Summer2006.html

Summer/Hurricane Forecast 2006

**Issued May 2nd 2006**

Welcome to yet another attempt at a seasonal forecast. My Winter forecast was good and bad and you can see the verification for that on the main page. Last Summer's forecast wasn't good, but the hurricane blurb was. This year I will issue a more specific hurricane forecast with reasons why I think things will pan out the way they will. The format will be to discuss different seasonal indices along with other factors and to try and relate them to how I think they will impact the summer weather across the US and the hurricane season. At the end of the discussion I will post monthly temperature and precipitation maps along with my bottom line Hurricane numbers and summary.

I think there is a market for long range/seasonal forecasting, and I think that as we begin to understand the interaction of different climate indices and discover new ones, the art of long range forecasting will continue to improve and become more grounded in solid science. As of now, it is a mix of statistics, limited understanding of the physical impacts of different indices to the global weather pattern, and art. But it is a worthy endeavor, one that some have become quite skilled at for different reasons, and it is alot of fun and educational as well. So off we go.

ENSO

Well I begin every seasonal forecast with an examination of the ENSO phase. Throughout much of the Winter the phase was a borderline negative-neutral/weak La Nina and as of March, the MEI came in at a value of -0.554 which was the lowest monthly reading since February 2001, and the region 3.4 anomaly for March was -0.6. However there are signs this cool phase has peaked and warming is occurring. Over the last few weeks region 3.4 has warmed considerably to a reading now of +0.1. Of course to add to the confusion region 1.2 has cooled over the last couple of weeks back down now to -1.4.

Current SST image.

However, if you look at subsurface temperatures in the equatorial Pacific:

It could be argued that the colder than normal waters are now confined to region 1.2 and are relatively shallow. While warmer water exists just below the surface throughout much of the central and western equatorial Pacific. This leads me to believe that there is a good chance that we will see conditions in at least a neutral phase for this summer and hurricane season. I don't think conditions will develop into El Nino conditions during the Summer, at least not right now. The available computer model forecasts for ENSO tend to agree with this assertion, holding off any El Nino formation until late Summer at the earliest, and remember the models tend to develop an El Nino too quickly. So my assumption for the Summer and Hurricane season of 2006 is that ENSO will be in a neutral phase. If one were to take a very general look at summers where ENSO conditions were neutral the west was warm for both neutral-negative and neutral-positive summers and it was wettest across the southern plains, Mississippi Valley, and Ohio Valley:

Neutral Negative Summer Temps

Neutral Negative Summer Precip

Neutral Positive Summer Temps

Neutral Positive Summer Precip

and another point to make is that with ENSO being in a neutral phase ENSO phase should not serve as an inhibiting factor on the 2006 Atlantic Hurricane Season. In fact when looking at hurricane seasons since 1948, the top 12 hurricane seasons with respect to # of named storms have all occurred when the (June,July,August,Sept) MEI averaged out to a neutral value (between -0.5 and +0.5). Nine of those twelve came with the MEI averaged between 0 and +0.5. I also must mention that the correlation between important atmospheric variables and ENSO conditions are weaker in the Summer then in the Winter, so there is more uncertainty when using ENSO for summer forecasting then with winter forecasting and it is harder to state a definitive physical cause for the analog composites.

PDO

The PDO switched to positive in the late Fall and was that way through the Winter and through the early Spring. However, the March number fell to 0.05 and the Pacific SST's look very much like a negative PDO right now, perhaps a strong one. Now, even though this is a multi-decadal oscillation, there can be significant short term changes, and it looks to me that there will be a -PDO for at least the early Summer and perhaps through the rest of the Summer and into the early Fall. So my assumption is that the PDO will at least be negative through the 1st half of the Summer, beyond this is difficult to predict, but I think it will return to neutral or weakly positive by late Summer into early Fall based off my interpretation that we are still in an overall warm phase of the PDO and based off the behavior of the PDO in this warm phase which has been underway since the Fall of 1991. It should be noted however there was an extended period of -PDO values particularly during the La Nina period from mid 1998 to mid 2000. If one were to look at summers that averaged out to have a moderate or weak negative PDO, again we see more warmth in the West when compared to the central and eastern US.

Weak Negative PDO Summer Temps

Moderate Negative PDO Summer Temps

Moderate Negative PDO Summer Temp Anomalies

Atlantic SSTs.

