Here are notes from previous weather discussions held
Fridays in the Meteorology Lab (Endeavour Room 139)
All discussion notes are in reverse chronological order.
For the current Friday Weather Discussion page, please click here.
Fall Semester 2014
December 5th 2014: Keith Roberts led a discussion titled, "Coming to a sky near you: A chance of rain"
The weather discussion led of December 5th began with a visualization of the 10-m wind field around the Northern hemisphere to demonstrate a tool that can visualize and aggregate real-time observations. At the time of the discussion, it was shown that there was anti-cyclonic flow over the northeast region with near seasonable temperatures and broken skies. As such, there was weak subsidence at 700 mb as indicated by 700 mb q-vector divergence. The past 2 weeks had been somewhat active with a southern stream cyclongenesis event around Thanksgiving followed by the passage of a Canadian maritime front on December 1-2.
Focus was turned to the present conditions. The flow at 500 mb over the eastern United States was zonal and progressive. Between Saturday-Sunday, a cold front was forecasted to clear the Eastern seaboard and set-up a baroclinic zone for potential cyclogenesis off the southeast U.S. The location and timing of the cyclogenesis event was considerably uncertain in model guidance. Deterministic model guidance was first analyzed. The differences between model guidance bifurcated into two camps: UKMET/ECMWF/CMC (more progressive and more intense cyclogenesis) and WRF/GFS (slower and weaker cyclogenesis) in the timing and phasing of the northern stream shortwave that was associated with the passage of the Arctic front on December 6-7. A spaghetti tool put up by the WPC demonstrated large differences at the 500 mb in the timing and amplitude of the northern stream shortwave between the two aforementioned model camps at rather short lead times (< 36 h). However, the majority of modeling systems eventually did develop a cyclone off the southeastern US coast by mid-week (Dec 9-10), but the movement of this cyclone and its intensity (1000-1020 hPa MSLP) had tremendous variance. To illustrate the variance in cyclone tracks, the cyclone tracks for various ensemble guidance systems (SREF,GEFS, etc.) was shown with cyclones tracking anywhere between Alabama and Boston at similar times. The large uncertainty was hypothesized to be attributed to diabatic effects associated with the western bound of the Gulf stream wall as upper level QG dynamics did not appear to be largely present.
To synthesize the large spread and uncertainty in model guidance, EOFs were analyzed for some of the GEFS and ECMWF guidance. The first EOF in both ensembles explained > 40 % of the ensemble variance in 500 mb. Day 4 (December 9th forecast) EOF1's sensitivity could be traced back to a disturbance north of the Arctic circle in the Aleutian sea. At this lead-time (+4 days), the sensitivity analysis tool was best suited for identifying where the storm's uncertainty was originating from rather than serving to be physically interpretable. This type of downstream sensitivity to disturbances north of the Arctic circle has been routinely observed in the northeast region since this tool has become operational. It motivates a more complete sampling of the Arctic weather conditions for a more accurate model initialization and theoretical work on the nature of uncertainty and its propagation in modeling systems. The largest sensible weather impacts were high wind on the backside of the Arctic front Sunday-Monday and heavy rain near the coast with a wet mix inland. The potential for coastal frontogensis mid-week as the cyclone moved up the coast was suggested by Dr. Brian Colle due to deformation between the cyclonic on-shore flow (maritime air) and the Arctic air inland associated with the anti-cyclonic over the north east. Due to the large uncertainty in the mid-week cyclogenesis event, it was difficult to say what the sensible weather implications would be, however, some guidance suggested a persistent on-shore flow in the NY/NJ Bight, which could lead to widespread coastal flooding.
October 3rd, 2014: Michael Erickson led a discussion titled, "A Glimpse at What This Upcoming Winter Might Have to Offer?"
Mike started weather discussion with a brief review of the past weeks weather. Long Island found itself on the east side of a strong 500 hPa ridge over the weekend, which resulted in high temperatures reaching the low to mid 80's on Saturday and Sunday. Thereafter a weak offshore low developed and persisted, resulting in cooler cloudier conditions with easterly flow. A weak upper level low propagated into the region, supporting the continued weak offshore surface cyclogenesis and easterly flow. The low finally propagated east today, with a return to something called sunshine.
Thereafter focus shifted to an attempt at a seasonal "forecast," although "experiment" might be a better word choice than "forecast." Mike would be lying if he considered himself a seasonal forecast in any sense. When making a seasonal forecast, Mike focused on lower frequency variability metrics which would likely change slowly such as El Nino Southern Oscillation (ENSO), the Quasi-Biennial Oscillation (QBO), the Atlantic Multi-decadal Oscillation (AMO) and the Pacific Decadial Oscillation (PDO). Similar previous years are identified using these low frequency metrics, and their winters are analyzed. Currently we are in ENSO neutral conditions leaning toward an El Nino with a 60-70% chance of an El Nino forming this fall or winter (according to the CPC). The QBO is negative and expected to peak in the negative phase this winter. The PDO is in a cold phase, although in the shorter term, most of the northern Pacific is experiencing above normal sea surface temperatures (SSTs). In fact, most of the Northern Hemispheres SSTs are above normal, which is an anomalous signal. Finally the AMO is in its warm phase.
All of these metrics are discussed in Mike's 2014-2015 "Experiment." A quick summary of the results show that years most similar to our current 2014 fall experienced below normal temperatures in the eastern 1/3 to 1/2 of the country. Of the most similar years, about 60% of them were below normal. A quick glimpse at precipitation shows above normal precipitation in most of the Southeast United States, normal precipitation for the Long Island region and below normal precipitation for Northern New England and New York. These results should be taken with a grain of salt, but there does appear to be a weak signal in the composites considered.
September 26, 2014: Sara Ganetis led a discussion titled, "Wonderful Weekend Weather, Anyone?"
Please see the ITPA website for a summary.
September 19, 2014: Weather discussion was cancelled for the New Student BBQ.
September 12, 2014: Michael Erickson led a discussion titled, "A Look at the Weather and All it's Glory."
Mike began his vaguely titled weather discussion by taking a look at the previous weeks weather, which was quite anomalous for the western half of the country. First on the board was the anomalous rain and flooding that occurred in the Southwestern United States on 9/8/2014. This anomalous event was likely caused by the circulation of Norbert modifying the flow to it's east, possibly in conjunction with some Atlantic influence. This resulted in severe flooding , particularly in Arizona.
Focus then shifted to an anomalous cold air outbreak in the Plains banking up against the Rockies combined with a lee cyclogenesis event. The result was an early season snow event along the front range on 9/11/2014. Snow totals ranged from a dusting or light coating in Boulder, CO and Rapid City, SD, to 14" in Story, WY.
Finally, focus shifted locally to Long Island where the weather has been quite pleasant, not just recently, but for most of the summer. A glimpse at the LGA timeseries of temperature compared to normal showed consistently normal summer temperatures with an absence of extended periods of anomalous warmth or cold. One slight exception to this occurred in early September where temperatures averaged a few degrees above normal. The long range forecast showed continued pleasant weather with the chance for a decent cool air intrusion from Canada by next Friday.
September 5, 2014: Keith Roberts led a discussion titled, "Fall Weather Outlook and Brief Summer Weather Recap."
Spring Semester 2014
May 2,2014: Keith Roberts led a discussion titled, "Central Plains Severe Weather Outbreak of April 27th - 28th"
The discussion was focused on the severe weather outbreak on the 27th and 28th of April, 2014 in the southeast United States. There were 108 reported tornadoes over the two-day period starting on April 27th. This severe weather caused 32 fatalities and 42 injured in Mississippi, Arkansas, and Georgia. Surface analysis charts valid for April 27 16 UTC and April 28th 12 UTC from the WPC were shown. There was a warm occluded low over the central Great Plains with a trailing cold front extending from northern OK to southeastern Texas on the 27th of April. However, this cold frontal boundary appeared weak and a well-defined temperature gradient could not easily be found at the surface nor a wind shift. At upper levels, the cold frontal boundary was more diffuse. 500 mb heights showed two main cut off lows over the CONUS and a blocky, stagnant pattern. The frontal structure resembled more of a dry line given the moisture gradient.
Radar reflectively showed precipitation initiation around local noon in the southeast US around Mississippi and Georgia. A sounding from Jackson field airport (JFD) in Mississippi (ahead of the cold front by about 200 km’s) revealed a strong capping layer around 950 mb with a significant dry layer aloft (inverted v-sounding). Nearby soundings showed a similar inverted v vertical thermal structure. Some severe weather parameters were analyzed. Sfc-6 km shear was around 60 kts with more than half of this shear contained in the lowest 3 km’s. Dewpoint depressions were on the order of 10-15 C ahead of the surface frontal boundary. CAPE was on the order of 3500 J/kg. Given the dewpoint depressions and low-level shear, it was suggested that linear structures would be the likely convective mode. Indeed the bulk Richardson number was in agreement with an average value around 40. Pure supercells are often supported with values around 20. Reflectivity revealed quasi-linear structures with embedded super cells initiated around 18 UTC. To the north in southeastern Illinois and Iowa, the convective regime was different. In this region, the majority of damage was associated with penny to quarter sized hail and heavy precipitation (HP) supercells. No tornadoes were reported in the northern Great Plains.
Thereafter the discussion focused briefly on the forecast for the local region. 18 UTC NAM and 12 UTC GFS combo showed that the weekend in the Northeast would feature a weak cold frontal passage Sunday afternoon associated with a northern stream 500 mb shortwave. Precipitation looked diffuse and mostly driven by large scale QG PVA. There was no forecasted severe weather potential for anywhere in the CONUS over the following three days.
April 25th: Zhenhai Zhang and Minghua Zheng led a discussion titled, "A brief summary of last winter's storms and the potential severe weather event over the Central U.S. this weekend"
The discussion began with a brief review of the winter storms during last winter (from November 2013 to March 2014). Comparing with the winter storm climatology (the mean of 1979-2012 winters), last winter is a normal winter overall. For example, the total number of cyclones during last winter over East Coast is 64, which is very close to the historical (1979-2012 winters) mean, 62. The intensity distribution (minimum central MSLP) of the cyclones doesn’t show any significant differences to the climatology either: there are slightly more (about 10%) weak (>1000hPa) and intense (<985hPa) cyclones, while there are about 15% fewer moderate (985-1000hPa) cyclones. However, more than 50% of these cyclones occur during February and March. Over the East Coast, during 2013 winter the cyclone frequency (cyclone center counts) decreases by about 30%, 23% and 10% for November, December and January separately, but it increases by about 18% and 50% for February and March comparing to the climatology. We can get a similar conclusion from the cyclone track map for individual month: there are significant more cyclones during later last winter, especially in March. One of the most important reasons for that is the temperature anomaly in that month. During 2014 March, the mean near surface temperature is 3~8K lower than the climatology (1979-2012 March mean) over most area of the continent, meanwhile, the near surface temperature is 1~2K higher than the climatology at low and middle latitude over Western Atlantic. As a result, the temperature gradient along East Coast, which is vital important to cyclone genesis and developing, is significantly larger in 2014 March than the climatology. To conclude, although 2013 winter is a normal winter for the total cyclone activities, most of these cyclones concentrate in later of this winter, especially in March, which is mainly caused by the significantly strong temperature gradient along East Coast. For early of this winter, there are fewer cyclone events, however, there are many wave packets over North America. We can see these nice wave packets in the figure of mean (30-55N) meridional wind, and many intense cyclone and heavy precipitation events occur over the downstream of coherent wave packets.
Attention then turned to the major weather systems for previous week. An animation of the surface SLP and fronts showed a frontal system enters Great Lakes on 21st April (Monday). It crossed central PA and NY in the evening of 22nd April (Tuesday). The surface low moved to New England on 23rd (Wednesday) with a minimum pressure of 994 mb in the morning. In the upper level, the 500 mb height developed a cut-off low over NY and PA in the morning of 23rd April. This H5 Low then moved northeastward and left New England on 24th (Thursday).
The discussion then focused on the system that may bring severe weather in the weekend. A short wave trough had move to the southern Plain in the evening of 24th. It then propagated eastward and reached the Southeast on 25th. At the same time, a surface low pressure system has developed over IN, OH and KY. The frontal system was crossing the western Southeast during the morning of 25th. The latest SPC soundings over TN and AL display the conditional instable environment, as well as the condition favoring supercell thunderstorms and hails. Based on SREF 12Z run, the shortwave trough and the surface low system will advance eastward in the afternoon and move through the northern Mid-Atlantic and the NE during this evening and Saturday morning. The NAM model shows that the low system will be over the southern NE. Considering the model consistency, heavy rain associated with strong to severe thunderstorms will develop across parts of the Mid-Atlantic and Southeast in the afternoon and the evening of Friday. On Saturday, the upper level low will move through and the cold from may trigger thunderstorms over Mid-Atlantic.
Out west, another shortwave trough from eastern Pacific had reached Northwest in the afternoon of Friday. Based on the regional models, the upper level dryline will mix eastward into central KS and across western TX in the southern Plains during Saturday evening. A strong low-level jet streak will favor the moisture advection through Saturday and aid in the development and provide a violent nature of updraft motion favorable for damaging hail, winds, and tornadoes over central and southern Plains. Due to the large-scale upward motion, the steep lapse rates will support a large hail threat across the southern Plains. The dew point will rise to 60s.The strong low pressure system will spawn supercell thunderstorms capable of generating large hail, damaging winds, and strong tornadoes. So, a large area of high populated regions in KS, OK and TX will be dealing with severe weather events on Saturday, Sunday, and Monday. The most dangerous day appears to be Sunday, and the SPC has issued their "Moderate Risk" forecast of severe weather over portions of OK, AR, TS, and LA that day.
Thereafter, the discussion focused on the Rossby wave packet moving across eastern Pacific. Based on GFS ensembles forecasts initialized at 00Z 25th, there was a wave packet over eastern Pacific, which will propagate enter southern Plains on 00Z 27th. This could be another case that the Rossby wave packets are associated with severe weather events: severe weather events tend to develop downstream of wave packets. This wave packet will be weakening on Monday and merge into the leading part of another robust wave packet which will be over the central Pacific on Tuesday. This new combined wave packet will sustain the existing short wave trough over the Central US and bring significant severe weather over central US and Mid-Atlantic.
April 18th: Matt Sienkiewicz led a discussion titled, "A look at this past week's April snow event"
April 11th: Sara Ganetis led a discussion titled "The Current Weather 'Down Under': Tropical Cyclone Ita's Assault on Australia"
The discussion began with an overview of the current weather over the Northeast U.S., mainly highlighting the well-received warmth experienced during the day prior to a cold frontal passage. Central Park reached 74F, for example. A topic that has been being addressed every week is the possible onset of El Nino based on the progression of a warm anomaly present near the thermocline slowly reaching the surface of the Eastern Pacific which can be seen by animating the temperature-depth plot from CPC. While it may seem apparent that El Nino is imminent, there is still the issue of the Spring predictability barrier that may mean that although the signal seems strong right now, there may be other atmospheric responses that aren't in-line with El Nino.
The discussion then switched gears to discuss a few notable events of the past week especially the spreading of wildfires in the region. Research is still ongoing at SBU looking into the predictability of high fire threat days which is tied to atmospheric moisture (or lack-thereof) and gusty winds. Last Sunday evening, 6 April, there was a wildfire at a State Forest in southern NJ that was identified via radar and discussed through social media outlets by local NWS WFO's (Mount Holly & NYC). What was interesting about this particular fire is that as the sun set, a strong temperature inversion was present that persisted overnight. That meant that the smoke and particulate matter could not be vertically dispersed but instead had to travel horizontally which it did with a strong southerly wind northward to NYC and points west. Smoke was able to be smelled far away from where the fire originated and an air quality alerts were issued for several counties. Some snapshots can be found here.
Tropical Cyclone Ita formed northeast of Australia in the South Pacific in waters with SSTs exceeding 30C and favorable environmental conditions (i.e. relatively moist air mass and far-removed from the reaches of extratropical waves that could enhance the environmental shear). At the time of the discussion, it had already crossed the Great Barrier Reef and made landfall (~10 UTC 11 April) in between the cities of Cooktown and Cairns on the eastern side of the Cape York Peninsula in Queensland, Australia with estimated wind speeds of 125 kts. A loop of the IR imagery shows that it developed a very small, symmetrical eye before making landfall and having that eye fill-in a bit due to frictional convergence, steep terrain and removal from surface fluxes. An ASCAT pass prior to landfall (2002 10 April) showed the fastest winds near the eye as well as along the edge of the continent in what looked like a barrier jet. The Cooktown station time series of pressure, sustained wind speed and wind gusts shows that the pressure decreased to around 983 mb around 15 UTC 11 April and the sustained wind speed (gusts) got to >40 (65) kts. The JTWC forecast track of TC Ita had the system recurving back out to sea in the South Pacific. A look at the GFS deterministic forecast showed that while Ita had persisted for so long and moved due west under the influence of a broad area of high pressure over the Coral Sea, a trough was moving eastward and expected to shift the track more southward and then eastward with time. A 250-hPa jet also developed poleward of Ita suggesting that it will have undergone extra-tropical transition. Ita was an interesting tropical system that produced some interesting discussion.
The weather discussion concluded with a local forecast for that Friday (11 April) evening and then the system forming over the weekend and expected to drag its cold front across the region Tuesday-Wednesday. The SBU-WRF was used to discuss the near-term forecast for that evening, mainly how much precipitation should be expected behind the cold front? The answer was different for both the NAM-WRF and GFS-WRF, with the NAM-WRF being the drier solution. Comparing the current regional radar with that of the NAM-WRF and GFS-WRF forecast reflectivity showed that the GFS had much larger areas of higher reflectivity than the NAM and the NAM was closer to reality so perhaps we wouldn't expect much rain with the cold front. The next system was expected to form lee of the Rockies in a developing trough out ahead of a cold, Canadian air mass and likely cause some severe weather in the Plains around Sunday. The fact that the southerly flow out ahead of this system is forecast to be strong and persistent and a high pressure off the East Coast should allow for an influx of moisture towards the coast means that there should be some anomalously high precipitable water values and thus very heavy rains Tuesday-Wednesday. We quickly compared the deterministic GFS with the ECMWF QPF forecasts to find that the ECMWF was slower and weaker with the precipitation totals than the GFS. That will have to be something we keep in mind as we get closer to the event. Check out the weather discussion page for a few visuals to go along with the above text.
April 4th: Nathan Korfe led a discussion titled "An Analysis of the Recent Extreme Weather Event in the Central U.S."
March 28th: Matt Sienkiewicz led a discussion titled "Inside a Bomb: This Week's Impressive Nor'easter"
The discussion was almost entirely devoted to the significant offshore cyclogenesis event of March 25th and 26th that was responsible for sustained blizzard conditions on Cape Cod, Martha's Vineyard, and Nantucket. Three short wave systems, two diving out of the Canadian provinces and one originating in the Gulf of Mexico, coincided off the east coast of Florida. Aided by divergence in the right entrance region of a very strong upper level jet streak, as well as the strong low-level baroclinicity over the Gulf Stream, rapid cyclogenesis occurred as the system developed into a very impressive extratropical cyclone. The Ocean Prediction Center's experimental lightning density product showed areas of strong convection associated with the cold front, as well as mesocyclone-like structures north and northeast of the low center. Satellite imagery available on the University of Wisconsin CIMSS blog showed relatively dry, ozone-rich stratospheric air diving southward and wrapping around the low center. Ship reports from around 12Z on the 26th, when the cyclone was due west of Virginia Beach, show was ship to the west of the low center experience 55 knot winds and 26 ft seas. ASCAT passes at 15Z/26th and 03Z/27th show the evolution of the hurricane force wind field southwest of the low center as the storm makes its way from the offshore Mid-Atlantic waters up to Nova Scotia. Satellite altimeter passes at 2230Z/26th showed 48+ ft seas generated by the area of hurricane-force winds produced, resulting from the cold conveyor belt.
Several NDBCstations showed interesting observations for this event. NDBC moored buoy 44027, located at the mouth of the Bay of Fundy, showed a wind gusts exceeding 100 knots. The National Weather Service office in Caribou, ME claims that while the buoy only had one working anemometer at the time, the instrument showed no signs of malfunction so the readings may quite possibly be legitimate. Moored buoy 44150, located south of Yarmouth, Nova Scotia, took almost a direct hit from the low center and recorded a minimum central pressure of 955 hPa. The swell from this system was predicted to be enormous, stretching across the entire North Atlantic basin. Buoys obs from 07Z/27th showed 54 ft seas southeast of Sable Island, and the OPC 24/48/96 hr wave forecasts predicted the swell to traverse the North Atlantic and arrive in Spain with significant wave heights exceeding 8 meters.
http://glacier.somas.stonybrook.edu/wxdisc
March 14th: Sara Ganetis led a discussion titled "Will we be shaking our snowflakes instead of our shamrocks this St. Patrick's Day?"
Before starting a discussion of the forecast for the potential snow event slated for St. Patrick's Day (17 March), Sara spoke about the mid-week event that was responsible for significant snow and ice throughout northern New England but only gave us 0.40 inches of rain (at ISP) and wind gusts up to 35 kt (at ISP). The wind was predominantly post-frontal as opposed to occurring due to the pre-frontal jet or momentum mixing via the precipitation along the front. The NOAA/GOES-R JPSS blog highlighted this by providing a discussion of the MODIS air mass imagery that shows a clear stratospheric intrusion of dry air that helped with the post-frontal mixing. Another note of interest that Sara discussed was the fact that the Climate Prediction Center (CPC) issued an El Nino Watch which meant that although the El Nino Southern Oscillation (ENSO) signal is neutral and will likely remain as such, there is a 50% chance that an El Nino (or warm phase of ENSO) will develop in the spring and summer months. We took a moment to look at the water temperature anomalies with depth which when animated showed a warm anomaly in the Central Pacific moving eastward and getting closer to the surface.
The discussion now turned to the next extratropical cyclone event as this pattern has been progressive and it seemed there was discussion of one in some form for the past several weeks! The cyclone was slated to form in the Deep South as a shortwave rounded the Rockies and enhanced a longwave trough in place. It was then likely to translate eastward towards the coast where it would encounter enhanced baroclinicity from the Gulf Stream and ample Atlantic moisture and upper-level dynamical support from an elongated jet would support broad ascent and significant snowfall. The question, however, was where would that northern periphery of the precipitation shield? Would we be shaking our snowflakes instead of our shamrocks on Long Island this St. Patrick's Day?
The key players were the southern stream shortwave and longwave trough, and upon looking at the current satellite imagery around 4:30 PM EDT (2030 UTC) 14 March, that shortwave was just about to reach the West Coast in the Pacific Northwest. The fact that this feature was entering our observational domain meant that the models would likely converge on a solution with subsequent model runs, which as this is being written post-event on 17 March, was true for the global models but the NAM flip-flopped a little bit. The precipitation is highly tied to the track of the system and so a discussion of the uncertainty of the track ensued. Mike Erickson's cyclone tracks page showed the considerable spread among the GEFS members. Dr. Colle provided some insight into the ensemble sensitivity analysis which showed that the northern stream played a role in that if the heights to the north of the cyclone were lower that it would favor a more southern track and miss Long Island. This northern stream sensitive region was unfortunately originating from the data-sparse region of northern Canada which may explain the differences between the global models (ECMWF & GFS) when compared to the regional, mesoscale models (e.g. NAM). Dr. Colle also showed that this system was on the leading edge of a well-defined wave packet originating from the Pacific.
To conclude the forecast discussion, we decided to take a look at a the WPC official forecast, few deterministic models and finally the short range ensemble forecast system (SREF). WPC called for about a 40% chance of > 1" snowfall for Long Island, with the gradient to much higher probabilities just to the south of the Island. Their model diagnostics discussion essentially discounted the NAM for being too weak, fast and northward than the rest of the available guidance. The GFS and ECMWF had some agreement with most of the precipitation remaining just to the south of the Island. Model uncertainty became clear when we decided to take a look at the SREF and parallel SREF, the plumes in particular. The ARW cores were much more aggressive with the forecast precipitation, giving ISP as much as 21 inches of snow. The mean was approximately 6 inches with just under an inch as the lowest amount. As another post-event thought and having received nothing at ISP, it's interesting that the 0900 UTC runs didn't have more members showing less, or even any showing zero. Sara concluded her discussion by showing the SBU-WRF and how at forecast hour 72, both solutions were closely tied to their initial and boundary conditions with the NAM-WRF being the wetter solution than the GFS-WRF from the 12 UTC runs on 14 March. Will we be snowed-in for St. Patrick's Day? Unlikely (and we weren't)!
