Model fitting to astronomical data

Astronomers are often faced with the problem of constructing models for systems which are not fully constrained by observational data. Multi-parameter model fitting is therefore an important component of the modelling process. In binary star research, time-dependent spectroscopic observations reveal the motions of the gas giving rise to the radiation through Doppler shifts of spectral lines (Doppler Tomography). The data set is then inverted to construct geometrical and physical models for the system.

Methods currently widely used for model fitting are the Maximum Entropy Method, and the Genetic Algorithm method. It is proposed that the student reviews these methods and tests them against accretion disc and magnetically confined accretion curtain and funnel models using simulated and real data.