|2008-12-24 ||A significant winter storm affected the eastern United States on 23-25 December 2008. Heavy snow impacted locations on the northwestern side of the storm. Over Pennsylvania freezing rain was observed over central and eastern sections. A surge of warm air turned the ice to rain and temperatures soared into the 40s and 50s from the Allegheny plateau westward. As this event ended, prolonged cold air damming event persisted over central Pennsylvania.
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 0000 UTC 24 and 0600 UTC 25 December (Fig. 1a). Over the eastern United States the event began on the 23rd and moved eastward. The precipitation along the northern edge of this event fell as snow. Heavy snow was observed from Illinois into Indiana, Michigan and portions of New York State. Heavy rain was observed in the middle-Mississippi Valley and freezing rain was observed over most of Pennsylvania.
In Pennsylvania, the precipitation began as sleet as far west as Pittsburgh but quickly changed to freezing rain over northern and central sections of the States. As mentioned earlier a surge of very warm air in western Pennsylvania quickly changed the precipitation to rain as temperatures rapidly rose into the 40s and 50s from the Allegheny plateau westward.
Over the eastern United States the event began on the 23rd and moved eastward. Most of the precipitation along the northern edge of this event fell as snow. Heavy snow was observed from Illinois into Indiana and Michigan. Heavy rain was observed in the middle-Mississippi Valley and freezing rain was observed over most of Pennsylvania.
|2008-12-19 ||A winter storm brought heavy snow, ice, and heavy rainfall from Iowa to New England from 18-20 December 2008. Area affected by the storm is clearly defined in the liquid equivalent precipitation for the 48 hour period ending at 1200 UTC 20 December (Figure 1a). A broad west to east precipitation shield is evident over the aforementioned. Heavy snow fell on the northern edge of the heavy precipitation shield, a band of mixed precipitation dominated the middle section of the band and heavy rains impacted the regions in the warm air.
Figures 3 & 7 showed the classic heavy snow signature identified by Stuart and Grumm (2006) including a strong anticyclone and an anomalous easterly jet in the cold air. The latter signal has been shown to be a predictor in major East Coast Storms and appears to function well in winter storms east of the Rocky Mountains. In this case, the heavy snow fell in and along the anomalous LLJ from the Great Lakes, eastward to southern New England.
|2008-12-12 ||A significant ice storm affected the eastern United States from 11 to 12 December 2008. Though freezing rain and accumulating ice was observed from Pennsylvania to Maine, warmer temperatures precluded heavy ice accretion over most of Pennsylvania. Temperatures in Pennsylvania were mainly in the 30-33F range and ice accumulations over 0.20 inches were generally limited to locations over 2000 ft. However, from northeastern Pennsylvania into New England, colder temperatures; many locations remained in the upper 20s; facilitated larger ice accumulations and thus damaging ice was observed over these areas Thus, from New York to Maine, significant ice accumulations produced widespread power outages affecting over 1 million customers.
The event was relatively well forecast by the NCEP models and ensemble prediction systems which generally showed considerable uncertainty with the timing and precipitation type. They did show a small area of heavy snow along the western edge of the precipitation shield which increased in probability after about 12/0300 UTC. Heavy snow was limited to a few locations in extreme northern Pennsylvania and a stripe across New York and Vermont on the western edge of the larger precipitation shield.
An examination of SREF planview precipitation type images, such as those shown in Figure 14, from 09/0300 through 10/1500 UTC seemed to indicate ice as the higher probability precipitation type until about 12/0300 UTC when probability of snow along the northern and western edge of the precipitation shield rapidly increased. Experiences with SREF and GEFS plumes and PTYPE forecasts typically suggest limited snowfall in areas of mixed precipitation. Over Pennsylvania and eastern New York mixed precipitation types dominated and freezing rain was the key forecast issue in terms of impact.
|2008-12-07 ||Lake Effect band on MODIS and in northwesterly flow from Lake Huron to Pennsylvania on 7 December 2008.|
|2008-11-22 ||Heavy rainfall from 21-23 November 2008 brought flooding and mudslides to the coastal sections of Santa Catarina State of Brazil. Locally, the floods occurred after a period of heavy rainfall in mid-late November 2008 and followed a significantly wet period from October through November of 2008. The heaviest rainfall fell on 22 November as shown in Figure 1. Note that the rainfall was confined to mainly coastal regions, suggesting that a combination of large scale and mesoscale forcing resulted in the localized heavy rainfall.
The rainfall that produced the was associated with a anomalously strong anticyclone over the western Atlantic. Anticyclones are not often associated with heavy rainfall. Aloft, a deep trough was present at 500 hPa over the affected region (Fig. 4) as was a pocket of relatively cold air at 850 (Fig. 4) and 700 hPa (not shown). In addition to the potential instability associated with the upper-level low, the return flow around the intensifying anticyclone (Fig. 6) produced a strong low-level jet along the coastal regions (Fig. 5).
The heavy rainfall produced by the 40km GFS (Fig. 7), mainly along the coast and over the adjacent western Atlantic was linked to this low-level jet.|
|2008-11-21 ||A period of cold weather settled in over Pennsylvania after 8 November 2008. A reinforcing cold front on 20 November 2008 brought a lake enhanced snow band to Pennsylvania late on the 20th and into the 21st. This band had a Lake Huron connection and it brought snow well east (Fig. 2) of traditional Lake Effect snow regions.
