Tuesday's storm took a remarkably late shift to the south in guidance on Monday. The level of shifting models did on this storm the day before the arrival of the storm was more typical of shifting you would see three or four days out. So, the questions are what led to this late shift, were there hints of a southward shift before it happened and should we have picked up on it sooner?
To begin to answer these questions, we need to look back at our forecasts for this storm. In our first forecast post for this storm published on February 8, we noted that the level of interaction (or phasing) between the northern stream energy and southern stream energy would be crucial in determining impacts on New England. We wrote in that post:
The low passing near New England's south coast (the southern stream energy) will likely interact with a trough crossing the Great Lakes (the northern stream energy). The amount of interaction (or phasing) between these two pieces will be key in how the storm amplifies and how heavy precipitation (snow or rain) may be.
A graphic we created for our February 8 article:
The next day, on Friday, February 9, we updated our forecast for the storm. Going into that forecast, we noted that confidence in what would happen did not increase from the previous day's forecast. We wrote that the interaction between the northern and southern streams was trending toward being minimal. We wrote:
The trends on the Tuesday storm have been shifting south with a less amplified system and less phasing between the northern and southern stream energy pieces. This would lead to a less impactful storm for much of New England. As of now, the further south you are in New England, the more likely you are to get into steadier precipitation.
So, at this point, it was looking like less interaction between the two streams. Little interaction favors a more southerly track (suppressing snowfall to the south) , a lot of interactions favors a northerly track (which would introduce more mild air and mixing in southern New England) and a middle ground favors a track in between (which would favor a good sized snowstorm for much of New England). Models pushed the storm further south compared to the previous day's run.
Despite this trend, we noted that recent snow events in New England this season initially looked like they were going well south before trending back north, most notably, the January 7th storm. We try to anticipate what models will do based on how recent storms played out. In that same article, we wrote:
This season, models have shown storms passing well to the south of New England before trending back northward as the event gets closer and being a few days out still, shifts in guidance will continue to be likely. We wouldn't be surprised at all if the storm did trend further north in the coming days
The next day, February 10, the northward trend on some models became a reality. Since we were anticipating this, we were quick to jump onto the trend.
Euro model runs on February 9 (1st image) and February 10 (2nd image):
In the February 10 article we mentioned that the chances of seeing a rapidly deepening low passing near the benchmark were becoming more likely. This would help the storm overcome a lack of cold air at the onset. We also brought up the critical phasing between the northern and southern streams. We wrote:
In previous posts, we talked about the importance of interaction (or phasing) between the northern stream energy and the southern stream energy. At this time, it does not look like the two energy pieces will interact all that much. That is helping to keep the storm's track suppressed offshore. That will also help keep the storm colder with the highest snow impacts in southern New England.
On Sunday, we published our first "impacts, timing" article for this storm. At this time, a major winter storm was looking likely for New England despite signs pointing toward less interaction between the streams. Whenever we publish a snowstorm forecast, we like to include ways the storm could bust. In this forecast, we wrote:
The main bust potential with this storm will come from potential track issues...if the storm isn't able to strengthen as quickly as anticipated due to a lack of interaction with the northern stream, snowfall rates will not get as intense.
National Weather Service alerts as of Sunday evening (February 11):
In our updated "impacts, timing" article published on Monday morning saw trends clearly beginning a southward shift. We titled that article "Strong Storm Trends South." In that article, we mentioned the lowering snowfall amounts, particularly for northern New England. We wrote:
Latest guidance has taken a rather notable southward shift this morning...As stated before, latest guidance has shifted the storm a bit further to the south. This is a deviation from how things have played out so far this winter, which was to continually trend the storm northward leading up to the event. This has led to snowfall amounts being pushed southward...amounts will likely drop off quickly heading further north amid dry air. We need to watch today's trends closely as further southward shifts will lower amounts even more. There is growing bust potential for all in northern New England.
HRRR showing expected weather in the early afternoon for Tuesday. The 1st image was initialized on February 12th at 7am, the 2nd image was initialized on February 13th at 7pm:
Monday's model runs went all-in on a southerly track. In an admittedly hastily written article, we went into the major changes in the forecast and why the changes happened. Instead of quoting basically the entire article here, we'll just use this one sentence we put into the article. We made this sentence bold to emphasize the changes:
Bust potential has become high with this storm.
National Weather Service alerts as of Monday evening (February 12):
So, that details the forecast saga with this storm. From the beginning, we brought up the importance of the interaction between the northern and southern streams. To put it very simply, less phasing equals a weaker storm suppressed south and more phasing equals a stronger storm pulled north. Despite trends toward less phasing, we continued to lean toward a more northerly track due to recent precedence, which was a mistake.
Another big factor with this storm was the lack of cold air ahead of the system. We mentioned that a stronger storm would be able to generate its own cold air and force snow. This happened last March when a nor'easter arrived and originally started as rain before switching to snow for most. The lack of phasing led to a weaker storm that could not overcome the warmer air in places (i.e. Boston).
Lastly, a large range of tracks remained a possibility based on major models. We will say that the CMC (Canadian model) remained locked on a more southerly track starting on Friday while other major models, GFS, ICON and particularly the Euro began to pull the storm north while trying their best to phase.
To answer the questions in the first paragraph directly: the reasons behind the late shift stem from the lack of phasing between the streams. There were subtle hints of the southward shift. We could have picked up on them sooner than Monday. Forecasting is all about interpreting. Based on what models were showing combined with recent precedence, we stuck with higher amounts further north until the last second. We put the emphasis on the wrong areas. We'll learn and we'll come back.