Next: Summary of Model Formulation
Up: Model Physics
Previous: Cloud Top Effect on
Archived sample figures of QTCM simulations for various
versions are available from http://www.atmos.ucla.edu/~csi/QTCM/
, along with
comparisons to the observed climatology for your reference.
For a number of climate features, you may notice that
the differences between simulations by different versions are modest changes in
intensity, slight shifts in position of particular features such as convection
zones, storm tracks, transition to westerlies, etc. The notes here simply
point out some aspects that we have noted as particular to each version.
This section was added with v2.3, so notes on earlier versions are sparse.
- V2.3
- Released August 2002
-
Many of the changes between v2.2 and v2.3 were motivated by ocean coupling.
Considerable attention was paid to
the climatological surface winds and stress which tended to be too strong
in v2.2. These are now weaker in the tropics,
in better agreement with observations.
We note that the zonal average wind stress is controlled by nonlinear
momentum transports, so this quantity is quite a challenging test.
Winds at 850mb are much less sensitive: a change in the smaller surface
winds is less noticable at this level or higher.
For the simulation at midlatitudes, the northern hemisphere surface wind
remains too strong and the southern hemisphere too weak for their respective
winter seasons.
- As a trade-off, the ENSO wind
stress anomalies are weaker than in v2.2, and it has been difficult to strengthen
these without also increasing the climatological stress.
-
A number of physics changes have contributed to the surface wind configuration.
The inclusion of vertical advection of momentum contributes substantially
to surface wind changes. The change of the boundaries to 80S-80N impacts
the midlatitude transients which in turn improves both tropical and
extratropical winds. The change to fourth order diffusion reduces the
effects of viscosity at large scales in the momentum
equations. The atmospheric boundary layer for winds influences
the wind turning and especially the equatorial wind strength.
- When the ABL is turned off, the overall climatology is
quite similar. Extratropical precipitation tends to be slightly stronger
when the ABL is on. The zonal mean zonal wind stress is very similar
but the zonally asymmetric zonal stress has differences, notably along the
equator where the ABL reduces the magnitude of the stress.
- The gross moist stability is smaller (more in line with
the estimated range from observations or reanalysis data sets). This
results in somewhat more intense convection zones and increased precipitation
anomalies in interannual variabilty, including ENSO.
- There is more tropical intraseasonal variability in this version
than in the last few versions (even without inclusion of the stochastic
precipitation parameterization explored in Lin and Neelin 2000 in v2.2).
In time series, this variability looks promising, but a spectral wavenumber
decomposition shows that the release version differs from observations
in some spectral characteristics and has too much westward propagating variance
at low wavenumbers along the equator. This behavior is sensitive to
parameters including the gross dry stability reference value and the
minimum wind speed in evaporation. Anyone interested in examining the
intraseasonal variability might consider exploring for an improved regime.
- V2.2
- Released November 2000
- This version has very low tropical intraseasonal
variability. This can actually be an asset for certain purposes. For instance,
Lin and Neelin (2000) used it to test what intraseasonal variability could
be maintained by a stochastic convective parameterization, with the
control having little variability maintained by other mechanisms.
Next: Summary of Model Formulation
Up: Model Physics
Previous: Cloud Top Effect on
Climate Systems Interaction Group
Sun Aug 25 00:58:46 PDT 2002