If one were to look at the correlation of 500mb heights with different climate indices, one would actually see that different Atlantic SST indices seem to have a higher degree of correlation than other more well know indices for the summer. Take for instance the TNA (Tropical North Atlantic Index) correlation to 500mb heights shown below. The TNA is a measure of the SST anomalies from 5.5N to 23.5N and 15W to 57.5W:

Keep in mind a correlation of 0.3 or higher is considered statistically important. Notice the correlation with heights across the Plains and Rockies, showing that in general when the TNA is positive, heights there will be higher. Also the WHWP, (Western Hemispheric Warm Pool) which is a measure of the area of ocean surface warmer than 28.5C (83F) in the Atlantic and eastern North Pacific, shows a correlation with 500mb heights over the southern US and the AMO, (Atlantic Multidecadal Oscillation) which measures SST anomalies in the Atlantic basin north of the equator, shows a correlation with a large area of the country outside of the western US and NE US.

WHWP 500mb Height Correlation:

WHWP June/Sept Correlation with 500mb Heights

AMO 500mb Height Correlation:

AMO June/Sept Correlation with 500mb Heights

Although the Atlantic has cooled with respect to anomalies the past couple of months and is cooler than the last few years at this time, the prime hurricane genesis regions (MDR, Caribbean Sea, and Gulf of Mexico) remain above normal, and I have no reason to think that they will not continue to remain above normal through the Summer and into the hurricane season, which also has implications with the hurricane forecast talked about later. When looking at values for the AMO and TNA indices, these indices recorded their highest values on record last year. And of course, which should have not come as a surprise considering ENSO was in a neutral state, the Atlantic hurricane season was the most active in recorded history. As of now I do not think that the Atlantic will average this warm again but will still be above normal.

QBO

As with ENSO the summer time correlation of the QBO to important atmospheric variables is not as great as it is in the Winter. However I do believe that physically this oscillation does have an effect on the upper tropospheric winds and thus should still be considered when one does a summer forecast although possibly not as great as other parameters. Also it has been noted by some that a westerly(positive) phase of the QBO is more favorable for hurricane formation due to relaxed shear in the Atlantic tropics. The QBO was strongly easterly (negative) during the Fall and early Winter but has quickly turned around to where it is now near 0 as of March and appears to be on its way to a moderate to strong positive (westerly) phase for the upcoming Summer and Hurricane season. A general look at Summers that averaged strongly westerly, again show heat for the western US and less heat for the central and eastern US.

Strong Westerly QBO Summer Temps

Summer Forecast

It is in my opinion that the met summer (June, July, August) of 2006 will feature the warmest weather across the western US. It will begin with the heat centered in the west while normal to below normal temps and rainier than normal weather will be present across portions of the central US and Ohio Valley into the NE US. As the summer progresses the heat will spread out particularly in the northern US but the west looks to stay warm the whole summer. By late in the Summer August/September most of the country will be normal to above normal except across the SE US, where I think tropical systems could play a role in keeping temperatures down overall and resulting in above average rainfall. Below is the monthly breakdown with temperatures and precip maps. So to sum up, the western US looks to be warm this summer, and the heat will spread across the northern US into the NE US as summer progresses, with August/September being the warmest months for those locations. The southern plains look to get drought relief particularly early in the summer, and tropical systems could play havoc across the SE US in August and September. I list below my main and secondary analogs. It should be noted that this forecast is not based solely off of these analogs, but a blend of them and also my own understanding and interpretation of the climate indices and my thoughts on how the pattern will play out. As shown above, most of the composites for different climate indices, point to a warmer western US when compared to the eastern US and wet first half of the summer for the parts of the central US and Ohio valley.

June Temps/Precip

July Temps/Precip

August Temps/Precip

September Temps/Precip

Main analogs:

1961
1981
1984
1996
2001
2004

Secondary analogs:
1960
1985
1952

Others looked at, but not weighted too heavily:
1986
1953
1990
1959

What Could Go Wrong?

Well when you are dealing with seasonal forecasting, plenty. But my main concerns, are:

1.) El Nino sets in earlier than I anticipate. Many SST models are showing an El Nino forming by late Fall and Winter. If it sets in earlier, that could screw up the ENSO forecast, which is a large driver of my thoughts.
2.) The Atlantic cools off. Late in the winter and into the Spring, the Atlantic SSTs have been cooling somewhat. If they cool further and/or don't rebound, this could be a warning sign that my forecast is off.
3.) The PDO switches to positive quicker than I think. The PDO looks to be in a very negative regime right now, if it goes positive, it would ruin my assumption.
4.) The southern plains drought is too much too overcome and it stays dryer there, then I forecast.

