The impact of convective parameterization on NAM forecasts for the February 25 2005 Winter Storm

The winter storm event of 27 February 2005 provided major challenges for forecasters in the mid-Atlantic and northeast, as the NAM and GFS models showed significant disagreement in the intensity and track of the primary area of surface low pressure. The discrepancies led to major differences in predictions of amounts and type of precipitation.

The roles of initial conditions and convective parameterization in the forecast differences will be examined. Experiments have been performed with the NAM using different analyses, and reruns of the model will be made with multiple convective schemes. Major differences arise when the Kain-Fritsch scheme is substituted for the Betts-Miller-Janjic (BMJ), and other schemes and various modifications to both the deep and shallow branches of the BMJ scheme will be tested.