Interesting developments with a couple recent models shifting landfall back north. What I am seeing is a gyre off the mid-atlantic coast seen in CMISS wind analysis. This gyre affects shallow systems mostly, with less affect on deep systems. What looks to be occurring is oscillations in how deep the models are forecasting the system to be when it starts interacting with the gyre (the point at which the eastward turn occurs, and NE movement starts). The result is directly related to how far out into the Atlantic Sandy gets before it turns back into the trough. In the case with current models (such as the 00Z GFS), a weaker, shallower Sandy is pushed further into the Atlantic resulting in a landfall near New York city. In comparison, the ECMWF and the GFDL are forecasting a deeper system, causing less affect from the gyre and resulting in a more southerly landfall, possibly even southern MD or into VA.

I think over the next 12-24 hours we will have a much better handle on the landfall location - the depth of the storm and it's interaction with this gyre will be very evident once it starts to make its turn to the NE.