Influence of Chinese New Year overlapping COVID-19 lockdown on HONO sources in Shijiazhuang

Abstract

Nitrous acid (HONO) is an important precursor of hydroxyl radical (OH) in the atmosphere. It is also toxic to human health. In this work, HONO concentrations were measured in Shijiazhuang using a Monitor for AeRosols and Gases in ambient Air (MARGA) from December 15, 2019 to March 15, 2020, which covered the heavy air pollution season, the Chinese New Year (CNY) vocation and the Corona Virus Disease-19 (COVID-19) lockdown period. During & after CNY overlapping COVID-19 lockdown, the air quality was significantly improved because of both the emission reduction and the increase in diffusion ability of air masses. The mean HONO concentration was 2.43 ± 1.08 ppbv before CNY, while it decreased to 1.53 ± 1.16 ppbv during CNY and 0.97 ± 0.76 ppbv after CNY. The lockdown during & after CNY reduced ~31% of ambient HONO along with ~62% of NO and ~36% of NO₂ compared with those before CNY after the improvement of diffusion ability had been taken into consideration. Heterogeneous reaction of NO₂ on ground surface dominated the nocturnal HONO sources, followed by heterogeneous reaction on aerosol surface, vehicle emission, reaction between NO and OH and emission from soil on pollution days throughout the observation. Except for elevated soil emission, other nighttime HONO sources and sinks decreased significantly during & after CNY. The relative importance of heterogeneous reaction of NO₂ on surfaces further increased because of both the decrease in vehicle emission and the increase in the heterogeneous conversion kinetics from NO₂ to HONO during & after CNY.

Graphical abstract

Highlights

• •High concentration of HONO was observed from 0.08 to 12.67 ppbv in Shijiazhuang.
• •Heterogeneous conversion of NO₂ to HONO dominated the nighttime HONO source.
• •The lockdown during & after CNY reduced ~31% of ambient HONO.
• •Heterogeneous conversion kinetics from NO₂ to HONO increased during & after CNY.
• •The T and aerosol properties promoted conversion of NO₂ to HONO during & after CNY.

1. Introduction

Nitrous acid (HONO) is a crucial atmospheric species because it is the vital precursor of hydroxyl radical (OH) (Alicke et al., 2003; Ren et al., 2006; Spataro and Ianniello, 2014), which is able to clean up primary air pollutants and produce secondary air pollutants in the atmosphere. Photolysis of H₂O₂, HCHO, O₃ and HONO and the reaction between NO and HO₂ are the major sources of OH radical in the atmosphere (Alicke et al., 2003; Volkamer et al., 2010), whereas photolysis of HONO is the dominant primary OH source in the early morning when other OH sources are still weak (Alicke et al., 2003; Spataro et al., 2013; Tan et al., 2018; Tang et al., 2015; Volkamer et al., 2010). Photolysis of HONO accounts for average up to 60% of OH production in the boundary layer in some cases (Alicke et al., 2003; Elshorbany et al., 2009; Tan et al., 2018). It has been found that HONO is capable of promoting secondary aerosols (Czader et al., 2015; Liu et al., 2020; Xing et al., 2019; Zhang et al., 2019a; Zhang et al., 2019c) and ozone formation (Zhang et al., 2019a). In addition, exposures to high concentration of HONO may damage mucous membranes and result into the respiratory system of asthmatics (Ohyama et al., 2019; Ohyama et al., 2010; Rasmussen et al., 1995). HONO is also a precursor of the mutagenic and carcinogenic nitrosamines via reaction with secondary and tertiary amines (Pitts et al., 1978; Sleiman et al., 2010). Therefore, it is important to investigate the atmospheric HONO from the point view of both atmospheric chemistry and toxicology, in particular, in the regions suffering from heavy air pollution.