March 7th: Mike Erickson led a discussion titled "A glimpse back at our meteorological winter and potential future happenings."
Meteorological winter (defined as the months of December through February) has ended, so the beginning of this discussion focused on analyzing the coldest months of the year. Mike started with a followup of his "seasonal forecast" from the weather discussion on November 22nd, 2013. In that Power Point, multiple low frequency signals such as the El Nino Southern Oscillation, the Atlantic Multi-Decadal Oscillation, the Pacific Decadal Oscillation and the Quasi-Biennial Oscillation were used to composite years with similar Autumn conditions. The winters following the similar Autumn conditions were then displayed as a proxy for the 2013-2014 seasonal forecast. The followup Power Point reminded everyone of that "forecast" and then compared it to what actually happened using the North American Regional Reanalysis (NARR) data.
The consensus seasonal forecast was very poor compared to reality, unless the user multiplied the forecast by -1. The strongest signal in the seasonal forecast was a negative height and temperature anomaly extending from the Gulf of Alaska to the Northwest Territories of Canada, with a positive height/temperature anomaly off the coast of California and in the vicinity of Greenland. There was little if any anomaly signal across most of the eastern two thirds of the United States, suggesting equal chances for warm or cold. In reality, Alaska had a strong positive height/temperature anomaly along with the western Atlantic, while a strong negative height/temperature anomaly prevailed in south central Canada southward into the Great Lakes and Ohio Valley of the United States. Daily animations of height/temperature throughout the winter showed frequent surges of cold air originating from central Canada plunging into the eastern half of the United States, especially in January and February. Overall, North Dakota extending eastward into the Great Lakes were the coldest with winter surface temperatures running greater than 5 degrees Celsius below normal. Closer to the east coast and further south, temperatures averaged 1-3 degrees Celsius below normal with occasional warm ups. Before continuing on to near real-time weather, Mike stressed that he is not a seasonal forecaster, and that the skill of his forecast does not reflect the skill of the current science.
Mike briefly discussed the storm "bust" of Monday March 3rd. Forecasts the previous Friday were suggesting the potential for significant snowfall. However, models shifted south throughout the weekend. This was shown in an animation of cyclone track positions for the storm (also in the PPT above). Mike stressed that given the difficulty of 36-60 hour forecast predictions this winter, caution should be applied when predicting future storms. This point was stressed when looking at our next storm potential between 3/12-13. Mike first looked at the forecast for the weekend and early next week, which showed dry, sunny and generally above normal weather after a coastal storm moved off to the south early Saturday morning. Cold air attempted to move back in later Tuesday while lee cyclogenesis was taking place around Nebraska. The Global Forecast System (GFS) had a relatively weak coastal move south of Long Island that, taken literally, would drop a few inches of snow. The European Center For Medium Range Forecasting (ECMWF) had a stronger storm slightly further north, suggesting more of a rain event for Long Island. A more in depth look suggested that the GFS was producing a storm primarily from an Arctic shortwave, while the ECMWF was phasing both a Pacific vortmax and an Arctic vortmax. Mike closed with the combined cyclone track forecasts from the GEFS, FNMOC and CMC ensembles. Not surprisingly, there was a very large spread in ensemble solutions, with many members saying rain, snow, or a miss to the south for Long Island. Like any forecast for a storm 6 days out, it is best to make a mental note of what could happen. After that, we watch and wait.
February 28th: Mike Layer led a discussion titled "The Winter Brigade Continues into March!"
The discussion began with a loop of the 500 hPa heights over the Northern Hemisphere. Over the past week, the loop showed a trough over the Midwest/Great Lakes region, which translated east and amplified as a piece of the polar vortex broke off and dove down into southeastern Canada. The loop also showed the large cutoff low over the Pacific moving towards California. Looking ahead the next several days, the cutoff low looks to dissipate as it moves into the SW U.S. and gets absorbed into the mean flow. Heights rise over the eastern U.S. as a new trough develops over the central U.S. and moves eastward.
SPC mesoanalysis was used to analyze the system that moved through the NE U.S. on Thursday. Low pressure moved through SE Canada, dragging a cold front through the region. Brief heavy snow showers affected the area during the early afternoon, associated with steep low level lapse rates (7 to 9 C/Km) and weak 700 hPa frontogenesis. The 700 hPa loop also indicated a very strong north-south temperature gradient, with temperatures ranging from around -5 C in southern VA to -35 in western NY state. The cold front moved through the local area around 7:30 PM, as indicated by the sharp temperature drop and increase in winds thereafter as measured by the HSC weather station. Temperatures by Friday morning dropped to 5-10 degrees on average across the area, setting records lows at KBDR, KISP, KJFK, and KLGA. These four sites also observed record low highs for the date, with temperatures only rising into the low to mid 20s.
Attention then shifted to the massive extratropical cyclone spinning off the California coast, which most models were initializing in the low 970s hPa. The San Diego NWS page illustrated numerous hazards in effect for the region, including Coastal Flood Warnings, Flash Flood Watches, High Wind Warnings, and even Winter Storm Warnings for the higher elevations inland. Widespread heavy precipitation was forecasted, with many areas expecting 3-5 inches of rain, and up to 7 to 10 inches of liquid equivalent in the mountains. Since the area has been experiencing a major drought for quite some time, all of this rainfall at once was expected to be a little too much of a good thing, resulting in flooding issues.
Looking ahead to the new month, the threat of another winter storm was taking shape in the Mid-Atlantic and NE U.S. for Sunday into Monday. This storm was expected to develop from the energy associated with the low affecting the SW U.S., spawning a new low around the southern Mississippi Valley this weekend. Models then showed this low riding ENE along an arctic boundary, remaining rather weak but spreading a large swath of precipitation to its north and west. Model spread was still considerable, with the CMC maintaining a further south track, and thus only giving the local area a light snowfall. The GFS, which had been showing a consistent 1+ inches of QPF across the area, trended southward and lighter on the 12z run. Meanwhile, the NAM and SREF were the warmest and furthest north solutions, indicating a mostly rain event for the area. Ensemble sensitivity analysis revealed that the storm track was actually highly influenced by a shortwave rotating around the Polar Vortex over the Hudson Bay area, and not so much related to the southern stream energy.
February 21st: K.R. led a discussion titled "An Analysis of the Feb. 20th Central Plains Cyclogenesis Event"
An animation of the H5 height field along with low level relative vorticity for the previous 4 days (extending back to the 15th of February) showed a H5 shortwave digging into the Central Great Plains from western Canada into the Dakotas spawning the development of a surface cyclone around Colorado on the 17th of Feb 2014. The focus of the discussion was on the rapid surface cyclogenesis associated with the aforementioned H5 shortwave. The 18z WPC surface analysis on Friday Feb. 21st showed a ~980 mb low centered over the Great Lakes with a trailing cold front extending the Appalachian mountain range and a warm frontal boundary extending eastward from the Great Lakes roughly parallel with the 40 N latitude line bending southward over the mid-atlantic states. The 18 UTC RAP model derived sounding for KOKX (Brookhaven) showed a well-defined warm frontal boundary which had passed northward of KOKX and the general LI region around 18UTC. The warm frontal boundary was nearly stationary as of 20 UTC over the LI Sound and southern Connecticut and, due to the widespread snow cover overlaid with strong low-level warm air advection, there was a large inversion present for the majority of the morning hours generating widespread fog with visibilities < ½ of mile for the majority of the tri-state area. University of Albany QG height tendency maps were analyzed and the amplification of the H5 shortwave was largely attributed to large values of differential warm air advection ahead of the trough axis. 700 mb q-vector convg. showed a well defined couplet which was forecasted by the 12z GFS to wrap around the surface feature associated with the low’s occlusion.
The discussion then focused on the cold frontal boundary that was forecasted to traverse the general LI region around 22-23 UTC. The SPC mesoanalysis site was used to analysis the kinematics. 850 mb miller frontogenesis revealed a strong area of frontogenesis displaced 100-150 miles behind the axis of greatest reflectively. Given the orientation of the isentropes and the axis of dilatation (< 45 degrees), it was argued that the cold frontal boundary was intensifying north of the Delmarva and weakening southward of Delmarva. Due to the intense thermal gradient ,a large vertical wind shear was present associated with the frontal boundary which triggered the concern for tornadogenesis over west VA, south NJ, and Maryland. forcing and marginal instability in the warm sector. There were no verified reports of tornadoes at the time of weather discussion. Surface obs. from the south P lot and the HSC building showed a vertical migration of the frontal boundary that was hypothesized to be attributed to the marine boundary layer as the frontal zone interacted with the cold winter S.S.Ts. 12z WRF-ARW from the (COMAP WRF page) cross sections (d’-d) of the frontal passage showed significant sensitivity to model resolution with the 12 km solution depicting a rapid decrease in surface frontogenesis with the primary forcing moving aloft (700 mb) as the cold frontal zone moved over the marine boundary layer.
Thereafter, the discussion focused onto the future weather. A weak frontal boundary was progged to pass through the northeast Sunday night into Monday morning, but both moisture and forcing was weak. Ensemble sensi. analysis showed there was large uncertainty in terms of timing. Precipitation looked to be all liquid given low-level southerly flow and changing over to snow due to evap cooling into a mid-level dry layer, but this depended heavily on the progressive nature of the model solution. The next event on the horizon was Tues/Wed/Thurs time period associated with a Miller B cyclone track (redevelopment of nrth. branch off Carolina coast). All global models showed coastal development however there was tremendous spread in terms of intensity and track as the result of a complex phasing process between both two surface features. All models showed the northern stream surface feature creating a weak area of low level convergence over the northeast region but few model solutions actually showed any precipitation associated with the actual coastal.
February 14th: Mike Layer led a discussion titled "Analysis of this Week's Winter Storm and the Potential for More Snow this Weekend"
The discussion began with a look at the 500 hPa height evolution over the Northern Hemisphere for the past week, showing an active pattern with shortwave troughs digging into the Great Lakes and Northeast U.S., while ridging persisted over the West. Looking ahead to the following week, the evolution indicated a pattern change, with heights building over the eastern U.S. and troughing developing over the western and central U.S. A major culprit of this upcoming change was attributed to a piece of the polar vortex breaking off and diving south towards Alaska, amplifying into a large cutoff low.
Current conditions over the Northeast showed the major winter storm that just affected the area now located over the Canadian Maritimes. The low was down to 972 hPa, and the tight pressure gradient in its wake was producing strong westerly winds with gusts up to 45 mph at KISP. However, mostly sunny skies and high temperatures approaching 40 degrees were allowing for some melting of the snow cover.
Attention then turned to a review of Thursday's winter storm. SPC Mesoanalysis archive showed heavy precipitation in the form of snow moving through the area during Thursday morning, aided by very strong 700 hPa frontogenesis. The area was then dryslotted for much of the afternoon and early evening. Warm air pushed in at 850 hPa, with temperatures approaching 7C in the evening over Long Island, while surface temperatures remained near 32 degrees. Persistent drizzle during this time allowed for around two to three tenths of an inch of ice accretion in parts of the region. More widespread precipitation returned as the storm rapidly deepened off the coast in the evening, leading to some heavy thunderstorms in the NYC Metro, and heavy snowfall to the north and west where temperatures were colder. As the storm pulled away and cold advection ensued, all areas turned back to snow and received additional minor accumulations. Total snowfall from the storm ranged generally from 8-16 inches across the area, with about a foot observed near Stony Brook, most of which fell within a few hours during the morning. The highest totals were observed across the Hudson Valley, where some locations saw 20+ inches. Season snowfall statistics provided by the NWS revealed that the area is running well above normal, with several climate sites already in the top 10 for snowiest winter.
The SPC Mesoanalysis archive was also used to look back at the same storm, which delivered a major snow and ice event to the southeast U.S. earlier in the week. With warm advection in the low levels, but surface temperatures remaining near or below freezing due to cold air damming over Georgia and the Carolinas, many areas saw a prolonged period of frozen precipitation. Some locations observed close to a foot of snow, with areas a little further south receiving major ice accumulations up to around an inch, resulting in widespread damage and power outages.
The discussion ended with a look at another potential snow event for Saturday. Most computer models indicated low pressure moving eastward from the Tennessee Valley, off the Mid-Atlantic coast, and then bombing out rapidly as it moved northeastward off the New England coast. Many models had the storm deepening into the 970s or even 960s as it approached the Canadian Maritimes by Sunday morning. There was some differences among the models in how close to the coast the cyclone would come, leading to a very volatile snowfall forecast for the local area. The 12z GFS was the closest to the coast, bringing QPF amounts over three quarters of an inch to eastern Long Island, while the GGEM and NAM were a bit further east, bringing only a couple tenths of an inch to the area. The NWS was forecasting 2-4 inches of snow for most of the area, with 4-6 inches for eastern Long Island and SE Connecticut. Areas with the best chance of seeing significant snow (12''+) were SE New England and eastern Maine.
January 31st: Michael Erickson led a discussion titled "We will warm up, but for how long?
Mike began with the 500 hPa Northern Hemisphere Geopotential height evolution over the previous seven days, revealing that the polar vortex was split in two, with anomalously low heights in Siberia and Southeast Canada. This resulted in a very pronounced and persistent trough over the past week, which helped set the stage for an anomalous snow and ice storm in the Deep South on 1/28/14. Mike also showed the 500 hPa height anomalies for the month of January, revealing that the pattern has become more blocked than in December 2013. This has resulted in more persistent warm or cold temperature anomalies in the United States rather than the more transient cold air outbreaks of December.
The topic then shifted to the impact and predictability of the 1/28/14 winter event in the Deep South. Atlanta, GA was significantly impacted by this storm since most people were leaving school and work during the peak of the storm, resulting in gridlock lasting over 24 hours. Numerous stories of people abandoning their cars after 12-24 hours and even of a childbirth in a car, resulted in a heated debate and finger pointing. Although some government officials blamed forecasters, Weather Research and Forecasting (WRF) runs 24 to 48 hour prior to the event showed a growing snow threat for the Atlanta Metro region. In addition, the NWS issued warnings for the region several hours before the snow began. It is likely that this gridlock could have been reduced or prevented with better communication or understanding given that the forecast was relatively accurate.
Discussion then shifted to "catch up mode" since there hasn't been a weather discussion in some time. The El Nino Southern Oscillation (ENSO) continues to remain quiet and likely is having little impact on our weather. Even so, the subtropical jet hasn't played much of a role in our weather with most of our snow events forming as a result of northern-stream shortwaves. In late December, the Arctic Oscillation (AO) shifted from a positive to negative phase, indicating a breakdown of the polar vortex with the potential for more blocking in the Mid-latitudes. Meanwhile, most of Long Island has been moved from D1 (moderate drought) to D0 (abnormally dry) as a result of more precipitation in December and January.
The last several minutes of discussion were reserved for the future. As promised in the title, this weekend will feature a low cutting up through the Great Lakes and a warm up for Long Island. Highs will get into the 40's before a cold front moves through on Sunday. A weak system will pass by to the south of the region on Monday while temperatures drop to slightly below normal levels. The models have been showing an active weather pattern for next week with two potential storms on Wednesday and Saturday. However, the intensity and track for both remain up in the air and could bring rain, snow, a wintry mix or nothing to the region.
Fall Semester 2013
November 22th: Michael Erickson led a discussion titled "A look at the upcoming big cool down, weather up to Turkey Day and a brief attempt at a seasonal forecast."
Mike began with a discussion of the weather pattern over the United States during the past several weeks. In general, recent weather has consisted of a storm moving northeast through the Great Lakes region (a so called "lakes cutter") followed by a strong Canadian high pressure system. As the high pressure moves offshore, a warm return flow from the southeast develops, another "lakes cutter" occurs and the pattern repeats. This has resulted in a relatively dry pattern for Long Island with large temperature swings from day to day. Given our consistent weather patterns of late, the discussion transitions into the winter.
Mike used a very simple "seasonal forecast method" by looking at years similar to November of 2013. These similarities are based on low frequency oscillations in the ocean (the El Nino Southern Oscillation, the Atlantic Multi-Decadal Oscillation and the Pacific Decadal Oscillation) and atmosphere (the Quasi-Biennial Oscillation). In general, composites of similar years reveal no significant deviation from normal with respect to temperature (precipitation was not analyzed). This sparked a debate as to whether a composited year with no temperature anomaly can have any forecast skill. The Climate Prediction Center (CPC) winter forecast was similar with equal chances for above and below normal temperatures.
The conversation transitioned to the weekend forecast, where an Arctic blast of cold air is expected to arrive on Sunday. There was little variability in model solutions for Sunday, with the potential that highs on Sunday afternoon might not get above freezing. Finally, the discussion concluded with the potential for an East Coast storm next Wednesday. This was unfortunate timing since Wednesday is the day before Thanksgiving and one of the biggest travel days of the year. Several global models and a couple of ensembles were shown. Although the models have recently trended toward a storm track very close to the coast (i.e. rain for the big cities but potential snow inland), the ensemble spread is still significant. Therefore confidence was low regarding any forecast specifics for this storm.
November 15th: Matt Sienkiewicz led a discussion titled "Super Typhoon Haiyan: A Retrospective Analysis."
The discussion was based around the frustrating lack of available in situ data for Super Typhoon Haiyan, the storm that decimated the Philippines before making landfall again just north of Hanoi, Vietnam. The cyclone tracked just south of Taiwan's Dropwindsonde Observations for Typhoon Surveillance near the TAiwan Region (DOTSTAR) flight reconnaissance region, and it is likely that gps dropsondes from their G-IV aircraft would not have found their way near the eyewall anyways, so maximum wind speed measurements were very unlikely to begin with. Published maximum wind speed estimates were made available from the Japan Meteorological Agency (JMA) and the Joint Typhoon Warning Center (JTWC) of ~144 mph and 195 mph, respectively. The estimates were made through analysis of satellite imagery, as described by Sara Ganetis in the previous week's weather discussion. However, the value of 195 mph is suspicious, and an attempt was made to analyze photos of the aftermath in the Philippines to lend credence to that value. Photographs showed trees stripped bare of all of their leaves, but the majority of the trees appeared to be still standing. Almost all destruction seemed to be due to the horrific storm surge, and less directly due to the wind.
November 8th: Sara Ganetis led a discussion titled "Super Typhoon Haiyan and a Discussion of Model Uncertainty Regarding Next Week's Possible(?) Mid-Atlantic Coastal Storm."
Sara began her discussion by commenting about the recent frontal passage that ushered in a much cooler air mass into the Northeast but how the sensible weather over most of the CONUS was quite quiet. Everyone's focus was shifted towards the Northwest Pacific where Super Typhoon Haiyan was wreaking havoc in the Philippines and was plotting a course for Vietnam. The typhoon formed over waters near or exceeding 30 C and tracked westward in an environment with very low shear (< 10 m/s) which allowed it to form a nearly perfectly symmetrical eye and central dense overcast (CDO) around the time of its estimated peak intensity before its first landfall. There was some discussion among the media that falsely claimed that this storm was the strongest tropical cyclone in history which isn't true. Also, there aren't usually aircraft flown into typhoons in the Northwest Pacific so all of the wind speeds had been estimated from satellite imagery with what is known as the Dvorak technique. Super Typhoon Haiyan did reach the highest level of 8 on the Dvorak Intensity Scale, but a lack of ground-based measurements will mean that it is unlikely that the strongest winds produced by Haiyan will not be measured and reported. Other data from satellite included lightning observations. Lightning was found to be very prominent in the eyewall of Super Typhoon Haiyan but it is unclear exactly why beyond the understanding that the updrafts must had been most intense that allowed for a lot of variability of lofted hydrometeors. The anticyclone in place north of the Philippines steered it straight westward towards the Islands, keeping it in reach of deep, tropical moisture (PWAT values > 50 mm). In addition to the winds exceeding 100 mph, heavy rains are a hazard with this system as can be seen from this radar loop. Once it passes through the Philippines, it should round the southern flank of the anticyclone and maintain its strength as a Super Typhoon due to the very warm waters and relatively low environmental shear (26-29 C and < 20 m/s) and start on a more northwestward track towards the rugged coast of Vietnam where it should rapidly weaken and will likely cause terrible flooding as its moisture is ringed out over the rugged terrain. This storm is incredibly powerful and our thoughts are with all of those affected.
The national and local media had already been reporting on a possible Mid-Atlantic coastal storm that one or two models had resolved for the middle of next week. The problem with the hype was that there was incredible variability among all of the different models (i.e. GFS, ECMWF, CMC, UKMET) as well as considerable run-to-run variability (i.e. between consecutive model iterations). One hypothesis as to why there was hype about a storm is because the one model to resolve it for a couple of model runs was the ECMWF (aka the Euro) which, since Sandy, has been considered to be the best available model, which may be true for some situations but not necessarily for a coastal low development off of the East Coast of the United States. The NWS NYC WFO issued an informative and entertaining graphic that addressed the uncertainty around the possible event. In looking at the models, the 00Z and 12Z Euro runs both had a coastal low that was trending to be more offshore with each subsequent run. The GEFS and therefore also the operational GFS had either no low or a surface low much farther eastward over the Atlantic Ocean. It's one thing to look at the model output and see the variability of possible solutions in the medium-range but that doesn't provide much insight into why there is variability. One can speculate that maybe it's all stemming from the lack of uncertainty in the initial conditions emanating from the Arctic, where observations are sparse. Another method to look at physical reasons for uncertainty is by ensemble sensitivity analysis (ESA) which is done in-house through the SBU/NWS CSTAR collaboration. From looking at the first principle component's sensitivity, one can deduce that a deeper and slower trough originating from NW Canada will favor more of a coastal storm. Therefore, with each subsequent model run a forecaster can have more of an idea of the speed and amplitude of the trough in question to determine whether or not the storm is likely. At the time of the discussion, a coastal storm looked unlikely for Wednesday and Thursday of next week (13-14 November).
November 1st: Keith Roberts led a discussion titled "Oct 31-Nov 1st Cyclogenesis Event over Eastern U.S. & UK Windstorm."
The weather discussion held on November 1st, 2013 was focused on a cyclogenesis event occurring over the central CONUS. The discussion began with a brief overview of visible and IR satellite imagery. These imagery showed a large comma-shaped cyclone centered over Michigan with a trailing cold front extending from the northern Great Lakes down to the southeast portion of the US as of 22Z. Despite the rapid intensification of the parent surface low, doppler reflectivity showed sporadic and poorly defined precipitation structures associated with the cold frontal boundary. It was shown through the SPC mesoanalysis site that the majority of precipitation was outrunning the surface temperature boundary and, as a result, the frontal boundary was experiencing frontolysis. The cold frontal boundaries precipitation structure appeared to be maintained by H7 quasi-geostrophic omega.
Model guidance leading up to this event prompted local WFO’s to issue wind adv. for a large portion of the tri-state region. However, wind advs. did not verify and precipitation forecasts were largely overpredicted by .3 to .5 inches. Highest average wind speeds and gusts occurred during mid-morning with the passage of a small albeit well-defined band of precipitation around 11 am EDT. This feature dropped around ~ .10 of an inch of rain along LI. A narrow rope-like band was associated with this precipitation, its progression in time series observations revealed a very weakly defined surface temperature and moisture gradient. Despite its weak surface presence in the basic state fields, this feature had maintained its intensity and structure over hundreds of miles eastward independent of the actual surface forcing mechanism; this structure was highlighted for further study.
NCEP Ensemble sensitivity analysis guidance was shown from a 2-day lead perspective. Guidance showed decent ensemble coherence in timing and intensity of the MSLP surface feature which at the time was over the Great Lakes. However, the mean MSLP feature was underpredicted in intensity by the ensemble spread by 5-7 mb. Some research questions were posed by Dr. Brian Colle regarding the physical mechanisms for the poor numerical guidance regarding the frontolysis of the frontal boundary. Dr. Colle hypothesized that the topographical interaction between the Appalachian mountains and the surface front is often the culprit for downstream errors in numerical guidance.
The overall synoptic evolution and forcing mechanisms driving the rapid cyclogenesis of this event were shown. In-house wave packet guidance revealed a large wave packet envelope extending from the central CONUS to the mid-pacific. As this large wave packet envelope propagated eastward with the mean flow, it encountered a favorable area for cyclogenesis. A large area of positive PV anomalies was present over the CONUS. The wave packet excited a pre-existing cut-off H5 low over the SW US. Rapid cyclogenesis was aided through an upper-level moisture source from the GOM with PW in this region exceeding 2.5 inches. The H3 jet orientated itself on a PV boundary on the right-side of the positively tilted H5 cut-off low in the SW US facilitating rapid baroclinic energy conversion.
