A major winter storm affected the central United States from 23-25
December 2009. The intrusion of cold air and a deep mid-tropospheric
cyclone produced snow as far south as north Texas. This deep cyclone
provided a rare white Christmas too much of the southern plains. The
cost of this White Christmas included blizzard conditions on Christmas
Eve. In the warm air, this strong storm produced heavy rainfall (Fig.
1), flooding, and severe weather (Fig. 2). Most of the heavy rain and
severe weather was observed in north-south band in the region affected
by the surge of high PW air and the anomalous LLJ southerly jet (Figs.
Heavy snow was observed north and west of the surface cyclone and 850
hPa cyclone (Figs. 4 & 7). Most of the heavier snow fell in the
strong low-level northeasterly jet north of the 850 hPa cyclone. The
high latitude block, as shown in Figure 9 and the anomalous 500 hPa
heights likely contributed to this unusual cut-off cyclone. The deep
slow moving cyclone led to the cold air and precipitation remaining in
place long enough for the cold air to change the rain to snow. Though
not shown here, many locations in Oklahoma, Kansas and Missouri had
rain, sleet and then snow. There are many interesting aspects to this
storm which were not addressed here. The focus here was on providing an
overview of an historic Winter Storm and the Southern Plains Christmas
Eve Blizzard of 2009.
Blizzard...Winter Storm...Christmas Eve
historic winter storm brought heavy snow and blizzard conditions to
portions of the East Coast of the United States from 18-20 December
2009. Heavy snow was observed from North Carolina to New England. As
the storm moved northward, it produced record snows and impacted the
densely populated megalopolitan corridor from Richmond to New York
City. Heavy and record snows fell over the major cities in this region.
The 19th of December saw many new snowfall records set for cities from
Washington to New York. Snowfall totals of 1 to 2 feet were quite
common with most of the snow falling in the Mid-Atlantic region to
southern New York on the 19th. The heavy snow lingered into the 20th
over southeastern New York and eastern New England. The record snows in
the New York City, Philadelphia, Washington and Baltimore regions are
clearly related to nearly ideal condition including:
? The block over northeastern North America and the high pressure over
eastern North America
? The surge of high PW air from the subtropics into the storm
? The anomalous LLJ aimed at the DC area.
retreating cold air mass produced ideal conditions for an ice event
over central Pennsylvania on 13 December 2009. Due to strong warm
advection, the event ended as light rain and drizzle and the
temperature went above freezing over the entire affected region between
1800 and 2200 UTC. This case had a strong jet entrance region which
likely helped maintain the low-level cold air. The warm air aloft was
strong and there was little doubt that snow or ice pellets were a
serious threat (Fig. 10 & 11). The 850 hPa temperatures were
forecast to be above 0C (not shown) and verified close to observed
The higher resolution NAM showed the extent of the low-level cold air
better than the coarser GFS and SREF. In short term forecasting, higher
resolution can play a very important role that our current global
models and coarser ensembles cannot. The SREF indicated a potential,
40-80% chance for a brief period of freezing rain. This got forecasters
thinking about the potential, but the finer resolution data in the NAM
clearly out performed the other systems in its ability to simulate the
low level cold air damming and the prolonged 4-7 hour period of
freezing rainfall. There are times when the finer scale models will
outperform the coarser global models and the coarser EFS?s. Forecasters
need to be aware of this. In the near future, some uncertainty
information will be obtained from finer resolution models by temporal
and spatial adjustments. Thus the 0C contour in the NAM could be used
to produce zones of uncertainty with respect to freezing rain. In the
interim finer scale models advantages should be realized at shorter
ranges leveraging their details and the probabilities from the
ensembles. In this case, the fact that the SREF showed 40-80% chance of
ice and that the finer resolution model showed strong cold air damming
should have provide confidence in the colder solution of the higher
resolution model. The courser GFS had no real chance at simulating the
extent of the cold air damming. |
strong jet and surface cyclone moved ashore along the West Coast of the
United States on 6-7 December 2009. This system moved rapidly eastward
producing significant weather from coast to coast. The
?transcontinental express? produced heavy snow in the mountains of the
western United States; spread the coldest air of the season into the
northwestern United States, and the plains; produced heavy snow from
the southern plains into the Great Lakes; heavy snow from the
Appalachians from West Pennsylvania to Maine; and severe weather and
heavy rain in the Gulf States. The system left few places in the lower
The strong energy came under and block which had developed over
western North America and the eastern Pacific. As the system moved
onshore it pulled anomalously high PW air into the mountains of the
western United States. This led to rain and mountain snows. As the
system moved across the mountains it tapped warm moist air over the
Gulf of Mexico and pulled it northward up the Mississippi Valley. The
strong cyclone and the anticyclones to the north led to a strong
low-level frontal circulation. A strong easterly jet with -3 to -4 SD
u-wind anomalies developed in this region. Many areas affected by this
LLJ saw heavy snow. The strong cold front, very warm and moist air
produced severe weather along and ahead of the cold from Louisiana to
South Carolina (Fig. 8). The surge of warm air led to the convection
well to the north. Thunderstorms developed as far north as Pennsylvania
and isolated severe weather was observed over western Pennsylvania and
New Jersey. Thunderstorms were reported as far north as Massachusetts
weak surface cyclone moved up the East Coast on 5 December 2009. This
system was associated with a strong north-south 250 hPa jet and had a
strong easterly 850 hPa jet associated with it. The system produced 16
to 4 mm of rainfall along portions of the East Coast from North
Carolina and Virginia eastward across Maryland, New Jersey, Long
Island, and southern New England (Fig 3). The western edge of the
precipitation shield was mainly snow with a good 2-7 inch snowfall
event from Virginia into Maine. Some locally heavier amounts were
observed. This was the first significant snowfall of the winter of
2009-2010 for many locations such as the Washington and Philadelphia
This storm had some positive and negative aspects to it. The
precipitation shield extended farther west and north than guidance
initially indicated. Few forecasts ever got the north edge of the
precipitation shield correct (Figs. 1 vs. Figs 13 &14). The NAM
under predicted the heavy rainfall along the coastal plain. The overall
synoptic pattern appeared to be relatively well predicted by the NAM,
GFS, and SREF (no SREF features shown). But the edges of the
precipitation shield were hard to predict. The 04/1400 UTC SREF QPF and
POPS (Figure 15) suggest it offered little more information than the
NAM or GFS. It did attempt to depict the band of heavier rainfall along
the coastal plain. These data imply that there were considerable clues
as to the potential for snow on the western edge of this system and for
the heavy rainfall along the coast. The clues included the pattern,
such as the strong jet and strong 850 hPa LLJ and u-wind anomalies.