Atmospheric HONO has been investigated since the late 1970s (Perner and Platt, 1979). Field experiments were carried out at different environments worldwide in the last few decades (Acker et al., 2006; Crilley et al., 2016; Lee et al., 2016; Levy et al., 2014; Michoud et al., 2014; Spataro and Ianniello, 2014; Zhang et al., 2012a). The HONO concentrations varied from a few pptv in remote areas (Beine et al., 2001; Gu et al., 2020; Honrath et al., 2002; Liao et al., 2006; Spataro et al., 2017; Zhang et al., 2012a; Zhou et al., 2001) to several ppbv in polluted urban environment (Crilley et al., 2016; Hu et al., 2002; Spataro et al., 2013; Zhang et al., 2019e). In China, the concentration and budget of atmospheric HONO have also been extensively studied (Spataro et al., 2013), such as in Beijing (Hendrick et al., 2014; Hu et al., 2002; Spataro et al., 2013; Wang et al., 2017; Zhang et al., 2019e), Shanghai (Wang et al., 2013; Zhang et al., 2019b), Guangzhou (Hu et al., 2002; Su et al., 2008a), Hongkong (Xu et al., 2015), Ji'nan (Li et al., 2018) and Xi'an (Huang et al., 2017). Overall, the HONO concentrations were higher in China than those observed in Europe and America. At the present time, direct emissions from combustion and soil, homogeneous reaction between NO and OH radical, heterogeneous reaction of NO₂ on aerosol and ground surfaces and photolysis reactions of nitrates have been identified as the major sources of atmospheric HONO (Spataro and Ianniello, 2014), whereas their relative contributions depend on the location and the season. For example, heterogeneous reaction of NO₂ was proposed to be an important HONO source in the night (Wang et al., 2017; Zhang et al., 2019c) and even in daytime in Beijing-Tianjin-Hebei (BTH) (Zhang et al., 2019c), but it was unimportant compared with the unknown sources and homogeneous reaction between NO and OH in Ji'nan (Li et al., 2018) or compared with traffic emission on haze days in Beijing (Zhang et al., 2019e). At the same time, the traffic emission was found to be an unimportant HONO source during nighttime in BTH (Zhang et al., 2019c), while it could contribute ~50% to HONO sources on haze days in Beijing (Liu et al., 2020; Meng et al., 2019; Zhang et al., 2019e). In addition, an unknown daytime HONO source, which is highly correlated with light intensity (Lee et al., 2016; Michoud et al., 2014), was frequently observed at various places. More and more studies proposed that it might be associated with photo-enhanced conversion of NO₂ (Michoud et al., 2014; Su et al., 2008b) and photolysis of surface nitrate or particulate nitrates although an uncertainty may reduce their importance (Liu et al., 2019b). These results mean that the study about atmospheric HONO budget is still far from closed, which requires a significant effort on both the HONO measurement and the determination of related kinetic parameters for its production pathways (Liu et al., 2019b).

Shijiazhuang is one of the cities suffering from heavy air pollution in China (Qin et al., 2017; Tan et al., 2019; Xie et al., 2019; Zhang et al., 2020). Fine particulate matter (PM) pollution is more serious in Shijiazhuang than the neighboring areas such as Beijing (Cheng et al., 2019; Qin et al., 2017; Zhang et al., 2019). However, the studies on air pollution regarding both fine particles and HONO are very limited in Shijiazhuang. To the best of our knowledge, there is no publication on the concentration and the source studies of atmospheric HONO in Shijiazhuang. Therefore, it is meaningful to investigate the HONO sources for understanding the complex atmospheric chemistry in Shijiazhuang. In particular, during the Chinese New Year (CNY) in 2020 overlapping the Corona Virus Disease-19 (COVID-19) epidemic, the COVID-19 lockdown led to significant reduction of traffic and industry emissions. This provides us a unique opportunity to understand the influence of anthropogenic activities on atmospheric HONO concentrations and confirm the relative importance among different sources. In this work, we performed continuous field observations of HONO and other air pollutants at an urban station in Shijiazhuang from December 15, 2019 to March 15, 2020. The changes of HONO concentrations and the sources have been discussed during the COVID-19 lockdown compared with that before CNY.

