9/5/17

The Ongoing Active Tropical 2017 Season Begs the Question: Hurricanes in Michigan??? Along with...Hurroncane!

 Hurricanes in Michigan??? Along with...Hurroncane!


Originally written September 1999 & Updated September 2004/2012/2013 and with more the latest reanalysis information - 2016. And in 2017; all tropical remnants listed in Table-1 have links to the paths of the storms along with attending narratives back nearly a century to 1923!

Monday, September 25th is the 76th anniversary of one of the most unusual and intense wind storms ever to hit Southeast Lower Michigan. Well over seven decades ago on September 25th, 1941; the remnants of a tropical storm hit the region. This storm was not that usual being in that the remnants of tropical systems do occasionally make it up to the Great Lakes region. But, the coming together of all the meteorological components made for this Astonishing Storm of September 25th, 1941

At first glance, when one reads that headline, one might say, "What? Hurricanes here in the Great Lakes?? No way!" Of course you'd be right, no actual hurricane has ever been observed in Michigan under the true definition of a hurricane. The definition of a hurricane, according to the Glossary of Weather and Climate edited by Ira W. Geer, is as follows: "A severe tropical cyclone with maximum 1-minute sustained surface wind speed greater than 64 knots (74 mph) in the North Atlantic Ocean, Caribbean Sea, Gulf of Mexico and the Eastern North Pacific off the west coast of Mexico to the International Dateline. West of the Dateline they are known as typhoons." Furthermore, the definition of a tropical cyclone is as follows: "A generic term for a non-frontal synoptic-scale cyclone originating over the tropical or subtropical waters with organized convection and a definite cyclonic surface wind circulation." Clearly, neither definition applies in the Great Lakes area, although remnants of hurricanes that have become extra-tropical (loses its tropical characteristics) occasionally do make their way into the Great Lakes region.

Scanning a century's (since 1917) worth of hurricane track data suggests that remnants of a hurricane or tropical storm make their way into the Great Lakes region on an average of nearly twice a decade, especially the southern Great Lakes area (see Table-1). Also, in the majority of instances, by the time they visited this region they diminished to an area of rain with maybe some squally winds with many of the rainfalls/winds on the tail-end of their "lives".. There have been a few instances, along the way, however, that do bear mentioning, and ONE STORM in particular that screams for attention!

TABLE - 1

REMNANTS OF HURRICANES/TROPICAL STORMS THAT HAVE AFFECTED THE GREAT LAKES AND SOUTHEAST LOWER MICHIGAN (SINCE 1921)
DATE (Storm's life cycle)
DETROIT RAINFALL/DATES (Inches)
MAXIMUM WIND (mph)

.55 - 10/18
SE - 22

1.11 - 8/1
E - 25

2.95 - 9/3-4
NE - 21

.23 - 9/21-22
SW - 20

Trace - 9/25
SW - 52 *

.82 - 9/6-8
N - 25

1.30 - 10/6-7
SW - 21
.69 - 8/13-14
W - 20
.31 - 6/28
SW - 23
.13 - 6/14
W - 29
2.55 - 6/24-26
NE - 30
Trace - 9/22-23
NW - 38
1.41 - 10/5-6
N - 38
.99 - 9/7
NW - 25
0.25 - 9/18-19 #
W - 33
            Ike - 9/1-9/15 2008                                 3.78 - 9/13-14                                           N - 43

* denotes officially at Detroit City Airport, but gusts were clocked up to 75 mph across the Metro      Detroit area.
** TD - Tropical Storm
# Up to 2.50" reported along the St. Clair River
$ 1938 New England Hurricane: Fast forward speed reached 70 mph on September 20th, making it the fastest-moving Atlantic tropical cyclone on record.

>>Note; Hurricane Sandy's path /2012/ did not pass over/near Southeast Lower Michigan<<

First off, under the "mention" category in 1932 (well before hurricanes were named), a hurricane that developed in the Caribbean on August 27th, tracked northwest into the Gulf of Mexico, then generally north, across Alabama, eastern Arkansas, southeast Missouri, into southern Illinois and then from there, headed northeast across Indiana into Southeast Lower Michigan. The storm approached Southeast Lower Michigan late on September 3rd. A light to moderate northeast wind proceeded the storm on the 2nd and 3rd averaging around 11 mph with gusts into the lower 20s, hardly anything noteworthy wind-wise. The storm did however, pass right over Detroit, causing the barometric pressure to fall from around 30.20 inches early on the 2nd to around 29.60 late on the 3rd. Rain began to fall lightly but steadily early in the morning on the 3rd, but from mid afternoon into the evening, moderate to heavy rain fell and by midnight, over two and a half (2.55) inches was dumped on the Detroit area. Close to an additional half inch or so of rain was added to that on the 4th for a total of nearly three inches.