This case shows that there is considerable uncertainly, even at
relatively short forecast intervals, in predicting QPF and the western
edge of precipitation shields using NCEP models and ensembles
strong cyclone came out of the Gulf of Mexico on 2 December 2009. This
storm system pulled up moist air which was connected to a region of
very moist air over the tropical eastern Pacific Ocean (Fig. 4). The
plume of moisture or the atmospheric river (AR), produced heavy
rainfall from the eastern Gulf States into New England. The system also
produced some severe weather (Fig. 1).
The pattern associated with this event was a pattern quite commonly
associated with heavy rainfall and severe weather in the eastern United
States. The key features in such events include anomalous southerly
winds, typified by a strong LLJ with strong v-wind anomalies and a
surge of high PW air. Interesting to this event is that this plume of
high PW was directly related to a potential source region. The PW
plumes and the satellite data suggest this event had a tropical Pacific
connection and is likely another AR case. The data shown here suggest
that the pattern and the area of rainfall were relatively well
predicted by the NCEP GFS, NAM, and SREF. Clearly there were details
these models missed but the large scale area impacted and the surge of
high PW air were shown to be well predicted by the NAM and GFS. This
can be inferred in the SREF by its QPFs which suggested it too, by
proxy, correctly predicted the pattern for heavy rainfall. |
overview of the warm dry November of 2009. Focus on the Mid-Atlantic
region and Pennsylvania. Anomalies and standardized anomalies are shown
for the month of November. These data are compared to the anoamlies and
pattern from 1 Sept to 30 November 2009. Key feature was the above
normal heights and ridge over eastern North America. Key words: JRA
Anomalies, November 2009.|
atmospheric river brought a surge of high precipitable water (PW) air
into Ireland and the United Kingdom on 18-20 November 2009 (Fig. 1).
Preliminary reports suggest record rainfall was observed over Ireland,
northwestern England and southwestern Scotland from 18-20 November
2009. The United Kingdom Meteorological Office (UKMO) indicated a storm
on 13-14 November and 19-20 November impacted Britain. The focus here
is on the event of 19-20 November. Initial reports suggest as much as
300 mm of rainfall may have fallen. The extreme heavy rainfall produced
flooding in the Cumbria area of Britain.
Similar atmospheric rivers (Ralph et al 2006 & 2005) have been
observed along the West Coast of the United States. When these plumes
of high PW air impinge upon land, interact with sloped frontal
boundaries or interact with elevated terrain, they often produced heavy
The impact of moisture laden air and above normal PW standardized
anomalies (Grumm and Hart 2001) into the West Coast of North America
was documented by Junker et al (2008). They demonstrated the value of
standardized anomalies of fields such as PW, winds, and moisture flux
(MF) in defining areas where heavy rainfall would fall. Stuart and
Grumm (2007) showed the value of standardized anomalies in defining and
identifying heavy snow events in the eastern United States.
remnants of hurricane Ida came ashore on the 10th of November along the
coast of Alabama and western Florida. The system tracked across Georgia
and over the western Atlantic on the 11th through the 13th and then
meandered to the east as it weakened. A strong area of high pressure
and a larger scale ridge precluded the storm from moving to the north.
The slow moving storm produced heavy rainfall from Alabama to the
Mid-Atlantic region (Fig. 1). Some areas locally received 200-300 mm of
rainfall as the storm lumbered up the coast.
This event was realitively well predicted by the NCEP models and
ensemble forecast systems.|
relatively well known cool season heavy rainfall pattern set up over
Mississippi 29-31 October 2009. This pattern produced a synoptic type
heavy rainfall event (Maddox et al. 1979) across the region with some
areas received over 4 inches (100 mm) of rainfall in about 24 hours.
The timing and amount of rainfall support previous studies of heavy
rainfall during the spring and fall in the lower Mississippi Valley
(Korty 1980; Corfidi et al 1990, and Bellville and Stewart 1983). This
event was well predicted and likely well anticipated by forecasters.
The event was relatively well predicted by the NCEP NAM, GFS and by
both NCEP EFS including both the SREF and GEFS. All systems forecast
the pattern quite well relative to the verifying analysis from the GFS
(shown). The QPF pattern produced by these systems clearly had the
threat area well predicted, the rainfall amounts well predicted and
thus all provided useful guidance and high confidence in a heavy
rainfall event of a known type.
