A method for the optimization of a Fabry-Perot interferometer (FPI) designed for the measurement of Doppler-broadened emission lines is presented. Assuming that the measurement values (counts) are Poisson distributed, the likelihood function is derived. Maximization of the likelihood function yields optimal estimates of temperature <i>T</i>, wind velocity <i>V</i>, and line intensity <i>I</i><sub>0</sub> and is accomplished by an iterative procedure of the Newton type. An optimal FPI design is obtained by a minimization of the calculated estimation errors Δ<i>T</i> and Δ<i>V</i> that represent the measurement quality. The method is appropriate for short-time measurements of weak emission lines, especially in space applications.
© 1982 Optical Society of AmericaPDF Article