Another mentionable was odd from the start because of its timing; occurring very early in the season during late June of 1968 (only one other June system was found since 1921 to have affected the Great Lakes: the weak remnants of a tropical storm that moved from the Gulf of Mexico to near Chicago, June 22-28th, 1960). On June 22, 1968, Tropical Storm Candy formed off the coast of eastern Mexico and then headed north into southeast Texas, just north of Corpus Christi. She then weakened, headed north-northeast through eastern Texas and Oklahoma, central Missouri and Illinois, then she pivoted on a more easterly track across northern Indiana into extreme northwest Ohio, over Toledo. Candy began to influence Southeast Lower Michigan's weather on the morning of the 25th. A nearly steady rain, interspersed with a few thunderstorms, continued through the day and evening, depositing nearly two and a quarter (2.17) inches. More scattered, lighter showers fell on the 26th, adding another .38 to give a grand total of 2.55. Through it all, an east to northeast wind blew averaging 10 to 14 mph with gusts into the 20 to 30 mph range.

During the busy mid 1990s; a fairly impressive tropical system wound its way north out of the Gulf of Mexico, through Alabama, Tennessee, Kentucky and into Northern Ohio. She was known as Hurricane Opal and had a 10-day life span from September 27th to October 6th, 1995. On the evening of Oct 5th at 8pm, the remnants of Opal were located over the eastern Great Lakes. What was left of the "eye", or center of the storm, was well indicated on the NWS Doppler radar in White Lake (DTX). The radar showed the "eye" over Sandusky, Ohio moving north out over western Lake Erie. Light to moderate rain, with isolated areas of heavy rain, extended from Southeast Lower Michigan east across Southwest Ontario and Northeast Ohio. Spotty light rain first made an appearance in Southeast Lower Michigan earlier in the day, during the forenoon hours. During the afternoon, steadier rainfall developed with the heaviest rain (.54) falling between 5pm and 7pm. The rain ended just after midnight with a total of 1.41 inches falling from the storm. With the approach of the storm on the 5th, a generally north wind picked up and averaged over 15 mph with gusts up into the 30s (peak wind gust was 38 mph)

One of the more recent Hurricane remnants to affect Southeast Lower Michigan was Isabel, which moved quickly through the Eastern Great Lakes on September 18-19th, 2003. Rainfall from the system was quite variable ranging from just a trace over far western areas of Southeast Michigan, to as much as 2 1/2 inches at local spots along the St. Clair River. In 2008, remnants of Hurricane Ike September 13-14th brought very heavy rains to the region along with strong winds and a brief EF2 over western Wayne County on the 13th. The EF2 tornado struck Plymouth Township and damaged an apartment complex, flipped over cars and knocked out power to several hundred residents.

The Astonishing Storm of September 25th, 1941

Earlier, I mentioned "one storm that screams for attention", but maybe "howls" would be more appropriate. Before researching this project, I expected to find the above case scenarios, but not the following...

 Recently /2013 changes/ the Hurricane Research Division /HRD/ did a reanalysis of the hurricane and its associated track and wind speeds. *Note the asterisks in the following storm account are adjusted for those changes. (Thanks to Hoosier, moderator of American Weather for the heads up on the reanalysis)
   
* The storm made landfall in Texas as a category 3

* After moving onshore and through its entire remaining path, the storm was upgraded by maintaining tropical force winds as she evolved from a tropical storm to extra-tropical.. Also note the shifted, slightly westward track depicted over upper Ohio Valley and Lower Great Lakes in Indiana and Michigan.

Comparing the two tracks;

Original


  Reanalyzed


A tropical storm (#2 on original analysis map & #17-under 19-21, unseen on reanalyzed) formed in mid September over the Gulf of Mexico on September 17th, 1941. As it formed it meandered over the Gulf, making a loop in its track, well south of New Orleans. By this time it was a hurricane, intensifying briefly to a category #3 storm (111-130 mph wind) offshore as it took aim on eastern Texas. The hurricane made landfall late on the 23rd near Freeport, Texas with an estimated wind of a category #3* (111-130 mph), extremely high tides of nearly 11 feet and a barometer reading of 28.31 inches (959 MB). Further to the northeast, a ship just offshore of Texas City recorded a lowest pressure of 28.66 inches and winds of 83 mph. Other wind gusts were estimated near 100 mph at several locations near the hurricane's center along the Texas Gulf Coast. The hurricane quickly weakened to a category #1 (74-95 mph) as it made landfall and by the time the storm pushed on north to Houston, wind gusts had already dropped to 75 mph. Four lives were lost from the storm in Texas and property damage was estimated at $6.5 million (1941 dollars).

