Aspen Center for Physics Workshop on Climate Modeling and Stochastic Flows Atmospheric general circulation in an idealized dry GCM without eddy-eddy interactions Farid Ait-Chaalal and Tapio Schneider California Institute of Technology [email protected]June 26, 2012 Farid Ait-Chaalal (Caltech) Second-Order Atm. Circulation June 26, 2012 1 / 19
35
Embed
Atmospheric general circulation in an idealized dry GCM without eddy-eddy interactions
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Aspen Center for PhysicsWorkshop on Climate Modeling and Stochastic Flows
Atmospheric general circulation in an idealized dry GCMwithout eddy-eddy interactions
No evidence for any inverse energy cascade to scales larger than theRossby deformation radius in the atmosphere.
Previous work suggests that this is due to the effect of barocliniceddies on the thermal stratification that inhibits strong eddy-eddyinteractions (Schneider and Walker, 2006).
How important are nonlinear eddy-eddy interactions for the zonallyaveraged meridional circulation, the scale of the energy containingeddies and the stratification?
First step: look at the climatology of an idealized dry GCM inwhich the eddy-eddy interactions are removed.
Long-term objective: build higher-order closures for the hierarchy ofmoments to solve for the flow statistics (work in progress with BradMarston, Brown University).
No evidence for any inverse energy cascade to scales larger than theRossby deformation radius in the atmosphere.
Previous work suggests that this is due to the effect of barocliniceddies on the thermal stratification that inhibits strong eddy-eddyinteractions (Schneider and Walker, 2006).
How important are nonlinear eddy-eddy interactions for the zonallyaveraged meridional circulation, the scale of the energy containingeddies and the stratification?
First step: look at the climatology of an idealized dry GCM inwhich the eddy-eddy interactions are removed.
Long-term objective: build higher-order closures for the hierarchy ofmoments to solve for the flow statistics (work in progress with BradMarston, Brown University).
No evidence for any inverse energy cascade to scales larger than theRossby deformation radius in the atmosphere.
Previous work suggests that this is due to the effect of barocliniceddies on the thermal stratification that inhibits strong eddy-eddyinteractions (Schneider and Walker, 2006).
How important are nonlinear eddy-eddy interactions for the zonallyaveraged meridional circulation, the scale of the energy containingeddies and the stratification?
First step: look at the climatology of an idealized dry GCM inwhich the eddy-eddy interactions are removed.
Long-term objective: build higher-order closures for the hierarchy ofmoments to solve for the flow statistics (work in progress with BradMarston, Brown University).
No evidence for any inverse energy cascade to scales larger than theRossby deformation radius in the atmosphere.
Previous work suggests that this is due to the effect of barocliniceddies on the thermal stratification that inhibits strong eddy-eddyinteractions (Schneider and Walker, 2006).
How important are nonlinear eddy-eddy interactions for the zonallyaveraged meridional circulation, the scale of the energy containingeddies and the stratification?
First step: look at the climatology of an idealized dry GCM inwhich the eddy-eddy interactions are removed.
Long-term objective: build higher-order closures for the hierarchy ofmoments to solve for the flow statistics (work in progress with BradMarston, Brown University).
No evidence for any inverse energy cascade to scales larger than theRossby deformation radius in the atmosphere.
Previous work suggests that this is due to the effect of barocliniceddies on the thermal stratification that inhibits strong eddy-eddyinteractions (Schneider and Walker, 2006).
How important are nonlinear eddy-eddy interactions for the zonallyaveraged meridional circulation, the scale of the energy containingeddies and the stratification?
First step: look at the climatology of an idealized dry GCM inwhich the eddy-eddy interactions are removed.
Long-term objective: build higher-order closures for the hierarchy ofmoments to solve for the flow statistics (work in progress with BradMarston, Brown University).
Based on the GFDL pseudospectral dynamical core (Schneider and Walker,2006). Uniform surface, no seasonal cycle.
Radiative parametrization: Newtonian relaxation toward aradiative-equilibrium profile with pole-to-equator surface temperaturecontrast ∆h (∆h=90K for an Earth-like climate).
The model is dry but mimics some aspects of moist convection in aconvection scheme that relaxes temperature toward a prescribedlapse rate (γ× dry adiabatic lapse rate, γ = 0.7 for an Earth-likeclimate).
Based on the GFDL pseudospectral dynamical core (Schneider and Walker,2006). Uniform surface, no seasonal cycle.
Radiative parametrization: Newtonian relaxation toward aradiative-equilibrium profile with pole-to-equator surface temperaturecontrast ∆h (∆h=90K for an Earth-like climate).
The model is dry but mimics some aspects of moist convection in aconvection scheme that relaxes temperature toward a prescribedlapse rate (γ× dry adiabatic lapse rate, γ = 0.7 for an Earth-likeclimate).
Based on the GFDL pseudospectral dynamical core (Schneider and Walker,2006). Uniform surface, no seasonal cycle.
Radiative parametrization: Newtonian relaxation toward aradiative-equilibrium profile with pole-to-equator surface temperaturecontrast ∆h (∆h=90K for an Earth-like climate).
The model is dry but mimics some aspects of moist convection in aconvection scheme that relaxes temperature toward a prescribedlapse rate (γ× dry adiabatic lapse rate, γ = 0.7 for an Earth-likeclimate).
