On the analysis of emission spectra and interference images

This dissertation addresses the analysis of emission spectra and interference images to determine spectroscopic temperature and ultra-short laser pulse width, respectively. Of particular interest are data fitting methods and techniques. The eigen-analysis technique is elaborated for the determination of the number of significant experimental data. Results of extensive Monte Carlo simulations are presented.

Emission spectra from 1064-nm Nd:YAG laser-induced air breakdown are compared with synthetic spectra to obtain the spectroscopic temperature of the transient plasma. Some 40-300 µs after optical breakdown, OH contributions dominate the emission spectrum in the wavelength range of 305-322 nm. Background emission and line profile effects are investigated in the data reduction of recorded plasma spectra. The background shows a strong correlation with temperature while the line profile is weakly correlated with temperature. The variance of the spectroscopic temperature is proportional to the magnitude of the variation of the data.

The autocorrelation function of ultra-short laser pulses is measured by systematically varying the time delay of crossed beams and recording interference patterns. The analysis is performed by the use of Fourier transform techniques to separate the interference cross term from the low-frequency background. A single autocorrelation coefficient is determined from each image. The average coherence length is obtained by fitting the autocorrelation coefficients with a model autocorrelation function.