The storm continued to roar on its northward path through Texas and by 7 am on the 24th, it was located near the city of Tyler, over extreme northeast Texas. From this point on, the storm's last 36 hours or so really grabs ones interest for peculiarity as it tracked through the Mississippi Valley and on into the Great Lakes. From 7AM on the 24th to 7am the 25th, the storm shot northeast from Tyler to near Battle Creek, Michigan, covering close to 1000 miles in 24 hours! Thus, the forward movement of the remnants of the hurricane averaged 40 to 45 mph as it approached Southern Lower Michigan. As the storm tracked into the Great Lakes, it merged with a fairly strong cold front that pushed across the upper Midwest into the Lakes. The combination of the strong push of cool fall air, strong upper level dynamics and the remnants of the hurricane created quite a storm (not unlike the more common intense late fall cyclones that are seen in the Great Lakes). In addition, the track and speed of our "hurricane" brings to mind that of the "Panhandle Low" type of low pressure system in the winter (more information). While the speed of the system was fairly quick, it's not uncommon for hurricanes to accelerate northeast as they become extra-tropical and get "picked up" by the mid-latitude upper winds or jet stream. Yet, what was really unusual and noteworthy was the surface wind that accompanied the storm as it moved through the Great Lakes. By the time hurricanes make it this far north, they usually have blown themselves out, at least to the extent that surface winds are only gusting to, at best, 30 or 40 mph. Note the following, taken from the Detroit weather records on September 25th, 1941:

Windstorm: An intense tropical cyclone moving up from the Gulf thru eastern Texas (causing great damage in Texas), along the Missip. Valley and thence Newd across Ill & Mich, passing W & NW of Detroit with gale force winds and gusts to 65 mph from 10:18 AM - 2:30 PM & gusts to 75 mph 12:30 PM - 2:00 PM (see envelope back of book for newspaper clippings).

Most of the above noted news clippings show widespread wind damage to trees and power lines that would be commonplace in severe thunderstorms or a derecho. A derecho is a widespread windstorm consisting of a complex of thunderstorms that develop into a long-lived squall line. But there also were some unusual or freakish happenings (as the Detroit Free Press deemed them) as well. The following is taken September 26th, 1941, the day after, from the Detroit Free Press about the storm:

"River goes dry"

There were many freakish effects of the wind, including baring of the
Detroit River "middle grounds" off Belle Isle when water was backed into Lake St. Clair. The southwest gale literally blew the water out of The Detroit River, reducing its level by three feet, and leaving hundreds of pleasure craft high and dry on the muddy bottom. Several yachts broke their mooring or were heeled over at the Detroit Yacht Club. Another odd effect was the noticeable swaying of Downtown skyscrapers as the full force of the gale struck. Office employees who left tall downtown skyscrapers, were later reassured by engineers.

The Free Press goes on to say that "shortly after noon, the wind blew steadily at 56 miles an hour, but a times gusts reached hurricane velocity of 75 miles an hour." Dozens of people were injured by falling glass from windows blown out or debris tossed by the wind. One woman was literally blown into a fire hydrant, suffering a possible skull fracture. Other reports of scattered damage to homes and businesses across the region were mentioned in the article. In addition, the fierce wind churned up giant waves on the Lakes, including Lake Huron into the St. Clair River where two barges were blown ashore from of the shipping channel, even after dropping their heaviest anchors. In Southeast Lower Michigan, Storm Warnings were posted on Lakes Huron, Erie and St Clair at 10:30, the morning of the 25th. Downed telephone lines caused a disruption of service and communication across the Great Lakes and elsewhere. The "dying" hurricane left a trail of damage from Texas clear up into the Great Lakes and Canada. The wind of the storm was equated to an intense fall low pressure system that hit the area on November 29th, 1919 in which the wind blew 67 mph in Detroit and to the "Black Friday" storm in November of 1913.

The fact that the remnants of the hurricane, after weakening and becoming extra-tropical, traveled over a thousand miles, while sustaining an impressive amount of wind itself was very noteworthy. This, and the important addition of a relatively strong polar front traveling southeast at the same time across the Midwest, made for quite the enhancement and gave the additional punch and life to the storm's "tropical" characteristics. The surface map for the morning of the 25th is available (first and second map, below) and  depict quite nicely our extra-tropical storm racing northeast. This, combined with a rather potent cold front pushing southeast out of the Midwest, wrapped-up into the storm of tropical origins! This type of strong interaction is more likely observed on the East Coast. I inverted and then, enlarged the shading so the data could be seen better (click on all maps).





