Abstract
A diagnostic analysis and an appraisal of the precipitation calculation by the GFDL (Geophysical Fluid Dynamics Laboratory) 1967 version prediction model are presented, using two-week forecasts of 12 January and 12 July cases. The geographical distribution of predicted rainfall, moisture, and snow over the Northern Hemisphere and the contiguous United States was investigated in comparison with climatological maps published by other authors. The agreement of precipitation and dew-point temperature is marginal. The major causes for the deficiencies are 1) a specification of excessive soil moisture over land, 2) probably an improper treatment of moisture diffusion associated with topography, and 3) an inadequate rain generation process in the model. However, the predicted snow distribution over the United States was reasonable.
