View
220
Download
0
Tags:
Embed Size (px)
Citation preview
The tribulations and exaltations in coupling models of the
magnetosphere with ionosphere-thermosphere models
Aaron RidleyDepartment of Atmospheric, Oceanic and Space Sciences
Aaron RidleyDepartment of Atmospheric, Oceanic and Space Sciences
GEM/CEDAR WorkshopJuly 1, 2005
Slide 2 of
Ionosphere Thermosphere Modeling and coupling
Ionosphere Thermosphere Modeling and coupling
A quick review. The ionosphere and thermosphere. High latitude electrodynamics.
Coupling the neutral winds to the magnetosphere
Ion outflow Other couplings
Some that work Some that may not be on the horizon, but should be.
Pontification time
A quick review. The ionosphere and thermosphere. High latitude electrodynamics.
Coupling the neutral winds to the magnetosphere
Ion outflow Other couplings
Some that work Some that may not be on the horizon, but should be.
Pontification time
39
GEM/CEDAR WorkshopJuly 1, 2005
Slide 3 of
[e-] and Tn[e-] and Tn
Many Thermosphere/Ionosphere plots “stolen” from my student Yue Deng!All T/I results from the global ionosphere thermosphere model (GITM)
309
GEM/CEDAR WorkshopJuly 1, 2005
Slide 4 of
Temperature Altitude Distribution
Temperature Altitude Distribution
noon midnight
465
GEM/CEDAR WorkshopJuly 1, 2005
Slide 5 of
Low Altitude Temperature Distribution
Low Altitude Temperature Distribution
739
GEM/CEDAR WorkshopJuly 1, 2005
Slide 6 of
High Altitude Temperature Distribution
High Altitude Temperature Distribution
1001
GEM/CEDAR WorkshopJuly 1, 2005
Slide 7 of
Electron Density Altitude Distribution
Electron Density Altitude Distribution
1304
GEM/CEDAR WorkshopJuly 1, 2005
Slide 8 of
Low Altitude Electron Distribution
Low Altitude Electron Distribution
1573
GEM/CEDAR WorkshopJuly 1, 2005
Slide 9 of
High Altitude Electron Distribution
High Altitude Electron Distribution
1846
GEM/CEDAR WorkshopJuly 1, 2005
Slide 10 of
High Altitude Electron Distribution
High Altitude Electron Distribution
2149
GEM/CEDAR WorkshopJuly 1, 2005
Slide 11 of
Vi and Vn with Bz = -1 nTVi and Vn with Bz = -1 nT
Ion flows driven primarily by potential Neutral winds driven by (a) Gradient in pressure; (b) Corriolis; (c) ion drag.
Note dawn/dusk differences
2638
GEM/CEDAR WorkshopJuly 1, 2005
Slide 12 of
Vi and Vn with Bz = -10 nTVi and Vn with Bz = -10 nT
Ion flows driven primarily by potential Neutral winds driven by (a) Gradient in pressure; (b) Corriolis; (c) ion drag.
Note dawn/dusk differences
3067
GEM/CEDAR WorkshopJuly 1, 2005
Slide 13 of
[e-] and Vn with HPI = 100 GW[e-] and Vn with HPI = 100 GW
Significant increase in the electron density causes much larger ion drag effect
Dawn cell “much” more defined.
3471
GEM/CEDAR WorkshopJuly 1, 2005
Slide 14 of
Vi, Vn, and how well they are coupled
Vi, Vn, and how well they are coupled
4132
GEM/CEDAR WorkshopJuly 1, 2005
Slide 15 of
Vi in F-region and E-regionVi in F-region and E-region
Rotation of Vectors Shortening of Vectors
Rotation of Vectors Shortening of Vectors
4578
GEM/CEDAR WorkshopJuly 1, 2005
Slide 16 of
Would the real Vi please step forward?
Would the real Vi please step forward?
As the collision frequency becomes large, most people think of the ion velocity rotating away from ExB to E.
That is not really true. Since there is a neutral wind, the ion velocity rotates towards a combination of E and Un.