A brief look at the chain of cyclones that had impacted the UK and the Scandinavia region earlier in the week (Oct 25-Oct 28) was shown. A powerful surface cyclone developed in the Labrador Sea (min. MSLP ~960 hpa) and propagated eastward along the North Atlantic largely maintaining its intensity. This cyclone had a powerful upper level jet streak associate a trailing frontal boundary. These winds were able to mix down to the surface and caused widespread wind damage in southern UK and impacted a large portion of central Europe. ASCAT satellite derived winds were shown that revealed the highest average winds to be around 70 kts in a region near the German Bight.
October 25th: Mike Layer led a discussion titled "A look at the cold shot affecting the Eastern U.S. and a potential storm threat around Halloween."
Mike began with a loop of the 500 mb heights across the Northern Hemisphere, showing the troughing over the Central U.S. shifting eastward over the past week. The dynamic tropopause loop showed the tropopause heights lowering as the trough amplified and moved eastward. Looping forward over the next week, tropopause heights are forecast to rise again over the East as ridging returns and another trough develops over the central U.S. A look at the current SPC mesoanalysis showed a large 1030+ hPa high pressure centered over the Tennessee Valley, keeping conditions dry and unseasonably cold for much of the region. Locally, 850 hPa temperatures were well below 0C, in the -4 to -6C range. Attention was also brought to an interesting equivalent barotropic feature spinning off the Pacific NW coast over the past week. An animation of the daily temperature departures for the ASOS sites over the CONUS was run from Oct 1st to the 23rd. It showed the persistent warmth over the East for much of the month, that switched to well below normal over the past few days.
A look at the teleconnections on the CPC website showed the AO forecast to soar positive this week, with the NAO, currently near -2 standard devations, expected to return to near neutral to slightly positive. The PNA, currently slightly positive, is expected to drop to around neutral. ENSO conditions remain generally neutral, with a slight drop in SST anomalies over Nino 3.4 in the past few weeks. Turning the focus more locally, the updated Drought Monitor indicated the development of D1 drought over much of Long Island, with D0 scattered around other parts of the region. The annual precipitation trace for LGA showed precipitation running nearly 10 inches below normal, with no significant rainfall since June, and barely any rainfall for the month of October.
Attention then shifted to the forecast for the next 7 days. Low pressure was expected to cut into SE Canada Saturday, bringing gusty SW winds to the area. Models indicated a warming trend through the week as heights increase over the East and low pressure develops over the Central U.S. by around Wednesday. There was considerable spread between the operational GFS, ECMWF, and GGEM, as well as the GFS ensembles concerning both the strength and track of the low pressure, but the general consensus is that the low will deepen over the Great Lakes and then cut into Canada, dragging a cold front into the NE around Halloween. A stronger low tracking closer to the area would increase the chance for a significant rain & wind event for the region. A look at the ensemble sensitivities indicated that the storm's predictability had some connection with the recurving typhoons over the Western Pacific.
October 18th: Keith Roberts led a discussion titled "Medium Range North American Forecast"
Keith led a discussion with a focus on the time period extending from the present to the end of October and emphasized on larger type atmospheric phenomena (time scale ~ 15 days). The discussion began with a recap of the past weather over the CoNUS (continental United States) for the previous week. An “Indian summer” type pattern was observed for most of October with H5 anomalies over both the eastern U.S. and the Aleutian region with near climo height field over the middle part of the U.S. Drought like conditions > D3 persist throughout the majority of western great plains and parts of the Rocky mountains. While the northeast was not under any considerable drought conditions, small pockets of D0 drought are still present throughout the majority of LI and Connecticut (which has been a recurring theme through Fall 2013) which have persisted since late July. Locally on LI, the region is about 1.8 inches below climo for the 30 day period ending on the 18th of October.
A couple relevant teleconnection patterns were analyzed with the help of this(http://www.kylemacritchie.com/) website. In particular, a focus was on the NAO and AO. Both the AO and NAO were trending to near neutral conditions and forecasted to peak in the positive phase for at the least the first half of November (according to the GEFS). However, it was noted that both these teleconnection indices generally have low predictability for monthly and seasonal time scales and statistical forecasts are more skillful than human intuition. The state of the equatorial Pacific region has been shown to have considerable relations with mid-latitude atmospheric phenomena in literature, however, it too was weakly-defined (near neutral ENSO) to conclude anything. Probabilistic forecast guidance using the NMME was shown (http://www.cpc.ncep.noaa.gov/products/NMME/NMME_description.html) which indicated strong confidence of above normal temperatures throughout the south western part of the U.S, but no apparent temperature or precipitation signal was verifiable for this region. With the absence of strong signals, it was concluded that the rest of October to the beginning of Nov would feature near climo conditions.
To conclude the discussion, a brief look into the upcoming weather for the next week was shown by referring to some figures from (http://jasoncordeira.weebly.com/maproom.html) of tropopause isentropic temperatures. A positive-tilted longwave trough was expected to amplify in the middle part of the country as a >100 kt jet rotated around the leeward side of this longwave trough. This large system would eventually drag a series of cold fronts through the Northeast (one late Sat. overnight & and one around Monday morning) with the period seeing mostly cloudy skies for the majority of the next 4 days. It was shown that the longwave trough was connected to an early-season arctic air mass, and it would become noticeably cooler throughout the next week with high’s in the mid to upper 50’s throughout most of the NE. The possibility of lake-effect snow was emphasized given the presence of an Arctic air mass very early in the season as lake temperatures remain relatively warm...
October 11th: Sara Ganetis led a discussion titled "A Tale of This Week's Two Storm Systems: Their Relation and Impacts."
Sara began her discussion with a loop of the past week's satellite and surface analyses from WPC. The main features of interest was the eastward progression of the system that Matt spoke about during last week's weather discussion and the severe weather that resulted from the passage of its strong cold front across our area on Monday, 7 Oct. The SPC issued a tornado watch for most counties west of NYC and conditions were favorable for strong winds at the very least. One EF1 tornado verified near Paramus, NJ. A topic of discussion was the fact that the cold front weakened in apparent intensity (weaker reflectivity, no severe reports) as it crossed Long Island. We discussed the different factors that may have been responsible that included insufficient instability because CAPE values were < 500 J/kg from a lack of solar heating due to the pre-frontal cloud cover, the convection at the forefront of the precipitation associated with the front outran the "actual front" which separated it from the dynamics and forcing for lift, among other reasons.
The next topic of discussion was how the remnants of Tropical Storm Karen merged with the cold front offshore of Georgia and with the help of a weak upper-level shortwave was able to form into a separate coastal system which meandered up the coast. While there wasn't much upper-level dynamical support to strengthen the system, there still was a lot of moisture that was being transported westward onshore. This led to a lot of accumulated rainfall at such locations as Harrisburg, PA. The high pressure to our north kept LI dry for the most part and the pattern set in place over the Northeast resembled a rex block.
The global tropical activity was briefly looked at because of the high-impact storms of Phalin and Nari that should make landfall as very powerful storms in India and the Philippines, respectively. Typhoon Wipha may impact Japan. With all of this activity in the Western Pacific and Indian Ocean it's not surprising that the flow in the Northern Hemisphere is pretty amplified. One of the systems, likely Wipha looks to recurve and merge with the jet stream and a wavepacket is forecast to result from this merge which may affect North American weather in the long-range period.
October 4th: Matt Sienkiewicz led a discussion titled "A look at Tropical Storm Karen, as well as other weather systems not affected by the government shutdown."
Matt began with an overview of the 500-hPa heights for the past week, highlighting the mostly zonal flow over the North American continent and touching on the climatological record-setting warmth in the region. Nowcasts and forecasts for 500-hPa heights suggested a regime shift to a more progressive, amplified pattern as cyclone after cyclone was predicted to form over the Great Plains. There was some discussion as to whether tropical cyclone activity in the western Pacific could have caused this amplified pattern, and Rossby Wave Packets were analyzed to try to connect the two.
Next up was Tropical Storm Karen, which was floundering in the Gulf of Mexico. It had been experiencing a lot of shear and convection had set up predominantly on its eastward side, hindering any further intensification. Most recently convection had started firing closer to the circulation center and the storm looked to be moving in to an area of slightly weaker shear. This, accompanied with an approaching upper-level trough, gave it an outside chance at re-intensifying before making landfall. Model tracks at that time had the storm grazing the Mississippi Delta and making landfall on the Alabama coast as a Tropical Storm. Also discussed was a recent post from the Albany Maplist that highlighted observations from buoy 42001 in the Gulf earlier that day. The buoy showed a 10+ degree Fahrenheit spike in temperature accompanied with severe convection, and the question discussed was if it was due to a convective heat burst or just bad observations.
Finally, the discussion was wrapped up with the cyclone making its way across the Great Plains. The low center was located in Nebraska, and heavy banded snow was found to the north-northwest in South Dakota. The Storm Prediction Center had posted a moderate risk for the areas out ahead of the eastward-moving low, and reports of large hail in eastern Nebraska and Iowa had started coming in. The discussion ended with a quick look at WPC QPFs for the next few days to get a feel for the precipitation expected in the region from the approaching system.
September 27th: Michael Layer led a discussion titled "A Look at the Upcoming Weekend Storm in the Pacific Northwest and the Forecast for the Next Week."
Mike began with an overview of the conditions across the lower 48. Conditions were dry across much of the country, with a large area of high pressure over the eastern U.S. Some showers were occurring in the western central Plains in association with a low pressure system over Colorado, but stronger thunderstorms were expected to develop ahead of a cold front as the low moved northeastward. SPC had a slight risk (5% tornado, 15% hail and wind) from the Texas panhandle, northward into Western Oklahoma and Kansas. SPC mesoanaylsis showed low-moderate CAPE (500-2000 J/kg) and 0-6 km shear of 30-40 kts moving in from the west, along with low-level moisture advection from the south.
Attention then shifted to the major storm system set to impact the Pacific Northwest over the weekend. Low pressure over the Gulf of Alaska was expected to move southeastward as the trough amplified over the region. The NWS issued High Wind Watches and Warnings for coastal Oregon and southern Washington for wind gusts of 60 to as high as 80 mph Saturday into Sunday. Heavy precipitation (in the form of snow in the higher elevations and rain elsewhere) was also expected over the region. The HPC 7-day precip accumulation graphic (valid through next Friday morning) showed 5 to locally over 10 inches in parts of Oregon and Washington, most of which would fall in the Saturday through Monday time period.
For the Northeastern U.S., a weak area of low pressure well offshore was forecast by some models, such as the GFS, to retrograde far enough west to bring clouds and a chance of light rain to coastal sections on Monday. Looking ahead to the remainder of the week, a ridge was forecast to build over the eastern half of the U.S., with dry conditions and a warming trend expected. Westerly flow was expected to advect unusually warm air into the region, with the GFS indicating 500 mb heights building to 585 dm and 850 mb temperatures reaching 17C by mid to late week. The NWS graphical forecast products showed high temperatures going from the low to mid 70s this weekend to near 80 in the NYC metro area by Wednesday and Thursday. There were some signs of the pattern becoming more amplified by the end of the week, with a trough and associated low pressure system deepening over the Great Lakes by Friday. However, the CPC 6-10 and 8-14 day outlooks indicated a good chance of above normal temperatures continuing for the east.
September 20th: Michael Erickson led a discussion titled "A Look at the Joyous and Wild Temperature Swings of Autumn.”
Mike began with a review of the Catastrophic Flooding Event in Colorado that took place between September 9th and 15th where some areas received over 20" of rain. Although the upper level height field during this time did not appear very threatening, monsoon moisture was advected in from the south, while low level topographic up slope and a stalled frontal boundary was in the vicinity. The worst of the flooding occurred between September 12-13 when numerous showers and thunderstorms were in the region. The discussion switched to Atlantic tropical activity, which remains generally quiet. Humberto has dissipated and there are two tropical interests in the Atlantic, although both of them have a low chance of formation. The main story is Super Typhoon Usagi which is forecast to head toward China and another region of interest to Usagi's southwest that should be watched. This particular region of interest is currently forecasted by many GEFS members to become a typhoon and recurve toward Alaska in several days.
Finally the discussion switched more toward the proposed title and focused on our temperature ups and downs in the Long Island area. More recently we started with a very warm early September and then switched to below average temperatures for the past week. The NYC area has also experienced a growing rain deficit over the past couple months because most of the inland convection has dissipated before reaching the coastal plain. A look toward the future shows that generally below normal temperatures will continue, but not before a brief warm up for today and tomorrow.
September 13th: Weather Discussion was cancelled on account of it conflicting with the SoMAS New Student BBQ.
September 6th: Sara Ganetis led a discussion titled "While we cool down in the Northeast, tropical activity is heating up."
Sara began discussion with a review of the local severe weather event of 3 September. Convection fired up along a cold front as it translated across the island in an environment with plenty of moisture and instability that allowed for heavy precipitation and severe hail (1.00 inch in diameter). The passage of the cold front and setup of the broad trough over the Northeast made for a comfortable end of the week with a decrease in temperature and humidity.
The discussion continued to the recent activity in the Tropics. Tropical Storm Gabrielle had since dissipated after crossing over the harsh terrain of Hispaniola and into an unfavorable environment of drier air and higher shear (just downstream of the upper-level trough). Tropical Storm Lorena was making landfall along the southern tip of the Baja California Peninsula and then would likely dissipate. There were a few areas of possible development, but nothing more than 10% within the next 48 hours because of the proximity to unfavorable environments (too much shear, not enough moisture). The most hope for development at the time of the discussion was for a fresh wave coming off the coast of Africa that was just scattered convection but expected to organize as it moved over the warm ocean waters. The CIMSS Tropical Cyclones page was used to examine the tropical environment.
The discussion concluded with a brief forecast for the area of a cold frontal passage for Sunday that may force some scattered showers in the area. Other than that, there wasn’t too much of interest to discuss. If a hurricane does not form by 11 September then that will break the previous record of the latest formed hurricane in the Atlantic. Hopefully by next week there will be an active TC in the Atlantic so stay tuned!
August 30th: Brian Colle led a discussion titled "Why has the Atlantic hurricane season been so quiet, and how long will it continue?"
Dr. Colle began his discussion by providing some statistics to put the current Atlantic hurricane season into perspective. He showed NOAA’s official forecast from May for 13-20 named storms with 11 hurricanes and 3-6 major storms. Their latest updated forecast called for 18 named storms with 8 hurricanes and 3 major storms. Despite the relatively “slow start” they are still calling for a lot of activity. The climatological peak for TC genesis in the North Atlantic is about the week of September 10th (NOAA/NHC), so it’s not like we already missed the expected peak period of activity. Dr. Colle discussed the quantity called accumulated cyclone energy (ACE) which measures the relative intensity of each storm by estimating the energy used by each storm and is proportional to the square of the wind velocity (more info here). On a per-month basis, ACE also peaks in September in the Atlantic (from Dr. Ryan Maue’s page). Given this information, perhaps this Atlantic hurricane season shouldn’t yet be written off but it looks like the date of the first hurricane formation may break records (McNoldy, CWG). ACE on a globally integrated scale has been decreasing since 2005 (Maue 2011) but that may be a result of there being less intense storms during the later period (Maue’s page).
What was the deal with the start of this Atlantic hurricane season? To date there have been 6 storms. ENSO isn’t a major player this season, so Dr. Colle showed some plots to explain why the activity had been low starting on 1 July. The first point that he made was that there was anomalously dry air above the boundary layer (-12% RH anomaly) stretching westward from Africa all the way across the Atlantic into the Caribbean especially during the period 1-15 August. The Saharan Air Layer (SAL) provided a harsh environment for TC development and growth. In analyzing vertical stability, the Tropics are more stable this year compared to climatology (McNoldy, CWG) so that would act to discourage convection. SST anomaly maps didn’t show too much of an explanation for why activity has been weak. Maps of 500-hPa shear anomalies did show that during the period 1-15 August there was a 2-3 m/s westerly shear anomaly in the tropical Atlantic basin. Therefore, the weak activity was shown to be likely tied to the dry air and westerly shear that created an unfavorable environment for TC development.
How long will this stifled activity last? Likely not long, Dr. Colle explained. Climatological shear values are back in place and the dry anomalies are starting to weaken. At the time of the discussion, there was an invest area with a 40% probability for the chance of development that should move westward into an area with weaker shear but still some residual dry air. If it makes it past the subtropical high, according to the GFS, it may encounter a trough that may recurve it and keep it away from the Caribbean and East Coast. We’ll see what happens. Looking beyond this one system and to the rest of the Atlantic hurricane season as a whole, most students at the discussion agreed that they were not ready to give up on the season yet. Another factor that was discussed was the MJO moving into Phase 8 and then Phase 1 which has been shown to increase precipitation in the tropical Atlantic (NOAA/CPC).Spring Semester 2013
Spring Semester 2013
May 3: Matt Sienkiewicz led a discussion titled, "Record May Snow in the Midwest."
Matt began his discussion by displaying a loop of the 500-hPa heights that shows how much the pattern had changed since early last week that ultimately led to the record May snow event in the Midwest. There was clearly a regime change due to persistent ridging along the East Coast and the deep trough that formed across the Central US. The 500-hPa height pattern suggests that this is the result of the downstream propagation of energy from a strong trough over the Eastern Pacific early on in the period. In looking into this further, neither the AO nor NAO could really be used to see a clear shift.
The system in the Midwest produced up to 18 inches of snow in parts of Wisconsin. Was the record May snowstorm accurately predicted? After some discussion, the consensus was that the models captured the trough in the medium range but the uncertainty was around the snowfall totals. We looked at the environment during the times of heaviest snowfalls and found that the precipitation was collocated with strong 700-hPa frontogenesis along the cold front. We looked at model-derived soundings to see that the isothermal layer around 0 C was present from near the surface to 850-hPa due to melting effects. The consensus of the group was that we were glad we weren’t forecasting precipitation type during this event.
A discussion about the connection of the extra-tropics to the tropics ensued after looking at the surface observations and analysis and seeing how far south the cold front was analyzed (i.e. southern Mexico). Dr. Colle stated that he and some colleagues had done some research that a rope cloud can be formed that propagates into the Gulf due to ageostrophic down-gradient flow caused by the cold pool of the cold front in that region of Mexico.
The system in the Central US and its persistence has had negative effects on flooding, especially for the Mississippi River. Some stages affected by the snow in the Upper Mississippi Valley were expected to come down from moderate flood levels on Friday 3 May, while others in the lower Mississippi Valley were expected to approach moderate flood levels due to the forecast precipitation. The SREF ensemble mean was showing 2 inches of rain in the next 12 h (12-00 UTC 3-4 May). Matt showed that the location of the surface low and subsequent circulation was allowing moisture to be advected by the mean flow from the Gulf of Mexico and even westward from the Atlantic.
Matt ended his discussion with a forecast for the next week. The GFS and ECMWF both agree that the low will cutoff and remain pretty stationary until coming up the coast sometime next Thursday-Friday. This may threaten our de Zafra Symposium with poor weather, but we’ll be inside anyway so it should all work out! A May “Ridge of Death” sets up over the Central US which makes for very poor conditions for anyone who may be trying to storm chase during what should be a very active time. The ensembles are not exhibiting much spread with the previously discussed patterns. Matt concluded his discussion by saying that the Knicks should have pleasant weather for their bus ride back to NY after getting destroyed by the Celtics. (Edit on 4 May: Matt ate his words after the Celtics lost).
April 26: Sara Ganetis led a discussion titled, “Oh Hi, Spring! A Look at the Large-scale Pattern and its Forecast of Persistence.”
Sara began her discussion with a look at the Northern Hemispheric infrared satellite loop to motivate her discussion. Her discussion focused on the pattern change from progressive troughing in the Northeast to a predominantly zonal flow and the forecast ridging to persist in the region until the middle of next week. The slight chance of severe weather in the Central Plains and the threat of fire weather over the Northeast were also discussed.
The previous week’s progressive pattern that led to a pair of cold frontal passages across the Northeast was shown by looping a plot of the potential temperature on the dynamic tropopause which also showed the forecast raising of the DT with the forecast pattern into next week before the 12 UTC 26 April-initialized GFS forecast resolves an amplified trough digging into the central U.S. and becoming cutoff. This trough and resultant cutoff will be something to pay attention to in subsequent model runs. Another thing that was noticed was an area of anomalously lower heights at 500-hPa in the Central Pacific. This is associated with a cutoff at 500-hPa that persisted since ~10 April and is thus the reason why the PNA began to trend negative. A negative PNA in April is associated with ridging and anomalously higher temperatures in the Northeast, as we are expecting.
Taking a closer look at the current weather, the 15 UTC surface analysis from 26 April showed a weak high pressure system over the Great Lakes but the main player is the low located over the Texas panhandle. There was showers and thunderstorms collocated along and north of the warm front. The slight risk is for most of Oklahoma and upon looking at the soundings from 12 UTC 26 April, there is a large elevated mixed layer and MUCAPE > 1000 J/kg. The shear seemed sufficient for the development of supercells. The low-level southerly flow out ahead of the 500-hPa shortwave located over New Mexico should aid to advect moisture from the Gulf into the region to increase the dew points. At the time of the discussion there were no storms fired up yet. Mike Erickson added to the discussion by showing his fire threat webpage and discussed how the low relative humidity and relatively strong winds caused the Haines index to exceed the moderate fire threat level.
The discussion was wrapped up by taking a look at this morning’s 12 UTC OKX sounding that showed a very strong surface inversion that was caused to the sounding location being located near the Pine Barrens that are very efficient at radiatively cooling overnight. A brief discussion of the forecast wrapped things up by saying that high pressure will dominate the Northeast for this weekend and should persist, according to the GFS, until the middle of next week. Sara advised everyone to go outside and enjoy it.
April 19: Matthew Souders led a discussion titled, "A Retrospective on the 2012/2013 Winter Forecast and a Look at Summer and the Tropics."
Matt began his discussion with a discussion of the strengths and weaknesses associated with his 2012-2013 winter forecast. Regarding snowfall, he shared some numbers with the group including that New England snowfall was 1.5 times that of normal snowfall totals. Matt had forecasted for a warmer than observed winter but mused that it had been a very difficult winter to forecast for. The leading driver was explained to be the quasi-biennial oscillation (QBO), or the descending branch of the easterly wind anomaly that reached the tropopause. In doing so, Matt explained that this would induce midlatitude blocking regimes and result in a negative phase of the Arctic Oscillation (AO). Some discussion ensued about the possibility that midlatitude wave breaking energy could reach the Tropics, or the opposite- that an oscillation confined to the Tropics could have much influence over midlatitude activity.. A few papers worth reviewing on this topic are as follows: Reasor and Montgomery (1999), Naito and Yoden (2006), Hitchman and Huesmann (2009), etc. While looking at the QBO by means of a time-height section of monthly mean zonal winds at an equatorial station Singapore, 1°N/104°E (since Jan 1976) (download pdf), we wondered where the sudden stratospheric warming (SSW) event of early late Jan/early Feb fit in. It was clear that our impactful blizzard of 8-9 Feb occurred as the AO was starting to trend strongly negative and also the descending branch of anomalous easterly winds reaching the tropopause. To summarize his review of his winter forecast, he showed that when you plot a winter composite anomaly map, the anomalies wash each other out and it seemed to be an average winter. However, March for example, exhibited areas of < -4 sigma for surface air temperature across the eastern half of the country. His final thoughts on his forecast was that due to the lack of ENSO activity and unexpected importance of the QBO, his forecast for a warm winter was slightly off.
Using his knowledge of the current state of specific climate indices, Matt provided his qualitative forecast for Atlantic tropical cyclone activity. He claims that if you use the current phase of the QBO and the lack of ENSO signal (termed “La Nada”) then analog years may provide some support of East Coast TC activity for the 2013 season. Another important factor is SST; Matt explained how the warm/cold/warm tripole in the Atlantic will likely be connected to deep tropical activity. He showed the current water vapor imagery of the Tropical Atlantic and discussed how the tropics are already quite moist. The location and strength of the High over the South Atlantic will aid to regulate the easterlies bringing dry, dust-laden Saharan air off of Africa versus the westerlies currently cutting across the Caribbean. Another interesting point is that the current solar cycle maximum was the weakest since ~1930 and looking at analog years of the next-weakest years provides further support for East Coast TC activity. Another point is that the East Coast activity will depend on the existence and persistence of cutoff troughs that could act to guide the systems up the coast, otherwise we’ll be dealing with a “ridge over troubled waters.”
