Two-Dimensional Simulations of Frontal Collapse
In our study of frontal convection we first simulate
the evolution of a cold front in two dimensions.
This "2D" cold front has along-front wind speed also predicted,
so it might be considered "2.5-D".
This forcing is then moved to 3D, where random (and possibly
other) perturbations are added to allow three-dimensional
convection to evolve (which they do, given the shear in both along-
and cross-front wind).
As part of our modeling of the 2D front, we have carried
out some reference simulations in which the expected behavior is
known, based on previous research. The interested reader is
invited to examine
Williams (1967),
Gall et al. (1987),
Nakamura and Held (1989), and
Snyder et al. (1993).
In the simulations shown here, an idealized environment is used,
one appropriate for studying the Eady wave, an instability in which
differential advection of warm and cold air develops and intensifies
a cold front, which then shrinks in horizontal scale to some multiple
of the grid scale being used (regardless of dx),
a process referred to as "frontal collapse".
Specifications
- "Slab" symmetry - 2D model domain in x-z
- Along-front (out of xz plane) fixed gradient of
potential temperature
(by which the along-front (V) wind represents warm
and cold advection)
- Constant vertical wind shear
- Constant vertical potential temperature lapse rate
- Initial perturbation is sinusoidal in V(x)
- Horizontal resolution: 25 km
- Vertical resolution: 250 m
- Simulations performed on the
NCSA
Convex C3.
Movies
Realistic environments
The 2D evolution of the vertical velocity field in an environment with
non-constant stratification and shear is depicted here
(mpeg, 225k) |
(MooV, 1.9 MB).
Rising motion is
red and sinking motion is shown in blue. The domain shown is
4000 km wide and 18 km tall - the outer grid (prior to having two
inner nested grids being placed). The initial broad area of
rising motion is replaced by a narrow updraft jet at the leading edge,
with a narrow region of weakened updraft above that and greater rising motion at
a higher level. This split updraft formation is discussed in
Snyder et al. (1993)
and is the result of the superposition of the larger-scale semi-geostrophic
updraft and the vertical velocity oscillations accompanying a stationary
gravity wave at the leading edge of the front. The consequence for our
work is a somewhat narrow region of rising motion which is responsible for
convective initiation when this field is moved to three dimensions and
used in
squall line simulations.
Relevant links:
see also Dr. Brian Fiedler's page on
nonlinear Eady waves.
Brian F. Jewett
| bjewett@ncsa.uiuc.edu
| homepage