We compare the output of the full model with that of the modelwithout eddy-eddy interactions for 0.6 ≤ γ ≤ 1.0, for 0 ≤ ∆h ≤ 180Kand for three planetary rotation rates (ΩEarth,2ΩEarth and 4ΩEarth)
Simulations are run at T85 (128 latitude bands) with 30 σ-levels.The climatology is obtained through an average over 400 days, after aspin-up of 1600 days.
We focus on the meridional zonally averaged circulation, and morespecifically on mid-latitudes.
We compare the output of the full model with that of the modelwithout eddy-eddy interactions for 0.6 ≤ γ ≤ 1.0, for 0 ≤ ∆h ≤ 180Kand for three planetary rotation rates (ΩEarth,2ΩEarth and 4ΩEarth)
Simulations are run at T85 (128 latitude bands) with 30 σ-levels.The climatology is obtained through an average over 400 days, after aspin-up of 1600 days.
We focus on the meridional zonally averaged circulation, and morespecifically on mid-latitudes.
We compare the output of the full model with that of the modelwithout eddy-eddy interactions for 0.6 ≤ γ ≤ 1.0, for 0 ≤ ∆h ≤ 180Kand for three planetary rotation rates (ΩEarth,2ΩEarth and 4ΩEarth)
Simulations are run at T85 (128 latitude bands) with 30 σ-levels.The climatology is obtained through an average over 400 days, after aspin-up of 1600 days.
We focus on the meridional zonally averaged circulation, and morespecifically on mid-latitudes.
For large rotation rates, small ∆h, or, to a lesser extent small γ, themodel without eddy-eddy interactions forms realistic (magnitude andlocation) subtropical and eddy-driven jets.
For moderate rotation rates, large ∆h or γ, the circulation iscompressed in the meridional direction, with the appearance ofsecondary eddy-driven jets. This might be related to less isotropiceddies. The subtropical jet is over-estimated and the verticalstructure of momentum flux is not captured.
For large rotation rates, small ∆h, or, to a lesser extent small γ, themodel without eddy-eddy interactions forms realistic (magnitude andlocation) subtropical and eddy-driven jets.
For moderate rotation rates, large ∆h or γ, the circulation iscompressed in the meridional direction, with the appearance ofsecondary eddy-driven jets. This might be related to less isotropiceddies. The subtropical jet is over-estimated and the verticalstructure of momentum flux is not captured.
Potential vorticity eddy fluxes (colors), or equivalently the divergence ofthe Eliassen-Palm vector F (red arrows)
F = a cosφ
−u′v ′
f v ′θ′/ ∂θ∂pFarid Ait-Chaalal (Caltech) Second-Order Atm. Circulation June 26, 2012 13 / 19
Supercriticality Sc
A non-dimensional measure of near-surface isentropes slopes. Estimate themean level (pressure pe) up to which baroclinic activity redistributesentropy received at the surface (Schneider and Walker, 2006).
The no eddy-eddy model performs better when the baroclinic activityis weak (small γ, small ∆h and fast rotation in our experiments).Realistic subtropical jet, eddy-driven jets (without any inversecascade) and stratification.
The zonal eddy length scale are close to be reproduced, with a linearscaling of EAPE with eddy EKE over a wide range of parameters.However, the no edd-eddy model does not achieve a realistichorizontal isotropisation of the baroclinic eddies.
When the baroclinic activity is strong, it is too shallow in the noeddy-eddy model. What is the role of the eddy-eddy interactionsin determining the vertical structure of the mid-latitudestroposphere?
The no eddy-eddy model performs better when the baroclinic activityis weak (small γ, small ∆h and fast rotation in our experiments).Realistic subtropical jet, eddy-driven jets (without any inversecascade) and stratification.
The zonal eddy length scale are close to be reproduced, with a linearscaling of EAPE with eddy EKE over a wide range of parameters.However, the no edd-eddy model does not achieve a realistichorizontal isotropisation of the baroclinic eddies.
When the baroclinic activity is strong, it is too shallow in the noeddy-eddy model. What is the role of the eddy-eddy interactionsin determining the vertical structure of the mid-latitudestroposphere?
The no eddy-eddy model performs better when the baroclinic activityis weak (small γ, small ∆h and fast rotation in our experiments).Realistic subtropical jet, eddy-driven jets (without any inversecascade) and stratification.
The zonal eddy length scale are close to be reproduced, with a linearscaling of EAPE with eddy EKE over a wide range of parameters.However, the no edd-eddy model does not achieve a realistichorizontal isotropisation of the baroclinic eddies.
When the baroclinic activity is strong, it is too shallow in the noeddy-eddy model.
What is the role of the eddy-eddy interactionsin determining the vertical structure of the mid-latitudestroposphere?
The no eddy-eddy model performs better when the baroclinic activityis weak (small γ, small ∆h and fast rotation in our experiments).Realistic subtropical jet, eddy-driven jets (without any inversecascade) and stratification.
The zonal eddy length scale are close to be reproduced, with a linearscaling of EAPE with eddy EKE over a wide range of parameters.However, the no edd-eddy model does not achieve a realistichorizontal isotropisation of the baroclinic eddies.
When the baroclinic activity is strong, it is too shallow in the noeddy-eddy model. What is the role of the eddy-eddy interactionsin determining the vertical structure of the mid-latitudestroposphere?