An undiscussed interesting current topic is the severe weather associated with a cold front that is extending along the entire Eastern Coast on Friday afternoon. SPC has issued tornado watches that cover from NC into PA-bordering NY counties. At the time of the discussion, one tornado warning was issued for Mecklenburg, VA associated with a pre-frontal isolated cell. The potential for more tornadic supercells exists out ahead of the main cold front in an environment of high shear and ample moisture and instability.
April 12: Keith Roberts led a discussion titled, "The Vagaries* of the Weather."
Keith’s weather discussion took place on a rainy Friday afternoon. He used the current weather to motivate his discussion that focused on the midlatitude cyclone that caused severe weather in the Central and Eastern U.S. followed by a forecast for the coming week and a general discussion of forecast tropical activity. Keith began his discussion with introducing an interesting forecasting tool that he found created by a company called Dark Sky. An interesting display tool is model comparison of temperature, precipitation, and other sensible weather quantities at any location.
The current and recent pattern of a midlatitude cyclone was discussed starting with the Weather Prediction Center’s analysis that showed how the precipitation had outrun the surface cold front. Looking at precipitation from the national mosaic NWS radar and comparing it with the IR satellite imagery, there is hypothesized to be some sort of cold front aloft. WPC analyzed the cyclone as an open-wave, whereas the cyclone appears more occluded at least at upper levels. At the time of the discussion, the warm front was positioned to our south which is supported by the 12 UTC OKX sounding which showed an 850-hPa inversion and veering winds with height. Another useful way of looking at frontal boundaries is plotting potential temperature and its associated gradient. This led into a discussion of a strong sea breeze boundary that developed over NJ yesterday (10 Apr). There was some discussion about why this strong sea breeze front didn’t initiate any convection. John Mak flew through it and found it to be approximately < 300 m, so a shallow front.
The convection of 10 April made its way to Long Island and prompted the issuance of a severe thunderstorm warning from the NWS for portions of Suffolk County. Most students reported seeing an impressive shelf cloud along the leading edge of the complex. There was a lot of surprise about how it held together despite there being a lack of MUCAPE. Upon looking at the 00 UTC OKX sounding from 11 April, there appears to be an elevated mixed layer beginning at 850-hPa. If that level could be saturated then there would be a lot more calculated CAPE.
Keith then provided a brief forecast discussion for the near-term. The GFS initialized at 12 UTC 12 April showed the potential for severe weather for LA/TX around 00 UTC 18 April, given forecast CAPE > 2000 J/kg, a relatively strong temperature gradient, and higher moisture values advected by the mean flow from the Gulf. The general pattern will be a large trough in the West with weak ridging in the East. The 8-10 day mean 500-hPa heights from the GFS and ECMWF capture this pattern and with little variability of the GEFS shows confidence in this forecast pattern which may spell below average temperatures in the Central U.S. and above average precipitation for the East Coast.
CSU announced their forecast for the coming hurricane season which prompted some discussion about their reasoning. They are expecting ENSO to trend to La Nina, however current trends are unsupportive of that and dynamical models are in the transition season so aren’t trustworthy yet. We looked at the cross-section of water temperature anomalies and saw a warm anomaly at ~150 m depth break off and progress eastward and break up the trending cold surface temperature anomalies in the East Pacific. We are hypothesizing that this is due to a Kelvin wave. Our group’s current thinking is that the SST and water temperature profile are warm enough to support an active TC season in the Atlantic and current ITCZ activity is robust which shows that the Tropics are fueled up with moisture, which will be something to pay attention to. We’ll see if their forecast for 18 named storms and 4 major hurricanes verifies.
*Vagary: an unpredictable or erratic action, occurrence, course, or instance: “the vagaries of weather; the vagaries of the economic scene.”
April 5: Michael Erickson led a discussion titled, "When Will it Feel Like Spring?"
Mike began with an overview of the general pattern of the past week by showing the GFS 500-hPa geopotential height analyses and forecast. The main points discussed were the coastal storm that remained offshore Friday which allowed for surprisingly pleasant local weather, the multiple days of high fire threat in the region, and a general discussion of some climate indices used to answer the title question of the discussion of when it will consistently feel like spring.
There was a forecast challenge regarding the inland extent of the precipitation associated with a coastal low forecast for Friday, 5 April. The NCEP operational models differed on the general position of the low forming offshore of the Carolinas and translating northeast with the NAM taking the low farther northwest than the GFS. Most forecasts called for rain at least in the morning. The 21 UTC SREF run from the 4 April showed 0.1” liquid precipitation extending into CT. However, the cyclone tracked farther south and the precipitation did not extend much even into Long Island; Islip (ISP) had a trace of precipitation, with a bit more at Westhampton Beach (FOK) located farther east. While the forecast rain amounts weren’t impactful, the lack of measurable precipitation and the ultimately pleasant Spring-like conditions were surprising.
Mike conducts research into high fire threat days in the Northeast U.S. so he provided some expert insight into the previous high fire threat day of Wednesday, 3 April and the surprise of Friday, 5 April. A high fire threat is defined when conditions are favorable for the development and spread of wildfires which includes relative humidity < 30% and wind gusts > 25 mph. When these conditions are met, the NWS typically issues a Red Flag Warning. During the discussion (3 PM Fri, 5 Apr), the NWS hadn’t issued a Red Flag Warning for our area and upon checking observations around the island, RH < 20% but the windspeed criterion hadn’t been met. Fires did break out on Wednesday and Thursday throughout the Island, especially out east near the Pine Barrens.
Mike then switched gears into discussing the general climate pattern. He showed two interesting plots of the precipitation time series at LGA and the temperature time series at LGA for the past year. From the precipitation time series, it can be seen that there has been a steady deficit since the Fall. From the temperature time series, it can been seen that below normal temperatures were not common until the onset of Spring of this year. This clearly shows that Spring has been delayed. Upon further reflection of the recent past winter, Mike discussed temperature anomalies with depth for each ENSO region. The current signal is sort of washed-out and is referred to as “La Nada.” Spring is notorious for being difficult to forecast ENSO, so that sparked discussion about whether there extratropical forcing may affect the Tropics. A vertical cross-section of the geopotential height anomaly shows the stratospheric warming that occurred in late January and has since propagated into the troposphere.
Mike wrapped up his discussion with a forecast for next week. There is an active pattern with a few hot spots, specifically a severe weather threat for the Plains on Wednesday. There should be the potential for weak ridging for our area to allow us to see temperature around 60F. To answer the question, “When will it feel like Spring?” the CPC 6-10 and 8-14 day forecasts for temperature show the East Coast in the above average category, likely due to the active pattern and potential for ridging and the resultant warm air advection. Mike concluded his discussion warning that this is cut-off season so the temperatures and pattern can fluctuate. He admitted that it should feel like Spring at least until next Wednesday.
March 29: Sara Ganetis led a discussion titled, "The Snowless Event of 25-26 March* and the Severe Weather Threat for TX/OK."
Sara began by showing the global stitched IR satellite imagery to highlight the cutoff cyclone off of the coast of California, general troughiness over the East Coast, and an interesting area of tropical convection over the northern shore of Australia. A look at the 500-hPa heights according to the GFS analyses and forecasts shows a the cutoff cyclone creating a rex-block pattern in the Eastern Pacific.
Sara quickly touched on the potential for above average temperatures for the East Coast for the month of April likely due to the PNA pattern trending negative and the three-month outlook of April-May-June potential for above average temperatures likely due to the MJO convection translating eastward into the West Pacific.
The main topic of discussion is the current threat of severe weather over the Southern Plains, specifically southern Oklahoma and northern Texas. In looking at the current radar, some convection had already begun to develop causing some hail to be reported in southern Oklahoma. Southerly flow brought moderate moisture up into Oklahoma (dew points at least in the upper 50’s F) and the solar insolation allowed for surface-based CAPE to be > 2000 J/kg. The main question was what was providing the lift to initiate the convection? Upon further analysis, there was a warm frontal boundary along and slightly north of the convection. There is a greater threat for severe weather for the same region for tomorrow. The threat is stronger likely due to the involvement of convection along a dry line as evident from the forecast low-level theta-e field. Looking at some forecast soundings for Norman, OK (KOUN) showed moderate shear and plenty of CAPE. Attention was brought back to the East Coast in a quick forecast discussion of the Easter showers forecast for this Sunday. A broad cyclone located in Eastern Canada collocated with a long-wave trough will pass to our north and any resulting precipitation will likely be from the cold front.
*Note: Due to the active severe weather, we didn't have time to discuss the lack of snow that was forecast to occur for the 25-26 coastal storm. The bottom line was that it was too dry initially and too warm.
March 22: No Discussion (Spring Break)
March 15: Matthew Souders led a discussion titled, "A Look into March and April: When Will Winter End?"
Matt began his discussion by showing a loop of 500-hPa geopotential height anomalies from the past 30 days to show the general negative anomalies in place over the East Coast from February through early March. He then showed the forecast 500-hPa geopotential heights that displayed a progressive pattern that supports the fact that winter is sticking around for the East Coast and that the ensemble of Ground Hogs that had predicted an early spring were, surprisingly, wrong. Although the potential exists for a high-impact system on Tuesday, Matt geared his discussion to be more of a large-scale overview of the general patterns in play and their forecast effects on sensible weather for the rest of this month and into April.
Matt walked us through the various weather and climate indices. He showed us that there exists a weak El Nino signal and that there is a subsurface warm pool sloshing easterly, but ENSO is not the real driver due to the weakness of the signal and he claims it is less an "El Nino" and more a "La Nada." Next we looked at the Arctic Oscillation (AO) and saw that while the AO has been consistently negative, forecasts show a “tanking” AO with the index values becoming less than -5, which is a very strong signal. The GEFS and ECMWF forecasts support the significant negative forecast AO. The North Atlantic Oscillation (NAO) didn’t have as strong of a signal hypothesized due to the fact that this index is Atlantic ocean-based and by nature more stable or that most of the amplification is in the Pacific Basin, which is captured by the AO. The Pacific-North American Pattern (PNA) is also forecast to go negative but has a weaker signal than the AO. The Madden/Julian Oscillation (MJO) is forecast to move into octants 7, 8 and 1 within the phase diagram which has been linked to active weather on the East Coast. The Blocking Strength signal is weak now but blocks may set up over Alaska and Greenland. Looking at the influence of the stratosphere, the 30-hPa temperature forecast shows no evidence of a sudden stratospheric warming event causing the AO to “tank” according to theory that the AO is regulated by downward control from the stratosphere. However, there is no well-defined and cold polar vortex as one would expect with a negative AO. The final key player discussed is the Quasi-Biennial Oscillation (QBO) and what is shown is the easterly phase descending towards the tropopause which means blocking is favored.
To summarize what all of the indices told us to provide a story of what to expect on the East Coast, Matt concluded by showing what the Climate Prediction Center (CPC) has stated in their 6-10 day outlook and 8-14 day outlook. He stated at the storm track will likely be more southward than than the existing maximum precipitation position, but agreed with the below normal forecast temperatures. To answer the question, “When will winter end?” Matt said that the main driving forces are the AO and QBO. He created composite anomaly maps for the month of March from the years during which there was the easterly phase of the QBO (2005, 2001, 1996, 1987, 1971, 1969, 1958, 1952) and found that the surface air temperatures are cold but for the month of April there should be a bit of a warming, or at least return to normal. He stated that winter will lock and hold through March with not much of a break until April. What will that mean for the severe weather season? One hypothesis is that it just may be shifted later in the season, but not necessarily be weaker.
We had some special guests during our discussion and would like to thank our alumni and prospective students for their participation and patience with the lack of “normal” seating!
March 8: No Discussion
March 1: Keith Roberts led a discussion titled, "The Contemplation for the Potential of a Major Coastal Storm Next Week (Mar 7, 2013)."
Keith began his discussion by looking at the current loop of the global stitched satellite imagery to identify the train of multiple cyclones extending across the North Atlantic. According to the 500-hPa GFS forecast, there will be a merger of two cyclones in the North Atlantic. This general pattern sets up a persistent trough over the East Coast extending into next week. The main topic of discussion is the possibility of a major coastal storm on Wednesday, 7 March. Because it is a medium-range forecast, there is uncertainty in the exact track and intensity of the storm, however all models resolve it.
General Evolution: Matthew Souders provided some input about the idea of a wave packet, or the effect of one cyclone on cyclones downstream. Because the system is the leading edge of the wave packet and the energy of the wave packet moves faster than any individual cyclone. Thus Keith explained that the activity in the Pacific is the driving force behind the amplified pattern and forecasted shortwave that should develop over British Columbia on Sunday and propagate across the US until it becomes phased with the broad trough in place, and intensifies off of the coast.
Model Discussion: After looking at some plots from the 12 UTC 1 March deterministic GFS run and 00 UTC ECMWF run, it became apparent that the position of the cyclone forecast for 12 UTC 7 March was literally up in the air. As we’ve seen with previous storms, the position of the low is important for the precipitation type received so the ECMWF was making the event look like a snow event, however the GFS looked like rain.
To further understand the sense of uncertainty with the event, we looked at Ensemble Sensitivity Analysis (ESA) with helpful discussion from Minghua Zheng and Dr. Colle. What we saw from the GEFS Day 6 Summary (verifying time = 00 UTC 7 Mar) that the ensemble mean had the cyclone was off of the coast of NC/VA and more than 56% of the variance was attributed to the 1st Mode (EOF 1) that showed lower pressure to the north of the mean low’s center location, which meant that a lot of members were resolving the system farther north. It will be interesting to note how this will change with time. A look at the ECMWF ensemble showed that the mean and spread was more threatening than the GEFS because it was farther north and most of the spread was still near the 40-70 benchmark. Looking at the 500-hPa heights associated with EOF 1 for both the GEFS and ECMWF showed that the ridge in the North Central US on the 1 March (Day +3.5) will likely determine the position of the low. This makes physical sense because if the ridge is more amplified, then the downstream trough may be more amplified which would bring the low farther south and vice versa.
Storm Impacts: There is the potential for strong winds with this event. Keith showed some forecast model soundings (12 UTC GFS 1 Mar) for near KISP that showed 65 kts at 850-hPa and 40 kts at 925-hPa at 12 UTC 7 Mar. He then showed some horizontal plots of MSLP and 10 m winds and pointed out the strong winds wrapping around the cyclone that are likely enhanced due to the presence of a strong high (<1036 hPa) over Eastern Canada. The forecasted intensity of the low according to the GFS was 992 hPa.
The precipitation type is still uncertain, but the GFS showed 48-h total liquid accumulation for the event to approach 2 inches mainly north and west of the cyclone center.
Due to the duration of the event and likely persistent easterly winds, there exists the potential for storm surge to be an impact. Keith described how this depends on the position of the low but the large fetch with the prolonged strong easterlies really are “perfect” for storm surge. He would have shown a hydrodynamical model but they are not run out that far into the future.
Keith concluded that this storm does have the potential to be a high-impact event, but we will just have to wait for that Pacific energy to reach the coast and the key players to form and be observed and input into the model.
February 22: Michael Layer led a discussion titled, "A Look at the Upcoming Active Weather Pattern Over the Next Week."
Mike began his discussion by looking at the three main items of interest of the current weather: the cutoff low in the Midwest, the low affecting the Pacific Northwest, and the broad area of convection over the Southeast. The cutoff low had brought record snowfall on 21 February to places such as Kansas City, MO mostly due to warm air advection creating a broad area of precipitation that lacked any banding structure. The low affecting the Pacific Northwest is bringing snow to the higher terrain of the Cascades. The convection in the southeast seems to be associated with a weak frontal boundary and with low MUCAPE values of 500-1000 J/kg.
Mike then went into a discussion of the teleconnection indices and their forecasts to discuss the CPC’s forecast for March. The AO and NAO are trending negative and coupled with the positive trending PNA could mean troughy weather for the East Coast. An excited mumbling of seeing a Nemo: Part II ensued. We have been having an ongoing discussion of the persistent drought throughout most of the country since last semester and Mike expressed that he hopes this active pattern will help alleviate the severe drought conditions in a few areas of the country. The new seasonal CPC discussion for the spring was brought up and their forecasting methodology was discussed.
Mike moved on with a discussion of his forecast for the weekend disturbance that will likely bring solely rain to the region. He showed the 1200 UTC 22 Feb deterministic runs of the NAM and GFS and showed how the cutoff low over the Great Lakes will aid to form a secondary low off of the coast of North Carolina which will move north-northeast. There was a discrepancy between the NAM and GFS over the position of the low and its intensity, the NAM forecasting a stronger low that was closer to the coast. The winds are expected to be offshore, according to the NAM. Also, due to the easterly flow north of the cyclone, warm air should ensure that this is not a snow event. The SREF is forecasting for between 0.75-1.00” of rain for Islip. Another thing to note is that due to the strong easterly flow, there is the potential for storm surge, however due to the speed of the system and the fact that it doesn’t directly correspond with a Full Moon, sites aren’t expected to exceed moderate surge levels.
Mike concluded his discussion with the forecast for the likely Tuesday, 26 Feb event, which has the potential to be a high wind event. The GFS is showing a developing coastal storm within the 40-70 benchmark with 70 kts at 925 hPa.
February 15: Keith Roberts led a discussion titled, "A Review of Feb 8-9th Blizzard & A Look into the Possibility of More Snow this Weekend."
Keith began his discussion with showing a loop from the past two weeks of the dynamic tropopause on the 2 PVU surface from which we could see the evolution of the northern stream and southern stream systems and rapid intensification of last weekend’s notable event, the 8-9 February 2013 Snowstorm. He led the the discussion on the significance of this event and provided material that stimulated the discussion of scientific questions.
Significance: This nor'easter was significant for two main reasons. Firstly, for the fact that it evolved from the merging of both the northern and southern stream, i.e. a coastal low interacted with an upper-level trough which further enhanced it. Looking at an archived loop of satellite imagery really shows this evolution. Secondly, an classic band of intense snowfall set up northwest of the cyclone center which caused localized higher snowfall rates and totals across portions of Central Long Island and Connecticut that forecasts had underestimated. Stony Brook measured 28.0 inches of snow and the highest measured on Long Island was in Medford with 33.5 in. The intense, heavy snow during the evening commute caused cars to be stranded and subsequently buried for days.
Science Questions:
What were the predictability issues? The deterministic and multi-model ensembles underpredicted the precipitation amounts which suggests that there is a microphysical challenge within the models. This event was interesting because the snow to liquid water ratios varied in time and space in relation to the band. If the microphysical schemes assume a constant density then that will result in a large model moisture error.
What is that warm tongue at 850-hPa around 00 UTC 9 Feb just south and east of NYC? Some hypotheses include that warm air between the two streams got “squeezed” in between as they merged, diabatic heating related to the intense snowband, warming via subsidence from the descending branch of the ageostrophic circulation that developed with the frontogenesis associated with the band. We discussed Matt Sienkiewicz’s WRF runs for this case to try to further isolate this issue if he tries to turn off latent heating. Another thought is that there is a stationary gravity wave in place due to the latent heating.
What was the nature of this band and how was it different from others observed? We discussed how the convection seemed to be more upright than slantwise, which could explain why there were reports of “thundersnow” associated with the band. Also, the microphysical evolution of the band was interesting because at one location within a few hours one would witness large aggregated flakes, then small grains of graupel, then fine, dry snowflakes. Dr. Colle is looking into this more after having spent the night taking pictures of the hydrometeors with a microscope.
Recommended Reading:
February 8: Sara Ganetis led a discussion titled "Northeast US Blizzard: Impacts on Long Island"
The newsmaker this week is the potentially crippling Northeast US (NEUS) blizzard of 8-9 February 2013. At the time of the discussion (11 AM EST to accommodate the cancellation of classes after 12 PM), the cyclone center was East of the DELMARVA peninsula and light snow was mixing with sleet at SBU. By the end of the discussion (11:30 AM) there was a complete changeover to rain. We started discussion by highlighting the factors that will lead to the cyclone rapidly intensifying. The key players behind the rapid intensification of the coastal cyclone are the upper-level trough situated over the Great Lakes and the southern stream energy. The phasing of the cyclone with the upper-level trough will align it with the jet cores for preferential dynamical intensification. We discussed that the blizzard conditions experienced across the Island would be sensitive to the timing of the changeover from rain to snow and the position of the band once it developed.
Timing of changeover: For forecasting total storm snowfall amounts, the timing of the expected changeover from rain to snow is critical. Once the precipitation gets going, we expect the diabatic cooling to cool the column completely. Exploring some model-derived forecast soundings showed that the NAM and GFS had a warm-nose bias just below 850 mb but the RAP showed a much cooler solution before 00 UTC 9 Feb. We favored the RAP for our forecast, mainly because that would mean more snow for us. Looking at the SREF plumes for precipitation type at Islip, the changeover is expected before 00 UTC on 9 Feb and forecasts about 10 inches of snow.
Location of band: The models were growing more consistent by the 12 UTC run on the location of the cyclone passing a bit farther to our east. The forecast location of the cyclone center is critical for the forecast of where the primary snowband will develop, typically NW of the cyclone center. The band develops where there is maximum frontogenesis, typically between the surface and 700 mb. The band also coincides with a region of conditional instability which means there can be more rapid ascent and more intense snowfall. We looked at the 12-km simulation for the MM5 which put the band right on the Nassau/Suffolk border.
Winds and Coastal Flooding: Low-level winds in each of the model were >35 kt sustained around 00 UTC 9 Feb. Higher momentum air at ~925 mb can be mixed down by precipitation drag and, given the model soundings showing that the surface is warmer than aloft, through convective motions as well. As far as coastal flooding, an astronomically high tide (New Moon is on the 10th) with a few feet of storm surge on top of that can cause flooding at some prone locations. In addition, violent wave action may also be an issue along the North and South Shores.
Sara concluded her forecast calling for Stony Brook, NY to receive 14" of snow and the area receiving the highest snowfalls will likely be to our north and east.
February 1: Michael Erickson led a discussion titled "A Brief Look at Yesterday's Storm and our Prospects for Snow this Weekend."
Previous Weather: Mike led the first discussion of the semester beginning with that since we last met in December 2012, we’ve experienced a changing pattern. The AO went negative following a stratospheric sudden warming (SSW) event. The ENSO signal is now more like La Nina with cold anomalies found in the Eastern Pacific. Drought has lessened slightly, but still persists across much of the country.
High Wind Event of 31 January 2013: A strong continental cyclone dragged a cold front across our area very early Thursday morning. This cold front was responsible for severe weather across much of the South during the preceding days. The main threat in our area was for high winds exceeding the NWS severe criterion of > 58 mph. The cold front was strong in the fact that there was a very tight horizontal temperature gradient. In response to this, there was a very strong low level jet (LLJ) with speeds of around 70 kt around 925 hPa according to the 00 UTC 31 Jan OKX sounding that only got stronger as the cold front approached from the west. This high momentum air can mix down through the drag of heavy precipitation (found in the narrow cold frontal rainband (NCFR)) or by gravity waves induced by isolated areas of convection that can propagate within a stable layer. The NWS storm reports indicate that the maximum gusts were ~65 kt and the heaviest rain fell north and west of the Island. Mike discussed the marine influence that likely saved us from more severe winds and heavier rain. The most important factor was the strong inversion extending from the surface to a bit above 850 hPa (SSTs are in the single digits in Celsius but 850 hPa temperatures were near 15C). This inversion was well-forecast by the models. The stability inhibited the free mixing of the higher winds to the surface. Some of the highest wind gusts reported were coincident with the passage of the NCFR, however analysis of the base reflectivity from 09 -11 UTC 31 Jan of the band shows it breaking apart as it reaches the Island, possibly due to decreased friction and its ability to accelerate. There was some discussion about whether some of the gusts reported were associated with a gravity wave, but insufficient temporal data of pressure, for instance leaves that undetermined. The main science issue discussed was how the front evolved as it approached the coast. There was a distinct lag between when the front passed, when the winds shifted, and when the temperature finally began to drop throughout the day on the 31 Jan. This development of a prefrontal trough has been identified before in previous cases but hasn’t been explored in any great detail to our knowledge.
Weekend Forecast: Mike concluded his discussion with the fact that Sunday may be potentially interesting (according to some SREF members) with a slight chance of snow but most other days are forecast to be cold and uneventful, with the current forecast pattern being unsupportive of storminess.
For the current Friday Weather Discussion page, please click here.
