Techniques for deriving Doppler temperatures from multiple-line Fabry–Perot profiles: an analysis

Abstract
Two techniques for recovery of the Doppler temperature from Fabry–Perot spectrometer observations of multiple-line profiles are tested. A direct fit of a modeled profile to the data is found to produce more consistent results and is less sensitive to noise than the method of least-squares fitting to low-order Fourier-transform coefficients derived from the data. This result is unexpected since the Fourier-transform method permits the reduction of noise by truncating the series of coefficients used to determine the temperature.