SREF/eyewall secondary website

Pennsylvania State University-National Weather Service Weather and Research Site

Quick Links
NWS State College Threats Page NCEP GEFS CMS EPS Plumes Current Cases PSU Meteorology  

Parallel Site

Current Case Display machine:SREF

If cases do not appear click here for alternate server

For Previous years and events:


Historic Central Pennsylvania Flash Floods of 21 October 2016:A slow moving frontal system produced heavy rainfall in Mid-Atlantic region (Fig. 1) with a small are of extreme rainfall over central Pennsylvania (Fig. 1b). Most of the rain fell in a very brief period of time with spotter gages showing rainfall rates in excess of 3-6 inches in 1 hour and one untrained spotter near Milesburg, PA recorded over 10 inches of rainfall. The Stage-IV rainfall compared to the 6-hour 100 year average recurrence interval (ARI) showed rainfall rates from 0000 to 0600 UTC 21 October 2016 at 100 to 175% of the 100 year 6-hour ARI (Fig. 2). The 24-hour ARI showed two elongated southwest to northeast bands of over 4 inches of rainfall with 100 to 150% of the 24-hour 100 year ARI. More impressive and related to the flash flood nature of the event was the large area of 100 to 175% of the 6-hour ARI values in central Pennsylvania, just west of State College. This paper presents the pattern, verification, and forecasts of the historic and devastating central Pennsylvania Flash Floods 21 October 2016. This paper will also examine the forecast issues in the NCEP models and several more advanced CAMs. .


Mid-Atlantic Rainfall associated a cut-off 500 hPa cyclone: A deep slow moving 500 hPa cyclone, a strong anticyclone to the north, a plume of deep moisture and an enhanced front along the coastal plain of the Mid-Atlantic region produced heavy rainfall from 29 September through 1 October 2016. The verifying QPE fields imply most of the heavier rainfall fell on 29 and 30 September. The heaviest rainfall was clearly along the coastal plain due to the early season cold air damming scenario and a strong coastal front. The overall pattern, the cut-off and strong southerly flow was a classic pattern for heavy rainfall. Thus forecasters anticipated the potential for heavy rainfall based on the pattern. The models, and in the case, the GEFS was able to reproduce the pattern and the general locations where the moisture and lifted were maximized in the each member of the GEFS. Thus despite some spatial errors and grossly under forecasting the observed maxima, the GEFS did reasonably well. From a temporal perspective errors were on the order of 6-12 hours as heavy rainfall continued well into 30 September. Getting the coastal front rainfall correct is unlikely a skill in the coarse GEFS.


Pennsylvania Severe Weather of September 2016:September 2016 was a very quiet late summer month for severe weather in Pennsylvania. However a plume of deep moisture, strong shear, and modest CAPE brought supercells to the region on 17 and 18 September. With lower LCL heights and strong shear the supercells in western Pennsylvania produced tornadoes in northwestern Pennsyvlaina.


Southwestern European Heat Wave of 4-7 September 2016 Record Heat in Spain: A late season heat wave affected southwestern Europe on 4-6 September 2016. The strong subtropical ridged and associated 5940 m high combined with +3s 850 hPa temperatures provided strong signals for the late season heat wave. This heat wave shared many of the characteristics associated with previous heat waves such as the 2003 European Heat wave and the central European heat wave of 2011. Unlike the 2011 heat wave, this event was associated with a far less enduring subtropical ridge. A heat wave of similar magnitude and duration affected France in July 2015 (Fig.7). The subtropical ridge brought above normal 500 hPa heights and a 5940 m contour into Spain and France. The event persisted for several days. Fortunately, large scale ridges are relatively predicable and the NCEP GEFS (Fig. 8) correctly forecast the large ridge over southeastern Europe. Longer ranges forecasts had a weaker ridge due to relatively large spread (Fig. 9) in the forecasts at longer ranges. The spread was low near and in the subtropical ridge and higher in the strong height gradient north of the ridge. The spread and thus uncertainty decreased with forecast length. This lead to a sharper and stronger ridge in shorter range forecasts. Despite the spread and uncertainty, the GEFS forecast a 5940 m contour over Spain with 5-6 days of lead time. The GEFS also correctly forecast above normal 850 hPa temperatures (Fig. 10) and temperature anomalies in the +range. Apparently across northern hemisphere mid-latitudes, the 5940m contour is a good indicator for a heat episodes and heat waves. Additionally, as shown in Figures 8 & 9 ridges are relatively predictable compared to troughs and regions of strong gradients.


