|
It has been recognized that
radiative heating caused by atmospheric dust
is one of the most important factors in determining
the thermal and circulation structure
of the Martian atmosphere.
The vertical profiles of temperature observed by spacecrafts
frequently give stable lapse rate compared to dry adiabat
(e.g., Lindal et
al., 1979).
Those stable profiles are considered to be caused by radiative
heating associated with dust as is demonstrated by
one-dimensional (1D) radiative convective models
(Gierasch and Goody,
1972; Pollack et
al., 1979).
Dust is also considered to intensify the magnitude of
the Martian large scale atmospheric circulation,
as is suggested by the comparison between
the simulation results
under dusty and clear sky (i.e., dust-free) conditions
obtained with general circulation models (GCMs)
(e.g.,
Pollack et al., 1990).
However, GCMs have not succeeded in
the prognostic calculation of the amount of dust,
the important element of the Martian atmosphere.
The global dust storm, which is one of the most striking
phenomena in the Martian atmosphere
(e.g., Briggs
et al., 1979),
has not been self consistently
simulated in the current GCMs.
Starting from a dusty atmosphere as the initial condition,
the GCMs produce intense large scale wind
to inject dust into the atmosphere to maintain the
dust amount in the atmosphere.
However, starting from a dust-free or
a small amount of dust atmosphere,
the intensity of the large scale winds calculated in the GCMs
is so weak that the model surface stress is not large
enough to raise dust from the surface.
Current GCMs can not describe
a spontaneous transition from the dust-free Mars to
the dusty Mars
(Joshi et al., 1997;
Wilson and Hamilton, 1996).
It is argued by
Wilson and Hamilton (1996)
that the amount of surface stress
which is necessary for dust injection into the atmosphere
could be supplemented by small-scale wind fluctuations
which are not resolved in GCMs.
However, they presented neither the nature nor the origin of
small-scale wind motions.
As a candidate of the small-scale motions which are not
represented in GCM, one can think of vertical convection
driven by radiative forcing and sensible heat from the
surface.
Actually, it has been pointed out that
the data obtained at the site of Viking Lander 1
demonstrates the existence of vertical convection
(Hess et al., 1977;
Ryan and Lucich, 1983).
According to the studies by the use of 1D radiative convective
models,
the depth of convection layer is estimated to be
about from 9 to 10 km under the dust-free condition
( Flasar and Goody, 1976;
Pollack et al., 1979),
and
is about from 3 to 4 km under the dusty condition
when the visible optical depth of dust is from 0.3 to 0.5
(Savijärvi, 1991b;
Haberle et al., 1993).
However, since the fluid dynamical aspects of vertical
convection in the Martian atmosphere has not been
intensely examined so far,
the fundamental characteristics such as cell patterns and
wind intensity associated with vertical convection
have not been well revealed.
Vertical convection in the Martian atmosphere can be regarded
as dry convection in general.
Since the amount of water vapor is quite small,
condensation heating of water vapor can be neglected
compared to radiative heating in the Martian atmosphere
(e.g.,
Zurek et al., 1992).
Condensation of CO2,
the major component of the Martian atmosphere,
does not have to be considered except in the poler
region.
Dry convection occurs also in the terrestrial atmosphere
in the planetary boundary layer near the surface.
However,
dry convection in the Martian atmosphere
covers the whole height of Martian troposphere.
The characteristics of such "deep" dry convection
has been little examined so far.
In this study, we construct a numerical model which can
explicitly represent convective fluid motion, and
investigate possible characteristics of the vertical
convection in the Martian atmosphere.
We perform the following two cases of numerical simulations;
-
Simulation of vertical convection without dust
(the dust-free case):
Focus is placed on the patterns of circulation and intensity
of wind, and the amount of surface friction due to the
wind.
By using the calculated values of surface friction,
we discuss the possibility of dust injection from the
surface, which has not been realized in GCMs under the
dust free condition.
-
Simulation of vertical convection without dust
(the dusty case):
Assuming that the convective wind injects dust into
the atmosphere,
we investigate the characteristics of dust mixing by thermal
convection and the effects of radiative heating due to
dust on the circulation structure of vertical convection.
The numerical model utilized here is spatially two-dimensional
(2D).
The advantage of restricting the computational domain
to two-dimensional space is that
we can employ both a large computational domain and
a sufficiently high spatial resolution that
resolves convective plumes explicitly.
Moreover,
by using a 2D model,
it is expected to be easier to recognize some
characteristic features of convection, if any, than by
using a three-dimensional (3D) model.
|
