Contrasting the impact of dynamic data assimilation on the numerical simulations of cyclogenesis during GALEIOP 10 and IOP 1

Abstract

The nudging assimilation scheme described in the companion paper by Brill et al. is applied to study oceanic cyclogenesis during GALEIOP 10 on 27–28 February 1986. A 36-h control simulation statically initialized from 0000 UTC 27 February 1986 data moves the cyclone too far north and east in the 12-h period of most rapid deepening limiting the usefulness of the simulation for diagnostic study. The use of nudging to dynamically assimilate special 3-h and routine 12-h rawindsonde and dropsonde data into the model during the entire 36-h forecast period failed to deepen the cyclone as it moved northeast off the Atlantic seaboard beyond the area covered 3-h by sounding data. Subjectively analyzed mean sea-level pressures (MSLP) were included in the data base to allow the model to nudge toward 3-h surface pressure analyses extended to cover the region of cyclogenesis over the ocean. The assimilation of 3-h surface data over the ocean is insufficient to produce a realistic simulation of cyclogenesis. This result motivated the use of the nudging technique to assimilate surface pressure and upper air data over land during the 12-h pre-cyclogenetic period (i.e. dynamic initialization) and compare the subsequent 24-h simulation with one initialized statically at the same synoptic time. Dynamic initialization produced the best simulation of the occanic cyclone based upon the standard statistical scores and positions of the MSLP minima. This simulation is used to diagnose differences between cyclogenesis during GALEIOP 1 and IOP 10. Isentropic analyses and vertical cross sections are derived from the model simulations and are used to contrast the strength of the upper tropospheric forcing and the low-level static stability associated with each case. The results of the diagnostic analyses reveal that stronger surface response (based upon MSLP minima) to weaker upper-level forcing during GALEIOP 10 (compared with GALEIOP 1) was associated with differences in the lower tropospheric static stability and thermal advection patterns and their interaction with upper tropospheric features.