Hurricane Hermine the little Cyclone that Could:Hermine developed in the Gulf of Mexico in late August and developed into a category I hurricane before coming ashore in northwest Florida on 2 September 2016. It came ashore with 752 mph winds and gusts of 67 mph. The storm produced heavy rain from Florida to North Carolina. Coastal flooding was observed in North Carolina. As the storm moved off the North Carolina coast, the strong winds and rains moved offshore with it. Despite forecasts for strong winds, heavy rainfall, and coastal flooding from the Mid-Atlantic region into southern New England, the storm was rather uneventful in that region. The GEFS did show a sharp northern edge to the QPF shield and the strong easterly winds. It was a close call for Long Island and eastern New England. It was a close call from a storm with very strong easterly flow. The close call in the Mid-Atlantic region and southern New England impacted millions of people on the busy holiday weekend to close out the summer of 2016. The lack of rain, wind, and flooding did not set well in the media and public. However, it was a close storm with considerable uncertainty and public forecast likely convey the uncertainty than threat as best as practical. It could be shown that shorter range models, such as the HRRR could have helped improve forecasts on timescales of 0.5 to 1 days. Storm following models may help with storms of this nature in the near future. But the state of forecasting today likely contributed to the poor public perception.


The enduring Louisiana rain and flooding of August 2016:Historic flooding impacted Louisiana on 12-15 August 2016 (TWC 2016, USA Today 2016) causing over a billion dollars in estimated damages. This event was the ninth billion-dollar weather related disaster in the United States of 2016. At least 13 deaths were attributed to the flooding (AP 2016b). The maximum reported rainfall was 31.39 inches at Watson, Louisiana. The range of extreme rainfall at observing sites in Florida and Louisiana ranged from 14.43 inches in Panama City, Florida to 27.47 in Brownfields Louisiana (NOAA/NCEP/WPC). Rivers and streams in Louisiana crested above flood stage with 11 gauges setting new record crests. During the event several locations had exceeded all-time record stages and forecasts indicated flood stages that had not been observed before on Comite and Amite rivers. Most of the record crests were archived on 13 and 14th of August. A slow moving tropical wave and a plume of deep moisture brought heavy rain and historic flooding to Louisiana on 12-14 August 2016. At 850 hPa a closed cyclone with strong winds around the cyclone lumbered across the State. The 850 hPa u- and v-wind anomalies were in excess of -3 and +4 respectively with this system. The strong 850 hPa cyclone and the above normal precipitable water likely contributed to the conditions and moisture to produce an enduring and intense rainfall event. Despite some spatial and temporal issues, the NCEP GEFS forecast some incredible rainfall amounts in the Gulf States with at least 5 days of lead time and got the general potential for extreme rainfall for locations in Louisiana with about 2-days lead-time. In addition to the forecast amounts, the GEFS forecast an intense and enduring rainfall event. The 6-hour Stage-IV data were used here to examine the extreme rainfall. The 6-,12-, and 24-hour maximum rainfall rates were impressive. May locations saw several periods of in excess of 100 mm of rainfall and 24-hour rates exceeded 300 mm. The 1-hour data could be used to find the periods of maximum 1-hour rates and the times of these extreme rates. They key finding examining the 6-,12-, and 24-hour rainfall rates was the intensity and the enduring nature of the event. The GEFS correctly forecast a heavy rainfall event in the central Gulf States. Longer range forecasts had difficulty with the location of the heavy rain with the axis initially located too far east. The axis slowly shifted over Louisiana and the GEFS attempted to show the potential for 100 to 200 mm of QPF over portions of Louisiana. It under forecast the extreme values and it could not get the location correct. However, the model provided excellent signals for a potential high end rainfall event and an enduring rainfall event. The GEFS point data showed that the GEFS was able to show the potential for extremely heavy rainfall for a point such as Baton Rouge. But due to temporal and spatial errors the signal was useful with about 1.5 to 2 days of lead time. This was best illustrated in Figure 16. The plan view data provided additional lead time to the potential for an extreme rainfall event in the general region but these data were unable to outline the exact regions. The rainfall forecast by the GEFS was quite large, with member forecasting over 200 mm of QPF. Good knowledge of the GEFS QPF climatology would go a long way to put these data into a more useful context. Clearly, when the model forecasts a record or near record event they can and often do occur in the real atmosphere.


