research summary
 
home research publications teaching group for the public
 
Summary of A Gray-Radiation Aquaplanet Moist GCM. Part I: Static Stability and Eddy Scales by Frierson, Held and Zurita-Gotor, which appears in Journal of the Atmospheric Sciences.

Primary arguments:
  • Development of a simplified moist general circulation model for studies of large scale dynamics.
  • Midlatitude static stability is controlled by moisture over a wide range of parameters.
  • Length scales vary remarkably little despite large changes in static stability.
Discussion:
Our first goal with this work was to design a simplified moist general circulation model, something along the lines of the Held-Suarez dry GCM, but with moisture. In particular, we wanted a model in which we could vary the atmospheric moisture content over a wide range, to get an idea of how the midlatitude jet stream, pole-to-equator temperature gradients, the Hadley circulation, the location of rainy and dry regions, etc., are influenced by latent heating.

We chose idealized physical parameterizations for the model, which allow for ease of interpretation of results and reproducibility. Since we wanted to design an energetically consistent system, the model ends up being quite a bit more complicated than the Held-Suarez model, but still is significantly simpler than a full GCM. You can read the paper for specifics on what we chose and why, but I'll highlight here a few of the choices that seem to surprising or confusing to people.

First, often people are surprised to see that the model can be run without a convection scheme, i.e., with grid scale condensation only (we referred to this as "large scale condensation only" in the paper, but "grid scale condensation only" is a better term. See the description of Frierson 2007a for reasons why). One might think that running with grid scale condensation only might cause the model to blow up, since the strong radiative heating near the surface creates strong convective instability, which goes to the smallest scale unless a convection scheme is present (see also Garner et al 2007). However, the model can indeed run despite the strong instability.

In retrospect, one part of the model that we probably should have simplified is the stable side of the drag coefficient and boundary layer parameterization. When the surface is stable, we reduce the drag coefficients and the corresponding boundary layer diffusivities according to Eqns. 11 and 18. However, in the simulations here, the surface is rarely in a stable state, and when the surface is stable, the reduction in drag coefficient is minimal. Therefore, those wishing to reproduce these calculations should be able to do so in a simpler manner by using the unstable drag coefficient and boundary layer parameterization (Eqns. 12 and 19) in all cases. Stable surface layers would be important for instance over high latitude land surfaces in the winter, so the model as described is more easily extended to cases with land.

One of the first things that one notices when changing the moisture content in this model is how easy it is to change the isentropic slope in midlatitudes: the midlatitude static stability increases greatly with moisture content. Typical theories for static stability/isentropic slope (Stone 1972, Held 1978, Schneider 2004) are based on dry baroclinic eddy dynamics, and do not predict such changes in static stability. I have investigated midlatitude static stability in two other papers, Frierson 2007c and Frierson 2006.

With a large change in midlatitude static stability, one might expect a large change in baroclinic eddy length scales, since these are often thought to be determined by the Rossby radius of deformation. However, we see no such change in eddy length scales in the midlatitudes, which stay remarkably fixed in these simulations. We instead interpret the change in terms of the Rhines scale, which may indicate that a turbulent cascade may be present in our simulations. The question of what determines baroclinic eddy length scales in the presence of moisture is one I am actively investigating with Ed Gerber and Lorenzo Polvani.

The figure above is an instantaneous plot of the boundary layer depth in the dry limit simulation we study here.  This plot is key to showing that convection reaches the tropopause well into the midlatitudes in this simulation.  

Full citation:
Frierson, D. M. W., Held, I. M. and P. Zurita-Gotor. A Gray-Radiation Aquaplanet Moist GCM. Part I: Static Stability and Eddy Scales. Journal of the Atmospheric Sciences, 63, 2548-2566, 2006.

The official journal link can be found here.

A PDF download of the full paper can be found here.

This download is courtesy of the AMS, who owns sole rights to it. The download is subject to copyright laws and statutes. For more information, please visit the AMS/Allen Press website.