This case is a classic example of how the patterns associated with
heavy rainfall and the ensemble probabilities can be used to have
confidence in a potentially significant rainfall event. The patterns
can be viewed as the inside thinking or meteorology and the
probabilities as the outside thinking or numerical values assigned to
the event potential.|
rare and historic early season snowfall impacted central Pennsylvania
on 15-16 October 2009. At the official observation site at the
Pennsylvania State University (PSU:STCP1) 4.7 inches of snow was
recorded for the date. This broke the old record for the first snow and
first inch of snowfall set on 17 October 1977. Officially, there have
been 2 other snow events in State College in recent years including the
events of 26 October 2005 and the 31 October 2002. It should be noted
the STCP1 had 4.1 inches of snow on 1 November 1993, most of the snow
fell on Halloween 31 October 1993 (Grumm and Nicosia 1996) fell on 30
and 31 October. The combination of heavy wet snow and leaves on the
trees can make early season snow fall extremely destructive. So as with
many early season snow events this record event was extremely
destructive. When snow totals reached around 2 inches leaf burdened
limbs and stressed branches began to snap and fall. As snow totals
reached about 4 inches, large leaf burdened trees lost limbs and some
fell over. It should be noted that most of the precipitation fell as
snow and over 1.3 inches of liquid equivalent water fell during the
event. Thus, the leaf and snow burdened trees produced power outages.
It was estimated that about 27000 customers lost power between about 6
PM Thursday 15 October and 6 AM Friday 16 October. Around 12000
customers in State College lost power and Centre County accounted for a
over 50% of the power issues . |
deep cyclone affected the West Coast of the United States 13-14 October
2009. The storm produced heavy rainfall in the mountains of California
and southern Oregon (Fig. 1). Over 128 mm of rainfall is shown in the
coarse data displayed in Figure 1. Reports in California of locally
heavy rainfall amounts over 250 mm (10 inches) and 300 mm (16 inches)
were reported. In addition to the heavy rainfall, the storm produced
strong winds in the mountains of California and western Nevada. Wind
gusts of 30 to 50 mph were common throughout the region. In Nevada,
higher winds with gusts over 75 mph were observed in a narrow corridor
between I-80 and US-50. Over the Siera Crest, peak winds over 100 mph
were common with a report of 135 mph near Lake Tahoe and Mammoth Lakes.
The higher winds were mainly confined to the higher elevations of
western Nevada and California. Snow also was observed mainly above 7000
ft. This strong storm had connection with Super Typhoon Melor.
As the cyclone associated with Melor moved toward the Pacific Coast of
the United States, it produced locally heavy rainfall from California
northward to Oregon (Fig. 1). The storm had anomalously deep mean
sea-level pressure anomalies (Fig. 2) comparable to the several
historic storms from the past, including the historic 1962 ?Columbus
Day? wind storm in the Pacific (Lynott, and Cramer, 1966). A recent
ranking of historic Storms in the Pacific Northwest ranked the famous
Columbus Day windstorm as the 10th strongest storm since 1948 (Graham
and Grumm 2010).
GEFS Ensemble anomalies
rainfall affected the Missouri Valley on 9 October 2009. The heaviest
rainfall was observed over Missouri and adjacent portions of Kansas,
Arkansas, and Oklahoma (Fig. 1). The heaviest rainfall fell on the 8
October over the region shown in Figure 1.
The large scale pattern was similar to the large scale patterns which
have been identified to be associated with heavy rainfall over the
Midwest (Junker et. al. 1999 and Junker and Schneider 1997). Key
features (Fig. 2) included a strong jet entrance region just downstream
of the affected region (Fig. 2b) and an approaching trough and jet exit
region (Fig. 2a). This pattern implies a couplet jet circulation over
the region. At 500 hPa there was a deep trough to the west (Fig. 2c)
and a strong ridge to the south and east. Warm mid troposphere
temperatures (Fig 2d) were present in the ridge. This note will
document the heavy rainfall event of October 2009 over the Missouri
Valley. The emphasis is on the pattern of this event relative to past
events. The focus is on the ability of the NCEP models and ensemble
forecast systems to predict the pattern and thus the potential for
Ensembles Anomalies GFS NAM SREF heavy rainfall|
deep cyclone moving through the Great Lakes and over southeastern
Canada produced high winds across portions of the Mid-Atlantic Region
and into southern New England. Wind damage was reported as far west as
Michigan and eastward to the coast of New Jersey and New England.
Over central Pennsylvania about 75000 people lost power due to trees
down on wires. Reported wind values were generally in the 40 to 53 mph
range, normally associated with a wind advisory category event. The
event did produce significant impact to a wide range of people. A
comparison of the pressure changes for the 12 February 2009 event (Fig.
8) revealed that pocket of 9 to 15 hPa pressure rises moved across
Pennsylvania during that event compared to the 2-4 hPa/6hr-1 observed
during this event. The French storm of 10 February 2009 showed 12 to 18
hPa rises with the wind event. The latter two events did produce more
widespread damage and the French event produced massive power outages
across northern France, a truly high impact event.
Pressure rises... wind anomalies ... nonconvective winds|
stalled east-west frontal boundary and a persistent area of low
pressure produced a prolonged period of heavy rainfall in the
southeastern United States from 19 to 22 September 2009. Rainfall
amounts in excess of 12 inches were observed in the multi-sensor
precipitation data (Fig. 1) and at observation locations in Georgia.
The National Weather Service Office in Peachtree City, Georgia reported
that ?the heavy rainfall caused significant runoff into area streams
and rivers, resulting in major to record flooding across the Atlanta
metropolitan area?. They indicated heavy rain fell on the 19th, 20th
and 21st with 11.80 inches in Douglas County and as much as 9-12 inches
reported in western sections of the Atlanta Metropolitan area. These
rainfall amounts like are within the heavy rain area shown in Figure 1.
