CARS spectral fitting with multiple resonant species using sparse libraries

Abstract

AbstractThe coherent anti‐Stokes Raman spectroscopy (CARS) technique is often used in the study of turbulent flames. Fast and accurate algorithms are needed for fitting CARS spectra for temperature and multiple chemical species. This paper describes the development of such an algorithm. The algorithm employs sparse libraries whose size grows more slowly with number of species than a regular library. It was demonstrated by fitting synthetic ‘experimental’ dual‐pump CARS spectra containing four resonant species (N2, O2, H2 and CO2), both with added noise and without it, and by fitting experimental spectra from a H2air flat flame produced by a Hencken burner. In the four‐species example, the library was nearly an order of magnitude smaller than the equivalent regular library (fitting times are correspondingly faster), and the fitting errors in the absence of added noise were negligible compared to the random errors associated with fitting noisy spectra. When fitting noisy spectra, weighted least squares fitting to signal intensity, as opposed to least squares fitting or least squares fitting to square root of intensity, minimized random and bias errors in fit parameters. Copyright © 2011 John Wiley & Sons, Ltd.