 |
| Titel |
A comparative analysis of Simplified General Circulation Models of Venus atmosphere |
| VerfasserIn |
Sébastien Lebonnois, Stephen Lewis, Masaru Yamamoto, Christopher Lee, Jon Dawson, Peter Read, Joao Mendonca, Helen Parish |
| Konferenz |
EGU General Assembly 2011
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 13 (2011) |
| Datensatznummer |
250050889
|
|
|
|
|
|
| Zusammenfassung |
| With the successful Venus Express mission, and future missions planned for Venus
exploration in the near future, study of the atmosphere of Venus has been a rapidly expanding
field in the last few years. The development of Global Circulation Models (GCMs) has
focused on helping researchers to understand the details of the super-rotation mechanism, the
large-scale planetary waves and the polar vortices which are seen in this complex atmospheric
system.
Several groups that have been developing such tools have joined together within the
framework of a working group supported by the International Space Science Institute (ISSI,
Berne, Switzerland), and have started to compare how the different models behave under the
same forcing conditions. The goal of this intercomparison project is to test how robust the
response of the different numerical models is to identical constraints. A similar project has
been conducted recently at CalTech (Lee and Richardson, JGR 115, 2010, hereafter
LR10) using three different dynamical cores within a common model frame, and
we wanted to build upon this first study. We developed a common protocol and
conducted many simulations of Venus atmospheric circulation with three additional
GCMs: the CCSR model developed in Japan, the LMD model developed in France
and the Open University model (OpU) developed in Great Britain. A new model
developed in UCLA contributed simulations under very similar conditions, and is
therefore added to the project. We add to these new simulations the results of the LR10
study, as well as the results obtained in Oxford by Lee et al. (JGR 112, 2007; Lee,
PhD, 2006). These models are using a range of different types of dynamical cores
(spectral, finite differences or finite volumes). The baseline common parameters
include resolution, initial conditions, planetary and atmospheric parameters as well as
several physical parameterizations: thermal forcing, upper and lower boundary
conditions. In this work, thermal forcing is reduced to a simple newtonian cooling
parameterization with diurnally averaged conditions and no orbital variation of solar
forcing.
Comparison between the models shows how the different models spin up from rest,
yielding different final states. Though all models do reach states with significantly positive
superrotation, the amplitude and shape of the zonal wind fields is highly variable between
different GCMs and with changes in model parameters. We have been varying the physical
parameters to study the mode sensitivity. The choice of lower boundary condtions (the
planetary boundary layer scheme and presence, or absence, of surface topography) at the
planet’s surface has a significant influence on the deep atmospheric winds. The vertical
resolution (number of levels) in the model is demonstrated to have a strong effect in most
models. Modifications in the horizontal resolution were also found to be significant,
both on the shape and strength of the peak winds and on the deep atmospheric
winds. The impact on the peak winds seems to be qualitatively consistent between
models, though the impact on the winds deeper in the atmosphere is not. We have also
investigated the impact of varying the initial conditions. Most previous experiments
have been initialised with an isothermal atmosphere at rest, but we find evidence of
different end states in models which were initialised with super-rotating winds,
even after very long integration times. These experiments have been conducted
with the CCSR, LMD and OpU models. The results vary in detail between the
models, though in each case the atmosphere tends to stabilize with much higher peak
winds.
Though this work is done using a simplified thermal forcing and therefore may not be
fully representative of the real Venus atmosphere, it offers some guidance to the community
concerning the degree of complexity and sensitivity of the GCMs currently developed for the
Venus atmosphere. It also illustrates interesting differences between dynamical model cores
of the type in common use in terrestrial GCMs under conditions which lead to small
residual differences becoming highly significant, providing a strong test of model
dynamics. |
|
|
|
|
|
|