A Semi-Blind Source Separation Method for Differential Optical Absorption Spectroscopy of Atmospheric Gas Mixtures

TL;DR

This paper introduces a semi-blind source separation method for atmospheric gas analysis using DOAS, combining multi-resolution analysis, convex minimization, and independent component analysis to identify known and unknown trace gases.

Contribution

The novel three-step approach effectively separates known and unknown gases in complex atmospheric spectra, improving upon traditional least squares methods.

Findings

Successfully extracted ozone from experimental data

Identified multiple trace gases from residuals

Enhanced gas detection accuracy

Abstract

Differential optical absorption spectroscopy (DOAS) is a powerful tool for detecting and quantifying trace gases in atmospheric chemistry \cite{Platt_Stutz08}. DOAS spectra consist of a linear combination of complex multi-peak multi-scale structures. Most DOAS analysis routines in use today are based on least squares techniques, for example, the approach developed in the 1970s uses polynomial fits to remove a slowly varying background, and known reference spectra to retrieve the identity and concentrations of reference gases. An open problem is to identify unknown gases in the fitting residuals for complex atmospheric mixtures. In this work, we develop a novel three step semi-blind source separation method. The first step uses a multi-resolution analysis to remove the slow-varying and fast-varying components in the DOAS spectral data matrix $X$. The second step decomposes the preprocessed data $\hat{X}$ in the first step into a linear combination of the reference spectra plus a remainder, or $\hat{X} = A\,S + R$, where columns of matrix $A$ are known reference spectra, and the matrix $S$ contains the unknown non-negative coefficients that are proportional to concentration. The second step is realized by a convex minimization problem $S = \mathrm{arg} \min \mathrm{norm}\,(\hat{X} - A\,S)$, where the norm is a hybrid $\ell_1/\ell_2$ norm (Huber estimator) that helps to maintain the non-negativity of $S$. The third step performs a blind independent component analysis of the remainder matrix $R$ to extract remnant gas components. We first illustrate the proposed method in processing a set of DOAS experimental data by a satisfactory blind extraction of an a-priori unknown trace gas (ozone) from the remainder matrix. Numerical results also show that the method can identify multiple trace gases from the residuals.