Turning into quite the close call for I-95 corridor (NJTPK in NJ), with enough precip for 12-24", especially north of Trenton, but with about half of the main models now showing some sleet mixing in for a few to several hours, which would keep accumulations down somewhat - although the impact would be the same, as it's the mass of frozen precip that matters for snow removal/traction etc. For areas towards the coast, especially for coastal Monmouth and even interior Ocean, there's now a decent possibility of this storm being 1/3 snow, 1/3 sleet and 1/3 rain with maybe 4-8" of slop that the rain falls into (and won't melt) - and it's possible that areas south of Toms River see mostly rain after a few inches of snow/sleet. The other thing snow lovers don't want to see is any further shift in the actual track (as said below, it's almost beyond the time where the global models are useful and we watch surface, upper level and radar trends of the actual storm forming and evolving).
The problem is that nobody on the planet can predict the exact placement of a gradient that will likely go from 24" snow to 12" snow/4" sleet to 8" snow/3" sleet/1"rain to 4" snow/2" sleet/1.5" rain to 2+" rain over maybe 50 miles. It's simply impossible. What you hope is that that gradient is offshore so nobody has to make that call and you only worry about where the deformation bands set up to give more or less snow. Doesn't look like that will be the case, though. Will be interesting to see what the NWS and media forecasters say now - don't envy them on this one. Here's something one of my favorite mets (Wxoutlooksblog) said just a few minutes ago:
"For the NYC Metro Region with all the models out, some showing western Long Island and NYC itself to get the most snow, and others including the European model showing an extremely close call with sleet/rain mixing over western Long Island and possibly the city, probably the only thing left to do is to watch this storm develop and move hour by hour and see how far north and west the rain/snow/sleet mix and changeover lines get. It is such a close call on the European model, for example, that some of the computer generated snowfall forecast maps off of this very same model differ by 6" or so for this region with NYC/northwestern Long Island on the edge of a swatch of 12-18" snowfall to the southeast and 18-24" snowfall to the northwest. So, we will have to watch radar and surface observations to the south to see if and when certain portions of the NYC Metro Region mix with or change to rain and/or sleet tomorrow morning. This is called nowcasting and I hereby declare this historic March blizzard is now officially in nowcasting mode."
Edit: interestingly, the NWS didn't officially decrease the snowfall amount for Middlesex County (or Somerset and NW of Middlesex), as the updated blizzard warnings show, which just came out, so after all the model input.was considered - still at 18-24", but notes amounts could decrease due to sleet. We'll have to just watch and see now.
http://forecast.weather.gov/wwamap/wwatxtget.php?cwa=phi&wwa=blizzard warning