2. Material and methods

2.1. Field measurements

Field measurements were performed at Hebei Atmospheric Supersite, which is located in the campus of Shijiazhuang University (Lat. 38.0281° and Lon. 114.6070°). The observation station (Fig. S1) is on a rooftop of the main teaching building (5 floors, ~23 m above the surface) which is around 250 m from the Zhujiang road. It is a typical urban observation station surrounded by traffic and residential emissions.

Mass concentration of PM_2.5 was measured by a beta attenuation mass monitor (BAM-1020, Met One Instruments) with a smart heater (Model BX-830, Met One Instruments Inc., USA) to control the RH of the incoming air to 35% and a PM_2.5 inlet (URG) to cut off the particles with diameter larger than 2.5 μm. Water-soluble ions (Cl⁻, NO₃ ⁻, SO₄ ²⁻, Na⁺, NH₄ ⁺, K⁺, Mg²⁺, and Ca²⁺) in PM_2.5 and gas pollutants (HCl, HONO, HNO₃, SO₂ and NH₃) were measured using an analyzer for Monitoring AeRrosols and Gases in ambient Air (MARGA, ADI 2080, Applikon Analytical B.V.) with 1 h of time resolution. Trace gases including NO_x, SO₂, CO and O₃ were measured with the corresponding analyzer (Thermo Scientific, 42i, 43i, 48i and 49i). Meteorological parameters including temperature, pressure, relative humidity (RH), wind speed and direction were measured using a weather station (WXT 520, Vaisala). Planetary boundary layer (PBL) height was measured using a Doppler Lidar (EV-Lidar-CAM, Everise Technology Ltd.). Particles size distribution from 10 to 760 nm was measured with a scanning mobility particle sizer (SMPS, TSI), which is consist of a differential mobility analyzer (DMA 3938, TSI) and a condensation particle counter (CPC 3776, TSI). The particles from 500 to 10,000 nm were measured by an APS (3321, TSI). OC and EC were measured according to the National Institute for Occupational Safety and Health (NIOSH) protocol using an OC/EC analyzer (Model 4, Sunset).

The MARGA was externally calibrated using anionic solutions (Cl⁻, Br⁻, NO₃ ⁻, SO₄ ²⁻) and cationic solutions (Li⁺, Na⁺, K⁺, Mg²⁺ and Ca²⁺) seasonally. Internal calibration was also hourly carried out using LiBr standard solutions. The detection limit of Cl⁻, NO₃ ⁻, SO₄ ²⁻, Na⁺, NH₄ ⁺, K⁺, Mg²⁺, and Ca²⁺ were 0.01, 0.05, 0.04, 0.05, 0.05, 0.09, 0.06 and 0.09 μg m⁻³, respectively. All of the instruments for trace gas measurements were calibrated using the corresponding standard gases weekly. The detection limits are 0.05, 0.05, 40, and 0.5 ppbv for NO_x, SO₂, CO and O₃, respectively. External calibration was performed biweekly for the OC/EC analyzer using sucrose solutions.

2.2. HONO budget calculation

The major sources of ambient HONO include direct emission from soil (E _soil) (Meusel et al., 2018; Oswald et al., 2015) and vehicle exhaust (E _vehicle) (Trinh et al., 2017), homogeneous reaction between NO and OH (P _NO-OH) in the atmosphere, photolysis of nitrate (P _nitrate) (Bao et al., 2018), heterogeneous reaction of NO₂ on aerosol surface (P _aerosol) (Liu et al., 2015) and ground surface (P _ground) (Li et al., 2018; Liu et al., 2019b; Wang et al., 2017). Photolysis of HNO₃ and nitrophenol were usually not considered in source budget analysis because they were believed as minor HONO sources (Lee et al., 2016). The sinks of HONO include photolysis (L _photolysis), the homogeneous reaction with OH radical (L _HONO-OH), dry deposition (L _deposition) (Liu et al., 2019b) and vertical and horizontal transport (T _trans) (Soergel et al., 2011). Thus, the HONO budget can be calculated by,