A particularly strong and persistent snow band set up late on the 20th (Fig 10) and produced locally heavy snowfall across central Pennsylvania (Fig. 2). Over 12 inches of snow was reported in Warren County, closer to Lake Erie and as much as 4 inches of snow fell over the Middle-Susquehanna Valley, relatively far removed from the Great Lakes and not a region prone to Lake Effect snows.
This unique snow band did show a connection to Lake Huron, a feature known to produce strong and persistent snow bands. The relatively warm lakes and the enhanced convergence may have played a role in this snow band and its extent across Pennsylvania. There was a clearly a larger scale signal present as both NCEP models, the NAM and GFS predicted a potential snow band over Pennsylvania at about the correct time and in a relatively correct position.
The NAM, GFS, and NCEP SREF all predicted the potential for a snow band over Pennsylvania (Fig. 6-9). From a synoptic scale perspective the forecasts were quite successful. On a more mesoscale perspective, the details of where and how much snow would fall were not as well predicted.
|2008-10-28 ||An early season snow storm affected portions of Pennsylvania and New York State on 27-28 October 2008. Snowfall amounts ranged from a few inches to over two feet of accumulation in New York State (Fig. 1). As shown in Table 1, October snow events occur about once every 3 years in Pennsylvania.
This early season snow event was not well forecast with more than about 24-36 hours of advanced warning for many of the affected areas. Both NCEP EPS were too slow to move the cyclone over the western Atlantic far enough to the west to impact New York and Pennsylvania. The NCEP NAM was the first to show the potential for precipitation, snowfall, and a cyclone closer to the coast. The NCEP NAM forecasts from 26/1800 and 27/0000 UTC provided clues and information that would have provided 36 to 42 hours advanced notice for the storm.
Previous events, such as the 17-18 October 1977 snow event (Fig. 15) reveal some common characteristics of amongst these event including:
? a deep upper-level cyclone and the snow areas are often under the track of the upper-level low;
? unseasonably cold low-level temperatures;
? a strong low-level easterly jet and a strong upper-level jet.
There is a clear signal in the pattern; a clear signal in the precipitation rates, and a clear signal with an elevation component. These elements could be used to anticipate these events .
|2008-10-25 ||A fast moving frontal system brought a short-lived period of moderate to heavy rainfall to the Mid-Atlantic region on 25-26 October 2008. The north-south frontal orientation and the strong low-level southerly winds identified this event as a Maddox Synoptic type event (Figs. 3-7). Overall rainfall was between 25 and 75 mm (1-3 inches) with a few areas in eastern Pennsylvania and south-central New York State where 100 mm (4 inches) of rainfall were observed.
The SREF forecasts showed that the SREF correctly predicted the pattern associated with a Maddox Synoptic type event at least 3 days prior to the onset of the rainfall (Figure 9). Though not shown, the GEFS was equally successful at predicting the pattern. When the GEFS, with a current horizontal of about 100 km, is improved to 70 km in December 2008 it should improve on its ability to predict these mesoscale event types.
The PW and wind anomalies in the SREF showed that as the uncertainty decreased, as a function of forecast length, the value of the anomalies increased. Thus, the anomalies are a gage of the uncertainty in the forecasts. Though not shown, the NAM and GFS forecasts showed the potential for 2-3 SD PW anomalies several days in advance as they were unencumbered by uncertainty issues.
Getting the timing and location of the QPF, in terms of location, start and end times is important for a wide range of reasons. During this event, the fast moving rainfall allowed for improving conditions and fair weather for the PSU-OSU football game in Columbus, Ohio. However, the east, the slow exit of the rainfall delayed the onset of the World Series in Philadelphia.|
|2008-09-28 ||An subtropical storm (Simpson 1952) brought locally heavy rainfall (Fig. 1) to portions of the Mid-Atlantic region from 26 to 29 September 2008. This storm formed under a cut-off upper-level low which formed a Rex block (Rex 1950a & 1950b) over eastern North America and the adjacent North Atlantic (Fig. 3). The block likely contributed to the slow movement of the system and the prolonged period of rainfall over the eastern United States. The interaction with hurricane Kyle was also evident, though Kyle remained well east of the Mid-Atlantic region.
This storm, especially during its development and progression westward near the Carolina coast, was quite similar to the subtropical cyclone described by Simpson (1952). It also developed in a pattern remarkably similar to that shown in Figure 13, where the upper-level close 500 hPa was between an upper-level anticyclone to its northeast and to its southwest. Though no satellite data was shown, the storm also looked similar to the patterns shown by Cochran (1976).
Of significance in the Mid-Atlantic region was the locally heavy rainfall and prolonged period of cloudy weather. As shown in Figure 1, the event produced significant rainfall over a significant portion of the eastern United States. And as shown in Figures 15 & 16, the forecasts of the rainfall were of value though they lacked some of the interesting if not important details associated with this event.
|2008-09-13 ||Hurricane Ike was an expansive category II hurricane when it came ashore along the Texas coast on 13 September 2008 between 4 and 5 AM. The storm was downgraded to a Tropical Storm by 5 PM 13 September 2008 over Texas and a tropical depression by 5 AM 14 September along the border of Missouri and Arkansas.
The storm was so large that winds and waves affected the Gulf Coast from Florida to Texas while the storm was well offshore. As the storm approached it pushed water into the eastern Texas coast and the storm surge destroyed homes along the coast (Fig. 1). The storm came ashore a few hours before sunrise on 13 September (Saturday) with 110 mph winds, large waves, and heavy rainfall. The water flooded and destroyed thousands of homes while the strong winds blew out windows in skyscrapers in downtown Houston. The winds and falling trees downed power lines cutting off power to more than 3 million people. As documented by hurricanes (Pielke and Pielke RA 1997), this large storm had significant societal impacts on the eastern coast of Texas.