The NAM provided poor QPFs relative to the GFS over the affected region.|
weak frontal boundary sagged southward on 19 August 2008. The boundary
stalled in southern Pennsylvania and remained in place on 20 August
2008. This boundary interacted with the warm moist unstable air along
and ahead of it producing some severe weather, wind damage, in
southeastern Pennsylvania (Fig. 1). A more widespread event is evident
well to the west of Pennsylvania over Ohio where a stronger frontal
system was progressing eastward.
One of the storms that developed over eastern York and western
Lancaster Counties produced wind damage and well photographed funnel
cloud (Fig. 2). The picture was taken along route 322 in Lower Swatara
Township in southern Dauphin County likely facing south into northern
Lancaster County... This funnel cloud produced no reported damage.
line which produced severe weather with bowing segments over eastern
Pennyslvania. A mesoscale short-wave moved across Ohio, West Virginia
and Pennsylvania on 18 August 2009. The wave triggered showers and
thunderstorms as it moved eastward. During the mid-afternoon hours, the
line of showers and thunderstorms blossomed into a line of severe
storms. These storms produced damaging and winds and a few reports of
hail as it raced across eastern Pennsylvania into New Jersey (Fig. 1)|
of 2009 will be remembered as a cool and wet month in the eastern
United States and a bit hot and dusty in the western United States.
Figure 1 shows the total monthly precipitation and the mean 500 hPa
heights over the United States. These data show the wet conditions over
southeastern New England, Florida, and portions of the Mississippi and
The 500 hPa height (Fig. 1) field showed a persistent trough over the
eastern United States and a ridge over the western United States. Dry
conditions dominated from Texas to the Pacific coast. It will be shown
that below normal temperatures were present at lower levels near the
well documented and recorded EF1 tornado which was mainly and EF0
tornado over most of its 1.25 to 1.50 mile long swath, affect Elk
County from 1821 to about 1825 UTC on 26 July 2009. Clearly cell phone
and digital technologies contributed to our knowledge about this event.
The track in Figures 1 & 3 was derived from these images.
The event clearly occurred in relatively weak large scale conditions
and was not impressive on radar. After the fact it is clear that this
storm was a weak, low topped supercell thunderstorm. Stronger storms
have affected this region in the recent past with no impacts. Storms
with rotation couplets in excess of 50 kts passed this region twice on
11 July and produce no reported damage. Why did this storm produce a
tornado and those clearly stronger storms did not is an important
Elk County Tornado |
rains fall over Pennsylvania and eastern New England on the 23rd of
July 2009 (Fig.1) ahead of an upper level trough. The heavy rainfall
over Rhode Island and Connecticut draw ones attention in Figure 1. Over
Pennsylvania, the rain fell mainly in the late afternoon and evening
hours from slow moving thunderstorms. The heaviest rainfall was
observed in Cumberland and York Counties where reports of 5.7 inches
were observed at Bloserville and Carlisle Springs and around 5.8 inches
near Siddonburg. Heavy rains and flooding from 3 to 4 inches of
rainfall was observed near and just west of State College, PA. The
official COOP sites showed 4.10 inches in Cumberland County. The COOP
data is heavily used in the UPD data shown in Figure 1.
The key meteorological players in this event included a deep and
unusually strong mid-tropospheric trough (Fig 1 &
7d), instability, above normal low-level moisture, and a trough moving
across the region. Clearly, instability and upright convection played a
role in the event over Pennsylvania. Though radar data was not
presented, radar showed low-core and slow moving thunderstorms over the
region. In New England, the event was better predicted and had stronger
forcing and patterns more typically associated with heavy rainfall.
This included the strong low-level easterly jet, the strong surface
cyclone, and the above normal PW surge into the region. Despite the
pattern being correct, the details in the heavy rainfall over New
England were not well predicted. Figure 10 clearly showed significant
variation in each successive 6-hour NAM forecast cycle. The shortest
forecast, initialized at 23/1800 UTC may have been particular poor
implying some model spin-up issues which caused an underestimation in
the rainfall. |
approaching cold front (Fig. 1b) and upper-level shortwave (Fig. 1a)
produced severe weather across the northeastern United States during
the afternoon and evening hours of 11 July 2009. The severe weather was
focused over Pennsylvania and included a tornado over northeastern
Pennsylvania (Fig. 1). In addition to the Pennsylvania, there was some
severe weather over Missouri associated with a large MCS which
developed on the flanks of a massive subtropical anticyclone and
accompanying heat over the southwestern United States. The note will
document the severe weather and tornadic event over Pennsylvania.
2009 was unusually cool and wet over the northeastern United States.
For cool weather, Boston had its 6th coldest June and many other
locations in the Northeast were unusually cool.
In terms of precipitation, most of the northeast and the Mid-Atlantic
region were quite wet experiencing above normal rainfall and an
unusually high number of days when rainfall was observed. So me
locations had long stretches of rainfall and it rained over 50% of the
30 days of June including Boston, MA (22), Hartford, CT (24),
Worcester, MA (22), Concord, NH (18), Portland, ME(21), New York City,
NY (23), Poughkeepsie (17) and to the west, State College (16). An
overview of the pattern for June and early July 2009 is presented here.
upper-level short-wave and cold front triggered convection over the
Mid-Atlantic Region and northeastern United States on 26 June 2009 (Fig
1). The majority of the severe weather was in the form of hail (blue
squares in Figure 1). There were some isolated wind reports in the
northeastern United States, but removed from the upper-level trough
wind reports dominated the severe weather over Tennessee and Kentucky
and in the southeastern United States.