Historic Ellicott City Flood of 30 July 2016:Locally heavy rainfall associated with convection brought devastating flooding to the Ellicott City, MD. Rainfall rates near 4 inches per hour were observed. Rainfall rates of this value are relatively rare occurring less than 0.1% of the time in most locations in the Mid-Atlantic region and if they occur only occur in the warm season. Such rates are not observed in the autumn through spring. The extreme rainfall rates and the urbanization produced severe flooding. The flooding produced extensive damage and led to at least 2 fatalities. The radar and satellite data showed the complex evolution of lines and clusters of thunderstorms. The implications based on radar velocity data was that the relatively weak flow allowed outflow boundaries to remain quasi-stationary and interact with more organized convection to the west. The slow moving boundary likely played the role to focusing the convection over the same region for a prolonged period of time. It fit the MULRMERGE model outlined by Jessup and Colucci (2012). The larger scale models forecast rainfall but they were incapable of the outlining the region where the record rainfall and rainfall rates were observed. They showed a signal at the coast of over forecasting the QPF and under forecasting the maximum QPF. These larger scale models and ensembles are hydrostatic models and require convective parameterization schemes (CPS). In convectively forced situations, hydrostatic models are of limited value. The SREF as shown had too much QPF over too large an area and its maximum from all members was well west of the observed location. The CAMS, represented by the HRRR and the derived HRRR-TLE had a signal here. The HRRR was shown because it is updated hourly and may be better suited to deal with environments which are modified by recent or ongoing convection. Other CAMS were examined but lacking rapid updating they were of limited value. Convection modifies the environment, thus rapid updating models have some signals which may be of value and of value over a smaller regions. In this case, the HRRR and HRRR-TLE showed the potential for heavy rainfall in the southern Pennsylvania and Maryland. Similar to the hydrostatic models, the HRRR could not get the location of the extreme rainfall. To the HRRR’s credit, it had a signal for the potential for heavy rainfall. How forecasters can leverage this signal to be more mindful of the potential of flooding is another issue and may require some training. These models will likely never forecast events of this nature. Thus, a forecast strategy requires using rapid updating CAMS. Noting the signals for regionally heavy and extreme rainfall. This may be identified by the probability of rainfall rates in excess of 3 inches in 1 and 3 hours. When the CAMS have a signal, real-time data from sources such as radar and spotters is still required. Only by monitoring the environment with radar and observational data can effective warnings be issued for events of this scale and nature.


The Persistent Warm month of July 2016:July 2016 will be remembered as a persistently warm month over much of the United States. Approximately 1470 record high temperature records were set or tied during the month (Table 1) and there were 3 distinct warm episodes (Fig. 1) with the most expansive warm episode occurring near the end of the month when 100 records were set or tied on 28 July. The mean pattern from 0000 UTC 22-29 July, computed from 6-hourly data, showed the persistent 5940m contour over the southwestern United States (Fig. 9a). During this period of time the mean 850 hPa temperatures were above normal across the entire United States with regional maximum anomalies over the western and eastern United States. During this period the axis of maximum moisture and PW anomalies extended from the Gulf of Mexico onto the Mid-Mississippi and Ohio Valleys. This case and the enhanced late July heat wave shows the utility of the 5940 m contour in identifying heat waves over the United States.