In addition to the damage caused by the Gulf of Mexico, Ike produced heavy rainfall (Fig. 2) along the Texas coast. As shown by the inset in Figure 2, some regions of eastern Texas received in excess of 400 mm (18 in) of rainfall. Heavy is a common impact of land falling hurricanes (Pielke and Pielke RA (1997).
In addition to the coastal flooding, wind damage, and heavy rainfall, Ike spawned tornadoes in Louisiana on the 12th and 13th and in Arkansas on the 13th. All the tornadoes and severe weather were on the east side of the storm. Severe weather was observed in the Ohio valley on the 14th and straight line wind damage was associated with system from the Ohio Valley to western New York on the 14th.
This note will document the weather and patterns associated with hurricane Ike. |
|2008-09-06 ||TS Hanna moved hastily up the East Coast on Saturday, 6 September 2008. The fast moving storm put down locally heavy rainfall (Fig. 1) along and just to the west of the track of the surface low center (Fig. 2). A frontal system, with lighter rainfall, produced light rain west of the rainfall associated with TS Hanna.
The NAM analysis was used to show the key features associated with TS Hanna. The tight circulation center remained independent from the frontal system to the west. Similar to many TS and Hurricanes, the heaviest rainfall was along and west of the track of the cyclone. This is quite common when tropical storms interact with a larger scale jet streak (Fig. 8). The band of heavy rain was very narrow and was just west of the cyclone center.
The NCEP models and EPS from 04 through 06 September clearly showed that TS Hanna would provide a narrow band of heavy rainfall, just west of the cyclone track. The details of where this heavy rain band would lie remained quite elusive. The devil was and always is, in the details. Thus, despite relatively successful forecasts of the area and shape of the rain band, the exact areas to be impacted were difficult to specify.
It should be noted that Hurricane Charlie (13-15 August 2004) was another storm which remained independent of a frontal system. The storm produced bands of heavy rainfall (Fig. 13) mainly of the coast and just west of the very compact surface cyclone position (Fig. 14).
|2008-09-01 ||Hurricane Gustav came ashore in Louisiana on 1 September 2008 approximately 72 miles southwest of New Orleans during the morning hours of 1 September 2008. The category 2 storm on the Saffir-Simpson scale produced power outage and a wide range of severe weather over the Gulf Coast States. The storm, as analyzed by the NAM 00-hour forecasts, lingered over Louisiana for over 2 days as it slowly weakened.
The storm produced considerable damage from winds, severe, weather and heavy rainfall. Rainfall in portions of Louisiana exceeded 250 mm. The heavy rains were clearly focused in the strong southeasterly flow on the north and east side of the storm. The NAM analysis appeared to capture some sense of the bands on the northeast side of the storm. Both the rainfall data (Fig. 1) and the severe weather patterns (Fig. 2) imply the bands on the northeast side of the storm.
SREF forecasts should the area where Gustav would make landfall on 1 Septemer 2008. SREF forecasts also showed a high probability of over 5 inches of rainfall.|
|2008-08-21 ||Tropical storm Fay took a leisurely tour of the southeastern United States from 18 to 24 August 2008. The storm dumped heavy rainfall along the East Coast of Florida on the 19-22 August 2008. The rainfall over Florida for this period is shown in Figure 1. The mid-coast area of eastern Florida showed over 300 mm of rainfall for this period.
Tropical storm produced rainfall amounts in excess of 200 mm in 24 hours for over 3 successive days. At it?s peak, rainfall amounts along the central areas of Florida?s East Coast exceeded 300 mm in 24 hours (Fig. 7) which led to the event total amounts in excess of 20 inches (Fig. 2) of that region of Florida.
The NAM 00-hour analysis show that the low-level wind anomalies at 850 and 925 hPa were in close proximity to the heavy rain areas. The PW anomalies were also closely aligned with these features. These high resolution data imply the potential use of wind and PW anomalies to predict record rain events in near real-time.
|2008-08-02 ||A series of thunderstorms moved across Pennsylvania on 2 August 2008 causing wind damage and reports of large hail (Table 1). Numerous photos of hail, to include hail the size of tennis balls, were received by TV stations in the lower Susquehanna Valley . In total, 12 reports were provided in a storm report and summarized by the Storm Prediction Center (SPC). There were over 35 reports of large hail received during the event (See Table in Appendix).
Overall, there was a significant severe weather event over the eastern United States on 2 August as a cold front pushed southeastward. There were over 412 reports of severe weather (Fig. 1) with a significant percentage of the reports along the East Coast from Massachusetts to Louisiana. Most of the severe weather in Pennsylvania was focused in Southeastern portions of the State. The lone report in central Pennsylvania occurred in the morning hours.
This paper will summarize the event of 2 August 2008. The meteorological setting is provided along with some radar imagery to put the event into perspective.
|2008-07-24 ||North south oriened frontal system with anomalous 500 hPa low produced heavy rains in the northeastern United States. Over Pennsylvania, the event produced hail beneath the 500 hPa low. Over NY and New England, the event produced locally heavy rainfall and severe weather. Most of the severe weather was confined to eastern New England with tornadoes in southern Maine and eastern New Hampshire.|
|2008-07-13 ||A convective weather event impacted several central Pennsylvania counties on Sunday July 13th 2008. The main threat was from several marginal supercells, and broken line segments along and pre-frontal during the late afternoon early evening hours. While most storms were sub-severe or marginal at best, a few line-segments produced wind gusts between 30-40kts.
