Following the previous blog post, significant changes have occurred in the model guidance forecast for Sunday, with most of the model guidance at some point over the last 2 days having depicted a much weaker low pressure producing little to no snow. With barely 24-30 hours left until the onset of precipitation, discrepancy in the model guidance has actually increased, with some models reverting back to a moderate snowstorm scenario. While at least some snow is expected to affect the area, especially east of NYC, the major snowstorm previously depicted on the model guidance is no longer expected.
Brief Review of Model Trends
The last few days featured frequent swings in the model guidance, ranging anywhere from over 15 inches of snow across the tri-state area to practically no snow anywhere in the Northeast US. These trends, and their potential causes and associated forecasting challenges, will be reviewed in more detail in a post-storm analysis to be posted this week. The main highlights of these trends, however, is that the model guidance struggled to handle the evolution of the upper-level flow and interaction between multiple troughs correctly, resulting in multiple shifts in the model guidance that ultimately brought back a somewhat snowier scenario to the area, but eliminated the possibility of a major snowstorm as the ECMWF depicted several days ago.
Late 3/17 into 3/18 GFS forecast 500 hPa vorticity and heights valid at 0000 UTC Sunday, 20 March, depicting the first trend towards deamplification of the northern US trough. Image from Tropical Tidbits.
Up through Thursday, 17 March, the ECMWF model, and to a lesser extent the UKMET and CMC models, were consistent in depicting a rapidly deepening low pressure close enough to the coast to produce a major snowstorm across the region, with snow accumulations upwards of a foot of snow. The GFS was farther east, but similarly depicted a rapidly deepening low pressure off the coast on Sunday night. With the 0000 UTC Friday, 18 March runs, the majority of the models trended towards a much lower impact event in the region, both trending east with the track and weaker with the low pressure. Several factors contributing to this trend appear to have been a shift towards a farther southwest upper level low over Canada and a flatter trough over the northern US, resulting in less amplification downstream of the trough and a more confluent flow over the northeast US suppressing the initial low pressure on Saturday into Sunday farther south, and less interaction between the northern US trough and an approaching vorticity maximum from central Canada, which delayed the onset of rapid intensification of the coastal low on Sunday night by several hours.
3/19 GFS forecast 500 hPa vorticity and heights valid at 1200 UTC Monday, 21 March, depicting the second trend towards a slower shortwave trough over the region.
The result of the first shift on the models was a substantially weaker and farther east low pressure, with weaker forcing resulting in lighter precipitation and a lack of strong deformation banding northwest of the low pressure. Over the last day, however, especially beginning with yesterday’s 1200 UTC runs (19 March), a gradual trend towards a farther west track began again, although this time in association with the northern US trough trending slower, allowing for more downstream amplification and more interaction with the approaching Canadian vorticity maximum. This trend slowed down the timing of the coastal low pressure, while also resulting in a farther west track and an earlier onset of rapid deepening. The positioning of these two upper-level features has shifted significantly compared to earlier ECMWF runs, to the extent that a major region-wide snowstorm cannot occur. However, this trend was sufficient to result in the potential for moderate snow accumulations in eastern parts of the area into coastal New England.
As of this morning, the model guidance is primarily split into two scenarios, both of which are substantially closer to each other than yesterday’s runs were: a sharp precipitation gradient east of NYC, in which northeast NJ and NYC would record less than 0.20″ of liquid-equivalent precipitation while Long Island and the eastern half of Connecticut mostly receive over 0.50″ QPF; or a farther west gradient near or just west of NYC, in which the city itself may receive several inches of snow. Considering that some of the models appear to have exaggerated the trend towards a slower trough and are slightly backing off, the former scenario was used for this forecast.
17z HRRR hour 15, valid at 0800 UTC Monday, 21 March, depicting 2m temperatures lingering just above freezing for most locations except for southern CT.
With the above forecast scenario considered, several factors influencing snowfall accumulations need to be taken into account. Snow for this event is unlikely to accumulate at 10:1 snow to liquid ratio, with numerous factors contributing to lower ratios and accordingly less accumulation than solely looking at the QPF would support. While the air mass aloft is sufficiently cold to produce snow with any precipitation that falls, with 850 hPa temperatures near -6C and 925 hPa temperatures near -4C, the delayed onset of precipitation as previously mentioned has allowed 2-meter temperatures to rise into the upper 30s to low 40s this afternoon, with a lingering warm boundary layer as precipitation begins to fall towards 8-11 PM tonight. Temperatures in the boundary layer will cool down later into the event as precipitation saturates the boundary layer and cools the air temperature closer to the wetbulb temperature, but this process will be slower to occur in locations with light snowfall rates, especially west of NYC. Additionally, given that temperatures in the last several days have been in the 50s and 60s, ground temperatures remain relatively warm which is a limiting factor for snow accumulations; heavy snow rates will likely overcome this issue, especially in Long Island and eastern Connecticut, but may lower accumulations farther west where lighter snowfall rates are expected, as snow would initially melt on contact with the ground.
Another aspect that will influence snowfall amounts is the potential for snow banding, with a narrow band of strong low-mid level frontogenesis expected to set up off the coast as the low pressure begins to rapidly deepen. The above scenario would support an approximate placement of the heaviest snow banding from eastern Long Island into southeast Massachusetts, which if comes to fruition would result in the potential for higher than forecast snow totals. The frontogenetic banding isn’t expected to be as strong as with other recent storms this winter, such as the January blizzard during which a heavy snow band persisted over NYC with strong subsidence and a very sharp snowfall gradient just to its north, although weaker forcing for ascent inland of the banding may nonetheless result in lighter snowfall rates and less snow accumulation.