Fall Semester 2014
December 5th 2014: Keith Roberts led a discussion titled, "Coming to a sky near you: A chance of rain"
The weather discussion led of December 5th began with a visualization of the 10-m wind field around the Northern hemisphere to demonstrate a tool that can visualize and aggregate real-time observations. At the time of the discussion, it was shown that there was anti-cyclonic flow over the northeast region with near seasonable temperatures and broken skies. As such, there was weak subsidence at 700 mb as indicated by 700 mb q-vector divergence. The past 2 weeks had been somewhat active with a southern stream cyclongenesis event around Thanksgiving followed by the passage of a Canadian maritime front on December 1-2.
Focus was turned to the present conditions. The flow at 500 mb over the eastern United States was zonal and progressive. Between Saturday-Sunday, a cold front was forecasted to clear the Eastern seaboard and set-up a baroclinic zone for potential cyclogenesis off the southeast U.S. The location and timing of the cyclogenesis event was considerably uncertain in model guidance. Deterministic model guidance was first analyzed. The differences between model guidance bifurcated into two camps: UKMET/ECMWF/CMC (more progressive and more intense cyclogenesis) and WRF/GFS (slower and weaker cyclogenesis) in the timing and phasing of the northern stream shortwave that was associated with the passage of the Arctic front on December 6-7. A spaghetti tool put up by the WPC demonstrated large differences at the 500 mb in the timing and amplitude of the northern stream shortwave between the two aforementioned model camps at rather short lead times (< 36 h). However, the majority of modeling systems eventually did develop a cyclone off the southeastern US coast by mid-week (Dec 9-10), but the movement of this cyclone and its intensity (1000-1020 hPa MSLP) had tremendous variance. To illustrate the variance in cyclone tracks, the cyclone tracks for various ensemble guidance systems (SREF,GEFS, etc.) was shown with cyclones tracking anywhere between Alabama and Boston at similar times. The large uncertainty was hypothesized to be attributed to diabatic effects associated with the western bound of the Gulf stream wall as upper level QG dynamics did not appear to be largely present.
To synthesize the large spread and uncertainty in model guidance, EOFs were analyzed for some of the GEFS and ECMWF guidance. The first EOF in both ensembles explained > 40 % of the ensemble variance in 500 mb. Day 4 (December 9th forecast) EOF1's sensitivity could be traced back to a disturbance north of the Arctic circle in the Aleutian sea. At this lead-time (+4 days), the sensitivity analysis tool was best suited for identifying where the storm's uncertainty was originating from rather than serving to be physically interpretable. This type of downstream sensitivity to disturbances north of the Arctic circle has been routinely observed in the northeast region since this tool has become operational. It motivates a more complete sampling of the Arctic weather conditions for a more accurate model initialization and theoretical work on the nature of uncertainty and its propagation in modeling systems. The largest sensible weather impacts were high wind on the backside of the Arctic front Sunday-Monday and heavy rain near the coast with a wet mix inland. The potential for coastal frontogensis mid-week as the cyclone moved up the coast was suggested by Dr. Brian Colle due to deformation between the cyclonic on-shore flow (maritime air) and the Arctic air inland associated with the anti-cyclonic over the north east. Due to the large uncertainty in the mid-week cyclogenesis event, it was difficult to say what the sensible weather implications would be, however, some guidance suggested a persistent on-shore flow in the NY/NJ Bight, which could lead to widespread coastal flooding.
October 3rd, 2014: Michael Erickson led a discussion titled, "A Glimpse at What This Upcoming Winter Might Have to Offer?"
Mike started weather discussion with a brief review of the past weeks weather. Long Island found itself on the east side of a strong 500 hPa ridge over the weekend, which resulted in high temperatures reaching the low to mid 80's on Saturday and Sunday. Thereafter a weak offshore low developed and persisted, resulting in cooler cloudier conditions with easterly flow. A weak upper level low propagated into the region, supporting the continued weak offshore surface cyclogenesis and easterly flow. The low finally propagated east today, with a return to something called sunshine.
Thereafter focus shifted to an attempt at a seasonal "forecast," although "experiment" might be a better word choice than "forecast." Mike would be lying if he considered himself a seasonal forecast in any sense. When making a seasonal forecast, Mike focused on lower frequency variability metrics which would likely change slowly such as El Nino Southern Oscillation (ENSO), the Quasi-Biennial Oscillation (QBO), the Atlantic Multi-decadal Oscillation (AMO) and the Pacific Decadial Oscillation (PDO). Similar previous years are identified using these low frequency metrics, and their winters are analyzed. Currently we are in ENSO neutral conditions leaning toward an El Nino with a 60-70% chance of an El Nino forming this fall or winter (according to the CPC). The QBO is negative and expected to peak in the negative phase this winter. The PDO is in a cold phase, although in the shorter term, most of the northern Pacific is experiencing above normal sea surface temperatures (SSTs). In fact, most of the Northern Hemispheres SSTs are above normal, which is an anomalous signal. Finally the AMO is in its warm phase.
All of these metrics are discussed in Mike's 2014-2015 "Experiment." A quick summary of the results show that years most similar to our current 2014 fall experienced below normal temperatures in the eastern 1/3 to 1/2 of the country. Of the most similar years, about 60% of them were below normal. A quick glimpse at precipitation shows above normal precipitation in most of the Southeast United States, normal precipitation for the Long Island region and below normal precipitation for Northern New England and New York. These results should be taken with a grain of salt, but there does appear to be a weak signal in the composites considered.
September 26, 2014: Sara Ganetis led a discussion titled, "Wonderful Weekend Weather, Anyone?"
Please see the ITPA website for a summary.
September 19, 2014: Weather discussion was cancelled for the New Student BBQ.
September 12, 2014: Michael Erickson led a discussion titled, "A Look at the Weather and All it's Glory."
Mike began his vaguely titled weather discussion by taking a look at the previous weeks weather, which was quite anomalous for the western half of the country. First on the board was the anomalous rain and flooding that occurred in the Southwestern United States on 9/8/2014. This anomalous event was likely caused by the circulation of Norbert modifying the flow to it's east, possibly in conjunction with some Atlantic influence. This resulted in severe flooding , particularly in Arizona.
Focus then shifted to an anomalous cold air outbreak in the Plains banking up against the Rockies combined with a lee cyclogenesis event. The result was an early season snow event along the front range on 9/11/2014. Snow totals ranged from a dusting or light coating in Boulder, CO and Rapid City, SD, to 14" in Story, WY.
Finally, focus shifted locally to Long Island where the weather has been quite pleasant, not just recently, but for most of the summer. A glimpse at the LGA timeseries of temperature compared to normal showed consistently normal summer temperatures with an absence of extended periods of anomalous warmth or cold. One slight exception to this occurred in early September where temperatures averaged a few degrees above normal. The long range forecast showed continued pleasant weather with the chance for a decent cool air intrusion from Canada by next Friday.
September 5, 2014: Keith Roberts led a discussion titled, "Fall Weather Outlook and Brief Summer Weather Recap."
Spring Semester 2014
May 2,2014: Keith Roberts led a discussion titled, "Central Plains Severe Weather Outbreak of April 27th - 28th"
The discussion was focused on the severe weather outbreak on the 27th and 28th of April, 2014 in the southeast United States. There were 108 reported tornadoes over the two-day period starting on April 27th. This severe weather caused 32 fatalities and 42 injured in Mississippi, Arkansas, and Georgia. Surface analysis charts valid for April 27 16 UTC and April 28th 12 UTC from the WPC were shown. There was a warm occluded low over the central Great Plains with a trailing cold front extending from northern OK to southeastern Texas on the 27th of April. However, this cold frontal boundary appeared weak and a well-defined temperature gradient could not easily be found at the surface nor a wind shift. At upper levels, the cold frontal boundary was more diffuse. 500 mb heights showed two main cut off lows over the CONUS and a blocky, stagnant pattern. The frontal structure resembled more of a dry line given the moisture gradient.
Radar reflectively showed precipitation initiation around local noon in the southeast US around Mississippi and Georgia. A sounding from Jackson field airport (JFD) in Mississippi (ahead of the cold front by about 200 km’s) revealed a strong capping layer around 950 mb with a significant dry layer aloft (inverted v-sounding). Nearby soundings showed a similar inverted v vertical thermal structure. Some severe weather parameters were analyzed. Sfc-6 km shear was around 60 kts with more than half of this shear contained in the lowest 3 km’s. Dewpoint depressions were on the order of 10-15 C ahead of the surface frontal boundary. CAPE was on the order of 3500 J/kg. Given the dewpoint depressions and low-level shear, it was suggested that linear structures would be the likely convective mode. Indeed the bulk Richardson number was in agreement with an average value around 40. Pure supercells are often supported with values around 20. Reflectivity revealed quasi-linear structures with embedded super cells initiated around 18 UTC. To the north in southeastern Illinois and Iowa, the convective regime was different. In this region, the majority of damage was associated with penny to quarter sized hail and heavy precipitation (HP) supercells. No tornadoes were reported in the northern Great Plains.
Thereafter the discussion focused briefly on the forecast for the local region. 18 UTC NAM and 12 UTC GFS combo showed that the weekend in the Northeast would feature a weak cold frontal passage Sunday afternoon associated with a northern stream 500 mb shortwave. Precipitation looked diffuse and mostly driven by large scale QG PVA. There was no forecasted severe weather potential for anywhere in the CONUS over the following three days.
April 25th: Zhenhai Zhang and Minghua Zheng led a discussion titled, "A brief summary of last winter's storms and the potential severe weather event over the Central U.S. this weekend"
The discussion began with a brief review of the winter storms during last winter (from November 2013 to March 2014). Comparing with the winter storm climatology (the mean of 1979-2012 winters), last winter is a normal winter overall. For example, the total number of cyclones during last winter over East Coast is 64, which is very close to the historical (1979-2012 winters) mean, 62. The intensity distribution (minimum central MSLP) of the cyclones doesn’t show any significant differences to the climatology either: there are slightly more (about 10%) weak (>1000hPa) and intense (<985hPa) cyclones, while there are about 15% fewer moderate (985-1000hPa) cyclones. However, more than 50% of these cyclones occur during February and March. Over the East Coast, during 2013 winter the cyclone frequency (cyclone center counts) decreases by about 30%, 23% and 10% for November, December and January separately, but it increases by about 18% and 50% for February and March comparing to the climatology. We can get a similar conclusion from the cyclone track map for individual month: there are significant more cyclones during later last winter, especially in March. One of the most important reasons for that is the temperature anomaly in that month. During 2014 March, the mean near surface temperature is 3~8K lower than the climatology (1979-2012 March mean) over most area of the continent, meanwhile, the near surface temperature is 1~2K higher than the climatology at low and middle latitude over Western Atlantic. As a result, the temperature gradient along East Coast, which is vital important to cyclone genesis and developing, is significantly larger in 2014 March than the climatology. To conclude, although 2013 winter is a normal winter for the total cyclone activities, most of these cyclones concentrate in later of this winter, especially in March, which is mainly caused by the significantly strong temperature gradient along East Coast. For early of this winter, there are fewer cyclone events, however, there are many wave packets over North America. We can see these nice wave packets in the figure of mean (30-55N) meridional wind, and many intense cyclone and heavy precipitation events occur over the downstream of coherent wave packets.
Attention then turned to the major weather systems for previous week. An animation of the surface SLP and fronts showed a frontal system enters Great Lakes on 21st April (Monday). It crossed central PA and NY in the evening of 22nd April (Tuesday). The surface low moved to New England on 23rd (Wednesday) with a minimum pressure of 994 mb in the morning. In the upper level, the 500 mb height developed a cut-off low over NY and PA in the morning of 23rd April. This H5 Low then moved northeastward and left New England on 24th (Thursday).
The discussion then focused on the system that may bring severe weather in the weekend. A short wave trough had move to the southern Plain in the evening of 24th. It then propagated eastward and reached the Southeast on 25th. At the same time, a surface low pressure system has developed over IN, OH and KY. The frontal system was crossing the western Southeast during the morning of 25th. The latest SPC soundings over TN and AL display the conditional instable environment, as well as the condition favoring supercell thunderstorms and hails. Based on SREF 12Z run, the shortwave trough and the surface low system will advance eastward in the afternoon and move through the northern Mid-Atlantic and the NE during this evening and Saturday morning. The NAM model shows that the low system will be over the southern NE. Considering the model consistency, heavy rain associated with strong to severe thunderstorms will develop across parts of the Mid-Atlantic and Southeast in the afternoon and the evening of Friday. On Saturday, the upper level low will move through and the cold from may trigger thunderstorms over Mid-Atlantic.
Out west, another shortwave trough from eastern Pacific had reached Northwest in the afternoon of Friday. Based on the regional models, the upper level dryline will mix eastward into central KS and across western TX in the southern Plains during Saturday evening. A strong low-level jet streak will favor the moisture advection through Saturday and aid in the development and provide a violent nature of updraft motion favorable for damaging hail, winds, and tornadoes over central and southern Plains. Due to the large-scale upward motion, the steep lapse rates will support a large hail threat across the southern Plains. The dew point will rise to 60s.The strong low pressure system will spawn supercell thunderstorms capable of generating large hail, damaging winds, and strong tornadoes. So, a large area of high populated regions in KS, OK and TX will be dealing with severe weather events on Saturday, Sunday, and Monday. The most dangerous day appears to be Sunday, and the SPC has issued their "Moderate Risk" forecast of severe weather over portions of OK, AR, TS, and LA that day.
Thereafter, the discussion focused on the Rossby wave packet moving across eastern Pacific. Based on GFS ensembles forecasts initialized at 00Z 25th, there was a wave packet over eastern Pacific, which will propagate enter southern Plains on 00Z 27th. This could be another case that the Rossby wave packets are associated with severe weather events: severe weather events tend to develop downstream of wave packets. This wave packet will be weakening on Monday and merge into the leading part of another robust wave packet which will be over the central Pacific on Tuesday. This new combined wave packet will sustain the existing short wave trough over the Central US and bring significant severe weather over central US and Mid-Atlantic.
April 18th: Matt Sienkiewicz led a discussion titled, "A look at this past week's April snow event"
April 11th: Sara Ganetis led a discussion titled "The Current Weather 'Down Under': Tropical Cyclone Ita's Assault on Australia"
The discussion began with an overview of the current weather over the Northeast U.S., mainly highlighting the well-received warmth experienced during the day prior to a cold frontal passage. Central Park reached 74F, for example. A topic that has been being addressed every week is the possible onset of El Nino based on the progression of a warm anomaly present near the thermocline slowly reaching the surface of the Eastern Pacific which can be seen by animating the temperature-depth plot from CPC. While it may seem apparent that El Nino is imminent, there is still the issue of the Spring predictability barrier that may mean that although the signal seems strong right now, there may be other atmospheric responses that aren't in-line with El Nino.
The discussion then switched gears to discuss a few notable events of the past week especially the spreading of wildfires in the region. Research is still ongoing at SBU looking into the predictability of high fire threat days which is tied to atmospheric moisture (or lack-thereof) and gusty winds. Last Sunday evening, 6 April, there was a wildfire at a State Forest in southern NJ that was identified via radar and discussed through social media outlets by local NWS WFO's (Mount Holly & NYC). What was interesting about this particular fire is that as the sun set, a strong temperature inversion was present that persisted overnight. That meant that the smoke and particulate matter could not be vertically dispersed but instead had to travel horizontally which it did with a strong southerly wind northward to NYC and points west. Smoke was able to be smelled far away from where the fire originated and an air quality alerts were issued for several counties. Some snapshots can be found here.
Tropical Cyclone Ita formed northeast of Australia in the South Pacific in waters with SSTs exceeding 30C and favorable environmental conditions (i.e. relatively moist air mass and far-removed from the reaches of extratropical waves that could enhance the environmental shear). At the time of the discussion, it had already crossed the Great Barrier Reef and made landfall (~10 UTC 11 April) in between the cities of Cooktown and Cairns on the eastern side of the Cape York Peninsula in Queensland, Australia with estimated wind speeds of 125 kts. A loop of the IR imagery shows that it developed a very small, symmetrical eye before making landfall and having that eye fill-in a bit due to frictional convergence, steep terrain and removal from surface fluxes. An ASCAT pass prior to landfall (2002 10 April) showed the fastest winds near the eye as well as along the edge of the continent in what looked like a barrier jet. The Cooktown station time series of pressure, sustained wind speed and wind gusts shows that the pressure decreased to around 983 mb around 15 UTC 11 April and the sustained wind speed (gusts) got to >40 (65) kts. The JTWC forecast track of TC Ita had the system recurving back out to sea in the South Pacific. A look at the GFS deterministic forecast showed that while Ita had persisted for so long and moved due west under the influence of a broad area of high pressure over the Coral Sea, a trough was moving eastward and expected to shift the track more southward and then eastward with time. A 250-hPa jet also developed poleward of Ita suggesting that it will have undergone extra-tropical transition. Ita was an interesting tropical system that produced some interesting discussion.
The weather discussion concluded with a local forecast for that Friday (11 April) evening and then the system forming over the weekend and expected to drag its cold front across the region Tuesday-Wednesday. The SBU-WRF was used to discuss the near-term forecast for that evening, mainly how much precipitation should be expected behind the cold front? The answer was different for both the NAM-WRF and GFS-WRF, with the NAM-WRF being the drier solution. Comparing the current regional radar with that of the NAM-WRF and GFS-WRF forecast reflectivity showed that the GFS had much larger areas of higher reflectivity than the NAM and the NAM was closer to reality so perhaps we wouldn't expect much rain with the cold front. The next system was expected to form lee of the Rockies in a developing trough out ahead of a cold, Canadian air mass and likely cause some severe weather in the Plains around Sunday. The fact that the southerly flow out ahead of this system is forecast to be strong and persistent and a high pressure off the East Coast should allow for an influx of moisture towards the coast means that there should be some anomalously high precipitable water values and thus very heavy rains Tuesday-Wednesday. We quickly compared the deterministic GFS with the ECMWF QPF forecasts to find that the ECMWF was slower and weaker with the precipitation totals than the GFS. That will have to be something we keep in mind as we get closer to the event. Check out the weather discussion page for a few visuals to go along with the above text.
April 4th: Nathan Korfe led a discussion titled "An Analysis of the Recent Extreme Weather Event in the Central U.S."
March 28th: Matt Sienkiewicz led a discussion titled "Inside a Bomb: This Week's Impressive Nor'easter"
The discussion was almost entirely devoted to the significant offshore cyclogenesis event of March 25th and 26th that was responsible for sustained blizzard conditions on Cape Cod, Martha's Vineyard, and Nantucket. Three short wave systems, two diving out of the Canadian provinces and one originating in the Gulf of Mexico, coincided off the east coast of Florida. Aided by divergence in the right entrance region of a very strong upper level jet streak, as well as the strong low-level baroclinicity over the Gulf Stream, rapid cyclogenesis occurred as the system developed into a very impressive extratropical cyclone. The Ocean Prediction Center's experimental lightning density product showed areas of strong convection associated with the cold front, as well as mesocyclone-like structures north and northeast of the low center. Satellite imagery available on the University of Wisconsin CIMSS blog showed relatively dry, ozone-rich stratospheric air diving southward and wrapping around the low center. Ship reports from around 12Z on the 26th, when the cyclone was due west of Virginia Beach, show was ship to the west of the low center experience 55 knot winds and 26 ft seas. ASCAT passes at 15Z/26th and 03Z/27th show the evolution of the hurricane force wind field southwest of the low center as the storm makes its way from the offshore Mid-Atlantic waters up to Nova Scotia. Satellite altimeter passes at 2230Z/26th showed 48+ ft seas generated by the area of hurricane-force winds produced, resulting from the cold conveyor belt.
Several NDBCstations showed interesting observations for this event. NDBC moored buoy 44027, located at the mouth of the Bay of Fundy, showed a wind gusts exceeding 100 knots. The National Weather Service office in Caribou, ME claims that while the buoy only had one working anemometer at the time, the instrument showed no signs of malfunction so the readings may quite possibly be legitimate. Moored buoy 44150, located south of Yarmouth, Nova Scotia, took almost a direct hit from the low center and recorded a minimum central pressure of 955 hPa. The swell from this system was predicted to be enormous, stretching across the entire North Atlantic basin. Buoys obs from 07Z/27th showed 54 ft seas southeast of Sable Island, and the OPC 24/48/96 hr wave forecasts predicted the swell to traverse the North Atlantic and arrive in Spain with significant wave heights exceeding 8 meters.
http://glacier.somas.stonybrook.edu/wxdisc
March 14th: Sara Ganetis led a discussion titled "Will we be shaking our snowflakes instead of our shamrocks this St. Patrick's Day?"
Before starting a discussion of the forecast for the potential snow event slated for St. Patrick's Day (17 March), Sara spoke about the mid-week event that was responsible for significant snow and ice throughout northern New England but only gave us 0.40 inches of rain (at ISP) and wind gusts up to 35 kt (at ISP). The wind was predominantly post-frontal as opposed to occurring due to the pre-frontal jet or momentum mixing via the precipitation along the front. The NOAA/GOES-R JPSS blog highlighted this by providing a discussion of the MODIS air mass imagery that shows a clear stratospheric intrusion of dry air that helped with the post-frontal mixing. Another note of interest that Sara discussed was the fact that the Climate Prediction Center (CPC) issued an El Nino Watch which meant that although the El Nino Southern Oscillation (ENSO) signal is neutral and will likely remain as such, there is a 50% chance that an El Nino (or warm phase of ENSO) will develop in the spring and summer months. We took a moment to look at the water temperature anomalies with depth which when animated showed a warm anomaly in the Central Pacific moving eastward and getting closer to the surface.
The discussion now turned to the next extratropical cyclone event as this pattern has been progressive and it seemed there was discussion of one in some form for the past several weeks! The cyclone was slated to form in the Deep South as a shortwave rounded the Rockies and enhanced a longwave trough in place. It was then likely to translate eastward towards the coast where it would encounter enhanced baroclinicity from the Gulf Stream and ample Atlantic moisture and upper-level dynamical support from an elongated jet would support broad ascent and significant snowfall. The question, however, was where would that northern periphery of the precipitation shield? Would we be shaking our snowflakes instead of our shamrocks on Long Island this St. Patrick's Day?
The key players were the southern stream shortwave and longwave trough, and upon looking at the current satellite imagery around 4:30 PM EDT (2030 UTC) 14 March, that shortwave was just about to reach the West Coast in the Pacific Northwest. The fact that this feature was entering our observational domain meant that the models would likely converge on a solution with subsequent model runs, which as this is being written post-event on 17 March, was true for the global models but the NAM flip-flopped a little bit. The precipitation is highly tied to the track of the system and so a discussion of the uncertainty of the track ensued. Mike Erickson's cyclone tracks page showed the considerable spread among the GEFS members. Dr. Colle provided some insight into the ensemble sensitivity analysis which showed that the northern stream played a role in that if the heights to the north of the cyclone were lower that it would favor a more southern track and miss Long Island. This northern stream sensitive region was unfortunately originating from the data-sparse region of northern Canada which may explain the differences between the global models (ECMWF & GFS) when compared to the regional, mesoscale models (e.g. NAM). Dr. Colle also showed that this system was on the leading edge of a well-defined wave packet originating from the Pacific.
To conclude the forecast discussion, we decided to take a look at a the WPC official forecast, few deterministic models and finally the short range ensemble forecast system (SREF). WPC called for about a 40% chance of > 1" snowfall for Long Island, with the gradient to much higher probabilities just to the south of the Island. Their model diagnostics discussion essentially discounted the NAM for being too weak, fast and northward than the rest of the available guidance. The GFS and ECMWF had some agreement with most of the precipitation remaining just to the south of the Island. Model uncertainty became clear when we decided to take a look at the SREF and parallel SREF, the plumes in particular. The ARW cores were much more aggressive with the forecast precipitation, giving ISP as much as 21 inches of snow. The mean was approximately 6 inches with just under an inch as the lowest amount. As another post-event thought and having received nothing at ISP, it's interesting that the 0900 UTC runs didn't have more members showing less, or even any showing zero. Sara concluded her discussion by showing the SBU-WRF and how at forecast hour 72, both solutions were closely tied to their initial and boundary conditions with the NAM-WRF being the wetter solution than the GFS-WRF from the 12 UTC runs on 14 March. Will we be snowed-in for St. Patrick's Day? Unlikely (and we weren't)!
March 7th: Mike Erickson led a discussion titled "A glimpse back at our meteorological winter and potential future happenings."