A weak EF0 was found a couple miles southeast of Jonestown in Lebanon County Pennsylvania from the event on 13th July 2008. This storm produced a small area of damage, impacting primarily farm land as shown in the images in Figures 11 &12 and in the map of the damage as shown in Figure 13.|
|2008-06-29 ||A severe weather event impacted central Pennsylvania on Sunday 29 June 2008. Several distinct echo types were observed including supercells, bow echoes, and pulse type storms. Despite the variety of storm event was rather modest over central Pennsylvania with only 24 severe reports across the State.
The first supercell over central Pennsylvania to affect the State College Forecast Area formed west of State College and moved over the north side of State College, passing over both the airport and the National Weather Service Forecast Office (WFO). A wall cloud was present and there were reports of funnel clouds. The storm tracked across Centre, southern Clinton, and Lycoming Counties. Despite the rotation in the storm, the storm produced no reports of severe weather.
Two bow echoes to the west of State College were also produced no reports of severe weather. The majority of storms which produced severe weather did have 50 dBZ cores over the -20C level, about 24 to 26KFT over central Pennsylvania on this particular date.
Perhaps the storms with the structures, to include distinct bows and cyclonic rotation were too shallow to produce severe weather. It is an interesting if not challenging problem to identify storms which are more likely to produce severe weather. But clearly on this day, elevated cores to the -20C level were better indicators than more classic storm structures.
|2008-06-27 ||A damaging wet microburst was observed near Newport, Pennsylvania. This lone storm produced heavy rainfall and damaging winds. There was nothing uniquely visible to help discriminate that this storm would be severe relative to any other storm on this particular day.
Rules on pulse storms are to maximize the probabilities and avoid over warning. Sometimes, the true randomness of these events become self evident. This event did not have 50dBZ cores to the -20C level. This storm had no signatures visible on radar to detect or increase the potential to predict a microburst. Despite these factors, this storm produced a damaging wet microburst.
Storm survey maps, pictures, and a brief overview of this wet microburst are presened.|
|2008-06-22 ||During the afternoon and early evening hours of 22 June 2008 two thunderstorms developed over southern Pennsylvania and moved north-northwestward over the region. The first storm produced large hail and wind damage between about 2048 and 2156 UTC (Fig. 1). A second storm developed on the heels of this storm and followed a track slightly to the west of the first storm. This storm also produced wind and hail damage (Table 1).
This paper will document the severe weather event of 23 June 2008. The focus is on the overall pattern and large scale set up over the northeastern United States. Severe weather images are focused over Pennsylvania.
|2008-06-20 ||During the afternoon hours of 20 June 2008 several thunderstorms developed over Pennsylvania. The flow over the region was quite weak and there were not distinguishing large scale features of note. Afternoon CAPE values were forecast to be in the 800 to 1200 JKG-1 range. There were 7 severe weather reports in Pennsylvania include 3 hail and 4 wind reports (Table 1).
The resulting convection produced two severe storms. The first storm was in southern Pennsylvania near the Maryland border. The storm produced wind damage near Mercersburg, Pennsylvania. The second storm formed in Lycoming County and produced nickel size hail near Biggertown and Penny size hail on the north side of Benton in Columbia Counties. Reports were obtained from untrained public observers.
|2008-06-16 ||A strong upper-level trough embedded in northwesterly flow brought a northwesterly flow severe weather event (Johns 1982, Johns 1984 and Fritsch and Giordano 1991) to the northeastern United States on 16 June 2008. The strong June sunshine and a surge of cold air at 700 hPa over the warmer low-level air to create showers and severe thunderstorms. The conditions were favorable for hail and wind gusts. Table 1 shows the reports by type over the eastern United States. There were 204 reports of hail of which 40 reports were observed in Pennsylvania. There were also a significant number of wind reports. As shown in Table 1 and in Figure 1 there were no tornadoes in Pennsylvania and the one tornado observation was in Newburgh, New York.|
|2008-06-10 ||A strong subtropical ridge dominated the weather over much of eastern North America from 5-10 June 2008. On the periphery of the ridge, warm moist tropical air collided with cooler air and produced a series of Mesocale Convective Systems (MCSs). Some of these systems produced record flooding. Under the ridge, hot weather dominated and many locations in the Mid-Atlantic region experienced successive days of 100F+ temperatures. Richmond, Virginia and Raleigh-Durham had several days where the maximum temperatures exceed 100F. In Pennsylvania many locations experienced 3 to 4 consecutive days with temperatures at or above 90F qualifying this event as a heat wave.
This paper will examine the conditions associated the early season heat wave of 6-10 June 2008. The focus is on the value of climatic anomalies to predict and characterize the heat wave. Data from previous heat waves, as analyzed by the GR data, are also presented. The focus is on the traditional features used to identify heat waves including the traditional fields, such as 500 hPa heights, 925 and 850 hPa temperatures and their departures from normal (climatic anomalies). Precipitable water (PWAT) anomalies are shown as additional tools to characterize a heat wave.|
|2008-06-10 ||A cold front moving into unseasonably warm air triggered thunderstorms and produce a significant severe weather event in the eastern United States on 10 June 2008 (Fig. 1). With over 512 severe reports this was significant severe weather day with the majority of the events, 373 (Table 1) associated with the cold front from North Carolina to Maine. The data in Table show indicated that New York State had 103 severe weather reports followed by Pennsylvania, and Virginia. Vermont had 33 reports which is likely a significant number owing to relative to its size compared to New York and Pennsylvania.
