- SUMMER RECAP -
This past summer featured an unusually high amount of severe weather activity across the Midwest, Great Lakes, Ohio Valley, and Northeast.
With a strong La Nina in place this past winter and going into spring, the severe weather / tornado season started off early.
An unusually early outbreak (even for a La Nina year) dropped 72 confirmed twisters on January 7-10. With that, we were off to the races with the Super Tuesday Outbreak (February 5-6) followed by significant outbreaks April 8-11, May 1-2, and then a series of tornadic systems May 7-15 and 22-31 ran the total up to 679 by the end of May.
As of July 31st, so far 1147 tornadoes have been confirmed this year in the continental United States. Compare this to 975 in 2007, 907 in 2006, 850 in 2005, 1155 in 2004, and 1369 in 2003. As the graphic shows below, the 2008 Jan-Jul total does not include preliminary reports which are in the process of being filtered through and added to the confirmation list. Once this is done, 2008 will likely exceed 2004 and 2003.
While frequent rounds of thunderstorms scoured the midwestern states, June featured the first heat wave of the season for parts of New England and in general a hot start to summer for areas east of the Mississippi River. The stormy pattern over central states kept temperatures below normal over the northern plains and produced some large precipitation anomalies.
This all changed as the pattern previously locked into the midwest shifted east. This kept an upper level low grounded over SE Canada July through mid August, funneling in frequent distubances and providing an anomalously cold air mass aloft to fuel thunderstorms, some severe, even when surface temperatures struggled to reach the 70's! While temperatures still reached above average for July supported by warm spells 7th through the 9th and the 18th and 19th, precipitation anomalies were notably larger over New England due to numerous thunderstorm days (5 in just 10 days between the 19th and 28th at KEEN).
Despite a resurgence of warm and dry weather in late August, much of the northeast ended with below normal temperatures and above normal precipitation. I'm sure we all remember the numerous cloudy days. The west continued high and dry, while the southeast ended very wet, largely due to tropical activity late in the month.
--- PACIFIC ---
Sea surface temperatures have been warming through much of the summer in the eastern Pacific in the Nino1+2 and east portion of Nino 3. Warm anomalies have spread westward however are low magnitude. Meanwhile, cold anomalies have held on in the central Pacific, Nino4. Forecasts for the upcoming winter cannot use trivial indices, such as the Nino 3.4 region because of the notable temperature gradient that is lingering and overall SSTAs really averaging neutral, however characterizing the current state of SSTA in the equatorial Pacific as "neutral" does not cut the cake either. The Spring and Summer temperature gradient along the Nino region of the Pacific has been the largest since 1998. It peaked in March before falling to just below 1.5 for JJA and the latest weekly SST index values return a 1.1 gradient. The eight years that exhibit the same high amplitude gradient, lean towards the El Nino and +PDO SSTA configuration in the Pacific, neither of which fit the current state but give some insight into at least some potential with regards to NPAC SSTAs. The last few months has featured a resurgence in the SOI (Southern Oscillation Index) which correlates to cooling in the equatorial Pacific. Latest depth analysis in the area show negative anomalies around 100m to 150m and propagating east and toward the surface. This may act to subdue eastern Pacific warm anomalies by later in the autumn. Although I am now confident in a negative ENSO through the end of 2008, I still do not anticipate an ENSO phase with magnitude greater than .5 through the winter. Regardless of the precise magnitude of ENSO this winter, not much change in the placement of forcing in the equatorial Pacific is likely.
Taking a closer look at mesoscale processes in the tropics, satellite imagery shows convection is now steadily shifting eastward over the Maritime continent. The MJO (Madden Julian Oscillation) phase space diagrams indicate wave progression into octant 5 as of 9/12. This is after a long lingering in stage 2 which influenced the burst of tropical activity in the last week of August. The signal is growing stronger with nearly a 2 sigma projection. The latest MJO wave is being amplified by constructive interference by phase 4 of the GWO (Global Wind Oscillation). Main question is whether it can progress into octant 6 which is currently not supported by the forecast models. At this point it appears it will pass through the <1 sigma region before emerging in phase 8 by the start of October. The continued positive phase of the IOD (Indian Ocean Dipole) should help keep the MJO wave at least somewhat apparent as it progresses into the western hemisphere.
In the northern Pacific, the increasing -PDO signal (-1.67 in July and -1.70 in August: lowest August PDO since 1955 and lowest overall since 2000) looks like it will hold on through the next 6 months. However, the latest satellite SSTA analysis indicates warm anomalies have shifted westward notably since August, from being centered around the dateline to now around 150E. While this still continues a -PDO, convective forcing is shifted west and may support an RNA pattern going into the winter with a +PNA developing as a result of growing wavelengths.
--- SPACE WEATHER ---
- This section involves all effects on our weather that take place outside of the troposphere -
The QBO (Quasi Biennial Oscillation) is a stratospheric wind anomaly that changes phases about once every 14 months. An easterly (negative) QBO can lead to a displacement / weakening of the polar vortex and correlates with colder winters, while a postive (westerly) QBO is the opposite. The QBO is in the positive state but ppears to have hit its peak in July and now a -dQBO/dt has commenced. At this point it looks like it will remain positive through at least December and January and Stratospheric temperatures remain below normal in the subtropics. The positive QBO does not automatically translate to a +AO (Arctic Oscillation) winter, but it limits our chances for any disruption of the polar vortex and -AO periods.
