Simulation of Concentration Measurement of SO(2), NO(2) and Particles Simultaneously by Differential Optical Absorption Spectroscopy

Abstract

Our daily life is disturbed seriously by the haze weather now. It is very important to measure the haze composition quantificationally. The main composition of haze is SO(2), NO(2) and particles. At present, the research of measuring gas and particle simultaneously is rare relatively. This paper use differential optical absorption spectroscopy (DOAS) to simulate the concentration measurement of gas and particle simultaneously and obtain some meaningful results. Absorption spectral of many groups of different concentration of SO(2), NO(2) and particle were simulated, and each concentration was inverted by DOAS. In the first group of single component, the concentration change from 100 to 1 000 ppm, the inverted error of SO(2) is not greater than 0.17%, and which is 0.64% for NO(2). When the diameter of particle change from 100 to 500 nm, the inverted error is not greater than 2.08%. In the second group of multiple gases, when the concentration ratio of SO(2) and NO(2) is at the range of 1 : 10 and 5 : 1, the error of SO(2) is not bigger than 8%, and 5% for NO(2), relatively. But when the concentration of SO(2) is 10 times than NO(2), the error is higher than 10% for NO(2). In the third group of gas and particle, the error of gas concentration is lower than 10%, but the concentration error of particle is depended on signal to noise ratio (SNR) greatly. When SNR is higher than 40 dB, error can lower than 10% and when SNR is lower than 30dB, the error is bigger than 20%. From these results, we can see that DOAS can measurement SO(2), NO(2) and particles simultaneously effectively, and can applied to measure and analyze haze composition. However, when the absorption strength of the gases is different greatly, the strong absorption gas influent the weak absorption gas largely. And the SNR is lower, the error of inverted particle concentration increased greatly. The solution of these problems need better filtering and noise reduction method.