By John Cifelli
NJTV News Meteorologist
Our surface low is wrapping up and sliding away, gradually moving east-northeast parallel to the coast of Long Island. It’s currently at 982 mb, and the tight gradient between the surface low and the 1034 mb high northwest of the Great Lakes is keeping the wind whipping across the state. Gusts in the last hour: 32 mph in Harrison, 33 mph in Pennington, 44 mph in Seaside Park, 39 mph at Cape May Courthouse. We’re drying out now, we just have some low level moisture responsible for freezing fog and snow showers. We could still pick up an inch or so as the back end of the storm moves through. We will chill down tonight, as temperatures drop into the low 20s and upper teens, before another day below freezing statewide tomorrow. There is a chance that anyone in the state, especially northern and central New Jersey, could see some flakes tomorrow as the upper level low passes across the state. Snow showers could be squally in nature, with a quick burst of snow and wind to put a coating down.
First, a summary of realized impacts:
1. One to two feet of snow and sleet fell from the 78 corridor north, six to 12 inches from 195 to Route 78 and three to six inches south of 195, some of which was washed away by rainfall afterwards.
2. Freezing rain was a bigger impact than expected. Where precipitation quickly changed to rain due to rapid intrusion of warm air aloft, freezing rain became a big problem in the southwestern part of the state, particularly in Salem, Gloucester and Camden counties. About 50,000 residents lost power due to icing, along with another 20,000 or so in Monmouth and Ocean who lost power due to winds. Currently, as of 3 p.m., about 20,000 still do not have power.
3. Coastal flooding met or exceeded expectations. Atlantic City reached the “major” threshold for flooding. Sandy Hook and Cape May also ran two to three feet above normal in their Tuesday morning high tides, and the Manasquan Inlet breached the breakwater.
4. Winds gusted to 60 mph along the coast and were strong enough for blizzard conditions to be realized in interior parts of New Jersey as well.
Of course, the part of the story that is missing, is the widespread higher snowfall totals. As the system slid up the coast, it became evident that the track would be on the western edge of possible paths. It also came in stronger than modeled by a few millibars. The model snapshots give a visual representation of this:
The maps show the surface low placement (red L) and the temperature aloft at 850 mb, a key layer for understanding precipitation type. The left was a projection for 2 p.m. this afternoon, created by one of the computer models late Sunday night. The second image shows the same elements at the same time, but it was generated this morning. In the 30 hours between these runs, the low placement was shifted 50 miles northwest towards the coast, and the temperature at this level went from about -8 degrees Celsius to -2 degrees Celsius along the Route 1 corridor. Temperatures above freezing at any level of the atmosphere will contribute to melting of the snowflake or water droplet as it falls through the column, and you wind up with sleet, freezing rain or rain as a result. As the forecast was honed on Monday, all guidance had temperatures below freezing aloft, but barely. Perhaps the biggest mistake was not representing the decreasing confidence in big snow totals for central New Jersey, and modest snow totals for southern New Jersey as the event approached. With these subtle shifts so late in the game, being able to definitively say that the forecast was going to bust for a large portion of New Jersey really wasn’t possible, especially as most guidance continued to indicate the big snow totals would be realized in the borderline air mass.
I’ve been asked why can a storm tick west like this, and what the mechanism is. The truth is that there are many possible reasons, and they can combine to compound an error in model guidance, or they can offset one another and make a correct model solution right by chance. The east coast of the United States is one of the parts of the world where mid-latitude cyclones are born. Cyclogenesis occurs near and along the Mid-Atlantic coast because of the temperature gradient that exists between the colder land, and the warmer water. Have you ever noticed that we don’t get these kinds of storms in the warmer six months of the year? The gradient isn’t there to the same extent, and it is never there in the middle of the ocean, and usually isn’t in the Ohio Valley. Our region is unique in this respect. It also makes it very difficult to forecast, and for computer models to resolve the countless subtle nuances that occur as storms develop. A forecaster in Ohio has a much easier (and in my opinion, less exciting and interesting) time forecasting for their area, as whatever weather is coming upstream is usually on the table as it approaches. For New Jersey, very often the dish is still being prepared just as it is being served.
In this particular case, there was a lot of convection — thunderstorms — firing off the coast of South Carolina Monday afternoon. The energy and heat released by the storms may have contributed to a rise in heights to the north and east of the storm, which would nudge the surface low west. Computer models — the key word is “model.” It is not an explicit representation of the atmosphere, and thunderstorms represent a dynamic, extreme variation from the mean state of the atmosphere. Any computer guidance that did not account for this convection when ingesting the initialization data, would give a flawed projection for all points in time forward. Resolution on modeling is improving, and so are verification scores for these programs. Regardless of whether or not this was the cause of the low, and the mixing line, being shifted northwest at the last minute, it’s an example of where forecasting still struggles. The Western Atlantic Ridge, a large scale pattern that steers east coast storms, was under modeled by guidance, being projected incrementally stronger as today’s weather event approached.
The bottom line is that there are still tools in the meteorological toolkit that need to be honed. Until then, experience, pattern recognition and climatology need to be exercised and used when forecasting for these kinds of storms. Even then, there will be examples where the greater forecasting community “busts.”