Being it was 1941, no upper air data/maps are available from the original time but a reanalysis of the day shows some interesting items (click on all maps). Note the strong (for the time of year) 500H /18kft/ trough digging into the upper Midwest. The 500H anomaly map below that map, depicts this nicely - the potency, depth and generally out of character of the trough for late September- very well. The last two maps are the Meridional (north/south) average wind at 500H and 850H /5kft/. Note the max-cores (depicted in meters/sec) over the eastern Lakes that is shown and the large upper trough over the Midwest at these levels in the 24 hour average (remembering that this is a reanalysis for the entire 24hour period - and thus, the average).

 



As the storm moved into Southern Lower Michigan, its center tracked northeast across Battle Creek, Lansing, Saginaw and then out over Lake Huron and into Ontario. Judging by the lowest pressure readings at Detroit (29.25 inches) and Flint (29.17 inches), where the wind gusted to 69 mph, its central pressure was estimated to around 29.10 inches (about 985 MB). Quite impressive for the remnants of a "dying" hurricane in the Great Lakes in September. In fact, this is the second lowest pressure reading ever recorded in Detroit during the month of September (the first being 29.21 inches on Sep 29th, 1966, during the passage of an intense early fall low pressure system).

One can only make a random guess as to the chances of another hurricane-force wind storm, from remnants of an actual hurricane, hitting the Great Lakes again. Since it was the only one of its kind in the record books at Detroit since records began in 1870, it may take several 100 years before another similar storm affects the region!

Hurroncane

One final, extremely interesting "hurricane" that affected the Great Lakes must be mentioned to make this article complete. While this storm was not from remnants of a tropical system, its development over Lake Huron had many uncanny likenesses to tropical systems...

The first likeness was its timing, forming over the Great Lakes right at the height of the typical hurricane season, September 11-15th, 1996. What started as a typical core-cold 500 MB low pressure system evolved into a warm-core system as it settled over the relatively warm waters of the Great Lakes, in particular, Lake Huron. The low pressure system actually had moved past Lake Huron but then retrograded, or was "drawn back", to the relatively warm waters of Lake Huron. (Similar to the tropics, the Great Lakes usually reach their warmest water temperatures late August into mid September.) The storm then deepened and intensified at the lower levels of the atmosphere compared to aloft, typical of a warm-core low. It is believed that the warm waters of Lake Huron and associated low level instability over the lake were, to a large extent, the major contributing factors in this storm's evolution. The storm went on to form a broad cyclonic circulation, including the "spiral bands and eye", typically seen in hurricanes! At one point, the cyclone produced tropical storm force winds (39 - 73 mph) and some of the spiral bands even had rainfall exceeding 10 cm (better than four inches), causing some flooding.

On satellite, the storm looked very much like the classic hurricane picture:



This "Hurroncane" reached its maximum intensity during the day on September 14th, when a central pressure of 29.34 inches (993 MB) was recorded in the late morning by a Lake Huron buoy that fortunately was positioned, at one point, in the "eye". By 2 PM, that "eye" measured close to 20 miles across and had a ring of tall convective clouds surrounding it, strongly resembling that of an "eye wall". The convective showers encircled the "eye" well out over 300 miles. As the "eye" moved to the southwest (retrograded), over the aforementioned buoy, the surface wind backed from west at close to 35 mph to the southeast, and then diminished to near 10 mph. After, the "eye" continued to track to the southwest, away from the buoy, and the surface wind backed further to the northeast, and briefly attained tropical storm force. A similar scenario but with varying wind speeds, would also be expected at the ocean's surface if a tropical system retrograded from northeast to the southwest overhead. In addition, the air temperature rose from 13°C (55°F) in the spiral shower bands, to near 18°C (64°F), which was also the lake temperature, in the clearing above the "eye". The storm weakened overnight as the surface lake temperature dropped 5°C (9°F) back into the mid 50s due to up-welling and mixing. The lower water temperature helped greatly in weakening the storm as a result of the lower latent heat supply.

For additional information on hurricanes, check in with the
National Hurricane Center.

Background on "Hurroncane" was provided by a paper entitled "Hurricane Huron" by Mr. Todd Miner of Pennsylvania State University along with Dr. Peter Sousounis, Mr. James Wallman and Dr. Greg Mann of the University of Michigan.


- Next up; Looking back at the Summer of  '17 along with trends of the Fall.

Making weather fun while we all learn,
Bill Deedler - SEMI_WeatherHistorian


6/24/17

Tale of Two Rainfall Patterns Emerging Thus Far across Southeast Lower Michigan This Summer of '17

Stark differences in rainfall amounts have begun the Summer of '17 across much of Michigan so far this June.

Flooding rains occurred over the Saginaw Valley north-northwest into Mid-Michigan this past week while further south across much of Southeast Lower Michigan (south of I-69); rainfalls remain mainly below normal, some areas well below for June. Driest areas and largest departures below normal extend from the Metro Detroit and Ann Arbor areas, south to the border with some areas seeing just 25% of the normal rain for the month.

