Lessons learned from model forecasts of the 30 Dec. 2000 East Coast snowstorm

• Accurate prediction of antecedent dynamics is a necessary but not sufficient requirement for accurate forecasts.
  
  

• The operational Eta and EtaX had virtually identical dynamical forcing for the predicted cyclogenesis.
  
  
• Accurate emulation of physics is necessary but again, not sufficient for accurate forecasts.
  
  
• It can be argued that the operational Eta and EtaX physics responded accurately to the lower boundary forcing (i.e., SSTs) provided. The differences in the forecasts resulted from differences in the boundary conditions.
  
  
• Accurate lower boundary conditions are necessary for accurate forecasts. The forecaster should assess the accuracy of both lower boundary values and gradients.
  
  
• In this case, the SSTs were incorrect in the operational Eta, and more accurately depicted in the EtaX.
• The maximum temperature contrast in the operational Eta was near the coast, rather than over the open water at the eastern edge of the cold coastal shelf waters along the Mid-Atlantic coast as in the EtaX and as observed.
• The orientation of the gradient was north-south rather than east-west as in the operational Eta, with a pool of warm water (~22°C) about 100 km east of Norfolk, VA.
  
  
• The forecaster should understand the advantages and limitations of the lower boundary conditions used in NWP models.

As model physics becomes increasingly realistic and model resolution improves, models will more closely mimic the variability and sensitivity of nature. This means the model forecast, especially its detailed structure, will become even more sensitive to the specification of lower boundary specifications such as SST, soil moisture, and vegetation type and fraction, some of which may be poorly observed or poorly specified. Forecasters must be alert to situations where these may be most likely to influence the forecast and be aware of model deficiencies in representing the actual conditions.