Statistics – Computation
Scientific paper
Aug 2006
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2006iaujd...8e..23k&link_type=abstract
Solar and Stellar Activity Cycles, 26th meeting of the IAU, Joint Discussion 8, 17-18 August 2006, Prague, Czech Republic, JD08,
Statistics
Computation
Scientific paper
We present a novel method for a fast LTE spectrum synthesis in magnetic stellar atmospheres suitable for Doppler-Imaging and Zeeman-Doppler-Imaging of active stars. The enormous computational demands for Doppler-Imaging and in particular Zeeman-Doppler-Imaging techniques require extensive calculations of local Stokes profiles over the entire stellar disk. Tabling techniques as used for conventional Doppler-Imaging to reduce the computational burden are not a feasible alternative for Zeeman-Doppler-Imaging of active stars with complex temperature and magnetic surface distributions. Also methods like the weak-field approximation which allows a fast estimate of Stokes V profiles do not provide the required accuracy in general. Artificial neural networks (ANN's) provide a new tool to approximate the full polarized radiative transfer under varying atmospheric conditions. Our method is based on a decomposition of Stokes line profiles, into their respective eigenspectra via a principal component analysis (PCA). A set of ANN's is then trained to calculate the reduced eigenspectrum of local Stokes line profiles as a function of the line of sight, effective temperature and the magnetic field configuration. The accuracy of the PCA-ANN synthesis is evaluated by comparison with a conventional numerical solution of the RTE which is based on a quadratic DELO method. Moreover, we present a number of test inversions with our Doppler-Imaging and Zeeman-Doppler-Imaging code to demonstrate the potential of this new method. The precision and the impressive speed of the PCA-ANN spectrum synthesis offers a viable and promising alternative for an on-the-fly calculation of local and disk integrated Stokes spectra.
Carroll Andrew T.
Kopf Markus
Strassmeier Klaus G.
No associations
LandOfFree
A Fast Stokes Profile Synthesis for Zeeman-Doppler-Imaging 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 A Fast Stokes Profile Synthesis for Zeeman-Doppler-Imaging, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and A Fast Stokes Profile Synthesis for Zeeman-Doppler-Imaging will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1784124