Noise reduction and background correction of overlapped UV absorption spectra based on EWT-ASG and airPLS

Abstract

High concentrations of SO₂ and NO emitted by vehicles not only endanger human health, but also cause a negative impact on the environment such as acid rain. The absorption intensities of NO and SO₂ at the ultraviolet region 201–230 nm are larger, which is theoretically more conducive to concentration inversion during optical monitoring. However, it is greatly limited by the spectrum overlapping of NO and SO₂ at this range, which was often accompanied by larger errors. Thus, few methods were used at this region for monitoring concentration. In this work, the higher sensitive detection method aimed at alleviated the problem of large retrieval error to the serious overlapping of UV-absorption spectrum was developed. EWT-ASG and airPLS were used for noise reduction and background correction. According to the theory of spectral superposition, the differential optical density (DOD) of NO, SO₂ and the gas mixture at the wavelength range of 201–230 nm were separated. Moreover, parameter correction method on gas concentration retrieval was applied, and the deviation caused by incomplete separation was corrected. The strength of absorption peaks and lower interference were comprehensively considered. Finally, the NO and SO₂ concentrations, the corresponding relative errors and the detection limits were calculated. Compared with the counterparts obtained by polynomial fit, the errors were decreased ∼30 %. This work can provide a reference for simultaneously monitoring of high concentrations of SO₂ and NO in complex environments with enhanced environmental adaptability.