A hydrostatic primitive equation model initialized ina highly baroclin
ically unstable state was used to simulate maritime cyclogenesis and f
rontogenesis. In order to identify boundary layer physical processes i
mportant in maritime frontogenesis, several different simulations were
performed. In an effort to isolate impacts due solely to the boundary
layer, moist processes were not included. An adiabatic and inviscid s
imulation provided the control for these experiments. Two different bo
undary layer parameterizations were used: a K-theory parameterization
featuring Richardson-number-dependent eddy diffusivity and a second-or
der closure scheme with prognostic equations for the turbulence quanti
ties. Results indicated that strong warm and cold fronts formed in the
adiabatic and inviscid case but that the vertical motion fields were
weak. In the K-theory simulation, the results were somewhat more reali
stic with stronger vertical motion. In both the K-theory and second-or
der closure simulations, the boundary layer in the cold air was highly
unstable and deep mixed layers formed in this region with a large gen
eration of turbulence. The largest cross-front temperature gradients e
xisted in the frontal zone above the mixed layer. These structures wer
e in qualitative agreement with observations of maritime cold fronts o
ver the northwest Pacific Ocean. The second-order closure simulations
produced a shallower mixed layer in the cold air with a stronger, more
narrow front and large vertical motion. These simulations were more c
onsistent with observations. Results from the second-order closure sim
ulations demonstrated that turbulent mixing of momentum was critical i
n reproducing the frontogenetic (and frontolytic) effects of the trans
verse secondary circulation.