This event shared many characteristics of eastern hail events, outside
a favorable supercell environment. This event was dominated by steep
mid-tropospheric lapse rates, a common feature in many eastern US hail
events when the shear is relatively week. With a prolonged period with
a deep trough over the eastern United States from mid-June to early
July, there were several hail dominated severe weather events in the
eastern United States. Dates of particular interest during this period
included 15 June (all hail), 25 June, 26 June, 30 June, 1 July, and 7
large subtropical ridge brought warm dry conditions from eastern Texas
to Alabama. Around this large subtropical ridge, a ring of moisture
produced the classic "ring of fire". Along the edges of the ridge there
were successive days of heavy rainfall and severe weather.
Ring of Fire
weak shear day with high CAPE produced large updraft thunderstorms over
the Mid-Atlantic region on 9 June 2009. The lack of shear precluded
organization of most of the storms until later in the event. The large
CAPE produced large cores. As shown in Figure 6, an interesting aspect
of this event was the lack of shear. This suggested severe weather
would be limited to big updraft storms due to the lack of shear. The
CAPE and storm type was well anticipated.
Despite the good forecasts and anticipating the event type. Warnings
were difficult as storms popped up in the unstable air. Some were over
sparsely populated areas and hail is poorly verified outside of
populated regions. The -20C level was near 20KFT during the event.
These storms typically pushed 50 dBZ cores to 30Kft and 60-65dBZ cores
occasionally reached 25-30kft in the storms that produced the largest
well predicted heavy rainfall event affected Botswana on 9 June 2009.
Observations in Botswana indicated over 100 mm (4 inches) of rainfall
in 24 hours. Estimated rainfall from the CMORPH and JMA are shown in
Figures 1a&b. The CMORPH data showed upwards of 128 mm of rainfall
while the JMA data indicated over 275 mm of rainfall.
This event occurred with a common pattern observed in other heavy
rainfall events found in other locations about the world. In this
event, an anomalous 500 hPa cut-off cyclone with mid-level cold air
combined with abnormally moist air to produce an ideal scenario for
heavy rainfall. PW and u-wind anomalies, often found to be associated
with heavy rainfall events, were on the order of 3 to 5 and -3 to -5
SDs respectively during this event. The anomalies clearly indicated a
potential heavy rainfall event.
The pattern associated with this event was well predicted and thus the
NCEP GEF provided excellent guidance in predicting the pattern and the
location for heavy rainfall. Using the concepts outlined by Junker et
al. (2009) forecasters could have leveraged the large anomalies to
better anticipate the heavy rainfall. It should be noted that the
coarse global model predicted a high probability of 50 mm in 24 hours
and 50 mm in 36 hours in Botswana (Figs 9-10).
episode over eastern Alaska.
A well predicted warm episode affected Alaska on 1-4 June 2009.
Temperatures in Juneau reached 80F on 3 June with a number of sites in
the Alaskan Panhandle reaching temperatures well into the 80s. One
remote observing site recorded a high temperature of 91F. Though the
warmest temperatures were achieved on 3 June, most of the week of 1-5
June 2009 was very warm across most of southeastern Alaska.
A well predicted warm episode affected eastern Alaska and the Yukon 2-6
June 2009. This warm episode shared the characteristics similar to
previous events over the region to include the May 2009 warm episode
(link). The large subtropical ridge, which was well predicted, produced
all the tell-tale signals of a warm episode. This included the above
normal 850 hPa temperatures west of the ridge axis and the surge of
high PW air north and west of the subtropical ridge.
The NCEP RMOP data showed that is was a well predicted and high
confidence event. Clearly, several days out the NCEP data predicted the
potential for a large and highly predictable ridge over Alaska and thus
confidence in the above normal 850 hPa temperatures indicated by the
RMOP GEFS anomalies....|
rainfall in southern and southeastern Pennsylvania. This was mainly a
heavy rain event with above normal PW values and strong southerly
winds. There were a few reports of severe weather.
warm episode with some record high temperatures. This well predicted
event was readily found in the NCEP RMOP data as being highly
predictable. Due to cold antecedent conditions, the onset of the warm
temperatures produced snow melt. The warm air and influx of water
caused ice in rivers to melt. Ice dams formed and did considerable
damage to places like Eagele, Alaska, on the Yukon River.
Anomalies and RMOP images showed the strong ridge and warm air
associated with this event.|
early season heat episode affected the eastern United States on 25-28
April 2009. This event was relatively well predicted at least 8 days in
advance. This event actually began in the western United States and
progressed eastward as the subtropical ridge moved eastward.
The NCEP GEFS provided good guidance and high confidence in the eastern
heat episode. Not attempt to quantify the western event was presented
here. The GEFS clearly predicted the 500 hPa heights and anomalies
associated with a strong subtropical ridge. Both the RMOP and
standardized anomaly data aided in predicting and diagnosing this
forecast rain event over the Mid-Atlantic region. Some banding and
locally more heavy rainfall over areas of Pennsylvania. The event
featured severe weather in the Ohio Valley and in the southeast. Well
forecast by the GFS, NAM, and SREF.|
early spring severe weather event affected the Mid-Atlantic region
during the afternoon hours of Sunday, 29 March 2009. The system brought
strong and damaging winds, large hail, and a tornado to the region
(Fig. 1). The Storm Prediction Center images does not depict the EF1
tornado that touch down in Ephrata and Lititz, Pennsylvania around 440
PM on 29 March . The tornado was on the ground for approximately 1.25
miles with peak winds of 85-95mph and it caused 3 injuries.