A picture (or two) says a thousand words: Look at the difference in rainfall and departures across Lower Michigan, especially from Mid Michigan to the Indiana and Ohio Borders!




An almost unbelievable contrast in rainfall amounts can be seen, especially with the torrential rains that just occurred in mid-Michigan. Rainfall totals of 10 to 15" for the month have been observed (normal rainfalls lie around 3 inches for the entire month). Rainfall departures in these areas mount in upwards of 300+% of normal! The heaviest corridor of rain extends from Mid Michigan across Lake Michigan into a large portion Upper Michigan and Wisconsin. The run-off of the heavy rains should help keep Upper Great Lakes into the Lower Great Lakes levels high for the summer, if not add to them.


Latest forecast for the Great Lakes, Lake St Clair and adjoining rivers from the Corp of Engineers:


Many areas are several feet above low water datum and also above their long term averages.

Making weather fun while we all learn,
Bill Deedler -SEMI_WeatherHistorian

6/8/17

Summer 2017 Outlook for Southeast Lower Michigan (and a look back at the warm Winter of 2016-17)


SUMMER OUTLOOK 2017

 

Spring 2017 Brief Synopsis


As we move into the summer months; a distinctive yet tolerable cool trend has dominated the weather across Southeast Michigan during May into early June. Temperatures have basically averaged normal to slightly below - a degree or less since the start of May - yet the entire spring itself averaged above normal fueled by a warm April.

It was a fairly wet spring across the land, especially around Metro Detroit where all three months averaged above normal. Early to mid spring was generally the wettest elsewhere with a notable drying trend taking place later spring across the Saginaw Valley into the Flint and Thumb region.


 

SUMMER 2017

 

TEMPERATURES

I look for overall normal during the summer with any "heat waves or warm periods" to be routinely tempered by cooler air pushing south out of Canada. I'm looking for an occasional -NAO to help deliver the cooler weather as discussed in Hemispheric Patterns below and in the Analogue section. On specific temperature departures; I expect temperatures to average from +1.5 to -1.5 of normal. 

 

RAINFALL

Latest model and atmospheric trends agree with Summer Analogues for generally below normal rainfall this summer as heaviest rains generally fall west and southeast/east of the region.  Of course; there will be pockets of heavier rains with convective activity but overall, below normal rains are expected.

 


HEMISPHERIC DISCUSSION


SST AND ENSO'S PRESENT STATE
 
ENSO's neutral state should have minimal effect on our summer's pattern. The outlook for the summer /JJA/ is one of neutrality to persist with just a slight edge toward the weak El Nino side of the equation. So basically, very little effect even if a weak El Nino did develop later since the downwind affects are not only negligible but the lag time would push any influence into autumn.

Interestingly, as a side note, the Hurricane Center also see any possible weak El Nino in the Pacfic as a non-event as better formed El Nino's tend to suppress activity in the Caribbean and Atlantic. They actually see an above average season for hurricanes and tropical storms in spite of a weak El Nino:

“The outlook reflects our expectation of a weak or non-existent El Nino, near- or above-average sea-surface temperatures across the tropical Atlantic Ocean and Caribbean Sea, and average or weaker-than-average vertical wind shear in that same region,” said Gerry Bell, Ph.D., lead seasonal hurricane forecaster with NOAA’s Climate Prediction Center.


Strong El Ninos and wind shear typically suppress development of Atlantic hurricanes, so the prediction for weak conditions points to more hurricane activity this year. Also, warmer sea surface temperatures tend to fuel hurricanes as they move across the ocean. However, the climate models are showing considerable uncertainty, which is reflected in the comparable probabilities for an above-normal and near-normal season.


Present Pacific SST's and Computer Projections


Jet Streams


The overall Jet Stream (Pacific, Polar) and the basic trend of the NAO will definitely lead the summer influence, as it has this past winter into spring with little affects from ENSO. Admittedly, the summer months bring weaker jets and NAO cycles but there has been a notable dominance of the Pacific jet stream the past several months with the -NAO (negative) actually becoming more of a player later this spring (and thus, our cooler weather). The $50K question is, will this continue or will this fade for the summer? More investigation is needed to make an educated forecast this summer.

Here is the NAO trend since February 2017:



Latest CFSv2 Jet projections at 10K FT /700 MB/  suggest ridging this summer over the Great Plains and Mississippi Valley to intermittently extend east and hook-up with the Bermuda High. This is a typical upper air summer pattern. At the same time however, troughing across eastern Canada is intimated to shift west and east through the summer, aided by the NAO cycle. How far west the troughing shifts at times during the summer will dictate our temperature pattern in conjunction with the movement of the ridging. At this time; there is agreement between many models and Local Analogues for a generally normal temperature summer with a slight negative departure bias.