In addition to creating a widespread severe weather event, the frontal system brought an end to a 5 day East Coast heat wave. The cooler and drier air behind the front brought welcome relief too much of the East Coast. Perhaps not surprisingly, one of the warmest and longest-lived early season heat waves ended with a significant severe weather event. |
|2008-06-08 ||A review of the pattern and rainfall patterns associated with the heavy rains in the upper Mississippi Valley. The focus is on the heavy rains of 8 June 2008 and a comparison of the pattern with the patterns associated with the previous record heavy rains and floods of 1993.
The anomalies of key fields showed similar patterns between the two events. The heavy rains of 5-13 June 2008 lead to the tragic flooding in Wisconsin and Iowa.|
|2008-06-07 ||Heavy rains brought flooding to portions of Illinois and Indiana on 7 June 2008. Portions of Indiana received over 225 mm (9 inches) of rain (Fig. 1). The swath of heavy rains extended from Missouri into eastern Indiana. The flooding was a result of a frontal boundary moving into and stalling as it encountered the subtropical ridge over the eastern United States.
It will be shown that the heavy rain fall was associated with a back building MCS (Maddox 1980;Maddox et al. 1986: Fritsch et al 1981:Weztel et al. 1983). MCC?s and MCS?s producing heavy rains are a common feature on the periphery of large subtropical ridges, which often produce heat waves or heat episodes in mid-latitudes. This event was another example of this inter-related process and relationship between MCS?s and heat episodes.
|2008-06-04 ||A building subtropical ridge over the eastern United States set up the classic ?ring of fire? convective pattern on 2-4 June 2008. These patterns having become known signals for existing strong subtropical ridges or building subtropical ridges associated with warm season heat episodes.
This strengthening subtropical ridge produced several days of severe weather with 275, 284, and 410 severe reports on June 2, 3 and 4 respectively. Most of these reports were associated with large MCS and some large MCC?s which developed in the moisture plumes moving up the west side of the ridge.
The severe weather reports for 4 June 2008 are shown in Figure 1. With 410 reports, this was the most active day and as shown in Figures 8a-8e it was a time of active production of MCS, several of which clearly developed in Derechoes. This reports focused on the conditions associated with the derecho that affected the Mid Atlantic region during the afternoon hours of 4 June 2008
|2008-05-12 ||An unusually deep and cold upper-level low moved across the eastern United States from 10 to 12 May 2008. This anomalous system produced severe weather and heavy rainfall over the eastern United States as it moved through the region. Over the higher terrain of central Pennsylvania, the storm produced a period of wet snow which covered the ground in some locations. It was one of the latest accumulating snow in several locations in central Pennsylvania.
This system was relatively well forecast at least 4 days in advance as indicated by the GEFS 500 hPa forecasts (Fig. 17) and the potential for heavy rainfall along the East Coast from forecasts initialized at least 4 days in advance (Fig. 16). These forecasts suggest something relatively predictable about the overall pattern the led to these relatively skillful 500 hPa and QPF?s.
Shorter term forecasts suggest that the SREF and GEFS both forecast the pattern over the eastern United States relatively accurately. This allowed the EPS?s to predict quite well the area of heavy rainfall in the Mid-West in close proximity and in relatively similar orientations to the observed patterns of rainfall (Fig. 2) and severe weather (Fig. 3). The mesoscale details were a bit elusive though the overall concept and area of the significant threat was well highlighted by the EPS?s. |
|2008-05-03 ||A line of showers and embedded thunderstorms moved across Pennsylvania during the afternoon and evening hours of 3 May 2008. The Stage-IV data proved useful in identifying the details of the event and this case demonstrate the utility of short-range high resolution models in the forecast process. |
|2008-04-28 ||A strong frontal system and upper-level wave produced rain and severe weather over the eastern United States on 28 April 2008. Behind the front, an unseasonably cold air moved into the eastern United States. As the cold air pushed through the coastal plain on the 28th it triggered a strong severe weather event in the Mid-Atlantic region to include an EF3 tornado.
The large scale forecast of the front and winds being correct likely contributed to the relatively accurate QPFs shown in Figure 10b verse the verification in Figure 6.|
|2008-04-27 ||A line of showers and thunderstorms moved across the State ending an April warm episode. Some locally heavy rains were observed. |
|2008-04-21 ||An upper-level low drifted over the Mid-Atlantic region on 20-21 April 2008. This system brought an end to a period of unseasonably dry and mild weather. Clouds and rain bands moved over the region on Sunday, 20 April 2008. Localized rainfall amounts of 3 to 4 inches were reported in isolated regions of Maryland and southern Pennsylvania. The larger scale rainfall amounts (Fig. 1) did not resolve some of the higher spotter reports.|
|2008-04-04 ||Heavy rains affected the central Mississippi and the Ohio River Valleys on 3-4 April 20008 (Fig. 1). Fortunately, this axis of heavy rains fell 100-200 km south of the axis of heavy rains which fell on 18-19 March 2008. The earlier event produced significant flooding along many streams and smaller rivers from Missouri to Ohio. Residual flooding was still present as of 2 April 2008 from this event.
This paper will document the heavy rainfall event of 3-4 April 2008. The goal include, showing the pattern associated with the event, documenting that the NCEP EPSs were able to forecast the pattern, and the value of the EPS QPF probability forecasts during potential extreme events, and show some limitations of any prediction system in getting at the details with significant lead times.
|2008-03-22 ||A fast moving ?Alberta Clipper? (Glickman 2000) brought snow from eastern North Dakota, across Minnesota, Wisconsin, northern Illinois, Michigan, northern Ohio, and Pennsylvania. The precipitation shield was shown in Figure 1 and it revealed, though displaced to the north, the generalized track of the Clipper. Most of the precipitation fell as snow with some significant snow fall over the Midwest. Snow amounts of 12-15.5 inches were reported in Wisconsin and 11 inches in northern Illinois, 10 inches was reported in southern Michigan and up to 7 inches in Pennsylvania.