We can then think of this in a couple of different ways: The current caused by E is divergenceless, but the current caused by
Un is not, so we have to force the total current to be: So, calculate the divergence of the neutral wind driven current (perpendicular to the magnetic field).
Integrate this current, to come up with a total wind driven current.
Solve a Poisson equation to find a potential that would cancel this current out.
The push the ions with the solved E-field. This the methodology used by all modeling groups for solving for equatorial electrojet and coupling to magnetospheric codes.
Pushing ions with Un will cause a polarization electric field. We could map this polarization electric field along field lines to higher altitudes.
Should be equivalent. Also applies to things like gravity and gradient pressure.
As the collision frequency becomes large, most people think of the ion velocity rotating away from ExB to E.
That is not really true. Since there is a neutral wind, the ion velocity rotates towards a combination of E and Un.
We can then think of this in a couple of different ways: The current caused by E is divergenceless, but the current caused by
Un is not, so we have to force the total current to be: So, calculate the divergence of the neutral wind driven current (perpendicular to the magnetic field).
Integrate this current, to come up with a total wind driven current.
Solve a Poisson equation to find a potential that would cancel this current out.
The push the ions with the solved E-field. This the methodology used by all modeling groups for solving for equatorial electrojet and coupling to magnetospheric codes.
Pushing ions with Un will cause a polarization electric field. We could map this polarization electric field along field lines to higher altitudes.
Should be equivalent. Also applies to things like gravity and gradient pressure.
4931
GEM/CEDAR WorkshopJuly 1, 2005
Slide 17 of
Test run of the Space Weather Modeling Framework.
IMF inputs shown. Look at potential. Look at currents caused
by neutral winds.
Test run of the Space Weather Modeling Framework.
IMF inputs shown. Look at potential. Look at currents caused
by neutral winds.
5577
GEM/CEDAR WorkshopJuly 1, 2005
Slide 20 of
Ionospheric outflowIonospheric outflow
Outflow is also very important in MI coupling.
Can control the density in the plasma sheet.
Oxygen outflow can significantly change the mass density in the magnetosphere. Lowers the Alfven velocity. Adds to the ring current.
Outflow is also very important in MI coupling.
Can control the density in the plasma sheet.
Oxygen outflow can significantly change the mass density in the magnetosphere. Lowers the Alfven velocity. Adds to the ring current.
5238
GEM/CEDAR WorkshopJuly 1, 2005
Slide 21 of
What controls Outflow?What controls Outflow?
It seems like outflow is a two step process: Raise the
ionospheric plasma up.
Suck it out into magnetosphere
Joule heating is one of the primary mechanisms thought to control the raising of the ionosphere.
It seems like outflow is a two step process: Raise the
ionospheric plasma up.
Suck it out into magnetosphere
Joule heating is one of the primary mechanisms thought to control the raising of the ionosphere.
QuickTime™ and aPhoto - JPEG decompressor
are needed to see this picture.
4872
GEM/CEDAR WorkshopJuly 1, 2005
Slide 22 of
Effect of heating on electron density
Effect of heating on electron density
QuickTime™ and aPhoto - JPEG decompressor
are needed to see this picture.
4321
GEM/CEDAR WorkshopJuly 1, 2005
Slide 23 of
Outflow ExperimentsOutflow Experiments
Examine what the influence of the ion outflow is on the magnetosphere
Use simple constant boundary conditions at the inner boundary of the magnetosphere diffusion lifts the density off the boundary a few cells
Gradient in pressure brings the plasma out into the magnetosphere
These experiments are meant to show what the most simple thing possible will do to the magnetosphere
Run to steady-state Northward IMF, flip to Southward IMF at t=0, and see what happens.
Examine what the influence of the ion outflow is on the magnetosphere
Use simple constant boundary conditions at the inner boundary of the magnetosphere diffusion lifts the density off the boundary a few cells
Gradient in pressure brings the plasma out into the magnetosphere
These experiments are meant to show what the most simple thing possible will do to the magnetosphere
Run to steady-state Northward IMF, flip to Southward IMF at t=0, and see what happens.