Hurricane Forecast

The reasoning for my hurricane prediction could be gathered from above. I believe that the climate indices are lining up for yet another active Atlantic hurricane season.

1. The MEI is forecast by me to be in a neutral state this Summer and early Fall. When one looks at hurricane seasons since 1950 there have been 10 seasons with 14 or more named storms, if I am counting right, 2005, 2004, 2003, 2001, 2000, 1998, 1995, 1990, 1969, 1953. In all of those 10 seasons, the MEI averaged for the June-Sept period to be neutral (between -0.5 and 0.5), with 9 of the 10 seasons being between 0 and 0.5. Below is a break down of each ENSO state determined by the imperfect method above:

ENSO State	# Storms
El Nino (June/Sept MEI average >= 0.50)	7.9
La Nina (June/Sept MEI average <= -0.5)	10.3
Neutral Negative (June/Sept MEI average between 0 and -0.5)	10.9
Neutral Positive (June/Sept MEI average between 0 and 0.5)	16.2

You can clearly see that statistically, a neutral ENSO state, particularly a neutral-positive one, corresponds with an active Atlantic Hurricane season. It is unclear whether or not the June-Sept period this year will average neutral-negative or neutral-positive, but my thought is that by August and September it will be in this state.

2. The Atlantic is forecast by me to be in a above normal state again this hurricane season, although not as warm as last year. It should be noted that in the above 10 seasons mentioned, that all of them featured an Atlantic that averaged out above normal for the June-Sept period. When the TNA has averaged positive for the June-Sept period the average # of storms since 1948 has been 11.8 and when negative it has been 9.4. When the TNA averaged positive and ENSO averaged neutral the average # of storms has been 13.7. The numbers are similar for the AMO, when positive an average of 12.1 storms, and when negative an average of 8.9 named storms, and when it is positive and MEI is neutral the average # has been 14.1 storms.

3. The QBO is forecast by me to be strongly positive(westerly) this hurricane season. There is a slight increase in named storms when the QBO for the June-Sept period averaged positive(westerly). Since 1948 the hurricane seasons that averaged a westerly QBO saw 11.8 named storms and those that were easterly saw 10.1 named storms. And those that averaged a value of 10 or greater (1997, 1999, 1990, 1985, 1995, 2002, 1980) saw an average of 12.4 storms. If you take out the strong El Nino of 1997 that average is 13.2.

So with a neutral ENSO, warmer than average Atlantic, and statistically favorable QBO phase, I do think we will see an above average hurricane season.

Average # of tropical storms/hurricane since 1948: 10.7
2006 Prediction # named storms: 16

Primary Analogs:
2004
1990
1961
Secondary Analogs:
1981
1996
2001
1960
1952
1953

Although it is fun to try and predict the number of hurricanes and major hurricanes and some have had success doing it, I will refrain this year.

Landfall Ideas

Predicting landfalls of tropical systems months in advance does seem a little silly, but I think we can point out potential danger zones if one assumes a certain upper air pattern will be present and also by looking at hurricane analogs. I will not use the analogs primarily for my threats, but as guidance along with my thoughts of how the upper pattern will evolve. My thoughts of the upper air pattern for the US could be inferred by looking at the surface temperature and precipitation forecasts. I think there is potential for a central or eastern gulf coast tropical system in June, as I think a trough will be centered across the Mississippi valley area. In my analog years above 2001 featured this (Allison) as did 1960 (Tropical Storm #1), and 1953 (Alice). As the season progresses I think the threat shifts east to Florida and the SE coast. Now keep in mind that several of the above analog years featured busy gulf seasons (2004, 1960, 1953), and I think the Gulf doesn't escape a landfall this year during the prime hurricane months. But I think the greatest threat this year lies for Florida and the SE coast (Ga,SC,NC). As we get into late Summer and early Fall, I think we could see a negative NAO which could prevent recurving storms from turning out to sea. I should also point out that many of the above analog years featured tropical systems effecting and threatening the NE US in at least a minimal way. Examples were (TS Hermine (2004), Hurricane Esther (1961), Hurricane Bertha (1996), Hurricane Fran (1996), Tropical Storm Brenda (1960), Hurricane Donna (1960), Hurricane Able (1952), and Hurricanes Barbara and Carol (1953). Again this is highly experimental, but it is just my thoughts for landfall danger zones, we shall see if they are valid.