In this event, most of the precipitation was observed north of the surface cyclone. Furthermore, most of the snow was observed in the area affected by the easterly winds and the larger low-level eastern wind anomalies. Stuart and Grumm (2006) showed the value of low-level u-wind anomalies in forecasting heavy snow in the eastern United States. This event followed that same model. Interestingly, the heaviest snowfall was observed in the area affected by the larger u-wind anomalies|
|2008-03-18 ||Heavy rains produced flooding over the central Mississippi and the Ohio River Valleys on 18-19 March 2008. Several streams and smaller rivers reached historic flood levels in Missouri and southern Illinois where some of the heaviest rainfall was observed (Fig. 1). Tragically, the flooding also produced several fatalities.
The heavy rainfall fell over and along a quasi-stationary frontal boundary which extended from northeast Texas into southern Ohio. Strong low-level southerly winds brought high precipitable water (PW) values into the region. The strong south/southwesterly flow and the persistent stream of moisture over the frontal boundary led to the heavy rainfall. It will be shown that both the pattern and potential for heavy rainfall was well forecast by the National Centers for Environmental Predictions (NCEP) Global Ensemble Forecast System (GEFS) and the NCEP Short-range ensemble forecast system (SREF).
This paper will document the heavy rainfall event of 18-19 March 2008. The goal include, showing the pattern associated with the event, documenting that the NCEP EPSs were able to forecast the pattern, and the value of the EPS QPF probability forecasts during potential extreme events. The use of the properly forecast pattern and the high probability of a significant heavy rainfall event will be shown to be reinforcing concepts.
|2008-03-08 ||A complex winter storm affected much of the eastern United States from 7 to 8 March 2008. The storm produced severe weather over the southeast on the 7th including tornadoes in Georgia and Florida and severe thunderstorms from Virginia to New on the 8th. Severe storm reports are summarized in Table 1. On the cold side of the storm, mainly north and west of the 850 hPa low track, the storm produced heavy snow. The heaviest snow was observed in the Ohio Valley from northern Alabama to northern Ohio. A records snowfall of 20.5 inches was observed in Columbus, OH, breaking a record that had stood for 98 years (Table 1). As much as 14 inches of snow was observed in the suburbs of Louisville, KY. On the warm side of the storm, heavy rainfall was observed from central Pennsylvania eastward into New Jersey and New York. Rainfall ranged from 1 to 3 inches across the region.
his event was relatively well forecast by the National Centers for Environmental Predictions (NCEP) Global Ensemble Forecast System (GEFS: Szunyogh and Toth 2002) 2-4 days in advance of the storm. There was considerable variation in the NCEP Global Forecast System (GFS) forecasts of this storm in the 3-8 range. The GFS is the model core of the GEFS. Thus, the NCEP GEFS has some limitations, relying on single model core, as outlined by Palmer et. al. 2008. To compensate for this single model limitation, NCEP has partnered with Meteorological Services of Canada (MSC) to share ensemble data between NCEP and the Canadian Meteorological Center (CMC). This allows for the production of a multi-model ensemble known as the North American Ensemble Forecast System (NAEFS).
|2008-03-04 ||A complex winter storm affected the eastern United States from 3 to 5 March 2008. The focus herein is on the impacts of the storm over the Mid Atlantic region on 4-5 March 2008. The storm produced severe weather over the south, along and ahead of the leading cold front. Snow and ice were observed to the north and west of the low track. An east of the primary low, the storm produced heavy rains.
The storm produced heavy rains over portions of the eastern United States. The rainfall is shown in Figure 2. The data do not include spotter reports and gages which indicated several locations with over 100 mm (4 inches) of rain in central Pennsylvania. The Cooperative data shows a broad region of 50 mm (2inches) over Pennsylvania with a large are of over 4 inches of rainfall. The heavy rains along with some snow melt produced flooding along rivers and streams.
Along the northern edges of the cyclone track the storm produced heavy snow, mainly in the Midwest from Missouri, across Illinois, and into Michigan. Shallow cold air produced freezing rain over portions of northeastern Ohio, northwestern Pennsylvania and western New York State.
There were 68, 88, and 284 reports of severe weather reports along the southern edge of the frontal system as is moved across the southern United States. There were 22 tornado reports associated with the event. The 284 reports on the 4th is a significantly large event for a wintertime severe weather event.
This was damaging late winter storm producing a mix of severe weather, heavy rainfall, ice, and heavy snow. Despite the complex nature of the storm, it was relatively well forecast by the NCEP operational EPS. The overall pattern was well forecast and heavy rain was extremely well forecast by the SREF and GEFS. Many runs showed over 2 inches of precipitation as a high probability forecast.
|2008-02-29 ||A relatively well forecast ?Alberta Clipper? (Glickman 2000) brought snow from Wisconsin, across Michigan, and Pennsylvania and into New England. The precipitation shield (Fig. 1) shows the track of the clipper, along the trailing front rain was observed. The data indicated the heaviest amounts of precipitation were observed in southwestern Pennsylvania. This is where the heaviest snow was observed (Fig. 2).
