A hydrostatic, primitive equation model is used to simulate an oceanic cyclone with idealized initial conditions. The model uses a pressure coordinate in the vertical with a grid spacing of 100 mb. In the horizontal a grid spacing of 25 km is used, which should be nearly sufficient to resolve slantwise convection. The model produces an explosive moist cyclone with an intense bent-back warm front. The thermal gradient in the bent-back warm frontal region exceeds 8 K/100 km, in agreement with recent observations.
Before rapid deepening, the model atmosphere becomes unstable to slantwise convection in the warm frontal region. After the spinup period, buckling in the angular momentum and θε surfaces are noted. It is suggested that the descending motion and the associated dry slot over the cyclone center may arise from the descending branch of the slantwise convection on the warm side of the warm front. The descent may be augmented by the evaporation of liquid water. After the explosive deepening period, the stratification in both the warm front and the bent-back warm front exhibits neutrality to slantwise convection.
The Ertel potential vorticity (EPV) inversion technique developed by Davis and Emanuel is used to obtain the perturbation geopotential at 900-mb, 500-mb, and 300-mb levels due to EPV anomalies at different levels. The inversion is applied at the mature stage of the cyclone at 45 h. It is found that there is a positive EPV anomaly along the regions of the warm front and bent-back warm front, and it accounts for 40% of the perturbation geopotential at 900 and 500 mb over the cyclone center. The contribution of low-level EPV anomaly in the moist cyclone to the perturbation geopotential at 500 mb over the cyclone center is twice that in the dry case. The circulation of the inverted nondivergent wind fields in the moist run shows a small-scale cyclonic vortex and the presence of cold advection in the bent-back warm frontal region.
The contribution of the upper-level EPV anomaly to the 900-mb perturbation geopotential is also significant in the moist cyclone. The physical mechanism for the latter effect can be traced to an increase in vorticity advection in the middle troposphere in association with the formation of the bent-back warm front. This finding is in agreement with the authors' recent two-layer model results, which show that the bent-back warm front represents a region of cold advection that can in turn lead to an intensification of the upper-level wave. The contribution of the 1000-mb θ anomaly in the mature moist cyclone is smaller than that of the dry run because of the convection-induced cold advection in the bent-back warm front.