Large hail was reported over many portions of Pennsylvania eastward to
New York. Dime size hail was reported as far east as Ridge, NY. Strong
winds also were observed and damaging straight line winds were observed
in York and Lancaster Counties causing damage to homes and trees.
ECWS brought heavy snow to portions of Tennessee, North Carolina, and
up the East Coast of the United States to Maine. This was the first
winter storm of the winter of 2008-2009 to bring widespread heavy snow
to heavily populated corridor of the eastern United States. Snowfall
totals of 6-10 inches were common from near the cities of Charlotte,
NC, Richmond, VA, Washington, DC, Baltimore, MD, Philadelphia, PA, New
York City, NY and Boston, MA. Some areas received 10-14 inches such as
the Long Island, New York.
The storm was a slow moving system and had a classic anomalous easterly
low level jet which did relatively well outlining the region of heavy
snowfall. Stuart and Grumm (2006) documented this association of the
anomalous LLJ and heavy snowfall in ECWS. The storm also brought much
need precipitation to the southeastern United States.
Express the severe weather and synoptic high wind event of 10-12
An intense cyclone zipped across the United States on 10-12 February
2009 producing widespread wind damage in its wake (Figure 1). The
severe weather was observed from 10 to 12 February while the synoptic
wind event was primarily on the 11 and 12th of February. This complex
storm caused many areas to experience high winds associated with
convection than, a few hours later, strong winds associated with a fast
moving isallobaric high. In some locations, it was difficult to
distinguish between the two events. Though each event was associated
with distinctly different meteorological conditions.
Narrow cold frontal rainbands and post frontal winds.
surge of warm air produced record high temperatures over much of
Pennsylvania and the eastern Unites State on 11 February 2009 (Fig. 1).
High temperatures range through the 50s and 60s over most of central
Pennsylvania (Table 1) and most locations set new record highs for the
day. It should be noted that NWS COOP sites record temperatures for the
period of 7 AM to 7 AM and thus the high for 11 February is recorded on
12 February. The values for record highs are thus from 12 February data
for all COOP sites.
The event was well predicted by the NCEP global models and ensembles
(GEFS) several days in advance. Well away from fronts and cyclones, the
spread between members was relatively low and this event represented a
high probability outcome event. None the less it represents a
successful forecast. Due to the limits of predictability, all forecasts
will show some degree of uncertainty and this event was no exception.
The GFS and GEFS got the pattern correct with subtle difference verse
the verifying analysis.
deepening Atlantic cyclone moved over the English Channel on 10
February 2009. The storm brought strong non-convective high winds and
rain to most of northern France. New accounts suggest hurricane force
winds of up to 87 mph (140kph) affected northern and western France.
The British Broadcast Company (BBC) report indicated that over 600,000
households across northern France lost power due to down lines and
trees on power lines. Advanced forecasts of winds over 100 kph prompted
the closing of Charles de Gaulle and Orly airports around Paris. High
seas caused the cancellation of cross Channel Ferry services.
As shown in the GFS and GEFS, the basic cyclone and strong winds were
relatively well predicted at 1-2 day forecast ranges. This event shows
the value of NWP and professional meteorologist in the decision support
activities associated with significant high impact weather events.
Ensemble probabilities of wind thresholds at 2 and 10m exceeding 50 and
70 kts would likely have been insightful though they were not produced
the evening hours, a mesoscale snow band set up over portions of
Chester and Lancaster Counties. Radar suggests a similar, if not
slightly weaker, snow band formed over portions of Adams and Cumberland
counties. The snow began around 0000 UTC 4 February (7 PM) and the
intensity picked up and after 0200 UTC (9 PM) snowfall rates of 3
inches an hour reported beneath this band. A video of this event was
made at 2 AM of the 12 inch snowfall in Lancaster, Pennsylvania . The
average snowfall in Manheim Township, Lancaster Pennsylvania suggested
that 12 inches fell in 5 hours, a sustained rate of about 2.5 inches
per hour. During the ?official? measurement it was claimed that this
was ?a truly remarkable event? and thus the title of this document. |
rare snow event struck Western Europe on 1-2 February 2009. This was
reported, by newspapers and on the United Kingdom Meteorological Office
(UKMO) to be largest snow fall over London and southern England in 18
years. The last event of this magnitude was observed on 7-9 February
The snowfall was observed in the United Kingdom and in portions of
France. New accounts implied that many business activities, schools,
and transportation systems had to be close. New accounts reported
?transportation? nightmares as trains and tubes slowed or came to a
halt. Snow being a relatively a rare phenomena, trains are not equipped
to remove snow from the tracks. Additionally, with significant snowfall
being so uncommon, snow removal equipment is normally not purchased or
on hand to remove snow.
The snow therefore was estimated to have cost about 1.3 billion British
Pounds in losses due to the closing of so many business and government
activities. The new report that the snow produced what was in essence a
National snow day.
The overall pattern was similar to the pattern observed during the snow
event o 7-9 February 1991. These two snow events included a strong
ridge at 500 hPa and surface anticyclone over Scandinavia and strong
easterly flow. Each event included the passage of a strong cold front
from the east with abnormally cold air behind the frontal boundary.