This clipper shared many of the characteristics of the 4-5 December 2007 Clipper and the 7 March 2007 Clipper. This event the low tracked farther north than the Clipper of 4-5 December 2007. An Alberta Clipper is defined (Glickman 2000; Hutchinson 1995; Thomas and Martin 2006) as a fast moving low in northwesterly flow that moves from the plains of Canada across the upper-Midwest, Great Lakes region, and over the eastern United States. It should be noted other names of northwesterly flow events such as this are also known as ?Saskatchewan Screamers" or ?Saskatchewan Schooners?. The name is normally referenced to the Canadian Province the upper-level wave traverses. The more common term Clipper is employed here.
The note will document the Clipper of 29 February 2008 focusing on the impact over Pennsylvania. A mix of model and ensemble data is presented.
|2008-02-22 ||A weak winter storm brought snow, sleet, freezing rain, and rain to the eastern United States and severe weather (Fig. 1) to the Gulf States 21-22 February 2008. The storm was the first significant snowfall (Fig. 2) in many coastal locations from a winter that to date has produced few significant snowfalls in the coastal plain from Philadelphia to New York City.
The heaviest snow, 6 to 9 inches, was observed from north-central New Jersey across southern New York into Connecticut (Fig. 2) and adjacent Massachusetts (not shown). Many of these locations had seen little significant snowfall this winter. Over Long Island the snow changed to rain which may have limited snowfall amounts. A few tenths of an inch of freezing rain impacted the Washington, DC area. Early in the event ice impacted cities in the Ohio Valley.
Forecasts implied that the general evolution of the event was well forecast. There were concerns with regard to precipitation types and the timing of the development of the coastal cyclone.
|2008-02-13 ||A strong winter storm affected the eastern United States from 11 through 13 February 2008. This storm brought snow from Illinois to New York. The snow changed to sleet and freezing rain from Kentucky, eastward across Pennsylvania. Eventually, many of these areas saw the freezing rain turn to rain as the warmer air moved in ahead of the initial surface cyclone. In the warm air over the Mid Atlantic region, a strong low-level jet and above normal PW air brought heavy rain and minor flooding to portions of eastern Pennsylvania, New Jersey, and southern New York State.
For the second time in about a week, a winter storm produced a tornado outbreak in the southern United States. As shown in Figure 1, over 158 reports of severe weather and 17 tornadoes were observed.
Over Pennsylvania, the broad area of 3-8 inches of snow was a rare heavy snow event without a strong easterly jet. Heavy snow with strong south-southwesterly flow also known as ?warm advection? snow is not a common phenomena. The snowfall totals were based on 24 hour amounts and few locations received more than 3-5 inches in any 12 hour period. Southern Pennsylvania saw the snow turn to sleet, freezing rain and rain. Central areas ended as freezing rain until the cold air transitioned the precipitation back to snow.
|2008-02-06 ||Super Tuesday and awful Wednesday: The tornado outbreak of 5-6 February 2008.
A winter storm brought one of the largest and most significant severe weather events to the central United States on 5-6 February 2008. This may have been one of the most significant mid-Winter severe outbreaks in recent history. The event produced 453 severe events, to include 91 tornadoes on the 5th and 59 severe weather reports and 2 tornadoes on the 6th of February (Fig. 1). Preliminary report suggest over 55 people lost their lives in this deadly tornado outbreak making it one of the most deadly tornado outbreaks in 23 years. The last deadly event of this magnitude was in May 1985.
Severe weather in association with winter storms is actually quite common. Galway and Pearson (1981) examined winter tornado outbreaks. Most of these events were associated with winter storms which produced snow in the cold air. They found that these events accounted for about 9% of all tornadoes over the course of any given year. Of the 23 outbreaks they examined, 19 were associated with storm which produced snow and ice in the colder air to the north. The noted that these events produced a large number of tornado related deaths.
The purpose of this paper is to document the severe weather event of 5-6 February 2008. Climate anomalies will be used to show that this storm contained signals, such as anomalously strong low-level winds and anomalous moisture which may have provided clues to the potential severity of this outbreak. Forecast products and analysis products will be presented.
|2008-02-01 ||A winter storm affected the eastern United States from 31 January through 2 February 2008. The storm produced heavy snow on its western edge, from Missouri into the Michigan (Fig. 1), sleet and freezing rain over the central Appalachians from Virginia into New York, and heavy rainfall from eatern Virginia into southern New England (Fig. 1). The snow fell primarily on the 31st into the first and the rain fell mainly on the 1st into early on the 2nd (Figs. 1b & 1c). Most of the precipitation in the Mid-Atlantic region ended in the afternoon hours of 1 February with some rain and snow continuing after 0000 UTC 2 February in New England.|
|2008-01-20 ||An unseasonably cold air mass engulfed much of the United States from 18 through 20 January 2008. As this air mass pushed southward, it produced snow in the southern United States. Snow was reported on Saturday 19 January in Mississippi, Alabama, Georgia, and South Carolina. The cold produced some of the lowest temperatures in the Mid-Atlantic to date for the winter of 2007-08. Overnight lows were in the single digits over most of Pennsylvania with several sub-zero readings in the northwest to the upper teens in the southeast.
For State College, the 7F reading was the 11th lowest for the date. The latest sub-zero readings were on 5 & 7 February 2007; 29 January 2005 and 10 January 2004. There have been 18 days with lows below -10F, 56 days with lows below 5F, and 221 days with sub-zero reading. The coldest period was 19-20 January when the low of -18 was reached and a high of -1 was observed. This event was cold but not as cold as the January 1994 event. There have only been 7 days were the high remained at or below 0F.
