Archives


Recent news

Edge-on disks
21st Nov. 2014

The previous version of the raytracing code was not really suited for edge-on disks. As soon as the line of sight would go below the upper layers of the disk, it would not work anymore. Since several interesting debris disks observed in scattered light are edge-on, or close to edge-on, it was a quite severe limitation of the code.

So I took the time to improve the code so that now it is possible to produce images of edge-on disks. It was actually quite a funny experience. First I really had to clean up the code because it was turning out to be a bit messy. And next I had to think of how to find the intercept between the disk geometry and the line of sight, which took me a bit of time but I am actually quite happy with the way it turned out. Below you can find an example of the kind of images I can produce (log-scale, H-band image, the star is not included).


Inclunding polarized light in raytracing
1st Sep. 2014

I spent the last couples of days trying to include the contribution of polarized light in my raytracing module of the code and it seems to be working quite well at the moment. I more or less had everything in place already since I was already computing the Mueller's matrix to get the first element of the diagonal (S11) to get the scattered light. I had to include the contribution of the S12 element in the raytracer and that was more or less it. I did have some interpolation problems which made it a bit more complicated than anticipated. Basically the phase function can oscillate betwenn positive and negative values as a function of the scattering angles. Since I do the interpolation in log space that was a bit troublesome and I had some nice "Not a Number" values. The workaround was to offset the phase function prior the interpolation and offset it down afterwards.
Above is an example of the Stokes Q vector for a "template" debris disk. Darker regions are negative values for the Q vector and lighter regions are positive values. Since I do not really do a lot of additional computations, the time to compute the image did not really increase by a lot and is still of about a few seconds (6 sec for this 301 x 301 pixels image at 0.66┬Ám). One should be aware that this is only valid for optically thin disks, but I hope this will be useful to model future SPHERE observations of debris disks.