SUMMER 2017 ANALOGUES  AND  SUMMER MODEL OUTLOOKS



TEMPERATURES:

Interestingly; Analogues and CPC's experimental model ECCA agree on a comfortable summer overall with temperatures normal to below. This would also give credence to a more intermittent -NAO this summer rather than the +NAO that dominated up until May, as seen on the graph above. Meanwhile, CPC's general public summer outlook goes for a better chance of above normal temperatures this summer (see further down on the page) siding against the ECCA model and the more typically used; the CFS model for temperatures.

Trends  in Analogues

Trends in the analogue summers show June has the better chance to average normal to above out of the three summer months for temperatures; while July and/or August has the best chance of averaging normal to below. Note; none of the July's were "hot' with well above normal temperatures and that also holds basically true for August with only one out of the nine analogues containing a hotter August - back in 2001. Overall; analogues depict a normal but the cooler side of the normal with departures /-1.1/ overall. This is nearly identical with the experimental ECCA below which also depicts close to normal but with a slight preference for a negative departure over Southeast Lower Michigan. The analogue months show on the monthly temperatures and departures and not intra-monthly. This is important since meteorological patterns reflecting upper wind cycles, come infrequently packed neatly in monthly trends. It's best to use summer months as trends and not hold any one month for specific numerical value.*

*For an example; looking at the Summer of 1890 we see June was a warmer than normal month, followed by a cooler than average July and then, a much cooler August. Therefore; you might interpret the trend of that summer was warmer than average at the beginning (but not necessarily June itself) followed by a normal, then a below normal temperature pattern mid and/or late summer. There may very well have been predominantly above normal readings the first six weeks of the summer (first half) and colder than normal the next six weeks (second half) to create that three month temperature pattern.





 E C C A Experimental Model  
              E C C A Experimental Model                       
The ECCA uses Canonical Correlation Analysis (CCA), an empirical statistical method that finds patterns of predictors (variables used to make the prediction) and predictands (variables to be predicted) that maximize the correlation between them. The most recent available predictor data for different atmospheric/oceanic variables are projected onto the loading patterns to create forecasts. The ensemble refers to forecasts produced by using each predictor separately to create a forecast. The final forecast is an equally weighted average of the ensemble of forecasts. The model is trained from 1953 to the year before the present year to create the loading patterns.

Predictor Selection :
The pool of possible predictors used in the forecasts are:


  • 200mb global velocity potential
  • global sea surface temperatures
  • sea level pressure (north of 40N)

  • The predictors selected to be used in the ECCA are based on factors such as :
    a) which climate signals/atmospheric variables play a large role in US temperature/precipitation for a certain season (ie. soil moisture is included in summer forecasts
    b) status or strength of climate signals that impact US temperature and precipitation. For example, seasons with current or expected relatively strong ENSO years typically include sea level pressure as one of the ECCA predictors because of its ability to represent the ENSO signal. 


     SUMMER TEMPS /CFS/
    CPC SUMMER 2017 TEMPERATURE OUTLOOK




    PRECIPITATION:

    The past analogues for the summer strongly suggest a drier summer overall with six drier summers, one normal summer and two slightly (less than an inch) above normal. The drier summers were so pronounced that they skewed the average rainfall of all the summers down to more than two inches below normal /-.2.25/. The analogues seem to suggest a notable dry spell mid to late summer, fairly typical of most summers yet more pronounced than usual. July and/or August has the best chance for the drier weather this summer but all three months averaged below normal rainfall. Only the Summer of  '51 contained rather steadily accumulating, plentiful rains.


    Our CFS Climate model for summer monthly rainfalls would seem to agree with the analogues; calling for each month to contain below normal rainfall. 




     SUMMER PCPN /CFS/

    The CPC favors Equal Chances for rainfall however; giving equal weighting to above, below or normal rainfall - therefore no rainfall preference forecast.






    Reiterating:

    TEMPERATURES

    I look for overall normal during the summer with any "heat waves or warm periods" to be routinely tempered by cooler air pushing south out of Canada. I'm looking for an occasional -NAO to help deliver the cooler weather as discussed in Hemispheric Patterns and in the Analogue section. On specific temperature departures; I expect temperatures to average from +1.5 to -1.5 of normal. 

    RAINFALL

    Latest model and atmospheric trends agree with Summer Analogues for generally below normal rainfall this summer as heaviest rains generally fall west and southeast/east of the region.  Of course; there will be pockets of heavier rains with convective activity but overall, below normal rains are expected.

    _____________________________________________________________________________

    A Look Back....

    Winter of 2016-17: The Roller-Coaster Ran Full Throttle!

    Obviously all seasons have their ups and downs in the temperature department but last winter was quite variable month to month and intra-month. In the end; a very mild winter was had by all with temperatures averaging well above normal. After a relatively normal start to the winter in December; readings jumped to well above normal for both January and February.