Historic and deadly West Virginia Floods of 23-24 June 2016:A large scale frontal system with flow over a strong subtropical ridge (Grumm 2016) set up a nearly classic Maddox frontal rainfall event (Maddox et. al. 1979) over the Ohio Valley on 23 June 2016. The larger scale pattern was relatively well predicted in the NCEP global forecast systems and thus these models showed a signal for the potential rain event. However, lacking the ability to produce convection these models failed to produce the QPF amounts relative to observed extreme rainfall. Several CAMS produced significantly higher QPF amounts in close proximity to the general locations where the heavier rainfall was observed. The probability matched mean products showed great promise in forecasting future extreme rainfall events. The NCEP GEFS (Fig. 11) correctly forecast the overall pattern and thus produced lift, as indicated by its QPFs in the correct geographic location. However; relying on convective parametrization schemes and the coarse resolution; the GEFS was incapable of forecasting the extreme rainfall as observed during this event. Despite these limitations, comparing the GEFS QPF to the internal GEFS QPF climate (M-Climate) it is clear that the GEFS was producing an extreme rainfall event within its climate space. It takes some experience for correctly use the GEFS and GEFS M-climate to identify potential extreme QPF events. These data, based on where the model has both lift and moisture, can indicate where in a specific pattern the ensemble thinks the QPF will occur. In this event it was on the warm side of the ensembles frontal boundary (based on 850 hPa temperatures and PW fields not shown). The NCEP GFS (Fig. 12) also produced a significant rainfall event in the Ohio valley. These data were displayed using the QPE/ARI ratios. Relative to the observed QPE (Fig. 7) the GFS had the correct idea of a heavy rainfall event and a weak signal in the QPF/ARI ratios showing ratios in the 50 to 70% of a 24-hour 100 year rainfall event. The observed QPE/ARI ratios reached as high as 125 to 150% of the 24-hour 100 year ARI in some areas of West Virginia. The GFS like the GEFS had a useful signal but could not generate the QPF relative to observed rainfall. The CAMS which are not as widely used had a better signal. The key point here is that the NCEP GFS was able to correctly forecast the larger scale area favorable (Fig. 12) for heavy rainfall. But like all models which cannot produce convection but rely on CPS’s, it was unable to forecast the higher amounts observed. The CAMS shown here including the HREF, HRRR, and NCAR Ensemble all showed great promise for the long term forecasting of extreme rainfall events. Using probability matched mean products these forecast systems correctly produced extreme rainfall amounts relative to forecast systems such as the GFS and GEFS. They also produced some extreme rainfall amounts in the general geographic area where it was observed. This event was strongly forced by a strong frontal system and sharp subtropical ridge. A strong 250 hPa jet was also present. Thus the overall success during this event may be related to the combination of the convective allowing systems and the stronger background synoptic signal. It is unclear how these systems will perform in weaker more convectively based events. These systems show great potential in improving flash flood forecasting the 18 to 30 hour window when there is strong forcing and the system has upright convection. Short-term forecasting of this event indicated that there was considerable training along a weak quasi-stationary frontal boundary. This training contributed to the heavy rainfall and combined with flow into the mountainous of West Virginia, the orography served to focus the extreme rainfall amounts where 200 to 250 mm of rain was observed. During this phase of the event, the rapid refresh models, like the HRRR had an advantage. As the event wound down on 24 June 2016, the HRRRV1 and HRRRV2 limited convection in and around the rain cooled areas of West Virginia . CAMS initiated during the ongoing convection at 0000 and 0600 UTC 24 June had too much QPF and a potential lingering rainfall threat in West Virginia. Correctly leveraging the power of rapid updating and CAMS could help provide improved short-term forecasts as the event reaches its peak and equally aid in better forecasting when the threat for heavy rainfall has passed.


Forecasts of 6000 m height and the heat wave of June 2016: A record early season subtropical ridge affected the western and United States on 19-21 June 2016. During this time the 500 hPa heights reached or exceeded 6000 m. This was likely the earliest occurrence of 500 hPa heights over 6000 m relative to the 1979-2016 CFSRV2 data set. It broke the previous record by about 6 days. This massive subtropical ridge brought hot weather from the western Plains to the Pacific Coast. In the southwestern United States several locations set or tied all-time high temperature records. The hot dry air contributed to massive wildfires which dominated the news from 19-23 June 2016 (AP 2016). This high-impact heat wave was well predicted by the NCEP GEFS and the biased corrected GEFS and NAEFS. All of these forecasts showed massive subtropical ridge and all of these systems forecasts several time periods where the 500 hPa heights would exceed 6000 m. They all successfully forecast a record event. It should be noted that these forecasts not all forecast a record event with the 500 hPa heights going over 6000 m, they forecast the earliest occurrence of a 6000 m contour at 500 hPa over the United States based on the CFSRV2 dataset from 1979-2016. It is also notable that the GEFS (Fig. 8-10), before bias correction also forecast a closed 6000 m contour at 500 hPa. As shown in Figure 10, most of the spread is near cyclones and in the regions of strong flow about the subtropical ridge. Thus the GEFS and the bias-corrected version of the GEFS produced a highest confidence of a record event. Subtropical ridges and ridges in general tend to have longer predictability horizons than troughs and areas of strong gradients. Thus the low spread in this event should have provided confidence in a record event.