Both events were characterized by strong low-level easterly flow and
the evolution of an upper-level closed cyclone. Finally, in each event,
a surface cyclone developed.|
significant differences between forecasts from models from different
modeling centers, inconsistencies in forecasts between successive runs
of the same model, and considerable uncertainty in more traditional EPS
data, a winter storm was grossly over forecast. The storm that verified
was far weaker than advertised and was farther north and east of the
location needed to have produced a significant event along the East
Coast. It would appear that many forecasters did not examine or
understand the uncertainty associated with this potential storm. This
storm was not guaranteed to have a high impact on the eastern U.S. The
data shown here, using simplistic ensemble techniques, suggests there
was considerable uncertainty associated with the evolution of this
The true LAF from the GFS and GEM (Figs 2 & 9) quantified the
differences. The large spread on the west side of the cyclone in all of
the LAFs was due to the more eastward forecasts with time. Better
methods of ensembling are available to forecast uncertainty; however
these data illustrate how each modeling system is sensitive to small
changes in initial conditions. Additionally, forecaster continue to
monitor, compare, and bias their forecasts toward specific solutions.
Thus, dProg/dt and LAF methods show how sensitive each modeling center
models are to uncertainty in initial
Megastorm, Groundhogvilla and other names followed this potential storm
which never evolved close to its hyped potential.|
heat wave struck southern Australia from 27 January to 02 February
2009. Clearly, the thermal anomalies show that the peak of the heat was
from about 28-30 January 2009. The ridge clearly moved eastward and the
thermal anomalies lessened after the 30th.
The re-analysis data suggests that a 5880 m contour at 500 hPa is a
good indicator of a subtropical ridge capable of producing record or
near record heat. This is lower than the 5940 m contour often
associated with heat waves over the United States. The re-analysis data
clearly showed that subsidence near the ridge likely played a key role
in the 2 to 3 SD thermal anomalies in the successive days from 27-31
January 2009. The low-level anomalies were quite close to the ridge.
The 4 January 1976 data showed the same pattern and similar, though
weaker ridge at 500 hPa. The data shown here imply the value in thermal
anomalies and the subtropical ridge in diagnosing and forecasting heat
episodes in Australia
Temperatures were well above normal over the region and many reporting
sites were 10 to 20C above normal. A record high of 41.5C was reported
at Flinders Airport setting a new record high for Tasmania breaking the
previous record of 40.8C set in Hobart on 4 January 1976. The 41.5C
reading crushed the old record of 38.8C set in January 2003.
storm Arkansas and Kentucky. Mixed precipitation over Pennsylvania.
A high impact ice storm struck the United States from Arkansas to
Pennsylvania. Severe icing caused massive tree damage and power outages
in many locations. News reports suggest Arkansas, Kentucky, and
southern Ohio were particularly hit hard by this significant winter
storm. Well over a million (1.3 million) customers lost power and
perhaps 35 deaths were attributed to the ice storm. Due to the
extensive damage to trees and power lines, many customers were informed
that it could take over a week to restore power to hundreds of
thousands of affected customers.
The high impact of the storm was mainly felt on 27-28 January. Figure 1
shows the high resolution precipitation data in 6-hour increments from
0000-1800 UTC 27 January 2009. These data show that the precipitation
arrived in Arkansas around 0000 UTC 27 January and then rapidly spread
eastward into Illinois, Indiana, Kentucky and Tennessee. Details of the
precipitation shield evolution are provided in section 3 of this
document and the total accumulated precipitation in shown in Figure 2.
This paper will document the ice storm of 27-28 January 2009. The goals
include putting this event into a meteorological context relative to
previously document ice storms and to show the value of ensembles in
forecasting potentially devastating ice storms.
deep cyclone brought strong winds to portions of southern France,
northern Spain and Italy on 24 January 2009. Tragically, the winds
caused considerable damage and resulted in about 21 deaths. Most of the
deaths were due to a roof collapsing near Barcelona, Spain.
The deep Biscay cyclone, which tracked across southern France and into
the eastern Mediterranean and into the Adriatic, brought the strong
winds to the region. This cyclone appeared to be quite predictable by
the NCEP GFS and GEFS. The both forecast the deep cyclone (shown only
from the GFS) and the strong 850 hPa winds quite well. The wind
probabilities suggested winds of 25ms-1 were a high probability
outcome. The observed winds were considerably stronger than the
forecasts. The east-west orientation of the Pyrenees Mountains may have
played a role in the locally high winds in both southern France and
northeastern Spain. Note the couplet of high winds (Fig.4b) north and
south of the Pyrenees in southeastern France and northeastern Spain.
Barcelona lies in close proximity to the v-wind maximum south of the
mountains. The maximum in the GFS likely lies closer to Tarragona than
American heat wave of January 2009 compared to the pattern of the
January 2003 heat episode.
he pattern of sub-tropical South American heat waves was presented
here. A comparison of the January 2003 and 2009 events was presented.
Key features included an anomalous 500 hPa ridge and anomalous 850 hPa
temperatures. It would appear a close 5880m contour at 500 hPa is a
good tell-tale sign of a heat wave over subtropical South America.
The data shown here suggests that this event was quite predictable and
was well predicted by the NCEP GEFS. The probability fields of 2m
temperatures suggest that 38C contour probability might be of
considerable value for heat wave prediction in this region.
unusually strong surge of arctic air descended upon the United States,
mainly east of the divide (Fig. 1), during the week of 12 January 2009.
This air mass produced widespread sub-zero (F) readings across the
United States from 13-17 January 2009. The coldest period was from the
15 to 18 January 2009 which is the period shown in Figure 1. The
sub-zero readings in Pennsylvania represented some of the coldest
reading in over 12 years and were likely the coldest readings, in many
locations in the 21st century to date.
The cold episode was relatively well forecast by the NCEP GEFS (Figs.
