Fig. 1. (a) Experimental setup of MAX-DOAS, (b) MAX-DOAS instrument and Observation site

Fig. 2. Example of deconvolution of the DOAS spectrum in evaluation (a) O₄, (b) NO₂ and (c) HONO slant column densities of 2°elevation spectrum at 9:00 (Local Time) in December 9, 2017
Bule lines represent the absorption signal and red lines represent the fit signal

Fig. 3. A totally average daveraging kernel of (a) aerosol, (b) NO₂ and (c) HONO at 09:00 (Local Time) on December 8, 2017
The thin line represents the average nucleus connection curve, and the different colors represent different selected height layers (both 200 m are taken as one layer). The thick bmack solid line represents the envelope obtained by the inversion

Fig. 4. The near-surface concentration (a) and vertical columndensity(b) of NO₂ and HONO the first line of each figure is the profile of NO₂ and HONO. The second line shows the time series for the ratio of HONO to NO₂; And the third line is the time series of the AQI data released by the National Control Site named Dongpu Reservoir
The red dashed box indicates the deletion of MAX-DOAS observation data due to rainy, and the red dashed oval box indicates the lack of data due to instrumental observations

Fig. 5. Time series of HONO (a) and NO₂ (b) VMRs in the height of 10, 50 and 100 m

Fig. 6. (a) Temporal series profiles of NO₂ and HONO VMRs for the period December 26—31, 2017; (b) The backward trajectories of winds from December 29 to December 31, 2017, simulated by the Hysplit model; (c) Wind rose during the entire pollution period