redictability Horizons and the cut-off low 21-22 May 2016:A late season cut-off 500 hPa low developed over the Mid-Atlantic region on 21-22 May 2016 (Fig. 1). Near the track of the cut-off clouds and rain kept temperatures unseasonably cold. Areas north of the cut-off were relatively warm. On 21 May temperatures in the Mid-Atlantic region and Pennsylvania were in the 50s while central and northern New England, near the larger scale ridge were in the 70s and 80s. Forecasts of the cut-off low from 6 GEFS forecast cycles indicated considerable uncertainty with the evolution of the 500 hPa wave and cyclone (Fig. 2). As will be shown, this had a significant impact on the weather near the surface, specifically the rainfall and temperature forecast. Many of the forecast issues associated with this event were quite similar to those often associated with a winter storm. The large uncertainty with the evolution of the cut-off cyclone affected the track of the precipitation shield. Thus the concept here apply to many events including winter storms. This paper will summarize the pattern and the forecast issues in the NCEP GEFS associated with cut-off cyclone and rainfall event of 21-22 May 2016. Focus is on the predictability horizon (PH) of this event and how to deal with uncertainty in the forecast process.


Recent Trends in Northern and Southern Hemispheric Cold and Warm Pockets:The climate forecast system was used to examine trends in warm and cold pools and trends in the precipitable water in both the northern and southern hemispheres from 1979-2010. The data were examined by season focusing on potential extremes in warm and cold season in both hemispheres. The data revealed a stronger signal in 850 hPa temperatures and 500 hPa heights in the northern hemisphere. Generally, areas covered by extreme cold air; defined as the areas where the 850 hPa temperatures were equal to or less than -30C (243K); in the northern hemisphere winter are decreasing. During the warm season, the areas covered by precipitable water values equal to or greater than 60mm are increasing. Similar but weaker 850 hPa temperature and 500 hPa height trends were found in the southern hemisphere. However, the trends in southern hemispheric precipitable water were similar to those in the northern hemisphere.


Southeastern Texas historic rain of 18 April 2016:Extremely heavy rainfall across southeastern Texas (Fig.1) produced major flooding in the Houston Metropolitan area during the morning hours of 18 April 2016. The flooding disrupted most human activities closing roads and causing the Houston areas schools to close for two days due to water on roads, and led to at least eight confirmed deaths (Houston Chronicle, AP 2016). Observed rainfall was likely over 17.60 inches (Table 1 based on NWPNS) based on gage data. The Stage-IV data (Fig. 1) indicated 250 to 300 mm (14 inches) of rain for the 24 hour period ending at 1800 UTC 18 April 2016. The Stage-IV gridded rainfall 24 hour total ending at 1800 UTC 18 April (Fig. 2) indicated areas in southeastern Texas had 75 to 125% of the 100 year recurrence interval in the 24 hour period. The Average Recurrence Interval (ARI) is based on the blended NOAA14 and NOAA40 data . The 6-hour period of heaviest rain was observed in the period of 0600 to 1200 UTC (Fig. 2 lower panel). The impacts of the event included extreme flooding, closed schools, and 8 fatalities. A critical issue in the Houston area is underpasses. The majority of the fatalities involved individuals driving into deeply flooded under passes. One individual died after she drove around a barrier before the steep drop-off into the under pass. The other 7 people drove into flood under passes. The City is investigating better means to prevent this type of tragedy. Many of the deaths occurred in under passes where similar deaths have occurred before (Houston Chronicle 2016). This paper will document the pattern in which the heavy rainfall occurred and present methods which may aide anticipating and characterizing this and similar events. The focus is on using climate data to identify the potential for heavy rainfall including the pattern in which the event occurred and the value of Average recurrence intervals (ARI: NOAA14) to characterize the observed and forecast precipitation. The estimate precipitation (QPE) and forecast precipitation (QPE) are shown along with ratios relative to the ARI values. Forecasts from several NCEP forecasts systems are also shown relative to the ARI values as ratios.