This event was relatively well predicted by the NCEP GEFS. As shown both 6 and 2 days forecasts indicated an anomalous 500 hPa trough over the eastern United States and -1.5 to -2SD anomalies in the 850 hPa temperature field. Though not shown, the GEFS also forecast the large and anomalous surface anticyclone associated with the cold outbreak. A large 1036 hPa closed contour was forecast at 21/1200 UTC with +2 to +2.5 SD pressure (Fig. 9 right side) anomalies over the eastern United States, quite similar, though lower pressures, to that shown in Figure 5. Forecasts initialized at 14/0000 UTC showed a weaker cyclone compared to observations and later forecasts.
The large anomalies at long ranges suggested a convergence of forecasts. The convergences in these forecasts suggest low uncertainty and high predictability (Toth et al 2001). This convergence to an abnormal forecast is similar to the results found by Van Den Dool and Toth (1991). The anomalies, specifically large anomalies serve as a relative measure of confidence in the forecasts. It should be noted that forecasters realized the relatively low uncertainty associated with this event and long range forecasts to alert the public to the cold snap were issued with exceptionalAs ensembles increase in use, measures of predictability will be more important. Relative measures of predictability (RMOP) such as those presented by Toth et al (2001) should allow forecasters to distinguish between a high uncertainty event verses a low uncertainty event, and thus recognize when predictability may be high and a 6-9 day forecast may be more skillful than a low predictability event at 1 day. lead-times.
|2008-01-17 ||A light snowfall event was observed on 17 January 2008 over central and eastern Pennsylvania. The snowfall affected the Mid-Atlantic region (Fig 1) with the heaviest snow in northern Virginia across Maryland and into extreme southern Pennsylvania. From a climatological aspect, heavy snow was likely observed in portions of Maryland and Virginia. The synoptic pattern associated with the event was well forecast.
The SREF and other NCEP models forecast the weakening surface cyclone quite well. They also captured the relatively strong low-level easterly jet north of the surface cyclone. The SREF under forecast this feature (Fig. 5) and its intensity relative to the NAM and GFS (not shown) and the NAM analyses (Fig. 3).|
|2008-01-14 ||A winter storm brought snow and areas of heavy snow (Fig 17) to New England on 14 January 2008. Other areas, such as New York, New Jersey, and eastern Pennsylvania were spared this mid-winter storm. Forecasts 3-5 days before the event at times indicated the threat for snow over much of the Mid-Atlantic region and the northeast. The potentially significant winter storm was forecast at some point by the NCEP GFS, the ECMWF, the UKMET, and the CMC deterministic models. This paper documented the event and presented methods to identify uncertain in the forecast process.
The National Centers for Environmental Prediction (NCEP) Global Forecast System (GFS) and the NCEP Global Ensemble Forecast System showed the potential for a major East Winter Storm (ECWS) between 0000 UTC 14-15 January 2008. Initial forecasts showed some run-to-run inconsistencies with the cyclone track. This ?model jumpiness? was a clear indication of uncertainty associated with this potential storm.
In addition to the NCEP models, models from other Centers, such as the Canadian Meteorological Center (CMC), United Kingdom Meteorological Office (UKMO), and the European Center for Medium Range forecasting (ECMWF) showed this potential storm. The solutions varied from run-to-run within each modeling system and there were clear differences between each centers forecasts of the storm.
The model jumpiness and inter-model differences were correctly interpreted to indicate a highly uncertain winter storm. This event served as another example of why a multi-model ensemble will outperform all individual models evaluated over a long period of time, such as a month, a season, or longer period of time. More importantly, cases such as this suggest the value of an ensemble of statistically post-processed output from a multi-model ensemble . In high uncertainty events, a statistically post-processed ensemble system should offer considerable information in forecasting the event.
|2008-01-08 ||Two days of unusually mild January weather culminated in a number of near-record or record high temperatures on Tuesday January 8, 2008 across central Pennsylvania. The "January Thaw" feature from last week on this website advertised the potential for record high temperatures one week in advance.
The pattern over North America changed abruptly in early January 2008. A massive storm brought record snows, rain, and winds to the West Coast and the intermountain West. To the eastern United States, the growing ridge brought unseasonably warm weather to the region on the 6 through 8 January 2008. Record and near record high temperatures were reported across Pennsylvania on the 7th and 8th of January 2008.
GMOS and downscaled GEFS data are shown as forecast aides. The GEFS anomalies predicted this event well in advance. This warm episode was well forecast over 7 days in advance.
|2008-01-03 ||An impressively large anticyclone moved over the United States 1 to 3 January 2008. This system set several pressure records for sites from Colorado to Florida and it brought cold air to the eastern United States. The large anticyclone at 1200 UTC 2 January 2008 is shown in Figure 1. Surface pressure anomalies of +1 or more standard deviations (SDs) above normal dominated the central United States with a large area of +2 and +3 SD anomalies. Pressure anomalies of over 5SDs were observed in south Texas. An addition to the anomalously high pressure the system brought dry air into most of the eastern United States as indicated by the low precipitable water values (Fig. 1b).
The surge of high pressure with the attendant cold and dry air brought a short-lived cold snap to most of the eastern Unite States. Temperatures fell to below zero on colder locations in Pennsylvania and some sites in Florida had the coldest low temperatures reported in over 5 years.
This event was relatively well forecast by the NCEP deterministic models and the NCEP ensemble prediction systems. Several forecasts from the NCEP Global Ensemble Prediction system will be presented to show how well this system was forecast. It begs the question as to why this system was so predictable.
Example using GMOS and down scaled GEFS 2m temperatures in the forecast process as well as the value of anomalies.|