    One thing that was persistent through the winter was the exaggerated roller coaster of temperatures with the above normal readings winning out. December by far was the snowiest and most "normal" of the winter months.


    TEMPS





    SNOWFALL

    RAINFALL




    DETROIT
    DEC
    JAN
    FEB
    WINTER







    NORM 30Y
    30.1
    25.6
    28.1
    27.9

    NORM 30Y
    42.5

    Norm
    6.44

    AVE
    29
    32.1
    38
    32.9
    8th Warm
    TOTAL
    37.9

    Total
    6.89

    DEP
    -1.1
    6.5
    9.9
    5

    DEP
    -4.6

    DEP
    0.45













    FLINT











    NORM 30Y
    27.5
    22.5
    24.9
    25.0

    NORM 30Y
    47.4

    Norm
    5.03

    AVE
    27.8
    28.9
    34.5
    30.2
    8th Warm
    TOTAL
    39.4

    Total
    7.14
    16th Wet
    DEP
    0.4
    6.5
    9.6
    5.2

    DEP
    -8

    DEP
    2.11













    SAGINAW











    NORM 30Y
    27.4
    22.3
    24.5
    24.7

    NORM 30Y
    44

    Norm
    5.10

    AVE
    27.1
    28.9
    34.5
    29.4
    9th Warm
    TOTAL
    37.4

    Total
    5.88

    DEP
    -0.3
    5.7
    9.1
    4.7

    DEP
    -6.6

    DEP
    0.78




    No one has to be told we've had some extreme winters in the past couple of decades since 2000,  just look at the frequent placements in the warmest or coldest winters with the warmest ahead in total.

    Top 20 Coldest/Warmest Winters in Southeast Lower Michigan


    Rank Detroit Area* Flint Bishop** Saginaw Area***
    Coldest Warmest Coldest Warmest Coldest Warmest
    Temp Year Temp Year Temp Year Temp Year Temp Year Temp Year
    1 18.8 1903-1904 36.9 1881-1882 16.7 1976-1977 33.4 1931-1932 15.7 1962-1963 33.3 1931-1932
    2 19.3 1874-1875 35.7 1931-1932 17.0 1962-1963 33.0 2015-2016 16.5 1917-1918 31.7 2015-2016
    3 19.7 1976-1977 35.2 1889-1890 17.8 1958-1959 32.2 1982-1983 16.6 1919-1920 30.9 1997-1998
    4 19.7 1917-1918 33.8 2001-2002 18.6 1978-1979 31.7 2001-2002 17.5 1978-1979 30.6 2011-2012
    5 20.4 1962-1963 33.8 1997-1998 19.0 2013-2014 31.3 1997-1998 18.0 1977-1978 30.5 2001-2002
    6 20.6 1977-1978 33.5 2015-2016 19.3 1977-1978 31.2 2011-2012 18.2 1976-1977 30.0 1936-1937
    7 20.8 1919-1920 33.0 2011-2012 19.8 1993-1994 30.3 1932-1933 18.2 2013-2014 29.6 1982-1983
    8 20.9 2013-2014 32.9 2016-2017 19.9 1944-1945 30.2 2016-2017 18.2 1958-1959 29.5 1920-1921
    9 21.0 1935-1936 32.5 1982-1983 20.9 1981-1982 29.1 1952-1953 18.7 1935-1936 29.4 2016-2017
    10 21.1 1892-1893 32.5 1879-1880 21.2 1935-1936 28.8 1936-1937 19.0 1993-1994 29.0 1918-1919
    11 21.4 1904-1905 32.3 1918-1919 21.3 1969-1970 28.6 1948-1949 19.2 1916-1917 28.8 1932-1933
    12 21.5 1978-1979 32.2 1952-1953 21.4 1985-1986 28.5 2012-2013 20.0 1961-1962 28.5 2012-2013
    13 21.7 1880-1881 31.6 1877-1878 21.5 1961-1962 28.3 1930-1931 20.1 1944-1945 28.3 1952-1953
    14 21.8 1911-1912 31.3 1948-1949 21.6 2008-2009 28.2 1991-1992 20.2 1969-1970 28.0 1930-1931
    15 21.8 1878-1879 31.2 1920-1921 21.6 1983-1984 28.2 1949-1950 20.3 1981-1982 27.8 1974-1975
    16 21.9 1981-1982 31.1 1875-1876 21.6 1947-1948 28.1 1974-1975 20.7 1947-1948 27.6 1948-1949
    17 22.4 1969-1970 31.0 1953-1954 22.0 2014-2015 28.0 1953-1954 21.2 2008-2009 27.2 1998-1999
    18 23.0 1958-1959 30.8 1949-1950 22.2 2010-2011 27.8 1998-1999 21.3 2002-2003 27.1 1986-1987
    19 23.2 1876-1877 30.7 2012-2013 22.4 1942-1943 27.7 1986-1987 21.4 1942-1943 27.1 1953-1954
    20 23.3 2014-2015 30.6 1932-1933 22.5 2002-2003 27.7 1940-1941 21.6 1985-1986 27.0 1991-1992
    * Detroit Area temperature records date back to November 1874.
    ** Flint Bishop temperature records date back to January 1921.
    *** Saginaw Area temperature records date back to January 1912.