Meteorological winter (defined as the months of December through February) has ended, so the beginning of this discussion focused on analyzing the coldest months of the year. Mike started with a followup of his "seasonal forecast" from the weather discussion on November 22nd, 2013. In that Power Point, multiple low frequency signals such as the El Nino Southern Oscillation, the Atlantic Multi-Decadal Oscillation, the Pacific Decadal Oscillation and the Quasi-Biennial Oscillation were used to composite years with similar Autumn conditions. The winters following the similar Autumn conditions were then displayed as a proxy for the 2013-2014 seasonal forecast. The followup Power Point reminded everyone of that "forecast" and then compared it to what actually happened using the North American Regional Reanalysis (NARR) data.
The consensus seasonal forecast was very poor compared to reality, unless the user multiplied the forecast by -1. The strongest signal in the seasonal forecast was a negative height and temperature anomaly extending from the Gulf of Alaska to the Northwest Territories of Canada, with a positive height/temperature anomaly off the coast of California and in the vicinity of Greenland. There was little if any anomaly signal across most of the eastern two thirds of the United States, suggesting equal chances for warm or cold. In reality, Alaska had a strong positive height/temperature anomaly along with the western Atlantic, while a strong negative height/temperature anomaly prevailed in south central Canada southward into the Great Lakes and Ohio Valley of the United States. Daily animations of height/temperature throughout the winter showed frequent surges of cold air originating from central Canada plunging into the eastern half of the United States, especially in January and February. Overall, North Dakota extending eastward into the Great Lakes were the coldest with winter surface temperatures running greater than 5 degrees Celsius below normal. Closer to the east coast and further south, temperatures averaged 1-3 degrees Celsius below normal with occasional warm ups. Before continuing on to near real-time weather, Mike stressed that he is not a seasonal forecaster, and that the skill of his forecast does not reflect the skill of the current science.
Mike briefly discussed the storm "bust" of Monday March 3rd. Forecasts the previous Friday were suggesting the potential for significant snowfall. However, models shifted south throughout the weekend. This was shown in an animation of cyclone track positions for the storm (also in the PPT above). Mike stressed that given the difficulty of 36-60 hour forecast predictions this winter, caution should be applied when predicting future storms. This point was stressed when looking at our next storm potential between 3/12-13. Mike first looked at the forecast for the weekend and early next week, which showed dry, sunny and generally above normal weather after a coastal storm moved off to the south early Saturday morning. Cold air attempted to move back in later Tuesday while lee cyclogenesis was taking place around Nebraska. The Global Forecast System (GFS) had a relatively weak coastal move south of Long Island that, taken literally, would drop a few inches of snow. The European Center For Medium Range Forecasting (ECMWF) had a stronger storm slightly further north, suggesting more of a rain event for Long Island. A more in depth look suggested that the GFS was producing a storm primarily from an Arctic shortwave, while the ECMWF was phasing both a Pacific vortmax and an Arctic vortmax. Mike closed with the combined cyclone track forecasts from the GEFS, FNMOC and CMC ensembles. Not surprisingly, there was a very large spread in ensemble solutions, with many members saying rain, snow, or a miss to the south for Long Island. Like any forecast for a storm 6 days out, it is best to make a mental note of what could happen. After that, we watch and wait.
February 28th: Mike Layer led a discussion titled "The Winter Brigade Continues into March!"
The discussion began with a loop of the 500 hPa heights over the Northern Hemisphere. Over the past week, the loop showed a trough over the Midwest/Great Lakes region, which translated east and amplified as a piece of the polar vortex broke off and dove down into southeastern Canada. The loop also showed the large cutoff low over the Pacific moving towards California. Looking ahead the next several days, the cutoff low looks to dissipate as it moves into the SW U.S. and gets absorbed into the mean flow. Heights rise over the eastern U.S. as a new trough develops over the central U.S. and moves eastward.
SPC mesoanalysis was used to analyze the system that moved through the NE U.S. on Thursday. Low pressure moved through SE Canada, dragging a cold front through the region. Brief heavy snow showers affected the area during the early afternoon, associated with steep low level lapse rates (7 to 9 C/Km) and weak 700 hPa frontogenesis. The 700 hPa loop also indicated a very strong north-south temperature gradient, with temperatures ranging from around -5 C in southern VA to -35 in western NY state. The cold front moved through the local area around 7:30 PM, as indicated by the sharp temperature drop and increase in winds thereafter as measured by the HSC weather station. Temperatures by Friday morning dropped to 5-10 degrees on average across the area, setting records lows at KBDR, KISP, KJFK, and KLGA. These four sites also observed record low highs for the date, with temperatures only rising into the low to mid 20s.
Attention then shifted to the massive extratropical cyclone spinning off the California coast, which most models were initializing in the low 970s hPa. The San Diego NWS page illustrated numerous hazards in effect for the region, including Coastal Flood Warnings, Flash Flood Watches, High Wind Warnings, and even Winter Storm Warnings for the higher elevations inland. Widespread heavy precipitation was forecasted, with many areas expecting 3-5 inches of rain, and up to 7 to 10 inches of liquid equivalent in the mountains. Since the area has been experiencing a major drought for quite some time, all of this rainfall at once was expected to be a little too much of a good thing, resulting in flooding issues.
Looking ahead to the new month, the threat of another winter storm was taking shape in the Mid-Atlantic and NE U.S. for Sunday into Monday. This storm was expected to develop from the energy associated with the low affecting the SW U.S., spawning a new low around the southern Mississippi Valley this weekend. Models then showed this low riding ENE along an arctic boundary, remaining rather weak but spreading a large swath of precipitation to its north and west. Model spread was still considerable, with the CMC maintaining a further south track, and thus only giving the local area a light snowfall. The GFS, which had been showing a consistent 1+ inches of QPF across the area, trended southward and lighter on the 12z run. Meanwhile, the NAM and SREF were the warmest and furthest north solutions, indicating a mostly rain event for the area. Ensemble sensitivity analysis revealed that the storm track was actually highly influenced by a shortwave rotating around the Polar Vortex over the Hudson Bay area, and not so much related to the southern stream energy.
February 21st: K.R. led a discussion titled "An Analysis of the Feb. 20th Central Plains Cyclogenesis Event"
An animation of the H5 height field along with low level relative vorticity for the previous 4 days (extending back to the 15th of February) showed a H5 shortwave digging into the Central Great Plains from western Canada into the Dakotas spawning the development of a surface cyclone around Colorado on the 17th of Feb 2014. The focus of the discussion was on the rapid surface cyclogenesis associated with the aforementioned H5 shortwave. The 18z WPC surface analysis on Friday Feb. 21st showed a ~980 mb low centered over the Great Lakes with a trailing cold front extending the Appalachian mountain range and a warm frontal boundary extending eastward from the Great Lakes roughly parallel with the 40 N latitude line bending southward over the mid-atlantic states. The 18 UTC RAP model derived sounding for KOKX (Brookhaven) showed a well-defined warm frontal boundary which had passed northward of KOKX and the general LI region around 18UTC. The warm frontal boundary was nearly stationary as of 20 UTC over the LI Sound and southern Connecticut and, due to the widespread snow cover overlaid with strong low-level warm air advection, there was a large inversion present for the majority of the morning hours generating widespread fog with visibilities < ½ of mile for the majority of the tri-state area. University of Albany QG height tendency maps were analyzed and the amplification of the H5 shortwave was largely attributed to large values of differential warm air advection ahead of the trough axis. 700 mb q-vector convg. showed a well defined couplet which was forecasted by the 12z GFS to wrap around the surface feature associated with the low’s occlusion.
The discussion then focused on the cold frontal boundary that was forecasted to traverse the general LI region around 22-23 UTC. The SPC mesoanalysis site was used to analysis the kinematics. 850 mb miller frontogenesis revealed a strong area of frontogenesis displaced 100-150 miles behind the axis of greatest reflectively. Given the orientation of the isentropes and the axis of dilatation (< 45 degrees), it was argued that the cold frontal boundary was intensifying north of the Delmarva and weakening southward of Delmarva. Due to the intense thermal gradient ,a large vertical wind shear was present associated with the frontal boundary which triggered the concern for tornadogenesis over west VA, south NJ, and Maryland. forcing and marginal instability in the warm sector. There were no verified reports of tornadoes at the time of weather discussion. Surface obs. from the south P lot and the HSC building showed a vertical migration of the frontal boundary that was hypothesized to be attributed to the marine boundary layer as the frontal zone interacted with the cold winter S.S.Ts. 12z WRF-ARW from the (COMAP WRF page) cross sections (d’-d) of the frontal passage showed significant sensitivity to model resolution with the 12 km solution depicting a rapid decrease in surface frontogenesis with the primary forcing moving aloft (700 mb) as the cold frontal zone moved over the marine boundary layer.
Thereafter, the discussion focused onto the future weather. A weak frontal boundary was progged to pass through the northeast Sunday night into Monday morning, but both moisture and forcing was weak. Ensemble sensi. analysis showed there was large uncertainty in terms of timing. Precipitation looked to be all liquid given low-level southerly flow and changing over to snow due to evap cooling into a mid-level dry layer, but this depended heavily on the progressive nature of the model solution. The next event on the horizon was Tues/Wed/Thurs time period associated with a Miller B cyclone track (redevelopment of nrth. branch off Carolina coast). All global models showed coastal development however there was tremendous spread in terms of intensity and track as the result of a complex phasing process between both two surface features. All models showed the northern stream surface feature creating a weak area of low level convergence over the northeast region but few model solutions actually showed any precipitation associated with the actual coastal.
February 14th: Mike Layer led a discussion titled "Analysis of this Week's Winter Storm and the Potential for More Snow this Weekend"
The discussion began with a look at the 500 hPa height evolution over the Northern Hemisphere for the past week, showing an active pattern with shortwave troughs digging into the Great Lakes and Northeast U.S., while ridging persisted over the West. Looking ahead to the following week, the evolution indicated a pattern change, with heights building over the eastern U.S. and troughing developing over the western and central U.S. A major culprit of this upcoming change was attributed to a piece of the polar vortex breaking off and diving south towards Alaska, amplifying into a large cutoff low.
Current conditions over the Northeast showed the major winter storm that just affected the area now located over the Canadian Maritimes. The low was down to 972 hPa, and the tight pressure gradient in its wake was producing strong westerly winds with gusts up to 45 mph at KISP. However, mostly sunny skies and high temperatures approaching 40 degrees were allowing for some melting of the snow cover.
Attention then turned to a review of Thursday's winter storm. SPC Mesoanalysis archive showed heavy precipitation in the form of snow moving through the area during Thursday morning, aided by very strong 700 hPa frontogenesis. The area was then dryslotted for much of the afternoon and early evening. Warm air pushed in at 850 hPa, with temperatures approaching 7C in the evening over Long Island, while surface temperatures remained near 32 degrees. Persistent drizzle during this time allowed for around two to three tenths of an inch of ice accretion in parts of the region. More widespread precipitation returned as the storm rapidly deepened off the coast in the evening, leading to some heavy thunderstorms in the NYC Metro, and heavy snowfall to the north and west where temperatures were colder. As the storm pulled away and cold advection ensued, all areas turned back to snow and received additional minor accumulations. Total snowfall from the storm ranged generally from 8-16 inches across the area, with about a foot observed near Stony Brook, most of which fell within a few hours during the morning. The highest totals were observed across the Hudson Valley, where some locations saw 20+ inches. Season snowfall statistics provided by the NWS revealed that the area is running well above normal, with several climate sites already in the top 10 for snowiest winter.
The SPC Mesoanalysis archive was also used to look back at the same storm, which delivered a major snow and ice event to the southeast U.S. earlier in the week. With warm advection in the low levels, but surface temperatures remaining near or below freezing due to cold air damming over Georgia and the Carolinas, many areas saw a prolonged period of frozen precipitation. Some locations observed close to a foot of snow, with areas a little further south receiving major ice accumulations up to around an inch, resulting in widespread damage and power outages.
The discussion ended with a look at another potential snow event for Saturday. Most computer models indicated low pressure moving eastward from the Tennessee Valley, off the Mid-Atlantic coast, and then bombing out rapidly as it moved northeastward off the New England coast. Many models had the storm deepening into the 970s or even 960s as it approached the Canadian Maritimes by Sunday morning. There was some differences among the models in how close to the coast the cyclone would come, leading to a very volatile snowfall forecast for the local area. The 12z GFS was the closest to the coast, bringing QPF amounts over three quarters of an inch to eastern Long Island, while the GGEM and NAM were a bit further east, bringing only a couple tenths of an inch to the area. The NWS was forecasting 2-4 inches of snow for most of the area, with 4-6 inches for eastern Long Island and SE Connecticut. Areas with the best chance of seeing significant snow (12''+) were SE New England and eastern Maine.
January 31st: Michael Erickson led a discussion titled "We will warm up, but for how long?
Mike began with the 500 hPa Northern Hemisphere Geopotential height evolution over the previous seven days, revealing that the polar vortex was split in two, with anomalously low heights in Siberia and Southeast Canada. This resulted in a very pronounced and persistent trough over the past week, which helped set the stage for an anomalous snow and ice storm in the Deep South on 1/28/14. Mike also showed the 500 hPa height anomalies for the month of January, revealing that the pattern has become more blocked than in December 2013. This has resulted in more persistent warm or cold temperature anomalies in the United States rather than the more transient cold air outbreaks of December.
The topic then shifted to the impact and predictability of the 1/28/14 winter event in the Deep South. Atlanta, GA was significantly impacted by this storm since most people were leaving school and work during the peak of the storm, resulting in gridlock lasting over 24 hours. Numerous stories of people abandoning their cars after 12-24 hours and even of a childbirth in a car, resulted in a heated debate and finger pointing. Although some government officials blamed forecasters, Weather Research and Forecasting (WRF) runs 24 to 48 hour prior to the event showed a growing snow threat for the Atlanta Metro region. In addition, the NWS issued warnings for the region several hours before the snow began. It is likely that this gridlock could have been reduced or prevented with better communication or understanding given that the forecast was relatively accurate.
Discussion then shifted to "catch up mode" since there hasn't been a weather discussion in some time. The El Nino Southern Oscillation (ENSO) continues to remain quiet and likely is having little impact on our weather. Even so, the subtropical jet hasn't played much of a role in our weather with most of our snow events forming as a result of northern-stream shortwaves. In late December, the Arctic Oscillation (AO) shifted from a positive to negative phase, indicating a breakdown of the polar vortex with the potential for more blocking in the Mid-latitudes. Meanwhile, most of Long Island has been moved from D1 (moderate drought) to D0 (abnormally dry) as a result of more precipitation in December and January.
The last several minutes of discussion were reserved for the future. As promised in the title, this weekend will feature a low cutting up through the Great Lakes and a warm up for Long Island. Highs will get into the 40's before a cold front moves through on Sunday. A weak system will pass by to the south of the region on Monday while temperatures drop to slightly below normal levels. The models have been showing an active weather pattern for next week with two potential storms on Wednesday and Saturday. However, the intensity and track for both remain up in the air and could bring rain, snow, a wintry mix or nothing to the region.
Fall Semester 2013
November 22th: Michael Erickson led a discussion titled "A look at the upcoming big cool down, weather up to Turkey Day and a brief attempt at a seasonal forecast."
Mike began with a discussion of the weather pattern over the United States during the past several weeks. In general, recent weather has consisted of a storm moving northeast through the Great Lakes region (a so called "lakes cutter") followed by a strong Canadian high pressure system. As the high pressure moves offshore, a warm return flow from the southeast develops, another "lakes cutter" occurs and the pattern repeats. This has resulted in a relatively dry pattern for Long Island with large temperature swings from day to day. Given our consistent weather patterns of late, the discussion transitions into the winter.
Mike used a very simple "seasonal forecast method" by looking at years similar to November of 2013. These similarities are based on low frequency oscillations in the ocean (the El Nino Southern Oscillation, the Atlantic Multi-Decadal Oscillation and the Pacific Decadal Oscillation) and atmosphere (the Quasi-Biennial Oscillation). In general, composites of similar years reveal no significant deviation from normal with respect to temperature (precipitation was not analyzed). This sparked a debate as to whether a composited year with no temperature anomaly can have any forecast skill. The Climate Prediction Center (CPC) winter forecast was similar with equal chances for above and below normal temperatures.
The conversation transitioned to the weekend forecast, where an Arctic blast of cold air is expected to arrive on Sunday. There was little variability in model solutions for Sunday, with the potential that highs on Sunday afternoon might not get above freezing. Finally, the discussion concluded with the potential for an East Coast storm next Wednesday. This was unfortunate timing since Wednesday is the day before Thanksgiving and one of the biggest travel days of the year. Several global models and a couple of ensembles were shown. Although the models have recently trended toward a storm track very close to the coast (i.e. rain for the big cities but potential snow inland), the ensemble spread is still significant. Therefore confidence was low regarding any forecast specifics for this storm.
November 15th: Matt Sienkiewicz led a discussion titled "Super Typhoon Haiyan: A Retrospective Analysis."
The discussion was based around the frustrating lack of available in situ data for Super Typhoon Haiyan, the storm that decimated the Philippines before making landfall again just north of Hanoi, Vietnam. The cyclone tracked just south of Taiwan's Dropwindsonde Observations for Typhoon Surveillance near the TAiwan Region (DOTSTAR) flight reconnaissance region, and it is likely that gps dropsondes from their G-IV aircraft would not have found their way near the eyewall anyways, so maximum wind speed measurements were very unlikely to begin with. Published maximum wind speed estimates were made available from the Japan Meteorological Agency (JMA) and the Joint Typhoon Warning Center (JTWC) of ~144 mph and 195 mph, respectively. The estimates were made through analysis of satellite imagery, as described by Sara Ganetis in the previous week's weather discussion. However, the value of 195 mph is suspicious, and an attempt was made to analyze photos of the aftermath in the Philippines to lend credence to that value. Photographs showed trees stripped bare of all of their leaves, but the majority of the trees appeared to be still standing. Almost all destruction seemed to be due to the horrific storm surge, and less directly due to the wind.
November 8th: Sara Ganetis led a discussion titled "Super Typhoon Haiyan and a Discussion of Model Uncertainty Regarding Next Week's Possible(?) Mid-Atlantic Coastal Storm."
Sara began her discussion by commenting about the recent frontal passage that ushered in a much cooler air mass into the Northeast but how the sensible weather over most of the CONUS was quite quiet. Everyone's focus was shifted towards the Northwest Pacific where Super Typhoon Haiyan was wreaking havoc in the Philippines and was plotting a course for Vietnam. The typhoon formed over waters near or exceeding 30 C and tracked westward in an environment with very low shear (< 10 m/s) which allowed it to form a nearly perfectly symmetrical eye and central dense overcast (CDO) around the time of its estimated peak intensity before its first landfall. There was some discussion among the media that falsely claimed that this storm was the strongest tropical cyclone in history which isn't true. Also, there aren't usually aircraft flown into typhoons in the Northwest Pacific so all of the wind speeds had been estimated from satellite imagery with what is known as the Dvorak technique. Super Typhoon Haiyan did reach the highest level of 8 on the Dvorak Intensity Scale, but a lack of ground-based measurements will mean that it is unlikely that the strongest winds produced by Haiyan will not be measured and reported. Other data from satellite included lightning observations. Lightning was found to be very prominent in the eyewall of Super Typhoon Haiyan but it is unclear exactly why beyond the understanding that the updrafts must had been most intense that allowed for a lot of variability of lofted hydrometeors. The anticyclone in place north of the Philippines steered it straight westward towards the Islands, keeping it in reach of deep, tropical moisture (PWAT values > 50 mm). In addition to the winds exceeding 100 mph, heavy rains are a hazard with this system as can be seen from this radar loop. Once it passes through the Philippines, it should round the southern flank of the anticyclone and maintain its strength as a Super Typhoon due to the very warm waters and relatively low environmental shear (26-29 C and < 20 m/s) and start on a more northwestward track towards the rugged coast of Vietnam where it should rapidly weaken and will likely cause terrible flooding as its moisture is ringed out over the rugged terrain. This storm is incredibly powerful and our thoughts are with all of those affected.
The national and local media had already been reporting on a possible Mid-Atlantic coastal storm that one or two models had resolved for the middle of next week. The problem with the hype was that there was incredible variability among all of the different models (i.e. GFS, ECMWF, CMC, UKMET) as well as considerable run-to-run variability (i.e. between consecutive model iterations). One hypothesis as to why there was hype about a storm is because the one model to resolve it for a couple of model runs was the ECMWF (aka the Euro) which, since Sandy, has been considered to be the best available model, which may be true for some situations but not necessarily for a coastal low development off of the East Coast of the United States. The NWS NYC WFO issued an informative and entertaining graphic that addressed the uncertainty around the possible event. In looking at the models, the 00Z and 12Z Euro runs both had a coastal low that was trending to be more offshore with each subsequent run. The GEFS and therefore also the operational GFS had either no low or a surface low much farther eastward over the Atlantic Ocean. It's one thing to look at the model output and see the variability of possible solutions in the medium-range but that doesn't provide much insight into why there is variability. One can speculate that maybe it's all stemming from the lack of uncertainty in the initial conditions emanating from the Arctic, where observations are sparse. Another method to look at physical reasons for uncertainty is by ensemble sensitivity analysis (ESA) which is done in-house through the SBU/NWS CSTAR collaboration. From looking at the first principle component's sensitivity, one can deduce that a deeper and slower trough originating from NW Canada will favor more of a coastal storm. Therefore, with each subsequent model run a forecaster can have more of an idea of the speed and amplitude of the trough in question to determine whether or not the storm is likely. At the time of the discussion, a coastal storm looked unlikely for Wednesday and Thursday of next week (13-14 November).
November 1st: Keith Roberts led a discussion titled "Oct 31-Nov 1st Cyclogenesis Event over Eastern U.S. & UK Windstorm."
The weather discussion held on November 1st, 2013 was focused on a cyclogenesis event occurring over the central CONUS. The discussion began with a brief overview of visible and IR satellite imagery. These imagery showed a large comma-shaped cyclone centered over Michigan with a trailing cold front extending from the northern Great Lakes down to the southeast portion of the US as of 22Z. Despite the rapid intensification of the parent surface low, doppler reflectivity showed sporadic and poorly defined precipitation structures associated with the cold frontal boundary. It was shown through the SPC mesoanalysis site that the majority of precipitation was outrunning the surface temperature boundary and, as a result, the frontal boundary was experiencing frontolysis. The cold frontal boundaries precipitation structure appeared to be maintained by H7 quasi-geostrophic omega.
Model guidance leading up to this event prompted local WFO’s to issue wind adv. for a large portion of the tri-state region. However, wind advs. did not verify and precipitation forecasts were largely overpredicted by .3 to .5 inches. Highest average wind speeds and gusts occurred during mid-morning with the passage of a small albeit well-defined band of precipitation around 11 am EDT. This feature dropped around ~ .10 of an inch of rain along LI. A narrow rope-like band was associated with this precipitation, its progression in time series observations revealed a very weakly defined surface temperature and moisture gradient. Despite its weak surface presence in the basic state fields, this feature had maintained its intensity and structure over hundreds of miles eastward independent of the actual surface forcing mechanism; this structure was highlighted for further study.
NCEP Ensemble sensitivity analysis guidance was shown from a 2-day lead perspective. Guidance showed decent ensemble coherence in timing and intensity of the MSLP surface feature which at the time was over the Great Lakes. However, the mean MSLP feature was underpredicted in intensity by the ensemble spread by 5-7 mb. Some research questions were posed by Dr. Brian Colle regarding the physical mechanisms for the poor numerical guidance regarding the frontolysis of the frontal boundary. Dr. Colle hypothesized that the topographical interaction between the Appalachian mountains and the surface front is often the culprit for downstream errors in numerical guidance.
The overall synoptic evolution and forcing mechanisms driving the rapid cyclogenesis of this event were shown. In-house wave packet guidance revealed a large wave packet envelope extending from the central CONUS to the mid-pacific. As this large wave packet envelope propagated eastward with the mean flow, it encountered a favorable area for cyclogenesis. A large area of positive PV anomalies was present over the CONUS. The wave packet excited a pre-existing cut-off H5 low over the SW US. Rapid cyclogenesis was aided through an upper-level moisture source from the GOM with PW in this region exceeding 2.5 inches. The H3 jet orientated itself on a PV boundary on the right-side of the positively tilted H5 cut-off low in the SW US facilitating rapid baroclinic energy conversion.
A brief look at the chain of cyclones that had impacted the UK and the Scandinavia region earlier in the week (Oct 25-Oct 28) was shown. A powerful surface cyclone developed in the Labrador Sea (min. MSLP ~960 hpa) and propagated eastward along the North Atlantic largely maintaining its intensity. This cyclone had a powerful upper level jet streak associate a trailing frontal boundary. These winds were able to mix down to the surface and caused widespread wind damage in southern UK and impacted a large portion of central Europe. ASCAT satellite derived winds were shown that revealed the highest average winds to be around 70 kts in a region near the German Bight.