$$\frac{d{c}_{\mathit{HONO}}}{\mathit{dt}}=E_{\mathit{soil}}+E_{\mathit{vehicle}}+P_{\mathit{NO}-\mathit{OH}}+P_{\mathit{nitrate}}+P_{\mathit{aerosol}}+P_{\mathit{ground}}+P_{\mathit{unknown}}-L_{\mathit{photolysis}}-L_{\mathit{HONO}-\mathit{OH}}-L_{\mathit{deposition}}+T_{\mathit{trans}}$$ (1)

where $\frac{d{c}_{\mathit{HONO}}}{\mathit{dt}}$ is the observed change rate of HONO mixing ratios (ppbv h⁻¹), P _unknown is the production rate of HONO from the unknown sources.

In our previous work (Liu et al., 2020), the calculation methods for these terms have been discussed in detail. Briefly, the emission rate of HONO (E _HONO, ppbv h⁻¹) from soil and vehicle were calculated based on the emission flux (F _HONO, g m⁻² s⁻¹) and the PBL height (H, m) according to the following equation,

$$E_{\mathit{HONO}}=\frac{a\bullet F_{\mathit{HONO}}}{H}$$ (2)

where, α is the conversion factor ($\mathrm{\alpha }=\frac{1\times {10}^9\bullet 3600\bullet R\bullet T}{M\bullet P}=\frac{2.99\times {10}^{13}\bullet T}{M\bullet P}$), M is the molecular weight (g mol⁻¹), T is the temperature (K) and P is the atmospheric pressure (Pa). For vehicle emission,

$$F_{\mathit{HONO},\mathit{vehicle}}=\frac{{\mathit{EI}}_{\mathit{HONO}},\mathit{vehice}}{A}=\frac{{\mathit{EI}}_{\mathit{NOx},\mathit{vehicle}}}{A}\bullet {\left(\frac{\mathit{HONO}}{\mathit{NOx}}\right)}_{\mathit{vehicle}}$$ (3)

where, EI _HONO is the emission inventory of HONO (g s⁻¹), A is the urban area of Shijiazhuang (496 km², measured based on Google map), EI _NOx,vehicle is the emission inventory of NO_x from vehicle exhaust (0.160 Gg day⁻¹ in Shijiazhuang before CNY) (Qi et al., 2017). The daily mean EI _NOx,vehicle was further converted into hourly mean emission inventory based on the hourly mean traffic index (www.nitrafficindex.com, Fig. S2A) before CNY. During & after CNY, the hourly mean emission inventory of NO_x was further calculated according to its reduction ratio from traffic emission (62% which will be discussed in Section 3.1). The NO_x emission inventories are shown in Fig. S2B and C. The emission ratio of HONO to NO_x (1.26%) was estimated using a low limit correlation method (Li et al., 2012). This value is very close to those derived values in Hongkong (1.2 ± 0.4%, (Xu et al., 2015) and 1.23 ± 0.35% (Liang et al., 2017)), Guangzhou (1.0%) (Li et al., 2012) and Beijing (1.17%, 1.3% and 1.41%) (Liu et al., 2020; Meng et al., 2019; Zhang et al., 2019e). The F _{HONO, soil} was calculated using the temperature-dependent emission flux of HONO from grassland with 35–45% of water content (Oswald et al., 2013). Homogeneous reaction between OH and NO was calculated based on the measured NO concentrations and the estimated OH concentrations in the light of second-order reaction. The second-order reaction rate constant (k _NO-OH) is 7.2 × 10⁻¹² cm³ molecule⁻¹ s⁻¹ (Li et al., 2012). The daytime OH concentration was estimated according to (Li et al., 2018),

$$c_{\mathit{OH}}=\frac{4.1\times {10}^9\times {\left(J_{O1D}\right)}^{0.83}\times {\left(J_{N{O}_2}\right)}^{0.19}\times \left(140c_{N{O}_2}+1\right)}{0.41c_{{\mathit{NO}}_2}^2+1.7c_{N{O}_2}+1}$$ (4)

