Solar flare CMOD prototype (work in progress)

[Cham]

Here what it gives for our Sun, with the default texture :







And here's the addon, for you to judge. Feedback would be appreciated :

http://nho.ohn.free.fr/celestia/Cham/Di ... Flares.zip (896 KB)

[t00fri]

What do you think of that flux tube (with proper color variations) ?

Martin,

OK, I understand all this quite well. Yet some important questions remain:

While the displayed magnetic field lines emerge from one dipole end and terminate in the other one, the apparent inhomogeneous distribution of neighboring field lines I cannot understand without assuming further ad hoc action! Why do you end up with flux tube type distributions, although your model is unrelated to super conductivity (at this "primitive" stage at least)??

Naively one may take a simple magnetic metal bar

N ===== S

as an intuitive example of your dipoles, making the field lines visible with metallic powder on a paper sheet, like many will remember from school.

These field lines also begin and end in the magnetic North (N) and South (S) poles, but you will NOT see such bundled tubes of magnetic flux as in your images. Of course from photographs we know that the solar flares tend to look flux tube like, yet one would wish to have a physically understandable mechanism as explanation!?

I hope you didn't simply "hand select" the field lines you want... from the lot?

Bye Fridger

[Cham]

Fridger,

don't forget it is about multi-dipoles distributions, with several dipoles at once and all interacting (field superposition). The resulting field lines can't be homogeneous. The lines can be heavily twisted and entangled. So your picture for the case of a single dipole do not apply here.

I'll try to find again the pages which describes the flux tubes formation mechanism on the sun (no superconductivity involved here). In brief, a flux tube (or string of entangled lines) is created when two dipolar fields are growing, one below the other one, and with some tilted angle. At some point of the process, the lines are rolled up the ones around the others, until there's a field reconnection occuring with a strong emission of charged particles to free space.

Of course, the flux tube I designed is a fake one (the small gray guy, shown above). I'm not very satisfied with it, so I removed it from my addon.

About the lines selection, I had to hand select some families of curves (not a curve at a time, of course !). This is consistent with the fact that the lines are "materialised" by the flow of hot plasma, which isn't flowing continuously in all directions.

Here are some web pages describing some flux tubes formation in astrophysics :

http://www.frascati.enea.it/ProtoSphera ... hysics.htm

http://web.hao.ucar.edu/public/asr/asr98/si/Figure5.gif

Watch this impressive movie :

http://www.asr.ucar.edu/2004/HAO/img/cme_movie1.gif

from this place :

http://www.nar.ucar.edu/2005/science8.jsp

Look also at the center of this web page, about the flux tube :

http://www.aip.org/tip/INPHFA/vol-9/iss-6/p18.html

This picture shows clearly what is happening with a flux tube :

http://spacescience.spaceref.com/newhom ... /loops.jpg

And look at the formation of some real tube on the sun. The part in the middle of the movie is really impressive :

http://www.nasa.gov/mpg/124373main_active_320px.mpg

There's a nice artistic animation from NASA showing a part of the process, here :

http://www.nasa.gov/mpg/125210main_flare_640x480.mpg

[cyber_space_doc]

wow, very nice addon and thanks for the mathematica work! Is there any similar work in a matlab .M file format?

I was thinking about the addon and I have several points:

1) perhaps sunspots should appear beneath the flares and flux tubes.

2) when I watch the sunspot video I am reminded of nvidia's floating point blending demo:-

http://http.download.nvidia.com/develop ... fp16_blend

3) perhaps adding point sprites similar to the volcano demo - without disturbing the cmod lines of course.

4) A high res version with extra field lines (possibly thinner lines?) would be very nice.

[Cham]

Is there any similar work in a matlab .M file format?

1) perhaps sunspots should appear beneath the flares and flux tubes.

4) A high res version with extra field lines (possibly thinner lines?) would be very nice.

It's probably possible to do something similar with MathLab. However, I don't have this program and I don't know how to use it. You're on your own there.

1) Yes, I intend to make a proper texture with sunspots beneath the lines. I'm also editing the addon (the addon I gave above was just a test version).

4) I'm trying to make more field lines which simulates what is occuring on the sun. Sadly, the modeling is pretty hard. Most of the time, I'm getting a mess of entangled lines (see the picture above with 5 dipoles).

You can adjust a bit the line thickness by editing the materials in the CMOD files. Just change all the opacity 0.5 to, say, opacity 0.4 or any other value you prefer. Currently, we can't really change the line thickness in Celestia. Just its transparency and color.

[Cham]

I've discovered that there are some very interesting movies here :

http://trace.lmsal.com/POD/bigmovies/DVDs/

These are complete, free, DVD movies of the sun magnetic activity. I downloaded one movie yet, and it's really worth the download time (well, if you are interested in that sort of things anyway).

[Fightspit]

Is there any similar work in a matlab .M file format?
...

It's probably possible to do something similar with MathLab. However, I don't have this program and I don't know how to use it. You're on your own there.

You can try Octave, it is a program (GNU) like and compatible with MathLab:

GNU Octave is a high-level language, primarily intended for numerical computations. It provides a convenient command line interface for solving linear and nonlinear problems numerically, and for performing other numerical experiments using a language that is mostly compatible with Matlab. It may also be used as a batch-oriented language.

For more information, see the page about Octave.

http://www.gnu.org/software/octave/

[Cham]

UPDATE :

While I now have a completely operational and perfectly satisfying addon version for the flares (just working on a texture, with Runar), I'm now testing a new procedure which may give another addon. This one will probably be of HIGH resolution :

Global magnetic map (Mathematica unfinished model)




Compare this model with the real thing, as shown below (the following picture comes from the TRACE web site) :


In brief, here's how I defined the model in Mathematica :

1- I defined a pure static dipole at the center of an unit sphere (Sun's surface). This will create a global dipolar field.

2- I distributed several magnetic monopoles just under the surface (in the previous model shown above, there are a total of 12 monopoles. I intend to put more of them). Each monopole represents a sunspot. Sunspots tend to be associated in pairs (North and South magnetic poles). Since there isn't any real magnetic monopoles out there and the field lines must be closed, the total magnetic charge must be 0.

3- All field lines are starting on the unit sphere and are returning to it. The only diifficulty in this approach is to define some proper field line initial conditions, so we could find a nicely looking model. This is where I need to make more tests, with more or less lines, etc.

[Fenerit]

Yup! Pratically is the same.

[Cham]

The global field map experiment is a failure. There are too much lines and it's a mess in Celestia. At some distance from the Sun, it isn't very satisfying. Also, it's slowing down the frame rate pretty much, since there are several hundreds of field lines (there were 14 monopoles and few dipoles in a single model).

Despite this, there are some very interesting results from this global experiment. I'm able to extract some small parts and make models of the local field, like the one I'm showing in the middle of the following picture :



This is the field around a sunspot. At close range :



The part inside the red circle should be a sunspot texture.

From the side view, the field is looking almost identical to what we can see on most of the TRACE pictures.