Computer Science – Performance
Scientific paper
Dec 2009
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2009agufm.p43e..03s&link_type=abstract
American Geophysical Union, Fall Meeting 2009, abstract #P43E-03
Computer Science
Performance
[1510] Geomagnetism And Paleomagnetism / Dynamo: Theories And Simulations
Scientific paper
In simulating planetary dynamos, it is important but practically difficult to decrease kinematic viscosity, ν, of model fluids. We usually take the Ekman number, E=ν/2Ωrc2, and the magnetic Prandtl number, Pm=ν/η, as measures of viscous effects in dynamo action, where rc is the radius of the core (the dynamo region), η is the magnetic diffusivity and Ω is the angular velocity. The Ekman number is generally much smaller than unity because of huge planetary length scales. The magnetic Prandtl number is also believed to be smaller than unity in either metallic or ionic fluids in interiors of magnetized planets. The smaller E is, the more difficult it is to resolve small length and time scales. The smaller Pm, the more difficult to obtain large enough magnetic Reynolds number that is a key parameter to excite the dynamo action. Previous studies have mainly covered the parameter space of E>10-6 and Pm>1. Recent developments in high performance computing have gradually enabled us to research the parameter space of E=O(10-7) and Pm=O(0.1). I review recent results of such low-viscosity dynamo simulations and discuss some implications for planetary dynamos. There are several differences and similarities between the low viscosity and previous higher viscosity models. First, the boundary condition for the core surface temperature may be more important in low-viscosity thermally-driven convective dynamos than has ever been thought (Sakuraba and Roberts, 2009). When the heat flux is laterally uniform at the core surface, large-scale vortices and strong magnetic fields are created, which are basically the same as obtained in previous higher viscosity models. However, when the surface temperature is uniform, as has been assumed in some other low-viscosity models, small-scale sheet-like flows dominate and the magnetic field (particularly the toroidal field) is weak. In the geodynamo, the uniform surface temperature model is inappropriate because of poor heat transport ability of the solid mantle. It seems that the uniform heat-flux model well simulated some geomagnetic field behaviors. Future planetary dynamo models should pay attention to physical properties of the overlying layer and what boundary condition is physically appropriate. Second, dynamos of small Pm seem to be qualitatively different from large-Pm dynamos. Even in small-Pm dynamos, it is possible to maintain large-scale vortices and strong magnetic fields. However, there is a tendency that the region where the magnetic field is strong is highly confined in space and small-scale turbulence takes place where the magnetic field is weak. In particular, the core surface magnetic field is significantly weaker than in the interior zonal magnetic field. This suggests that some attention should be paid in estimating planetary internal magnetic fields and structures by using surface magnetic field observations.
No associations
LandOfFree
Characteristics of low-viscosity models for planetary dynamos (Invited) does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Characteristics of low-viscosity models for planetary dynamos (Invited), we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Characteristics of low-viscosity models for planetary dynamos (Invited) will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1772894