bluewave

The lack of solid Atlantic blocking has been a big issue so far for us this winter. While we were able to score the 6-10 snow with the early January PNA rise and tail end of the brief late December -AO, the Pacific flow has been too overpowering since then. The fast northern stream combined with +AO and been suppressing the southern stream. A solid Greenland block would have forced the stronger northern stream to buckle underneath getting us closer to the most active storm track. It’s always more of a challenge to rely exclusively on a North Pacific block without help from the Greenland -AO block. Hopefully, we can put together a nice snow event near the end of January and start of February just on the lingering +PNA and -EPO.

Simon Lee has a great tweet illustrating the differences so far between this year and last.

NYC came into February 2006 with 11.7” but had a SSW event which produced the KU. We had Dec and Feb -AO patterns that year with record warmth in January. 2013 was another SSW winter which had the KU in February. Last February was great due to the SSW and record -AO. Since we won’t have a SSW and a great -AO going forward, we need a nice snow event next weekend before we lose the +PNA block.

The next 10 days or so will probably determine whether we make it to normal snowfall on the season. Winter is a bit like a football game. We can’t keep missing opportunities and hope to make it all up near the end of the game. We would need a nice snowstorm next week plus any snow we can pick up in February and March in order to reach near to above normal snowfall with the new higher 30 year averages.

30 day patterns have been the norm for us since the start of November. There have been regime shifts near the beginning of each new month . Our last change occurred in early January and was accompanied by the best snowstorm of the season. Let’s hope we can put together a nice snowstorm near the end of January as we move to the next winter pattern in February.

From all the historical accounts, the winter of 1779-1780 was the coldest winter on record during that era. https://www.nps.gov/morr/learn/education/classrooms/upload/HARD-WINTER-P.pdf ALL REPORTS INDICATE THAT THE WINTER OF 1779-80 IS THE WORST WINTER EVER KNOWN IN THE NORTHEAST UNITED STATES David M. Ludlum in his book Early American Winters 1604-1820, in the chapter on “The Hard Winter of 1779- 80” writes that it was“the most hard difficult winter....that ever was known by any person living. There has been only one winter in recorded American history during which the waters surrounding New York City have frozen over and remained closed to all navigation for weeks at a time...the Hard Winter of 1780.” According to Ludlum, January 1780 rated as the most persistently cold calendar month in the history of the eastern United States and “The severity of the 1780 season reached all parts of the colonies Reports from Maine southward along the seaboard to Georgia, and from Detroit down through the interior waterways to New Orleans, all chronicled tales of deep snow, and widespread suffering.” Ludlum believed that his research definitely showed that the season of 1779-80 in the eastern United States well deserved the name given to it by the people that experienced its effects....The Hard Winter EYEWITNESS ACCOUNTS SUPPORT THIS BELIEF William Smith (a loyali<t living in New York City) record< in his diary that on January 16, 1780, people were walking across the frozen Hudson River from New York to Paulus Hook (today Jersey City) New Jersey. A Hessian <oldier, Johann Dohla recorded in hi< diary on January 30, “The North (Hudson) and East rivers were frozen solid. The ice was checked and found to be eighteen feet thick. All ships were frozen in, and it was possible to cross over the North (Hudson) River on foot, riding or driving, without fear.” Later, (on February 22) Dohla wrote “Today the North River ice began to break, after having been frozen for nearly seven weeks.” A German officer, Major Baurmei<ter wrote, “The severe winter covered the North River with ice early in January; even where the current of the rivers <wifte<t, the ice was eleven feet thick, in spite of the fact that it is 1,800 yard< wide between Fort George (today Battery Park, New York City) and Powles Hook.” February 1780 seems to have been even colder: On February 10, William Smith mentionsin his diary that a few days earlier a “24 Pounder” (that i<, a cannon that fired a solid ball weighing 24 pounds—the entire cannon weighed three tons) was rolled across the Hudson River to PaulusHook (today part of Jersey City, New Jersey.) Smith goes on to say that even a heavy load as this it made no impression on the ice. On the night of February 10th, Smith heard that four to five hundred British cavalry rode their horses across the river IIIIIIIIIII MANY ACCOUNTS MENTION UNPRECEDENTED CONDITIONS A teacher in Yale College (New Haven, Connecticut) recorded approximately twenty days with snowfall, and a total of 95 inches of snow that winter. People walked across the Sound from Stanford, Connecticut to Long Island . Others walked from Rhode Island mainland to Block Island. Chesapeake Bay and the York River in Virginia froze over for the first time since Europeans settled there. Many people mentioned in letters that they could not remember a winter as bad.