October 25th: Mike Layer led a discussion titled "A look at the cold shot affecting the Eastern U.S. and a potential storm threat around Halloween."
Mike began with a loop of the 500 mb heights across the Northern Hemisphere, showing the troughing over the Central U.S. shifting eastward over the past week. The dynamic tropopause loop showed the tropopause heights lowering as the trough amplified and moved eastward. Looping forward over the next week, tropopause heights are forecast to rise again over the East as ridging returns and another trough develops over the central U.S. A look at the current SPC mesoanalysis showed a large 1030+ hPa high pressure centered over the Tennessee Valley, keeping conditions dry and unseasonably cold for much of the region. Locally, 850 hPa temperatures were well below 0C, in the -4 to -6C range. Attention was also brought to an interesting equivalent barotropic feature spinning off the Pacific NW coast over the past week. An animation of the daily temperature departures for the ASOS sites over the CONUS was run from Oct 1st to the 23rd. It showed the persistent warmth over the East for much of the month, that switched to well below normal over the past few days.
A look at the teleconnections on the CPC website showed the AO forecast to soar positive this week, with the NAO, currently near -2 standard devations, expected to return to near neutral to slightly positive. The PNA, currently slightly positive, is expected to drop to around neutral. ENSO conditions remain generally neutral, with a slight drop in SST anomalies over Nino 3.4 in the past few weeks. Turning the focus more locally, the updated Drought Monitor indicated the development of D1 drought over much of Long Island, with D0 scattered around other parts of the region. The annual precipitation trace for LGA showed precipitation running nearly 10 inches below normal, with no significant rainfall since June, and barely any rainfall for the month of October.
Attention then shifted to the forecast for the next 7 days. Low pressure was expected to cut into SE Canada Saturday, bringing gusty SW winds to the area. Models indicated a warming trend through the week as heights increase over the East and low pressure develops over the Central U.S. by around Wednesday. There was considerable spread between the operational GFS, ECMWF, and GGEM, as well as the GFS ensembles concerning both the strength and track of the low pressure, but the general consensus is that the low will deepen over the Great Lakes and then cut into Canada, dragging a cold front into the NE around Halloween. A stronger low tracking closer to the area would increase the chance for a significant rain & wind event for the region. A look at the ensemble sensitivities indicated that the storm's predictability had some connection with the recurving typhoons over the Western Pacific.
October 18th: Keith Roberts led a discussion titled "Medium Range North American Forecast"
Keith led a discussion with a focus on the time period extending from the present to the end of October and emphasized on larger type atmospheric phenomena (time scale ~ 15 days). The discussion began with a recap of the past weather over the CoNUS (continental United States) for the previous week. An “Indian summer” type pattern was observed for most of October with H5 anomalies over both the eastern U.S. and the Aleutian region with near climo height field over the middle part of the U.S. Drought like conditions > D3 persist throughout the majority of western great plains and parts of the Rocky mountains. While the northeast was not under any considerable drought conditions, small pockets of D0 drought are still present throughout the majority of LI and Connecticut (which has been a recurring theme through Fall 2013) which have persisted since late July. Locally on LI, the region is about 1.8 inches below climo for the 30 day period ending on the 18th of October.
A couple relevant teleconnection patterns were analyzed with the help of this(http://www.kylemacritchie.com/) website. In particular, a focus was on the NAO and AO. Both the AO and NAO were trending to near neutral conditions and forecasted to peak in the positive phase for at the least the first half of November (according to the GEFS). However, it was noted that both these teleconnection indices generally have low predictability for monthly and seasonal time scales and statistical forecasts are more skillful than human intuition. The state of the equatorial Pacific region has been shown to have considerable relations with mid-latitude atmospheric phenomena in literature, however, it too was weakly-defined (near neutral ENSO) to conclude anything. Probabilistic forecast guidance using the NMME was shown (http://www.cpc.ncep.noaa.gov/products/NMME/NMME_description.html) which indicated strong confidence of above normal temperatures throughout the south western part of the U.S, but no apparent temperature or precipitation signal was verifiable for this region. With the absence of strong signals, it was concluded that the rest of October to the beginning of Nov would feature near climo conditions.
To conclude the discussion, a brief look into the upcoming weather for the next week was shown by referring to some figures from (http://jasoncordeira.weebly.com/maproom.html) of tropopause isentropic temperatures. A positive-tilted longwave trough was expected to amplify in the middle part of the country as a >100 kt jet rotated around the leeward side of this longwave trough. This large system would eventually drag a series of cold fronts through the Northeast (one late Sat. overnight & and one around Monday morning) with the period seeing mostly cloudy skies for the majority of the next 4 days. It was shown that the longwave trough was connected to an early-season arctic air mass, and it would become noticeably cooler throughout the next week with high’s in the mid to upper 50’s throughout most of the NE. The possibility of lake-effect snow was emphasized given the presence of an Arctic air mass very early in the season as lake temperatures remain relatively warm...
October 11th: Sara Ganetis led a discussion titled "A Tale of This Week's Two Storm Systems: Their Relation and Impacts."
Sara began her discussion with a loop of the past week's satellite and surface analyses from WPC. The main features of interest was the eastward progression of the system that Matt spoke about during last week's weather discussion and the severe weather that resulted from the passage of its strong cold front across our area on Monday, 7 Oct. The SPC issued a tornado watch for most counties west of NYC and conditions were favorable for strong winds at the very least. One EF1 tornado verified near Paramus, NJ. A topic of discussion was the fact that the cold front weakened in apparent intensity (weaker reflectivity, no severe reports) as it crossed Long Island. We discussed the different factors that may have been responsible that included insufficient instability because CAPE values were < 500 J/kg from a lack of solar heating due to the pre-frontal cloud cover, the convection at the forefront of the precipitation associated with the front outran the "actual front" which separated it from the dynamics and forcing for lift, among other reasons.
The next topic of discussion was how the remnants of Tropical Storm Karen merged with the cold front offshore of Georgia and with the help of a weak upper-level shortwave was able to form into a separate coastal system which meandered up the coast. While there wasn't much upper-level dynamical support to strengthen the system, there still was a lot of moisture that was being transported westward onshore. This led to a lot of accumulated rainfall at such locations as Harrisburg, PA. The high pressure to our north kept LI dry for the most part and the pattern set in place over the Northeast resembled a rex block.
The global tropical activity was briefly looked at because of the high-impact storms of Phalin and Nari that should make landfall as very powerful storms in India and the Philippines, respectively. Typhoon Wipha may impact Japan. With all of this activity in the Western Pacific and Indian Ocean it's not surprising that the flow in the Northern Hemisphere is pretty amplified. One of the systems, likely Wipha looks to recurve and merge with the jet stream and a wavepacket is forecast to result from this merge which may affect North American weather in the long-range period.
October 4th: Matt Sienkiewicz led a discussion titled "A look at Tropical Storm Karen, as well as other weather systems not affected by the government shutdown."
Matt began with an overview of the 500-hPa heights for the past week, highlighting the mostly zonal flow over the North American continent and touching on the climatological record-setting warmth in the region. Nowcasts and forecasts for 500-hPa heights suggested a regime shift to a more progressive, amplified pattern as cyclone after cyclone was predicted to form over the Great Plains. There was some discussion as to whether tropical cyclone activity in the western Pacific could have caused this amplified pattern, and Rossby Wave Packets were analyzed to try to connect the two.
Next up was Tropical Storm Karen, which was floundering in the Gulf of Mexico. It had been experiencing a lot of shear and convection had set up predominantly on its eastward side, hindering any further intensification. Most recently convection had started firing closer to the circulation center and the storm looked to be moving in to an area of slightly weaker shear. This, accompanied with an approaching upper-level trough, gave it an outside chance at re-intensifying before making landfall. Model tracks at that time had the storm grazing the Mississippi Delta and making landfall on the Alabama coast as a Tropical Storm. Also discussed was a recent post from the Albany Maplist that highlighted observations from buoy 42001 in the Gulf earlier that day. The buoy showed a 10+ degree Fahrenheit spike in temperature accompanied with severe convection, and the question discussed was if it was due to a convective heat burst or just bad observations.
Finally, the discussion was wrapped up with the cyclone making its way across the Great Plains. The low center was located in Nebraska, and heavy banded snow was found to the north-northwest in South Dakota. The Storm Prediction Center had posted a moderate risk for the areas out ahead of the eastward-moving low, and reports of large hail in eastern Nebraska and Iowa had started coming in. The discussion ended with a quick look at WPC QPFs for the next few days to get a feel for the precipitation expected in the region from the approaching system.
September 27th: Michael Layer led a discussion titled "A Look at the Upcoming Weekend Storm in the Pacific Northwest and the Forecast for the Next Week."
Mike began with an overview of the conditions across the lower 48. Conditions were dry across much of the country, with a large area of high pressure over the eastern U.S. Some showers were occurring in the western central Plains in association with a low pressure system over Colorado, but stronger thunderstorms were expected to develop ahead of a cold front as the low moved northeastward. SPC had a slight risk (5% tornado, 15% hail and wind) from the Texas panhandle, northward into Western Oklahoma and Kansas. SPC mesoanaylsis showed low-moderate CAPE (500-2000 J/kg) and 0-6 km shear of 30-40 kts moving in from the west, along with low-level moisture advection from the south.
Attention then shifted to the major storm system set to impact the Pacific Northwest over the weekend. Low pressure over the Gulf of Alaska was expected to move southeastward as the trough amplified over the region. The NWS issued High Wind Watches and Warnings for coastal Oregon and southern Washington for wind gusts of 60 to as high as 80 mph Saturday into Sunday. Heavy precipitation (in the form of snow in the higher elevations and rain elsewhere) was also expected over the region. The HPC 7-day precip accumulation graphic (valid through next Friday morning) showed 5 to locally over 10 inches in parts of Oregon and Washington, most of which would fall in the Saturday through Monday time period.
For the Northeastern U.S., a weak area of low pressure well offshore was forecast by some models, such as the GFS, to retrograde far enough west to bring clouds and a chance of light rain to coastal sections on Monday. Looking ahead to the remainder of the week, a ridge was forecast to build over the eastern half of the U.S., with dry conditions and a warming trend expected. Westerly flow was expected to advect unusually warm air into the region, with the GFS indicating 500 mb heights building to 585 dm and 850 mb temperatures reaching 17C by mid to late week. The NWS graphical forecast products showed high temperatures going from the low to mid 70s this weekend to near 80 in the NYC metro area by Wednesday and Thursday. There were some signs of the pattern becoming more amplified by the end of the week, with a trough and associated low pressure system deepening over the Great Lakes by Friday. However, the CPC 6-10 and 8-14 day outlooks indicated a good chance of above normal temperatures continuing for the east.
September 20th: Michael Erickson led a discussion titled "A Look at the Joyous and Wild Temperature Swings of Autumn.”
Mike began with a review of the Catastrophic Flooding Event in Colorado that took place between September 9th and 15th where some areas received over 20" of rain. Although the upper level height field during this time did not appear very threatening, monsoon moisture was advected in from the south, while low level topographic up slope and a stalled frontal boundary was in the vicinity. The worst of the flooding occurred between September 12-13 when numerous showers and thunderstorms were in the region. The discussion switched to Atlantic tropical activity, which remains generally quiet. Humberto has dissipated and there are two tropical interests in the Atlantic, although both of them have a low chance of formation. The main story is Super Typhoon Usagi which is forecast to head toward China and another region of interest to Usagi's southwest that should be watched. This particular region of interest is currently forecasted by many GEFS members to become a typhoon and recurve toward Alaska in several days.
Finally the discussion switched more toward the proposed title and focused on our temperature ups and downs in the Long Island area. More recently we started with a very warm early September and then switched to below average temperatures for the past week. The NYC area has also experienced a growing rain deficit over the past couple months because most of the inland convection has dissipated before reaching the coastal plain. A look toward the future shows that generally below normal temperatures will continue, but not before a brief warm up for today and tomorrow.
September 13th: Weather Discussion was cancelled on account of it conflicting with the SoMAS New Student BBQ.
September 6th: Sara Ganetis led a discussion titled "While we cool down in the Northeast, tropical activity is heating up."
Sara began discussion with a review of the local severe weather event of 3 September. Convection fired up along a cold front as it translated across the island in an environment with plenty of moisture and instability that allowed for heavy precipitation and severe hail (1.00 inch in diameter). The passage of the cold front and setup of the broad trough over the Northeast made for a comfortable end of the week with a decrease in temperature and humidity.
The discussion continued to the recent activity in the Tropics. Tropical Storm Gabrielle had since dissipated after crossing over the harsh terrain of Hispaniola and into an unfavorable environment of drier air and higher shear (just downstream of the upper-level trough). Tropical Storm Lorena was making landfall along the southern tip of the Baja California Peninsula and then would likely dissipate. There were a few areas of possible development, but nothing more than 10% within the next 48 hours because of the proximity to unfavorable environments (too much shear, not enough moisture). The most hope for development at the time of the discussion was for a fresh wave coming off the coast of Africa that was just scattered convection but expected to organize as it moved over the warm ocean waters. The CIMSS Tropical Cyclones page was used to examine the tropical environment.
The discussion concluded with a brief forecast for the area of a cold frontal passage for Sunday that may force some scattered showers in the area. Other than that, there wasn’t too much of interest to discuss. If a hurricane does not form by 11 September then that will break the previous record of the latest formed hurricane in the Atlantic. Hopefully by next week there will be an active TC in the Atlantic so stay tuned!
August 30th: Brian Colle led a discussion titled "Why has the Atlantic hurricane season been so quiet, and how long will it continue?"
Dr. Colle began his discussion by providing some statistics to put the current Atlantic hurricane season into perspective. He showed NOAA’s official forecast from May for 13-20 named storms with 11 hurricanes and 3-6 major storms. Their latest updated forecast called for 18 named storms with 8 hurricanes and 3 major storms. Despite the relatively “slow start” they are still calling for a lot of activity. The climatological peak for TC genesis in the North Atlantic is about the week of September 10th (NOAA/NHC), so it’s not like we already missed the expected peak period of activity. Dr. Colle discussed the quantity called accumulated cyclone energy (ACE) which measures the relative intensity of each storm by estimating the energy used by each storm and is proportional to the square of the wind velocity (more info here). On a per-month basis, ACE also peaks in September in the Atlantic (from Dr. Ryan Maue’s page). Given this information, perhaps this Atlantic hurricane season shouldn’t yet be written off but it looks like the date of the first hurricane formation may break records (McNoldy, CWG). ACE on a globally integrated scale has been decreasing since 2005 (Maue 2011) but that may be a result of there being less intense storms during the later period (Maue’s page).
What was the deal with the start of this Atlantic hurricane season? To date there have been 6 storms. ENSO isn’t a major player this season, so Dr. Colle showed some plots to explain why the activity had been low starting on 1 July. The first point that he made was that there was anomalously dry air above the boundary layer (-12% RH anomaly) stretching westward from Africa all the way across the Atlantic into the Caribbean especially during the period 1-15 August. The Saharan Air Layer (SAL) provided a harsh environment for TC development and growth. In analyzing vertical stability, the Tropics are more stable this year compared to climatology (McNoldy, CWG) so that would act to discourage convection. SST anomaly maps didn’t show too much of an explanation for why activity has been weak. Maps of 500-hPa shear anomalies did show that during the period 1-15 August there was a 2-3 m/s westerly shear anomaly in the tropical Atlantic basin. Therefore, the weak activity was shown to be likely tied to the dry air and westerly shear that created an unfavorable environment for TC development.
How long will this stifled activity last? Likely not long, Dr. Colle explained. Climatological shear values are back in place and the dry anomalies are starting to weaken. At the time of the discussion, there was an invest area with a 40% probability for the chance of development that should move westward into an area with weaker shear but still some residual dry air. If it makes it past the subtropical high, according to the GFS, it may encounter a trough that may recurve it and keep it away from the Caribbean and East Coast. We’ll see what happens. Looking beyond this one system and to the rest of the Atlantic hurricane season as a whole, most students at the discussion agreed that they were not ready to give up on the season yet. Another factor that was discussed was the MJO moving into Phase 8 and then Phase 1 which has been shown to increase precipitation in the tropical Atlantic (NOAA/CPC).Spring Semester 2013
Spring Semester 2013
May 3: Matt Sienkiewicz led a discussion titled, "Record May Snow in the Midwest."
Matt began his discussion by displaying a loop of the 500-hPa heights that shows how much the pattern had changed since early last week that ultimately led to the record May snow event in the Midwest. There was clearly a regime change due to persistent ridging along the East Coast and the deep trough that formed across the Central US. The 500-hPa height pattern suggests that this is the result of the downstream propagation of energy from a strong trough over the Eastern Pacific early on in the period. In looking into this further, neither the AO nor NAO could really be used to see a clear shift.
The system in the Midwest produced up to 18 inches of snow in parts of Wisconsin. Was the record May snowstorm accurately predicted? After some discussion, the consensus was that the models captured the trough in the medium range but the uncertainty was around the snowfall totals. We looked at the environment during the times of heaviest snowfalls and found that the precipitation was collocated with strong 700-hPa frontogenesis along the cold front. We looked at model-derived soundings to see that the isothermal layer around 0 C was present from near the surface to 850-hPa due to melting effects. The consensus of the group was that we were glad we weren’t forecasting precipitation type during this event.
A discussion about the connection of the extra-tropics to the tropics ensued after looking at the surface observations and analysis and seeing how far south the cold front was analyzed (i.e. southern Mexico). Dr. Colle stated that he and some colleagues had done some research that a rope cloud can be formed that propagates into the Gulf due to ageostrophic down-gradient flow caused by the cold pool of the cold front in that region of Mexico.
The system in the Central US and its persistence has had negative effects on flooding, especially for the Mississippi River. Some stages affected by the snow in the Upper Mississippi Valley were expected to come down from moderate flood levels on Friday 3 May, while others in the lower Mississippi Valley were expected to approach moderate flood levels due to the forecast precipitation. The SREF ensemble mean was showing 2 inches of rain in the next 12 h (12-00 UTC 3-4 May). Matt showed that the location of the surface low and subsequent circulation was allowing moisture to be advected by the mean flow from the Gulf of Mexico and even westward from the Atlantic.
Matt ended his discussion with a forecast for the next week. The GFS and ECMWF both agree that the low will cutoff and remain pretty stationary until coming up the coast sometime next Thursday-Friday. This may threaten our de Zafra Symposium with poor weather, but we’ll be inside anyway so it should all work out! A May “Ridge of Death” sets up over the Central US which makes for very poor conditions for anyone who may be trying to storm chase during what should be a very active time. The ensembles are not exhibiting much spread with the previously discussed patterns. Matt concluded his discussion by saying that the Knicks should have pleasant weather for their bus ride back to NY after getting destroyed by the Celtics. (Edit on 4 May: Matt ate his words after the Celtics lost).
April 26: Sara Ganetis led a discussion titled, “Oh Hi, Spring! A Look at the Large-scale Pattern and its Forecast of Persistence.”
Sara began her discussion with a look at the Northern Hemispheric infrared satellite loop to motivate her discussion. Her discussion focused on the pattern change from progressive troughing in the Northeast to a predominantly zonal flow and the forecast ridging to persist in the region until the middle of next week. The slight chance of severe weather in the Central Plains and the threat of fire weather over the Northeast were also discussed.
The previous week’s progressive pattern that led to a pair of cold frontal passages across the Northeast was shown by looping a plot of the potential temperature on the dynamic tropopause which also showed the forecast raising of the DT with the forecast pattern into next week before the 12 UTC 26 April-initialized GFS forecast resolves an amplified trough digging into the central U.S. and becoming cutoff. This trough and resultant cutoff will be something to pay attention to in subsequent model runs. Another thing that was noticed was an area of anomalously lower heights at 500-hPa in the Central Pacific. This is associated with a cutoff at 500-hPa that persisted since ~10 April and is thus the reason why the PNA began to trend negative. A negative PNA in April is associated with ridging and anomalously higher temperatures in the Northeast, as we are expecting.
Taking a closer look at the current weather, the 15 UTC surface analysis from 26 April showed a weak high pressure system over the Great Lakes but the main player is the low located over the Texas panhandle. There was showers and thunderstorms collocated along and north of the warm front. The slight risk is for most of Oklahoma and upon looking at the soundings from 12 UTC 26 April, there is a large elevated mixed layer and MUCAPE > 1000 J/kg. The shear seemed sufficient for the development of supercells. The low-level southerly flow out ahead of the 500-hPa shortwave located over New Mexico should aid to advect moisture from the Gulf into the region to increase the dew points. At the time of the discussion there were no storms fired up yet. Mike Erickson added to the discussion by showing his fire threat webpage and discussed how the low relative humidity and relatively strong winds caused the Haines index to exceed the moderate fire threat level.
The discussion was wrapped up by taking a look at this morning’s 12 UTC OKX sounding that showed a very strong surface inversion that was caused to the sounding location being located near the Pine Barrens that are very efficient at radiatively cooling overnight. A brief discussion of the forecast wrapped things up by saying that high pressure will dominate the Northeast for this weekend and should persist, according to the GFS, until the middle of next week. Sara advised everyone to go outside and enjoy it.
April 19: Matthew Souders led a discussion titled, "A Retrospective on the 2012/2013 Winter Forecast and a Look at Summer and the Tropics."
Matt began his discussion with a discussion of the strengths and weaknesses associated with his 2012-2013 winter forecast. Regarding snowfall, he shared some numbers with the group including that New England snowfall was 1.5 times that of normal snowfall totals. Matt had forecasted for a warmer than observed winter but mused that it had been a very difficult winter to forecast for. The leading driver was explained to be the quasi-biennial oscillation (QBO), or the descending branch of the easterly wind anomaly that reached the tropopause. In doing so, Matt explained that this would induce midlatitude blocking regimes and result in a negative phase of the Arctic Oscillation (AO). Some discussion ensued about the possibility that midlatitude wave breaking energy could reach the Tropics, or the opposite- that an oscillation confined to the Tropics could have much influence over midlatitude activity.. A few papers worth reviewing on this topic are as follows: Reasor and Montgomery (1999), Naito and Yoden (2006), Hitchman and Huesmann (2009), etc. While looking at the QBO by means of a time-height section of monthly mean zonal winds at an equatorial station Singapore, 1°N/104°E (since Jan 1976) (download pdf), we wondered where the sudden stratospheric warming (SSW) event of early late Jan/early Feb fit in. It was clear that our impactful blizzard of 8-9 Feb occurred as the AO was starting to trend strongly negative and also the descending branch of anomalous easterly winds reaching the tropopause. To summarize his review of his winter forecast, he showed that when you plot a winter composite anomaly map, the anomalies wash each other out and it seemed to be an average winter. However, March for example, exhibited areas of < -4 sigma for surface air temperature across the eastern half of the country. His final thoughts on his forecast was that due to the lack of ENSO activity and unexpected importance of the QBO, his forecast for a warm winter was slightly off.
Using his knowledge of the current state of specific climate indices, Matt provided his qualitative forecast for Atlantic tropical cyclone activity. He claims that if you use the current phase of the QBO and the lack of ENSO signal (termed “La Nada”) then analog years may provide some support of East Coast TC activity for the 2013 season. Another important factor is SST; Matt explained how the warm/cold/warm tripole in the Atlantic will likely be connected to deep tropical activity. He showed the current water vapor imagery of the Tropical Atlantic and discussed how the tropics are already quite moist. The location and strength of the High over the South Atlantic will aid to regulate the easterlies bringing dry, dust-laden Saharan air off of Africa versus the westerlies currently cutting across the Caribbean. Another interesting point is that the current solar cycle maximum was the weakest since ~1930 and looking at analog years of the next-weakest years provides further support for East Coast TC activity. Another point is that the East Coast activity will depend on the existence and persistence of cutoff troughs that could act to guide the systems up the coast, otherwise we’ll be dealing with a “ridge over troubled waters.”
An undiscussed interesting current topic is the severe weather associated with a cold front that is extending along the entire Eastern Coast on Friday afternoon. SPC has issued tornado watches that cover from NC into PA-bordering NY counties. At the time of the discussion, one tornado warning was issued for Mecklenburg, VA associated with a pre-frontal isolated cell. The potential for more tornadic supercells exists out ahead of the main cold front in an environment of high shear and ample moisture and instability.
April 12: Keith Roberts led a discussion titled, "The Vagaries* of the Weather."