    WINTER 2016-17 ANALOGUES

    How did the Analogues do this past winter?

    Lousy. The actual raw analogues has called for a normal to below temperature winter and one of the few times the direction for temperature (warmer/colder) was wrong. Fortunately, I disagreed with the analogues and went for a warmer winter, mainly due to the ongoing pattern. 

    In the end, the Pacific Jet was more dominant, especially mid-late winter reflecting at times, a psuedo El Nino. This pattern was actually more like an El Nino across the entire country than our actual El Nino was the year before (Winter of 2015-16). The analogies weren't completely without merit however, there were a few mild to warm winters, especially 1889-90 where the that exceptionally warm winter with an average temperature of 35.1 at Detroit (3rd warmest). 

    However; an even better analogue winter to dig further into was the Winter of 1998-99 after the very strong El Nino of 1997-98 and not unlike the ENSO trend of our past two winters of 2015-16 and 2016-17. The successive Winter of 1998-99 (like this past winter) also averaged above normal in temperatures /30.4/,  just missing our 20th warmest winter ranking. Back then, only one winter month /January/ out of the three contained normal temperatures; while December and February were considerably above normal. Of course, this past winter it was December that contained more typical or normal temperatures while January and February were warmer than normal leading to an average of 32.9 degrees /8th warmest/.

    The SOI 's of the similar winters also were similar in trend.


    Winter 2016-17 SOI:

     

    Comparing the SOI between 2016-17 and 1998-99 (highlighted in yellow) shows a similar pattern overall. The reflected weak La Nina after the strong El Nino of 97-98 (below) was a bit more pronounced and persevered longer than our past La Nina of 2016-17 (above).


    Winter 1998-99 SOI

     
    Winter Outlook 2016-17 

    While I deviated from my analogues for the Winter Outlook by going for a normal to above normal winter, I didn't go warm enough regarding the numerical departures (+2.0F). Actual departures were 4-5.0F above normal; placing it in the top 20 warmest winters.

    From the Winter 2016-17 Outlook...

    Temperatures:  Normal to Above


    Expect temperatures during the 2016-2017 winter to be more variable than usual with conflicting data in research. This will be due to a variable upper wind jet stream. I look for both a split flow and phasing of jets as much over the Central and Western US as in the East. Guidance and Analogues suggest the heart of the winter /Jan-Feb/ seeing the most of the action (though that doesn't mean there won't be any, other times) but winters tended to start somewhat later than average (see more in Analogues). This will be guided by the anticipated weak La Nina to Neutral state interacting with the North Atlantic and subset Arctic Oscillation /NAO, AO/. Just as important this winter will be the Pacific Decadal Oscillation /PDO/ and associated Eastern Pacific Oscillation /EPO/.



    In the final analysis, I look for the Southeast Lower Michigan winter to average around normal to above normal.  Temperature departures should average from -1.0F to + 2.0F.


    WINTER 2016-17 SNOWFALL

     DETROIT
    NORM 30Y
    42.5

    TOTAL
    37.9

    DEP
    -4.6




    FLINT


    NORM 30Y
    47.4

    TOTAL
    39.4

    DEP
    -8




    SAGINAW


    NORM 30Y
    44

    TOTAL
    37.4

    DEP
    -6.6



    From the Winter 2016-17 Outlook...

    Snowfall and RainfallNormal to Above depending on location


    Because of the temperature variability and associated storm tracks observed in many of the analogue winters, snowfall in those winters ranged widely from above normal to below. This would be expected since the variance of temperatures hint at the variability of the upper atmospheric patterns and storm tracks. Therefore, pinpointing the perceived prevailing storm tracks this winter will make a significant difference in regard to seasonal snowfall. 

    Using the expected dominant storm tracks for the upcoming winter, it is likely much of the region will experience normal to above normal precipitation with above normal snowfall /5.0"+/ around Detroit's northern and western suburbs north into the Flint and Saginaw Valley; to around normal /within 5.0"/ over the far Southeast corner of Lower Michigan (south of Detroit). Mixed precipitation events seem to be a higher than normal risk.

    Making weather fun while we all learn,
    Bill Deedler -SEMI_WeatherHistorian