The temperatures this month are certainly feeling more winter-like for a change. NYC already has 5 days with a low under 20 this January with more to come. This is the most for any winter month in 3 years. These colder winter months have been few and far between since 15-16. Colder than average months bolded with 5 out of 20 Monthly Number of Days Min Temperature < 20 for NY CITY CENTRAL PARK, NY Click column heading to sort ascending, click again to sort descending. Year Dec Jan Feb Season 2021-2022 0 5 M 4 2020-2021 0 2 1 3 2019-2020 1 0 2 3 2018-2019 0 5 2 7 2017-2018 5 13 2 20 2016-2017 2 2 2 6 2015-2016 0 4 5 9

Yeah, March 2001 was pretty bad. But the writing was on the wall for January 2008. Each run from 3 days out was cutting back on the snowfall. The morning of the storm several models had much less snowfall than they were showing from the days before.

Since I have a frame of reference back to the 70s, I consider a model bust a complete miss of the P-Type or storm track from the day before for a high impact event. January 2000 fits that description with the storm coming hundreds of miles west from the day before. There was no storm forecast for many sections that got warning level to historic snows. I would consider January 2015 a large model error with a 60 mile + shift in the track to the east from the day before. We still got warning level snows around NYC and Western LI, but the jackpot shifted out to our east. Had the storm shifted hundreds of miles east and dropped no snow in NYC, then I would call that a bust. This happened several times in the 70s and 80s. We haven’t seen a complete P-type bust from the day before for a major event in a while. The most memorable one for me was 1-20-78. The forecast from the day before was for rain heavy at times. The next morning we had 12-18” on the ground. But I guess the term bust can be highly subjective.

It will be interesting to see if we can end the pattern with a nice snowstorm like we had at the beginning of it on January 7th. Sometimes we get bookend snows at the beginning and end of new patterns. Since the start of November, it has been a steady 30 day pattern followed by a shift at the beginning of each new month.

The RGEM has really been improving in recent years with continued upgrades. Unfortunately, there have been no NAM upgrades since 2017 when they decided to move toward developing a FV3 based CAM. But we have seen several problems with the the FV3 GFS changes. So we are stuck with an older version of the NAM until a replacement is ready. If the FV3 issues with the GFS are any indication, it may take a while for a suitable NAM replacement. Meanwhile, models like the RGEM will continue to pull ahead. The HRRR is often hit or miss in these situations with too much of a cold bias. https://vlab.noaa.gov/web/environmental-modeling-center/fv3-convective-allowing-forecast-system NAM development was frozen after the March 2017 upgrade, and RAP/HRRR development with the WRF-ARW model will cease after the RAPv5/HRRRv4 upgrade in 2020. However, operational execution of these modeling and associated DA systems will continue until comparable FV3-based systems are able to give similar performance.

34° and light snow here in SW Suffolk.

The models have become much more accurate than they were in the past. But in those days, we didn’t have weather forums and twitter posting every long range model snowstorm forecast. Models never have done particularly well with East Coast storm tracks past 3-5 days. Every once in a while a model like the Euro would score a long range win. This happened in January 1996 and February 2013. But we often forget how many long range misses the models had between the few long range successes. In the 70s and 80s, we often didn’t even know the storm track or P types a day before the storm. The forecast a day before the Jan 78 snowstorm with 12-18” was for rain heavy at times. That was my first school snow day in the 1970s that I can remember. Numerous snowstorm forecasts like in 1980 from a day before went OTS during nowcast time. The January 1987 forecast was for snow changing to rain hours before the storm started. Instead we got a heavy wet 10” of snow in about 5-6 hours. Our last actual model bust from a day before was January 2000. We went from no snow to record snows in the Mid-Atlantic. Nothing like that has happened in over 20 years. So that represents a great improvement. Perhaps the eventual Euro and EPS upgrade to convection permitting 3-5 km resolution will extend the forecast range for East Coast snowstorms. But we may not know until they actually try it.

We got the colder Pacific Ridge pattern this month. But record Pacific Jet from December never fully relaxed. So it served to suppress the STJ too far to our south. Hopefully, we can get a larger snow event before the retrogression of the ridge back to the Aleutians and milder Western Trough returns in early February.

The historic +13.3 departure in December 2015 marked the shift to much warmer winters. This January is only the 5th winter month out of the last 20 with below normal temperatures. The coldest departure month was December 2017 at only -2.5. NYC Jan 22….-1.4 Dec 21….+4.7 Feb 21….-1.1 Jan 21….+2.2 Dec 20…+1.7 Feb 20…+4.8 Jan 20….+6.5 Dec 19….+0.8 Feb 19….+0.9 Jan 19….-0.1 Dec 18…+2.6 Feb 18…+6.7 Jan 18….-0.9 Dec 17…..-2.5 Feb 17…..+6.3 Jan 17….+5.4 Dec 16….+0.8 Feb 16….+2.4 Jan 16….+1.9 Dec 15….+13.3

The EPS is great for general 500 mb teleconnections days 5 to 15. But the OP and EPS doesn’t have great skill with coastal storm tracks beyond 3-5 days. It does seem to have a slightly better long range track record with cutters like we had on Monday. Maybe the future 3-5 KM convection allowing resolution upgrade on the OP and EPS will extend the range of coastal storm track prediction.