3965
GEM/CEDAR WorkshopJuly 1, 2005
Slide 24 of
N=1000; Grid 4; No RCMN=1000; Grid 4; No RCM
QuickTime™ and aPNG decompressor
are needed to see this picture.
GEM/CEDAR WorkshopJuly 1, 2005
Slide 25 of
CPCP variations for 3 runsCPCP variations for 3 runs
Changing the density seems to:• Increase the cross polar cap potential• Make the transition take longer
N=10 N=100
N=1000
GEM/CEDAR WorkshopJuly 1, 2005
Slide 26 of
But……But……
The cross polar cap increasing doesn’t make much sense. Why does it do this???? After thinking a bit…
Our numerical solver has to add diffusion for stability. That diffusion is controlled by the fastest wave speed in
the cell… roughly the Alfven speed. Which is controlled by the density. So, turning the density up means turning the diffusion
down. Turning the diffusion down allows more current to make it
to the inner boundary, and hence to the ionosphere. The cross polar cap potential goes up. Purely numerical. Crap. The funny thing is that this is true for (a) grid
resolution, (b) where you put the boundary, and (c) Artificially reducing the speed of light (Boris) also.
The cross polar cap increasing doesn’t make much sense. Why does it do this???? After thinking a bit…
Our numerical solver has to add diffusion for stability. That diffusion is controlled by the fastest wave speed in
the cell… roughly the Alfven speed. Which is controlled by the density. So, turning the density up means turning the diffusion
down. Turning the diffusion down allows more current to make it
to the inner boundary, and hence to the ionosphere. The cross polar cap potential goes up. Purely numerical. Crap. The funny thing is that this is true for (a) grid
resolution, (b) where you put the boundary, and (c) Artificially reducing the speed of light (Boris) also.
3180
GEM/CEDAR WorkshopJuly 1, 2005
Slide 27 of
What Coupling Should BeWhat Coupling Should Be
Magnetosphere Model
Field-aligned Currents
Heat FluxElectron & Ion Precipitation
Plasmasphere Density
Potential
Electrodynamics Model
Ionosphere-Thermosphere Model
Neutral wind FACs
Conductances
Upward Ion Fluxes
Tides Gravity Waves
Solar Inputs
Photoelectron Flux
2713
GEM/CEDAR WorkshopJuly 1, 2005
Slide 28 of
What we have discussed so farWhat we have discussed so far
Magnetosphere Model
Field-aligned Currents
Potential
Electrodynamics Model
Ionosphere-Thermosphere Model
Neutral wind FACsUpward Ion Fluxes
2525
GEM/CEDAR WorkshopJuly 1, 2005
Slide 29 of
Electron and Ion PrecipitationElectron and Ion Precipitation
Magnetosphere Model
Electron & Ion Precipitation
Electrodynamics Model
Ionosphere-Thermosphere Model
Conductances
This is the hardest part of the coupling
T-I models use energy deposition codes to determine ionization and heating rates as a function of altitude, given input (ion and electron) spectra at the top of the model. This is sort of a major weakness if not done well, or if distributions are assumed to be Maxwellian and are not.
Need to have both ion and neutral densities correct to get conductances
2189
GEM/CEDAR WorkshopJuly 1, 2005
Slide 30 of
PhotoelectronsPhotoelectrons
Magnetosphere Model
Ionosphere-Thermosphere Model
Photoelectron Flux
Photoelectron are created by sunlight. These electrons flow along field lines from the sunlit hemisphere to the dark hemisphere, causing soft electron precipitation. This can effect the F-region density in the winter hemisphere.
Photoelectron codes are relatively “expensive” to run, so they are typically ignored.
Photoelectron flux could be parameterized with a transmission coefficient through the plasmasphere.
1939
GEM/CEDAR WorkshopJuly 1, 2005
Slide 31 of
Plasmaspheric DensityPlasmaspheric Density
Magnetosphere Model
Plasmasphere Density
Ionosphere-Thermosphere Model
Many global circulation models have a hard time getting the F-region densities correct, because the pressure gradient at the top of the model is unknown. With an accurate plasmaspheric model, the gradient could be determined and an inflow or outflow would be self-consistently derived.