where, J _O1D and J _NO2 are the photolysis frequency of O₃ and NO₂ (s⁻¹), respectively, c _OH and c _NO2 are the concentration of OH and NO₂ (molecules cm⁻³), respectively. The J _O1D, J _NO2 and J _HONO (photolysis frequency of HONO) were calculated using the hourly mean solar zenith angle, the longitude and latitude of the observation station under clear sky condition using a box model, then calibrated according to the measured UV light intensity (Liu et al., 2020). The nighttime OH concentration was assumed to be 1.0 × 10⁵ molecules cm⁻³ (Li et al., 2012; Tan et al., 2018). The calculated J values and OH concentrations were shown in Fig. S3. Production rate of HONO from photolysis of nitrate was calculated based the measured nitrate concentrations and the photolysis frequency of nitrate (J _nitrate). The mean J _nitrate value of 8.24 × 10⁻⁵ s⁻¹ (Bao et al., 2018) was normalized to the measured UV light intensity in this work. Heterogeneous conversion of NO₂ on aerosol and ground surfaces was calculated according to (Huang et al., 2017; Li et al., 2018),

$$k_{\mathit{het}}=\frac{c_{\mathit{HONO},\mathit{corr},t_2}-c_{\mathit{HONO},\mathit{corr},t_1}}{\overline{c_{\mathit{NO}2}}\left(t_2-t_1\right)}$$ (5)

$$c_{\mathit{HONO},\mathit{corr}}=c_{\mathit{HONO}}-0.0126c_{\mathit{NOx}}$$ (6)

where, k _het is the quasi first-order reaction rate constant for heterogeneous conversion from NO₂ to HONO (s⁻¹), c _HONO,corr is the corrected HONO concentration after the HONO emitted from vehicle exhaust has been subtracted (ppbv), $\overline{c_{\mathit{NO}2}}$ is the mean NO₂ concentration from t ₁ to t ₂. The calculated nighttime k _het was 0.016 ± 0.006 h⁻¹ from December 15, 2019 to March 15, 2020, which is comparable with those derived in urban environment such as Guangzhou (0.016 h⁻¹), Milan (0.012 h⁻¹) and Kathmandu (0.014 h⁻¹), and higher than that in Ji'nan (0.0068 ± 0.0045 h⁻¹) (Li et al., 2018; Xu et al., 2015). Interestingly, the nighttime k _het before CNY was 0.012 ± 0.005 h⁻¹, which was significantly lower than that during and after CNY (0.017 ± 0.003 h⁻¹, P < 0.05). Therefore, the production rates of HONO from heterogeneous reaction in different periods were calculated using the corresponding k _het. The contributions of aerosol and ground surfaces to heterogeneous conversion from NO₂ to HONO were further calculated based on the measured surface area to volume ratio of aerosol (S _a, 0.0017 ± 0.0013 m⁻¹) and the estimated ground surface to volume ratio (S _g, 0.0043 ± 0.0018 m⁻¹) according to,

$$S_g=\frac{2.2}{H}$$ (7)

where 2.2 is a roughness of urban ground surface (Li et al., 2018). It should be pointed out that direct emission from soil should also be considered in Eq. (6). However, we did not consider it similar to the previous work (Huang et al., 2017; Li et al., 2018) because the ambient concentration of HONO contributed by soil could not be calculated like that from vehicle emission using NO_x as a reference. This might result into some uncertainty for calculating the k _het and will be discussed in Section 3.3.

The loss rates of HONO by photolysis (L _photolysis), homogeneous reaction with OH radicals (L _HONO-OH) and dry deposition (L _deposition) (Liu et al., 2019b) were calculated according to the following equations.