Keith’s weather discussion took place on a rainy Friday afternoon. He used the current weather to motivate his discussion that focused on the midlatitude cyclone that caused severe weather in the Central and Eastern U.S. followed by a forecast for the coming week and a general discussion of forecast tropical activity. Keith began his discussion with introducing an interesting forecasting tool that he found created by a company called Dark Sky. An interesting display tool is model comparison of temperature, precipitation, and other sensible weather quantities at any location.
The current and recent pattern of a midlatitude cyclone was discussed starting with the Weather Prediction Center’s analysis that showed how the precipitation had outrun the surface cold front. Looking at precipitation from the national mosaic NWS radar and comparing it with the IR satellite imagery, there is hypothesized to be some sort of cold front aloft. WPC analyzed the cyclone as an open-wave, whereas the cyclone appears more occluded at least at upper levels. At the time of the discussion, the warm front was positioned to our south which is supported by the 12 UTC OKX sounding which showed an 850-hPa inversion and veering winds with height. Another useful way of looking at frontal boundaries is plotting potential temperature and its associated gradient. This led into a discussion of a strong sea breeze boundary that developed over NJ yesterday (10 Apr). There was some discussion about why this strong sea breeze front didn’t initiate any convection. John Mak flew through it and found it to be approximately < 300 m, so a shallow front.
The convection of 10 April made its way to Long Island and prompted the issuance of a severe thunderstorm warning from the NWS for portions of Suffolk County. Most students reported seeing an impressive shelf cloud along the leading edge of the complex. There was a lot of surprise about how it held together despite there being a lack of MUCAPE. Upon looking at the 00 UTC OKX sounding from 11 April, there appears to be an elevated mixed layer beginning at 850-hPa. If that level could be saturated then there would be a lot more calculated CAPE.
Keith then provided a brief forecast discussion for the near-term. The GFS initialized at 12 UTC 12 April showed the potential for severe weather for LA/TX around 00 UTC 18 April, given forecast CAPE > 2000 J/kg, a relatively strong temperature gradient, and higher moisture values advected by the mean flow from the Gulf. The general pattern will be a large trough in the West with weak ridging in the East. The 8-10 day mean 500-hPa heights from the GFS and ECMWF capture this pattern and with little variability of the GEFS shows confidence in this forecast pattern which may spell below average temperatures in the Central U.S. and above average precipitation for the East Coast.
CSU announced their forecast for the coming hurricane season which prompted some discussion about their reasoning. They are expecting ENSO to trend to La Nina, however current trends are unsupportive of that and dynamical models are in the transition season so aren’t trustworthy yet. We looked at the cross-section of water temperature anomalies and saw a warm anomaly at ~150 m depth break off and progress eastward and break up the trending cold surface temperature anomalies in the East Pacific. We are hypothesizing that this is due to a Kelvin wave. Our group’s current thinking is that the SST and water temperature profile are warm enough to support an active TC season in the Atlantic and current ITCZ activity is robust which shows that the Tropics are fueled up with moisture, which will be something to pay attention to. We’ll see if their forecast for 18 named storms and 4 major hurricanes verifies.
*Vagary: an unpredictable or erratic action, occurrence, course, or instance: “the vagaries of weather; the vagaries of the economic scene.”
April 5: Michael Erickson led a discussion titled, "When Will it Feel Like Spring?"
Mike began with an overview of the general pattern of the past week by showing the GFS 500-hPa geopotential height analyses and forecast. The main points discussed were the coastal storm that remained offshore Friday which allowed for surprisingly pleasant local weather, the multiple days of high fire threat in the region, and a general discussion of some climate indices used to answer the title question of the discussion of when it will consistently feel like spring.
There was a forecast challenge regarding the inland extent of the precipitation associated with a coastal low forecast for Friday, 5 April. The NCEP operational models differed on the general position of the low forming offshore of the Carolinas and translating northeast with the NAM taking the low farther northwest than the GFS. Most forecasts called for rain at least in the morning. The 21 UTC SREF run from the 4 April showed 0.1” liquid precipitation extending into CT. However, the cyclone tracked farther south and the precipitation did not extend much even into Long Island; Islip (ISP) had a trace of precipitation, with a bit more at Westhampton Beach (FOK) located farther east. While the forecast rain amounts weren’t impactful, the lack of measurable precipitation and the ultimately pleasant Spring-like conditions were surprising.
Mike conducts research into high fire threat days in the Northeast U.S. so he provided some expert insight into the previous high fire threat day of Wednesday, 3 April and the surprise of Friday, 5 April. A high fire threat is defined when conditions are favorable for the development and spread of wildfires which includes relative humidity < 30% and wind gusts > 25 mph. When these conditions are met, the NWS typically issues a Red Flag Warning. During the discussion (3 PM Fri, 5 Apr), the NWS hadn’t issued a Red Flag Warning for our area and upon checking observations around the island, RH < 20% but the windspeed criterion hadn’t been met. Fires did break out on Wednesday and Thursday throughout the Island, especially out east near the Pine Barrens.
Mike then switched gears into discussing the general climate pattern. He showed two interesting plots of the precipitation time series at LGA and the temperature time series at LGA for the past year. From the precipitation time series, it can be seen that there has been a steady deficit since the Fall. From the temperature time series, it can been seen that below normal temperatures were not common until the onset of Spring of this year. This clearly shows that Spring has been delayed. Upon further reflection of the recent past winter, Mike discussed temperature anomalies with depth for each ENSO region. The current signal is sort of washed-out and is referred to as “La Nada.” Spring is notorious for being difficult to forecast ENSO, so that sparked discussion about whether there extratropical forcing may affect the Tropics. A vertical cross-section of the geopotential height anomaly shows the stratospheric warming that occurred in late January and has since propagated into the troposphere.
Mike wrapped up his discussion with a forecast for next week. There is an active pattern with a few hot spots, specifically a severe weather threat for the Plains on Wednesday. There should be the potential for weak ridging for our area to allow us to see temperature around 60F. To answer the question, “When will it feel like Spring?” the CPC 6-10 and 8-14 day forecasts for temperature show the East Coast in the above average category, likely due to the active pattern and potential for ridging and the resultant warm air advection. Mike concluded his discussion warning that this is cut-off season so the temperatures and pattern can fluctuate. He admitted that it should feel like Spring at least until next Wednesday.
March 29: Sara Ganetis led a discussion titled, "The Snowless Event of 25-26 March* and the Severe Weather Threat for TX/OK."
Sara began by showing the global stitched IR satellite imagery to highlight the cutoff cyclone off of the coast of California, general troughiness over the East Coast, and an interesting area of tropical convection over the northern shore of Australia. A look at the 500-hPa heights according to the GFS analyses and forecasts shows a the cutoff cyclone creating a rex-block pattern in the Eastern Pacific.
Sara quickly touched on the potential for above average temperatures for the East Coast for the month of April likely due to the PNA pattern trending negative and the three-month outlook of April-May-June potential for above average temperatures likely due to the MJO convection translating eastward into the West Pacific.
The main topic of discussion is the current threat of severe weather over the Southern Plains, specifically southern Oklahoma and northern Texas. In looking at the current radar, some convection had already begun to develop causing some hail to be reported in southern Oklahoma. Southerly flow brought moderate moisture up into Oklahoma (dew points at least in the upper 50’s F) and the solar insolation allowed for surface-based CAPE to be > 2000 J/kg. The main question was what was providing the lift to initiate the convection? Upon further analysis, there was a warm frontal boundary along and slightly north of the convection. There is a greater threat for severe weather for the same region for tomorrow. The threat is stronger likely due to the involvement of convection along a dry line as evident from the forecast low-level theta-e field. Looking at some forecast soundings for Norman, OK (KOUN) showed moderate shear and plenty of CAPE. Attention was brought back to the East Coast in a quick forecast discussion of the Easter showers forecast for this Sunday. A broad cyclone located in Eastern Canada collocated with a long-wave trough will pass to our north and any resulting precipitation will likely be from the cold front.
*Note: Due to the active severe weather, we didn't have time to discuss the lack of snow that was forecast to occur for the 25-26 coastal storm. The bottom line was that it was too dry initially and too warm.
March 22: No Discussion (Spring Break)
March 15: Matthew Souders led a discussion titled, "A Look into March and April: When Will Winter End?"
Matt began his discussion by showing a loop of 500-hPa geopotential height anomalies from the past 30 days to show the general negative anomalies in place over the East Coast from February through early March. He then showed the forecast 500-hPa geopotential heights that displayed a progressive pattern that supports the fact that winter is sticking around for the East Coast and that the ensemble of Ground Hogs that had predicted an early spring were, surprisingly, wrong. Although the potential exists for a high-impact system on Tuesday, Matt geared his discussion to be more of a large-scale overview of the general patterns in play and their forecast effects on sensible weather for the rest of this month and into April.
Matt walked us through the various weather and climate indices. He showed us that there exists a weak El Nino signal and that there is a subsurface warm pool sloshing easterly, but ENSO is not the real driver due to the weakness of the signal and he claims it is less an "El Nino" and more a "La Nada." Next we looked at the Arctic Oscillation (AO) and saw that while the AO has been consistently negative, forecasts show a “tanking” AO with the index values becoming less than -5, which is a very strong signal. The GEFS and ECMWF forecasts support the significant negative forecast AO. The North Atlantic Oscillation (NAO) didn’t have as strong of a signal hypothesized due to the fact that this index is Atlantic ocean-based and by nature more stable or that most of the amplification is in the Pacific Basin, which is captured by the AO. The Pacific-North American Pattern (PNA) is also forecast to go negative but has a weaker signal than the AO. The Madden/Julian Oscillation (MJO) is forecast to move into octants 7, 8 and 1 within the phase diagram which has been linked to active weather on the East Coast. The Blocking Strength signal is weak now but blocks may set up over Alaska and Greenland. Looking at the influence of the stratosphere, the 30-hPa temperature forecast shows no evidence of a sudden stratospheric warming event causing the AO to “tank” according to theory that the AO is regulated by downward control from the stratosphere. However, there is no well-defined and cold polar vortex as one would expect with a negative AO. The final key player discussed is the Quasi-Biennial Oscillation (QBO) and what is shown is the easterly phase descending towards the tropopause which means blocking is favored.
To summarize what all of the indices told us to provide a story of what to expect on the East Coast, Matt concluded by showing what the Climate Prediction Center (CPC) has stated in their 6-10 day outlook and 8-14 day outlook. He stated at the storm track will likely be more southward than than the existing maximum precipitation position, but agreed with the below normal forecast temperatures. To answer the question, “When will winter end?” Matt said that the main driving forces are the AO and QBO. He created composite anomaly maps for the month of March from the years during which there was the easterly phase of the QBO (2005, 2001, 1996, 1987, 1971, 1969, 1958, 1952) and found that the surface air temperatures are cold but for the month of April there should be a bit of a warming, or at least return to normal. He stated that winter will lock and hold through March with not much of a break until April. What will that mean for the severe weather season? One hypothesis is that it just may be shifted later in the season, but not necessarily be weaker.
We had some special guests during our discussion and would like to thank our alumni and prospective students for their participation and patience with the lack of “normal” seating!
March 8: No Discussion
March 1: Keith Roberts led a discussion titled, "The Contemplation for the Potential of a Major Coastal Storm Next Week (Mar 7, 2013)."
Keith began his discussion by looking at the current loop of the global stitched satellite imagery to identify the train of multiple cyclones extending across the North Atlantic. According to the 500-hPa GFS forecast, there will be a merger of two cyclones in the North Atlantic. This general pattern sets up a persistent trough over the East Coast extending into next week. The main topic of discussion is the possibility of a major coastal storm on Wednesday, 7 March. Because it is a medium-range forecast, there is uncertainty in the exact track and intensity of the storm, however all models resolve it.
General Evolution: Matthew Souders provided some input about the idea of a wave packet, or the effect of one cyclone on cyclones downstream. Because the system is the leading edge of the wave packet and the energy of the wave packet moves faster than any individual cyclone. Thus Keith explained that the activity in the Pacific is the driving force behind the amplified pattern and forecasted shortwave that should develop over British Columbia on Sunday and propagate across the US until it becomes phased with the broad trough in place, and intensifies off of the coast.
Model Discussion: After looking at some plots from the 12 UTC 1 March deterministic GFS run and 00 UTC ECMWF run, it became apparent that the position of the cyclone forecast for 12 UTC 7 March was literally up in the air. As we’ve seen with previous storms, the position of the low is important for the precipitation type received so the ECMWF was making the event look like a snow event, however the GFS looked like rain.
To further understand the sense of uncertainty with the event, we looked at Ensemble Sensitivity Analysis (ESA) with helpful discussion from Minghua Zheng and Dr. Colle. What we saw from the GEFS Day 6 Summary (verifying time = 00 UTC 7 Mar) that the ensemble mean had the cyclone was off of the coast of NC/VA and more than 56% of the variance was attributed to the 1st Mode (EOF 1) that showed lower pressure to the north of the mean low’s center location, which meant that a lot of members were resolving the system farther north. It will be interesting to note how this will change with time. A look at the ECMWF ensemble showed that the mean and spread was more threatening than the GEFS because it was farther north and most of the spread was still near the 40-70 benchmark. Looking at the 500-hPa heights associated with EOF 1 for both the GEFS and ECMWF showed that the ridge in the North Central US on the 1 March (Day +3.5) will likely determine the position of the low. This makes physical sense because if the ridge is more amplified, then the downstream trough may be more amplified which would bring the low farther south and vice versa.
Storm Impacts: There is the potential for strong winds with this event. Keith showed some forecast model soundings (12 UTC GFS 1 Mar) for near KISP that showed 65 kts at 850-hPa and 40 kts at 925-hPa at 12 UTC 7 Mar. He then showed some horizontal plots of MSLP and 10 m winds and pointed out the strong winds wrapping around the cyclone that are likely enhanced due to the presence of a strong high (<1036 hPa) over Eastern Canada. The forecasted intensity of the low according to the GFS was 992 hPa.
The precipitation type is still uncertain, but the GFS showed 48-h total liquid accumulation for the event to approach 2 inches mainly north and west of the cyclone center.
Due to the duration of the event and likely persistent easterly winds, there exists the potential for storm surge to be an impact. Keith described how this depends on the position of the low but the large fetch with the prolonged strong easterlies really are “perfect” for storm surge. He would have shown a hydrodynamical model but they are not run out that far into the future.
Keith concluded that this storm does have the potential to be a high-impact event, but we will just have to wait for that Pacific energy to reach the coast and the key players to form and be observed and input into the model.
February 22: Michael Layer led a discussion titled, "A Look at the Upcoming Active Weather Pattern Over the Next Week."
Mike began his discussion by looking at the three main items of interest of the current weather: the cutoff low in the Midwest, the low affecting the Pacific Northwest, and the broad area of convection over the Southeast. The cutoff low had brought record snowfall on 21 February to places such as Kansas City, MO mostly due to warm air advection creating a broad area of precipitation that lacked any banding structure. The low affecting the Pacific Northwest is bringing snow to the higher terrain of the Cascades. The convection in the southeast seems to be associated with a weak frontal boundary and with low MUCAPE values of 500-1000 J/kg.
Mike then went into a discussion of the teleconnection indices and their forecasts to discuss the CPC’s forecast for March. The AO and NAO are trending negative and coupled with the positive trending PNA could mean troughy weather for the East Coast. An excited mumbling of seeing a Nemo: Part II ensued. We have been having an ongoing discussion of the persistent drought throughout most of the country since last semester and Mike expressed that he hopes this active pattern will help alleviate the severe drought conditions in a few areas of the country. The new seasonal CPC discussion for the spring was brought up and their forecasting methodology was discussed.
Mike moved on with a discussion of his forecast for the weekend disturbance that will likely bring solely rain to the region. He showed the 1200 UTC 22 Feb deterministic runs of the NAM and GFS and showed how the cutoff low over the Great Lakes will aid to form a secondary low off of the coast of North Carolina which will move north-northeast. There was a discrepancy between the NAM and GFS over the position of the low and its intensity, the NAM forecasting a stronger low that was closer to the coast. The winds are expected to be offshore, according to the NAM. Also, due to the easterly flow north of the cyclone, warm air should ensure that this is not a snow event. The SREF is forecasting for between 0.75-1.00” of rain for Islip. Another thing to note is that due to the strong easterly flow, there is the potential for storm surge, however due to the speed of the system and the fact that it doesn’t directly correspond with a Full Moon, sites aren’t expected to exceed moderate surge levels.
Mike concluded his discussion with the forecast for the likely Tuesday, 26 Feb event, which has the potential to be a high wind event. The GFS is showing a developing coastal storm within the 40-70 benchmark with 70 kts at 925 hPa.
February 15: Keith Roberts led a discussion titled, "A Review of Feb 8-9th Blizzard & A Look into the Possibility of More Snow this Weekend."
Keith began his discussion with showing a loop from the past two weeks of the dynamic tropopause on the 2 PVU surface from which we could see the evolution of the northern stream and southern stream systems and rapid intensification of last weekend’s notable event, the 8-9 February 2013 Snowstorm. He led the the discussion on the significance of this event and provided material that stimulated the discussion of scientific questions.
Significance: This nor'easter was significant for two main reasons. Firstly, for the fact that it evolved from the merging of both the northern and southern stream, i.e. a coastal low interacted with an upper-level trough which further enhanced it. Looking at an archived loop of satellite imagery really shows this evolution. Secondly, an classic band of intense snowfall set up northwest of the cyclone center which caused localized higher snowfall rates and totals across portions of Central Long Island and Connecticut that forecasts had underestimated. Stony Brook measured 28.0 inches of snow and the highest measured on Long Island was in Medford with 33.5 in. The intense, heavy snow during the evening commute caused cars to be stranded and subsequently buried for days.
Science Questions:
What were the predictability issues? The deterministic and multi-model ensembles underpredicted the precipitation amounts which suggests that there is a microphysical challenge within the models. This event was interesting because the snow to liquid water ratios varied in time and space in relation to the band. If the microphysical schemes assume a constant density then that will result in a large model moisture error.
What is that warm tongue at 850-hPa around 00 UTC 9 Feb just south and east of NYC? Some hypotheses include that warm air between the two streams got “squeezed” in between as they merged, diabatic heating related to the intense snowband, warming via subsidence from the descending branch of the ageostrophic circulation that developed with the frontogenesis associated with the band. We discussed Matt Sienkiewicz’s WRF runs for this case to try to further isolate this issue if he tries to turn off latent heating. Another thought is that there is a stationary gravity wave in place due to the latent heating.
What was the nature of this band and how was it different from others observed? We discussed how the convection seemed to be more upright than slantwise, which could explain why there were reports of “thundersnow” associated with the band. Also, the microphysical evolution of the band was interesting because at one location within a few hours one would witness large aggregated flakes, then small grains of graupel, then fine, dry snowflakes. Dr. Colle is looking into this more after having spent the night taking pictures of the hydrometeors with a microscope.
Recommended Reading:
- Rich Grumm of NWS State College has already put together a great post-mortem of the event found here.
- Hakim et al. 1995, “The Ohio Valley wave-merger cyclogenesis event of 25-26 January 1978. Part I: multiscale case study.”
- Novak et al. 2004, “An observational study of cold season-banded precipitation in Northeast U.S. cyclones.”
- Novak et al. 2006, “A forecast strategy for anticipating cold season mesoscale band formation within Eastern U.S. cyclones.”
- Novak et al. 2008, “High-Resolution observations and model simulations of the life cycle of an intense mesoscale snowband over the Northeastern United States.”
- Novak et al. 2012, “Diagnosing snowband predictability using a multimodel ensemble system.”
- Stark et al. 2013, “Observed microphysical evolution for two East Coast winter storms and the associated snow bands.”
February 8: Sara Ganetis led a discussion titled "Northeast US Blizzard: Impacts on Long Island"
The newsmaker this week is the potentially crippling Northeast US (NEUS) blizzard of 8-9 February 2013. At the time of the discussion (11 AM EST to accommodate the cancellation of classes after 12 PM), the cyclone center was East of the DELMARVA peninsula and light snow was mixing with sleet at SBU. By the end of the discussion (11:30 AM) there was a complete changeover to rain. We started discussion by highlighting the factors that will lead to the cyclone rapidly intensifying. The key players behind the rapid intensification of the coastal cyclone are the upper-level trough situated over the Great Lakes and the southern stream energy. The phasing of the cyclone with the upper-level trough will align it with the jet cores for preferential dynamical intensification. We discussed that the blizzard conditions experienced across the Island would be sensitive to the timing of the changeover from rain to snow and the position of the band once it developed.
Timing of changeover: For forecasting total storm snowfall amounts, the timing of the expected changeover from rain to snow is critical. Once the precipitation gets going, we expect the diabatic cooling to cool the column completely. Exploring some model-derived forecast soundings showed that the NAM and GFS had a warm-nose bias just below 850 mb but the RAP showed a much cooler solution before 00 UTC 9 Feb. We favored the RAP for our forecast, mainly because that would mean more snow for us. Looking at the SREF plumes for precipitation type at Islip, the changeover is expected before 00 UTC on 9 Feb and forecasts about 10 inches of snow.
Location of band: The models were growing more consistent by the 12 UTC run on the location of the cyclone passing a bit farther to our east. The forecast location of the cyclone center is critical for the forecast of where the primary snowband will develop, typically NW of the cyclone center. The band develops where there is maximum frontogenesis, typically between the surface and 700 mb. The band also coincides with a region of conditional instability which means there can be more rapid ascent and more intense snowfall. We looked at the 12-km simulation for the MM5 which put the band right on the Nassau/Suffolk border.
Winds and Coastal Flooding: Low-level winds in each of the model were >35 kt sustained around 00 UTC 9 Feb. Higher momentum air at ~925 mb can be mixed down by precipitation drag and, given the model soundings showing that the surface is warmer than aloft, through convective motions as well. As far as coastal flooding, an astronomically high tide (New Moon is on the 10th) with a few feet of storm surge on top of that can cause flooding at some prone locations. In addition, violent wave action may also be an issue along the North and South Shores.
Sara concluded her forecast calling for Stony Brook, NY to receive 14" of snow and the area receiving the highest snowfalls will likely be to our north and east.
February 1: Michael Erickson led a discussion titled "A Brief Look at Yesterday's Storm and our Prospects for Snow this Weekend."
Previous Weather: Mike led the first discussion of the semester beginning with that since we last met in December 2012, we’ve experienced a changing pattern. The AO went negative following a stratospheric sudden warming (SSW) event. The ENSO signal is now more like La Nina with cold anomalies found in the Eastern Pacific. Drought has lessened slightly, but still persists across much of the country.
High Wind Event of 31 January 2013: A strong continental cyclone dragged a cold front across our area very early Thursday morning. This cold front was responsible for severe weather across much of the South during the preceding days. The main threat in our area was for high winds exceeding the NWS severe criterion of > 58 mph. The cold front was strong in the fact that there was a very tight horizontal temperature gradient. In response to this, there was a very strong low level jet (LLJ) with speeds of around 70 kt around 925 hPa according to the 00 UTC 31 Jan OKX sounding that only got stronger as the cold front approached from the west. This high momentum air can mix down through the drag of heavy precipitation (found in the narrow cold frontal rainband (NCFR)) or by gravity waves induced by isolated areas of convection that can propagate within a stable layer. The NWS storm reports indicate that the maximum gusts were ~65 kt and the heaviest rain fell north and west of the Island. Mike discussed the marine influence that likely saved us from more severe winds and heavier rain. The most important factor was the strong inversion extending from the surface to a bit above 850 hPa (SSTs are in the single digits in Celsius but 850 hPa temperatures were near 15C). This inversion was well-forecast by the models. The stability inhibited the free mixing of the higher winds to the surface. Some of the highest wind gusts reported were coincident with the passage of the NCFR, however analysis of the base reflectivity from 09 -11 UTC 31 Jan of the band shows it breaking apart as it reaches the Island, possibly due to decreased friction and its ability to accelerate. There was some discussion about whether some of the gusts reported were associated with a gravity wave, but insufficient temporal data of pressure, for instance leaves that undetermined. The main science issue discussed was how the front evolved as it approached the coast. There was a distinct lag between when the front passed, when the winds shifted, and when the temperature finally began to drop throughout the day on the 31 Jan. This development of a prefrontal trough has been identified before in previous cases but hasn’t been explored in any great detail to our knowledge.
Weekend Forecast: Mike concluded his discussion with the fact that Sunday may be potentially interesting (according to some SREF members) with a slight chance of snow but most other days are forecast to be cold and uneventful, with the current forecast pattern being unsupportive of storminess.