The 1 AM run of the NAM reduced the QPF considerably and we are back to the original forecast from Friday of 1-2 inches with sharp cut off north of Philadelphia.

Light snow ends shortly after noon today for Philadelphia but continues in southern Delaware and Chester counties and southern NJ.

So… this was a boundary condition forecast which was poorly handled by the NAM model. The NAM has done well in these situations in the past, and it was prudent to report its forecast as a high probability.

The NAM forecast was predicated on a secondary low intensification at the coastline that will take place, but slightly further south.

Looking back, the GFS did better with the timing (starting earlier Saturday evening) and with the lower QPF.  I can think of many past storms where the NAM did better than the GFS.

The NBM did well with the QPF, but poorly with the onset of the snow. It had the snow starting 10-midnight which is what I used.   It also had forecast the secondary low intensification for Sunday night, which appears to be the same error made by the NAM.