Moving even further away from the troposphere, the solar flux, a function of the sunspot number, continued its minimum state through the summer. It was in fact the lowest recorded in the post-1948 data with a recent 52 day sunspotless streak and August being the first month without a sunspot since 1913. It will surely go into winter on the low end as it begins Cycle 24 (a bit delayed). A low solar flux correlates with enhanced storminess and below normal heights over the midwest and eastern US. Some examples of past winters featuring a solar minimum are 1995-96, 1986-87, and the infamous late 1970's and early 1960's. All of these winters also featured a negative NAO (North Atlantic Oscillation) and above normal snowfall in the east.
--- ATLANTIC and POLAR ---
The two variables discussed in this section have been aluded to above. They are the AO (Artic Oscillation) and NAO (North Atlantic Oscillation). These two indices are a function of high latitude blocking, and how amplified a pattern is. The negative phase of the NAO corresponds to increased heights over Greendland/Iceland and lower than normal heights near the Azores region while the negative phase of the AO corresponds to increased heights around the north pole and a displacement of the polar vortex. The domain of the NAO and AO overlap and therefore frequently are in phase. The negative phases of the NAO and AO have a significant correlation to below normal temperatures and above normal storminess in the eastern US. However, into central and northern New England, they matter less. I have developed a method of predicting the average wintertime NAO using a combination of SST anomalies in the Atlantic, and OLR anomalies in equatorial South America.
- SST anomalies
SSTAs have a strong influence on the distribution of geopotential height anomalies. Because of this, they play a significant role in seasonal forecasts, as can be shown with ENSO, PDO, etc. The NAO, being an Atlantic teleconnection, is influenced by Atlantic SSTAs. Due to the nature of the thermohaline circulation, these anomalies can be traced back in time, providing the ability to make rough prediction of the wintertime NAO. The two focus areas are the northern portion of the Gulf Stream, and the Greenland current. Using correlation tests, nodes are found in MJJ (May, June, July) at 38-43N, 68-40W (Area A); and 52-68N, 70-45W (Area B). The combination of the two (Area A - Area B) results in a correlation of .53 to the DJF NAO and a standard deviation of .53.
- OLR anomalies
I was originally intrigued by some research done on long term global OLR (Outgoing Longwave Radiation) anomalies and how they correlated with long term oscillations in the PDO and NAO. Looking at the distribution of OLRA across the Earth, it becomes evident that anomalies near the poles are lower magnitude but represent the global anomaly, while equatorial anomalies are concentrated higher magnitude nodes. The main area that I focused my attention on was equatorial South America, which has higher OLR variations and negatively correlates to the global OLRA. A negative OLRA in that region, indicative of tropical forcing, would result in a wave train extending northward leading to time delayed positive NAO height anomalies, and vice versa. Using October anomalies from 90-65W, 5N - 10S, the proceeding DJF NAO has a correlation of -.61, and using a linear regression, has a standard deviation of just .43.
Making small adjustments using Autumnal teleconnection states, the forecast method results in a .71 correlation to the DJF NAO and a standard deviation of .39.
Alright, so how is it looking this year? The MJJ Area A anomaly was -.018 and the Area B anomaly was .141 yielding a preliminary -.070. The South America OLR anomaly has been rising since early summer with the August value at -8.1 which as is would yield .636. However, with the MJO wave in octant 5, phases 6 - 8 should occur through week-2 (start of October) which would support positive anomalies between 60W and 90W. Using an extrapolated estimate of .5, the final NAO prediction would be .26. This is just under the DJF NAO values in the previous two years.
- SENSIBLE WEATHER FORECAST FOR NEW ENGLAND -
...ANALOGS...
Using mainly the current and expected state of equatorial SSTA distribution, the two most prominent analogs that come up are 1998-99 and 1956-57.
DECEMBER
JANUARY
FEBRUARY
...THE FORECAST...
Using these analogs and other parameters that I have gone over, this is my monthly temperature forecast:
After a period of below normal temperatures in mid October, the first 2 to 3 weeks of November will be warm with likely no snowfall. Our best shot for early season snow will come in late November (23 - 26) with a NAO minimum. A fairly inactive December will follow. As much as I do not see temperatures straying much from normal, there arent any blaring signals for snow events. Precipitation will be at or below normal and any shots of cold will be transient. However by the end of the month, maybe just in time for Christmas, the NAO will take a tumble and with that we could have a Miller B type storm. Colder than normal temperatures should stick around through the first half of Janaury. By mid January (15 - 18) the PNA should fall back below zero while the -NAO lingers at least a little longer. This could be a recipe for a significant Miller A storm track and a notable snow storm from DCA north. Could get interesting in my opinion. However, I dont see much promise after that through most of February. A major thaw period in the east with our main snow chances similar to last year with front end dumps before a change to rain, and even that isnt looking too promising right now. March is obviously a part of winter, and a contributor to the season snowfall totals, but I cant say much with confidence for what to expect this time. Could be a pattern switch going into the month allowing enough cold to support a couple snow events.
No comments:
Post a Comment