Tunable diode laser spectrometry signal de-noising using discrete wavelet transform for molecular spectroscopy study

Wavelet-based digital signal processing technique was applied successfully to tunable diode laser absorption spectroscopy (TDLAS) in the infrared spectral range. By applying the wavelet-based digital signal processing technique to simulated and real TDLAS signals, spectral signal-to-noise ratios have been significantly enhanced, which is very useful for atmospheric trace gas detection and molecular spectroscopy study. It is worth specially pointing out that the determined precision of spectroscopic parameters by wavelet-based de-noising technique was distinctly improved, especially for the spectra corrupted by the annoying etalon fringes. Typically, the fitting precision of 2-5 times was achieved for the selected absorption line P10e of CO2 near 2.064 mu m. The primary results show that the proposed filtering technique have a great potential for various laser spectroscopy applications.