1776
GEM/CEDAR WorkshopJuly 1, 2005
Slide 32 of
Electron Heat FluxElectron Heat Flux
Magnetosphere Model
Heat Flux
Ionosphere-Thermosphere Model
Magnetospheric electron heat flux causes the electron to heat up in the ionosphere. This changes the height distribution of the electron pressure, which causes the ions to lift.
1492
GEM/CEDAR WorkshopJuly 1, 2005
Slide 33 of
Electron Heat FluxElectron Heat Flux
Magnetosphere Model
Heat Flux
Ionosphere-Thermosphere Model
Wait. Did you say lift?
1100
GEM/CEDAR WorkshopJuly 1, 2005
Slide 34 of
Electron Heat FluxElectron Heat Flux
Magnetosphere Model
Heat Flux
Ionosphere-Thermosphere Model
The electron energy heat flux may cause changes in the amount of ion outflow.
Upward Ion Fluxes
Therefore, passing the heat flux from magnetospheric codes (that are capable of computing it - like RAM) to the IT models may be crucial for accurately specifying outflow regions
999
GEM/CEDAR WorkshopJuly 1, 2005
Slide 35 of
Electron heat flux experimentElectron heat flux experiment
Simulations done by Alex Glocer, a graduate student at UM.
Using updated version of the Gombosi et al. [1645, I think] polar wind code.
Do two ion outflow runs 80o latitude noon Summer conditions low f10.7
Run 1 nominal heat flux Run 2 double heat flux
Simulations done by Alex Glocer, a graduate student at UM.
Using updated version of the Gombosi et al. [1645, I think] polar wind code.
Do two ion outflow runs 80o latitude noon Summer conditions low f10.7
Run 1 nominal heat flux Run 2 double heat flux
675
GEM/CEDAR WorkshopJuly 1, 2005
Slide 36 of
Electron heat flux experimentElectron heat flux experiment
By changing the electron heat flux by a factor of two: increase H+ outflow by a
little bit. Increase O+ by a factor
of two. While the polar wind code
is still being developed and validated, the results are intriguing.
By changing the electron heat flux by a factor of two: increase H+ outflow by a
little bit. Increase O+ by a factor
of two. While the polar wind code
is still being developed and validated, the results are intriguing.
472
GEM/CEDAR WorkshopJuly 1, 2005
Slide 37 of
What Coupling Should BeWhat Coupling Should Be
Magnetosphere Model
Field-aligned Currents
Heat FluxElectron & Ion Precipitation
Plasmasphere Density
Potential
Electrodynamics Model
Ionosphere-Thermosphere Model
Neutral wind FACs
Conductances
Upward Ion Fluxes
Tides Gravity Waves
Solar Inputs
Photoelectron Flux
281
GEM/CEDAR WorkshopJuly 1, 2005
Slide 38 of
SummarySummary
The thermosphere and ionosphere are overlapping, tightly coupled regions of space that do influence the magnetosphere. And Vise-versa.
We sort of understand the neutral wind coupling to the ion flows.
We sort of understand what happens to electrons and ions from the magnetosphere (if the magnetosphere could specify them correctly…)
We really don’t understand outflow Joule heating effects can last a LONG time. Electron energy flux could play a role - no one has
coupled this yet. Plasmasphere? Photoelectrons? Wouldn’t it be great is we could model the system
without the numerics getting in the way?
The thermosphere and ionosphere are overlapping, tightly coupled regions of space that do influence the magnetosphere. And Vise-versa.
We sort of understand the neutral wind coupling to the ion flows.
We sort of understand what happens to electrons and ions from the magnetosphere (if the magnetosphere could specify them correctly…)
We really don’t understand outflow Joule heating effects can last a LONG time. Electron energy flux could play a role - no one has
coupled this yet. Plasmasphere? Photoelectrons? Wouldn’t it be great is we could model the system
without the numerics getting in the way?
78