Western Plains heavy rain and Mountain snows of 15-17 April 201: A significant precipitation event affected the western Plains on 15-17 April 2016 (Fig. 1). The event included heavy snows along the front range of the Rocky Mountains and heavy rains in the plains of eastern Colorado and western Kansas and Nebraska. The multi-day event in the plains was the result of an amplifying ridge in the eastern North America (Fig. 2) as a deep trough moved into the southern Rocky Mountains. Deep southerly flow developed between the large ridge to the east and the trough to the west. High precipitable water (PW) air, with PW anomalies over 2above normal moved into the Plains from the south (Fig. 3) with the maximum PW anomalies peak at over above 3normal. This paper will document the pattern in which the heavy precipitation occurred and present methods which may aide anticipating and characterizing this and similar events. The focus is on using climate data to identify the potential for heavy rainfall including the pattern in which the event occurred and the value of Average recurrence intervals (ARI: NOAA14) to characterize the observed and forecast precipitation. The estimate precipitation (QPE) and forecast precipitation (QPE) are shown along with ratios relative to the ARI values. Forecasts from several NCEP forecasts systems are also shown relative to the ARI values as ratios.


A Review of warmth of March 2016 Globally averaged, the month of March 2016 was the warmest March on record dating back to the 1880, the first year of the NOAA global temperature dataset (NOAA, 2016). Based on the NOAA data (Fig.1) most of the record warmth was focused over the Indian and Pacific oceans. It was much above normal over a large extent of North America. The pattern over North America (Fig. 2) for the month of March showed a trough (Fig. 2a) over eastern Canada and a ridge over the southwestern Atlantic and Caribbean. Positive height anomalies were located in the ridge and these anomalies extended into the eastern United States. This note summarized the pattern and conditions of March 2016.


Southern Heavy rain and floods of 8-10 March 2016:Heavy rains (Fig. 1) produced record flooding in northeastern Texas and northwestern Louisiana 9-11 March 2016. The period of heaviest rain was the 24 hour period from about 1800 UTC 8-9 March 2016. Some areas of northwestern Louisiana had over 300 mm (11 inches). Point maximum data indicated during the longer period of time rainfall may have exceed 20 inches in some locations. The heavy rains produced record flooding along the Sabine River which runs along the Texas and Louisiana border. Record crests along the Sabine were measured at several locations including Burkeville, Deweyville, and Bon Wier (TWC, USA Today, AP 2016). Several of the sites broke records which were had stood for over 100 years. This paper will document the pattern in which the heavy rainfall occurred and present methods which may aide anticipating this and similar events. The focus is on using climate data to identify the potential for heavy rainfall including the pattern in which the event occurred and the value of Average recurrence intervals (ARI: NOAA14) to characterize the observed and forecast precipitation. The estimate precipitation (QPE) and forecast precipitation (QPE) are shown along with ratios relative to the ARI values. Forecasts from several NCEP forecasts systems are also shown relative to the ARI values as ratios.


Eastern United States Severe and Tornado Event of 24 February 2016:A rare late winter severe weather event produced widespread severe weather from the Carolinas to New England on 24 February 2016. The strong frontal system produced over 20 tornadoes and basically tripled the recorded February tornado climatology for Pennsylvania. The first February EF2 tornado and the first tornados since February 1990 were observed on 24 February 2016. The tornadoes occurred in an environment which rarely develops in February. The surge of deep moisture and anomalous PW along with a strong low-level jet contributed to the instability and shear necessary for severe weather. The PW, 850 hPa winds, and temperatures were all well above normal for February. Surface temperatures were in the 50s and 60s over much of the mid-Atlantic region where the convection was observed and farther south temperatures were in the 70s. The CAPE was relatively low but the strong shear allowed for significant convection to develop. The satellite and lightning data showed a line of storms which produced lightning from Pennsylvania to southern Virginia. The WV image implied a dry slot moving over the low-level moist air may have contributed to the event. The radar data indicated an extremely strong mesocyclone with winds varying by over 80kts across the mesocyclone which crossed Lancaster County, PA. The strong shear in the mesocyclone was detected by the 3 WSR-88D which cover Lancaster County. Base V in the area was observed to be over 100 to 110kts at times. Strong winds, unseasonably warm and moist air came together to produce a rare February tornado event across Pennsylvania and the Mid-Atlantic region.


