Abstract
An unusual early-season snowstorm dumped more than 50 cm of snow over portions of interior eastern New York and western New England on 4 October 1987. This was associated with poorly forecasted cyclogenesis. In the wake of atmospheric cooling associated with upward motion and cold advection, additional cooling associated with melting snow in a heavy precipitation region aided in the creation of a low-level dome of cold air to the west of the surface cyclone track. This allowed heavy snow to fall at low elevations. The absence of low-level warm-air advection over the snow region was crucial to the maintenance of the cold dome as relatively warm maritime air to the east was prevented from reaching the area, eroding the cold dome and changing the snow to rain.
The storm was also noteworthy because its development was in accordance with well-known qualitative quasi-geostrophic principles, but the NMC operational prediction models, while simulating the 500-mb circulation well, failed to predict accurately the storm development off the mid-Atlantic coast and its associated precipitation. An investigation of the Regional Analysis and Forecast System (RAFS) initialized fields revealed a critical underestimate of the vertically integrated moisture and lower-troposphere vorticity and divergence in the coastal baroclinic zone of the incipient storm environment at 1200 UTC 3 October 1987. Despite plentiful forcing aloft, the RAFS was never able to simulate the surface development successfully with the improperly analyzed low-level wind field and vertically integrated moisture field.