$$L_{\mathit{photolysis}}=3600\bullet J_{\mathit{HONO}}\bullet c_{\mathit{HONO}}$$ (8)

$$L_{\mathit{HONO}-\mathit{OH}}=3600\bullet k_{\mathit{HONO}-\mathit{OH}}\bullet c_{\mathit{OH}}\bullet c_{\mathit{HONO}}$$ (9)

$$L_{\mathit{deposition}}=\frac{3600\bullet v_d\bullet c_{\mathit{HONO}}}{H}$$ (10)

where, J _HONO is the photolysis rate of HONO (s⁻¹), k _HONO-OH is the second-order reaction rate constant between HONO and OH (6 × 10⁻¹² cm³ molecule⁻¹ s⁻¹) (Atkinson et al., 2004), and v _d is the dry deposition rate of HONO (0.001 m s⁻¹) (Han et al., 2017). The vertical and horizontal transport was estimate according to Eq. (11),

$$T_{\text{trans}}=k_{\text{dilution}}\left(c_{\text{HONO}}-{c_{\text{HONO}}}_{\text{background}}\right)$$ (11)

where k _dilution is a dilution rate (0.23 h⁻¹, including both vertical and horizontal transport) (Dillon et al., 2002), c _HONO and c _HONO,_background is the HONO concentration at the observation site and background site, respectively (Dillon et al., 2002). In this work, the lowest nighttime HONO concentration was taken as the c _HONO,_background.

3. Results and discussion

3.1. Overview of the air quality during observation

Fig. 1 shows the time series of selected parameters including the concentrations of PM_2.5, CO, NO_x and HONO (Fig. 1A and B) and the meteorological parameters such as temperature, pressure, wind speed, wind direction and PBL height (Fig. 1C and D) during the observation. The CNY vacation from January 23 to February 2 was highlighted by the light purple column. The temperature increased gradually and the pressure varied from 998 to 1034 hPa during the whole observation (Fig. 1C). Before and during CNY, the frequency of stagnant weather conditions characterized by low wind speed and low PBL height was higher, while the UV light intensity was lower than those after CNY (Fig. 1D). Five light precipitations occurred on December 15, 2019, February 14 and 15, 24 and March 8, 2020. These datasets will be ruled out in following discussion.

Fig. 1.

The time series of (A) PM_2.5 and CO concentrations, (B) HONO and NO_x concentrations, (C) temperature and pressure and (D) wind direction, wind speed and ultraviolet light intensity from December 15, 2019 to March 15, 2020. The concentrations of PM_2.5 and HONO are colored by RH and NO concentration, respectively. The wind direction and wind speed are colored by planetary boundary layer (PBL) height. The highlighted period is during Chinese New Year (CNY).

Six pollution events before CNY, one during CNY and another six after CNY overlapping the COVID-19 lockdown were identified according to the concentration of PM_2.5 (Fig. 1A). After strong wind (regardless of wind direction) accompanied with relatively high PBL height cleaning up the air masses (Fig. 1D), a pollution episode gradually built up with a typical duration of 3– 5 days. Interestingly, the PM_2.5 concentration well kept pace of the reciprocal of the PBL height (Fig. S4A), in particular, before CNY. This means that the sources of PM_2.5 should be quite stable and the mass loading of fine PM should be greatly determined by the diffusion ability of air masses in Shijiazhuang. As shown in Fig. 1A, the high PM_2.5 concentration usually coincided with the high RH (> 50%) which is favorable of heterogeneous conversion of gas phase precursors (Hodas et al., 2014; Wang et al., 2016; Wu et al., 2019). In addition, the PM_2.5 evolved synchronously with CO, especially, before CNY. These results mean that both secondary production and primary emissions contribute to PM_2.5 accumulation in Shijiazhuang.

Before CNY, the hourly mean PM_2.5 concentration varied from 7 to 403 μg m⁻³ with a mean value of 137.9 ± 85.8 μg m⁻³ (Fig. 2A). 70.3% of hourly PM_2.5 concentration was higher than the daily mean air quality standard of China (75 μg m⁻³) before CNY. This indicates the serious air pollution in Shijiazhuang when compared with other cities, such as Beijing and Tianjin (Fu et al., 2014). During CNY, a pollution event occurred continuously from January 22 to February 1. Although the maximum PM_2.5 concentration during CNY was lower than that before CNY, the mean PM_2.5 concentration (173.9 ± 72.7 μg m⁻³) was higher than that before CNY because of the long-term favorable metrological conditions for pollutants accumulation including high RH (Fig. 1A), low wind speed and low PBL height (Fig. 1D). After CNY, pollution event still happened frequently. However, both the mass concentration of PM_2.5 and the frequency of pollution event decreased significantly because of both the reduction of primary emission (indicated by CO, Fig. 1A) and the increase in wind speed and PBL height (Fig. 1D). The mean concentration of PM_2.5 was 76.7 ± 61.4 μg m⁻³, and 36.1% of hourly mean PM_2.5 concentration were higher than 75 μg m⁻³ after CNY.