Eastern United States Big Chill of 13-15 February 2016: A surge of arctic air with a strong and deep 500 hPa trough brought a short-lived shot of extremely cold air to the eastern United States on 13-14 February 2016. The pocket of coldest air had 850 hPa temperatures values between -30 and 32C. In recent years, there has been a dearth of large pockets of sub -30C air at 850 hPa. The crème de la crème of the cold air at 850 hPa used to denote arctic is in North America has been fleeting in recent years. This contraction of the cold air has been document by Martin (2014). This short-lived but intense arctic air mass focused over New York and New England. This led to temperatures below 0F at New York City, Providence, and Boston. New York had not seen a sub-zero reading in decades. Temperatures were well below zero over much of the eastern United States from northern Pennsylvania to Maine with readings in the -20 to -30F range over northern New York and New England. The NCEP GEFS and the joint NCEP-CMC NAEFS should considerable skill in predicting this intense arctic outbreak to include the short duration of the event. The larger scale features to include the large ridge to the west were relatively well forecast. The cold pool over the Northern hemisphere was examined using Python to find the areas -20 and -30C. Total area covered by the -20C contour peaked over North America at 0000 UTC 14 February with a deep cold pocket over eastern North America and Greenland. The total area was about 6.46% of the area of the Northern Hemisphere. The area covered by the -30C air was smaller, about 0.74 percent of the entire Northern Hemisphere. The two notable cold pockets are over Siberia and eastern North America with a third smaller area over northern Greenland. The area of Siberia persisted for several days while the pool over North America quickly dissipated.


Historic Eastern United States Winter Storm of 22-24 January 2016: Record snow along the megalopolitan corridor:The storm developed as a strong 500 hPa shortwave moved into the southern Plains (Fig. 3) and across the Gulf States. This short-wave originated over the North Pacific Ocean (not shown). As this wave moved eastward, strong convergence developed over the northeastern United States (Fig. 3a-d). A strong 250 hPa jet developed in this region (Fig. 6). At 1200 UTC 22 January 2016 (Fig. 6b) a well-developed “banana jet” was present over the eastern United states with an implied jet entrance over the Ohio Valley. At 850 hPa on 22 January a strong 850 hPa LLJ was present with -4850 hPa u-wind anomalies (Fig. 5a-c). The heavy snow (Fig. 2) in south-central Kentucky fell in close proximity to this strong 850 hPa easterly jet. As the pattern shifted eastward, the 250 hPa winds peaked along the East Coast around 0000 UTC 23 January 2016 (Fig. 7c). Between 0000 and 0600 UTC the 850 hPa jet had strengthened along the Mid-Atlantic coast (Fig. 8a-c) with -6850 hPa u-wind anomalies. This strong LLJ was associated with the QPE and snow maximum across the Mid-Atlantic region from West Virginia across northern Maryland and southern Pennsylvania. During this period the coastal cyclone deepened as it moved up the coast (Fig. 9). The surface cyclone pressure in the CFS was about -4 below normal. The strong 250 hPa jet and the strong easterly flow at 850 hPa are well known features associated with many major and historic East Coast Winter storms. This storm with -6 850 hPa jet was one of the strongest ECWS in recent memory and the 850 hPa u-wind anomalies were on the same scale as those observed during Hurricane Sandy in October 2012. The strong 850 hPa winds were a known signal for a potential record setting event. The storm was generally well forecast with the notable exception of the heavy snow which fell well north of many forecast system predictions. web 1


Insert a new Case/Event Protected insert