Model simulations and satellite microwave observations of moist processes in extratropical oceanic cyclones

• Main Author: Miller, Douglas Kirby
• Other Authors: Petty, Grant W.
• Format: Text
• Language: English
• Published: Purdue University 1996
• Subjects: Atmosphere|Remote sensing; Pacific; North Atlantic
• Online Access: https://docs.lib.purdue.edu/dissertations/AAI9721860

Coincident satellite passive microwave (SSM/I) observations and 48 h numerical simulations using a hydrostatic limited-area mesoscale model of 23 intensifying extratropical cyclones located over the North Atlantic or North Pacific Oceans during a single cold season have been examined in an attempt to discern systematic differences in the moist processes of storms exhibiting rapid and ordinary intensification rates. Analysis of the observations and simulations focused on the 24 h period of most rapid intensification for each case as determined from European Centre for Medium-Range Weather Forecasts (ECMWF) 12 h mean sea level pressure analyses. Comparisons between the two sets of data highlight inadequacies in model moisture physics and suggest possibilities for improvement. Multiple tests of a single cyclone showed that the final forecast cyclone intensity and position was highly sensitive to the chosen convective parameterization scheme, which determines sub-grid scale warming and drying processes and their effects on storm evolution. SSM/I observations of area-averaged precipitation and an index that responds to cold-cloud (convective) precipitation to the northeast of surface cyclone centers correlated well ($\sim$0.80) with the latitude-normalized deepening rate (NDR) of the study sample. This large correlation was replicated by the numerical model, although the area-averaged precipitation region yielding the maximum coefficient differed significantly from that determined using microwave imagery. A similar correlation emerged between model-derived area- and vertically-averaged vertical motion fields and NDR. The similarity of these correlations for nearly coincident averaging regions relative to the storm center implicates unrealistic patterns in vertical motion fields as the reason for the failure of the model to accurately capture the observed optimal area-averaging region. This region was located near the storm triple point and occluded (bent-back) front, both potentially strongly convective environments. ...