Fig. 2.

(A)–(D) Changes of pollutants concentration and meteorological parameters before CNY (BCNY), during CNY and after CNY (ACNY) and (E) relative changes of pollutants and meteorological parameters during & after CNY compared with that before CNY.

From the point view of chemical composition, the fraction of both nitrate and sulfate increased during and after CNY compared with that before CNY (Fig. S5). The fractions of secondary inorganic aerosol (SIA, including NO₃ ⁻, SO₄ ²⁻ and NH₄ ⁺) increased significantly from 48.6 ± 13.8% to 53.4 ± 13.6% (P < 0.05). The mass ratio of nitrate to sulfate decreased from 1.78 ± 0.78 before CNY to 1.38 ± 0.48 during CNY and then recovered to 1.88 ± 1.08 after CNY (Fig. S4A). This implies that reduction of traffic emission should be prominent during CNY, while both traffic and industry sectors contribute to the emission reduction after CNY overlapping COVID-19 lockdown. In addition, higher conversion ratios of nitrate (NOR) were observed during (0.44 ± 0.11) and after CNY (0.29 ± 0.17) compared with that before CNY (0.17 ± 0.10) (Fig. S6). The above results mean that air quality was greatly improved, while chemical conversion of the precursors to PM was slightly enhanced during and after CNY overlapping COVID-19 lockdown when compared with the counterpart.

The relative change of other pollutants and meteorological parameters including OC, EC, nitrate, SO₂, CO, O₃, NH₃, HONO, NO, NO₂, NO_x, visibility, wind speed, PBL height and CO₂ are shown in Fig. 2. The absolute concentrations of PM_2.5 components and SO₂ increased significantly during CNY compared with that before CNY, while followed by a decease after CNY. For example, the mean concentration of OC, EC, nitrate and SO₂ were 22.2 ± 7.7, 5.4 ± 2.5, 37.7 ± 17.5 μg m⁻³ and 7.2 ± 2.4 ppbv during CNY, respectively, while the corresponding mean values were 16.1 ± 7.9, 5.1 ± 3.1, 32.7 ± 22.2 μg m⁻³ and 6.4 ± 3.1 ppbv before CNY and 9.8 ± 6.6, 2.3 ± 1.7, 21.5 ± 24.9 μg m⁻³ and 4.5 ± 2.3 ppbv after CNY. The concentration of CO and CO₂ decreased from 1.7 ± 0.9 and 0.49 ± 0.04 ppm before CNY to 1.4 ± 0.6 and 0.47 ± 0.03 during CNY, and further decreased to 1.0 ± 0.5 and 0.45 ± 0.02 ppm after CNY. The concentrations of NO_x well followed PM_2.5 before CNY, while it slightly increased in the pollution events during and after CNY (Fig. 1). The mean NO_x concentrations were 58.0 ± 33.7, 17.7 ± 12.4 and 15.7 ± 10.4 ppbv before, during and after CNY, respectively. As a more effective tracer of traffic emission, the mean NO concentration decreased from 26.3 ± 26.2 ppbv before CNY to 4.2 ± 6.6 ppbv during CNY and to 3.4 ± 4.4 ppbv after CNY. The concentration of NH₃ was almost constant with the mean concentration of 19.6 ± 10.0, 19.2 ± 5.6, 19.9 ± 11.4 ppbv, respectively, during the three periods. However, the concentration of O₃ increased significantly during CNY (27.5 ± 24.4 ppbv) and after CNY (24.4 ± 13.4 ppbv) compared with that before CNY (7.2 ± 7.9 ppbv) due to increase in UV light (Fig. 1D). The above results indicate that emission reduction from traffic sector is prominent during CNY and after CNY, while industry sector might further contribute to the emission reduction after CNY overlapping the COVID-19 lockdown in Shijiazhuang. To simplify the discussion, the datasets during and after CNY will be combined and marked as during & after CNY in the following section because both periods experienced reduction of anthropogenic activities.