10-12) and other models (not shown). The guidance provided a long
lead-time on the cold surge.
The pattern associated with this event was quite similar to patterns
observed in other cold Januarys from the past. The patterns at 700 hPa
and 850 hPa for the 3 cold winters from 1977 to 1979 and the cold
winter of 1994 were shown for comparison. They shared common
characteristics to the event that unfolded in January of 2009.
surge of warm air into Alaska produced unseasonably warm weather over
much of the State from 14-19 January 2009. Table 1 lists the
temperatures and records for Anchorage from 8-20 January 2009. These
data show that temperatures were below normal early in the month.
Additionally, there was considerable snow cover over the State. The
temperatures went above freezing on the 14th reaching a high of 37F. At
the official observing site, temperatures remained above freezing until
the 18th. Locally some areas went to freezing and the antecedent cold
ground conditions produced severe ice problems on local roadways in and
A record high temperature of 50F was achieved on 16 January 2009. In
addition to the warm daytime highs, several high overnight lows
temperature records were achieved. Table 2 shows the temperature data
for Fairbanks, Alaska. The onset of the warm temperatures was delayed a
few days due to the more northerly latitude. The highest maximum
temperature of 52F was achieved on the 16th. Similar to Anchorage,
Fairbanks showed a complete reversal from cold to warm.
An examination of previous days with record high temperatures and
record high, high temperatures revealed certain common features. The
more obvious feature was above normal 850 hPa temperatures during each
event. Other features included a negative MSLP and 500 hPa height
anomalies over the Aleutians and Bering Sea. Positive MSLP and 500 hPa
anomalies were present over the western Canada. In most cases, above
normal PW with +1 or greater anomalies were present over portions of
surge of warm air into Alaska produced unseasonably warm weather over
much of the State from 14-19 January 2009. Table 1 lists the
temperatures and records for Anchorage from 8-20 January 2009. Tables 2
& 3 show temperature data for Juneau and Fairbanks respectively.
These data show that temperatures were below normal early in the month.
Additionally, there was considerable snow cover over the State. The
temperatures went above freezing on the 14th reaching a high of 37F. At
the official observing site, temperatures remained above freezing until
the 18th. Locally some areas went to freezing and the antecedent cold
ground conditions produced severe ice problems on local roadways in and
The RMOP images (Figs 9-11) showed just how predictable this event was.
The large 500 hPa ridge, which was the dominant weather feature over
the region, was highly predictable. The trough to the west, over the
Aleutians was also relatively highly predictable. Thus, despite the
lower predictability at 500 hPa over much of Alaska, the sensible
weather was likely highly predictable. The key point is the RMOP shows
the feature which is highly predictable. Thus the sensible weather
between the relatively predictable Aleutian low and the very
predictable ridge was likely more predictable than normal. |
clipper brought snow from the Midwest to the East Coast on 9-11 January
2009. Snow fall totals generally were in the 3-7 inch range with this
system, with some locally higher amounts in northern Illinois, southern
Wisconsin, and Ohio where snowfall totals around 12 inches were
observed in some isolated locations, to include Chicago. In
Pennsylvania, the storm produced a widespread 3-7 inch snowfall over
central and northern areas with some locally higher amounts in a few
The storm followed the generalized rules associated with winter storms
and specifically Clipper events including:
? Snow fall with a Clipper generally requires that the cyclone pass to
? Snowfall generally falls north and west of the 850 hPa cyclone, as in
the ?Younkin rule? (Goree and Younkin 1966; Younkin, 1968; and Brown
and Younkin 1970). This rule worked will in the Midwest with this
event. The 850 hPa low actually tracked along the NY/PA border in the
? Regions along the track and south of the Clipper get little snow. In
this case there was a strip of ice in Pennsylvania along and south of
the track of the surface cyclone.
fourth significant ice storm of the winter of 2007-2008 affected the
eastern United States on 6-7 January 2009. This storm produced ice, in
the form of freezing rain and ice pellets, over much of central
Pennsylvania. This event shared many of the characteristics common to
recent ice storms, including an intrusion of cold dry air at low-levels
before the precipitation arrived and a sharp east-west baroclinic zone.
This was the fourth event in which the NCEP short-range ensemble
forecast system (SREF) correctly identified the problem with regards to
the precipitation type. The SREF clearly showed that freezing rain and
ice pellets would be the dominant precipitation type. The few SREF
members which forecast snow showed the snow was of short duration
before a change over to mixed precipitation. Longer range forecasts
indicated a slightly higher potential for rainfall. The event was
generally well forecasts, in and of the fact that the longer range NCEP
Global Ensemble Forecast System (GEFS) showed a high potential for a
winter storm on our about 6-7 January 2009. Longer range forecasts
ranged from a snow event, with the main cyclone tracking to the south
to a nearly all rain event as forecasts began trending the storm track
well to the north and west. The total liquid equivalent precipitation
over Pennsylvania and the eastern United States is shown in Figure 1.
Most of the precipitation over Pennsylvania fell between 2100 UTC 06 to
1500 UTC 07 January 2009 (Fig. 1a). This event was primarily a heavy
rain event in the warm sector over the Ohio Valley and an ice event in
from northern Maryland to southern New York State (Fig. 2). The
heaviest rainfall was over the central Appalachian Mountains where over
100 mm of rain was observed in the mountains of western Tennessee. It
will be shown that the higher amounts of precipitation were aligned
with the PW and PW anomaly fields. Additionally, over Pennsylvania, the
boundary between the warm air and low-level cold air focused the heavy
precipitation over Pennsylvania (Fig. 1a).