The rule this season has been needing more than one model showing the event. The GFS was the overamped model on January 3rd initially showing the heavy totals near ACY in our area. But the mesos showed the confluence over our area and us missing the storm. This event it was the Euros turn to show several overamped runs before it backed off and joined the other models. It would be great to have the old days back like February 2013 and January 1996 when amped up Euro run was all we needed for a KU without any other models on board. But these days we need multiple globals on board beyond 72 hrs and then have the mesos hone in on the finer details like banding locations.

My guess is there have been several factors contributing to the record snowfall since 02-03. Much stronger blocking patterns both in the North Atlantic and North Pacific. Warming winters allow the atmosphere to hold more water vapor. Record SSTs also provide more moisture for snowstorms. The last 20 years will stand against any historic 20 year period for snowfall. But the one caveat is that older eras before the 1980s measured snowfall differently. They measured after snowstorms which allowed the snows to compact. So those older totals would be higher if they used the modern methods. https://news.ucar.edu/14009/snowfall-measurement-flaky-history

We can get a KU without NYC reaching 10”. But the lowest rank of cat 1 usually requires widespread accumulations over 10” somewhere in the NE corridor. Most KUs also produce numerous 10” + totals in our area. Sometimes the 10”+ zones are east of NYC like Nemo in Feb 2013. With Stella in March 2017, the jackpot was NW of NYC. So the normal or better snowfall winters in NYC need storms that can reach these levels somewhere in our area.

The bar was lower for normal snowfall in the 1980s since the 1951-1980 30 year normal was around 24.7” in NYC. So NYC had around 5 seasons in the 1980s with close to normal snowfall. We had 1KU in February 83…1 KU in April 82…2 KUs a few days apart in January 87. The 2nd event did much better in Central NJ to Eastern Long Island. There was another KU in February 87 that favored Central NJ. Monthly Total Snowfall for NY CITY CENTRAL PARK, NY Click column heading to sort ascending, click again to sort descending. Year Nov Dec Jan Feb Mar Apr Season 1980-1981 T 2.8 8.0 T 8.6 0.0 19.4 1981-1982 0.0 2.1 11.8 0.4 0.7 9.6 24.6 1982-1983 0.0 3.0 1.9 21.5 T 0.8 27.2 1983-1984 T 1.6 11.7 0.2 11.9 0.0 25.4 1984-1985 T 5.5 8.4 10.0 0.2 T 24.1 1985-1986 T 0.9 2.2 9.9 T T 13.0 1986-1987 T 0.6 13.6 7.0 1.9 0.0 23.1 1987-1988 1.1 2.6 13.9 1.5 T 0.0 19.1 1988-1989 0.0 0.3 5.0 0.3 2.5 0.0 8.1 1989-1990 4.7 1.4 1.8 1.8 3.1 0.6 13.4

I sure hope so. NYC is going to need at least 1 KU event before the season is over to reach normal snowfall. This is the first season with the new higher 29.8” normal seasonal snowfall in NYC. Every season since 09-10 needed at least 1 KU event in the region to make it to at least normal snowfall. Time Series Summary for NY CITY CENTRAL PARK, NY Click column heading to sort ascending, click again to sort descending. Ending Date Total Snowfall Oct 1 to Apr 30 KU Storms 2022-04-30 6.8 0 2021-04-30 38.6 2 2020-04-30 4.8 0 2019-04-30 20.5 0 2018-04-30 40.9 4 2017-04-30 30.2 2 2016-04-30 32.8 1 2015-04-30 50.3 2 2014-04-30 57.4 3 2013-04-30 26.1 2 2012-04-30 7.4 0 2011-04-30 61.9 3 2010-04-30 51.4 3

While it has gotten much colder than December, we still have the overpowering Pacific Jet. So we have had several events get suppressed. The cutter was our most intense storm but the lack of -AO and -NAO hurt us. We generally need the northern branch to back off when there are STJ disturbances. So places to our south like ACY already have reached their normal snowfall for the entire season. At least we are close to average for January so far which is a big win coming off December.

Models seem to lock in cutter tracks like yesterday earlier than coastal storm tracks. Especially like the OP runs did over the ensembles such as the GEFS which were way too suppressed and cold. My guess is that the complexity created by the thermal contrast between the Eastern US and Gulf Stream is just too hard for the longer range runs to resolve.So it’s no surprise that the mesos like the NAM often take the lead with coastals like we saw in January 2016. But they only go out 84 hrs so we need to be patient until the storm track gets within their better range under 36-48 hrs.

The GFS hasn’t been doing so great recently with storm tracks either like we saw with it being overamped on January 3rd.

I believe Nemo in February 2013 was the last high profile snowstorm that only the Euro showed to work out. But that was before the series of upgrades and the January 2015 and 2016 misses. So now we need more than just one model showing any given solution to have enough confidence in an event from more than 72 hrs out.