1 TS 1 H 1 H
1978 12 7 5 2 4 1 H 3  TS 1H  1TS 1 0
1988 12 6 5 3 2 4  H 
3 TS
2 TS 1 1
1995 19 8 11 5 3 4H  3TS 1 6 2

I  do not agree with some of the analog years that Dr Gray has selected in his  MAY 30  report / forecast. The years of   1952  1966  and 1998  are NOT close matches to the current  La Nina  conditions that are developing.  While  1998  summer / fall was a La Nina is was a strong one and came AFTER  the strong  1997-1998  El Nino of the precedding winter.  The summer of 1966 was a warm  weak  El Nino  and 1952 was a Neutral  season.  1954 and 1964  are fairly close but those were  NEUTRAL winters   that preceeded those Hurricane seasons.



june2003.jpg (169250 bytes)   july2003.jpg (169140 bytes)  august.jpg (212720 bytes)

In my selection of the 1995 and 1988 La Nina as the closest analog years I am NOT  forecasting a severe Heat wave or a to extreme drought which those particular analog years featured. So let's be clear about that point. I am however forecasting Below or Much Below normal precipitation and Above or Much Above normal temperatures over good portion of the Plains  and Midwest  and a cool wet SUMMER over the Great Lakes and Northeast.

JUNE 2003: It appears that latest data continuous to show that there will be a pronounced blocking feature ( known as the Greenland Block) still holding in position through the first week of June. This this means more below normal temperatures across the Northeast US Great Lakes and for much of the Midwest with very little opportunity for a sustainable heat to develop across these regions through the 10th or 12th of the month.

There will be numerous rain events but probably not the major Rains we saw in the middle of April to the middle of May. However with the below normal temperatures and frequent moderate rain events there may be more concerns about the corn and soybean emergence as opposed to any sort of planting delays.

TOURISM:  In the Northeast United States coastal communities will have problems as the continued below normal temperatures and excessive rain amounts may lead to reduced tourism and increase the threat of localized flooding and communities from Virginia to Massachusetts due to the saturated ground.

ENERGY: this sort of pattern means below normal CDD over the Eastern very United States and Above Normal CDD over the western U.S.. With the mean Trough position holding just off the West Coast... I also anticipate Above Normal rainfall and Below Normal temperatures across the immediate Coast of the Pacific Northwest. Inland however just across the Cascades eastern Washington state... Oregon... and Idaho will be warmer than Normal -- possibly significantly warmer than normal.


JULY 2003

GRAINS: Assuming that the moderate La Nina continues to develop in the month of June.... JULY 2003 should see a classic La Nina pattern set up in the jet stream. I expect to see a major reduction in the amount of Rain across the Plains and the Midwest with Above and Much Above normal temperatures... along with Increase crop and cattle stress. In terms of actual DROUGHT per se much will depend on how much beneficial rain fell in the SPRING 2003 and the month of June. While this SPRING has been quite wet the long-term Soil moisture is going to be a problem and two or three wet months do not replace the moisture which is the missing from two or three dry years.


ENERGY: CDD degree days will be Above normal in the areas under the Ridge in the nations midsection and I continue to see Below normal CDD across the Northeast where the mean trough of the located in the Ohio Valley.

AUGUST 2003: at this point August 2003 continues to appear as a hot dry month for the western portions of the Midwest and most of the Plains. The mean trough position will be somewhere between the Ohio Valley and the East Coast and as a result the Ohio Valley areas may be significantly cooler than the rest of the Midwest. As the month wears on the mean trough position over the Northeast United States will break down and the middle and late August will probably be the hottest period in the summer 2003 for the Northeast states from Virginia to Maine -relative to normal.

TROPICAL ACTIVITY: like many La Nina years the CAPE VERDE  (East coast Hurricanes)  hurricane season will get also slow start with little tropical cyclone activity / development from the TROPICAL ATLANTIC region until the middle or late August. However with the mean trough located on or  just off the East Coast of United States this pattern would favor the development of systems that break off from cold fronts or Upper level low pressure areas that stall over the Bahamas or subtropical Atlantic region... and "transform" themselves into tropical cyclones.