During & after CNY, the reduction ratios of NO, NO₂, NO_x, SO₂, CO, CO₂, PM_2.5, OC, EC and nitrate were 86.4%, 60.5%, 72.2%, 20.6%, 36.3%, 7.7%, 29.8%, 23.1%, 41.0% and 23.9%, respectively, compared with those before CNY (Fig. 2E). When the background concentrations of CO (235 ppbv) (Kim et al., 2010) and CO₂ (413 ppm, www.esrl.noaa.gov) were taken into consideration, reduction fraction of CO and CO₂ were 42.2% and 47.1%, respectively. At the same time, the concentration of NH₃ and O₃ increased 7.2% and 71.4%, respectively, during & after CNY compared with before CNY. The visibility increased 54.3% due to the improvement of air quality. The diffusion ability of air masses also increased obviously (~24.1%). For example, the wind speed increased 25.5% and the PBL height increased 22.7% during & after CNY compared with those before CNY. When all of these measured sulfur species (SO₂ and sulfate), nitrogen species (NO, NO₂, HNO₃, HONO and nitrate) and carbon species (CO, CO₂, OC and EC) were taken into consideration, the reduction ratios of total measured sulfur, nitrogen and carbon were 22.7%, 62.8% and 40.8%, respectively, during & after CNY compared with before CNY. This means that ~40% of total nitrogen and ~20% of total carbon reduction might be related to decrease of anthropogenic activities, while reduction of total sulfur might be mainly related to improvement of diffusion ability during & after CNY overlapping COVID-19 lockdown. According to the emission inventory of Hebei province (MEIC v1.3) (Li et al., 2017), transportation sector accounted for 32.6%, 2.6% and 9.1% of NO_x, SO₂ and carbon species emission in 2016, respectively, while the corresponding contribution from industry sector were 51.5%, 72.3% and 61.1%. If we assume that NO measured at our observation site is mainly from traffic emission and the emission profile of NO_x in Shijiazhuang is the same as that in Hebei province, the lockdown during & after CNY overlapping COVID-19 may reduce ~62% (=86.4%-24.1%) of NO_x from traffic emissions and ~ 22% of NO_x from industry emissions after the improvement of diffusion ability has been considered.

3.2. Influence of COVID-19 lockdown on HONO concentration

As shown in Fig. 1B, elevated HONO concentrations were frequently observed during the observation. HONO well followed the PM_2.5 pollution events before CNY. Increase of HONO concentration was still discernable in each PM_2.5 pollution event during & after CNY although it was significantly lower than that before CNY. Throughout the observation, the HONO concentration ranged from 0.08 to 12.67 ppbv, with a mean value of 1.64 ± 1.41 ppbv. The maximum hourly value of 12.67 ppbv was recorded in the morning of January 11, 2020. The mean HONO concentration was 2.43 ± 1.08 ppbv before CNY, which was significantly higher than that during CNY (1.53 ± 1.16 ppbv) and after CNY (0.97 ± 0.76 ppbv). 49.7%, 22.4% and 9.1% of hourly mean values exceeding 2.0 ppbv before, during and after CNY, respectively. The nighttime mean HONO concentration was 1.96 ± 1.32 ppbv, compared with that in daytime (1.26 ± 1.43 ppbv) throughout the observation. Before CNY, their corresponding nighttime and daytime mean values were 2.84 ± 1.36 and 1.93 ± 1.76 ppbv, while they were 2.84 ± 1.08 and 1.09 ± 1.10 ppbv during CNY, and 1.19 ± 0.69 and 0.71 ± 0.76 ppbv after CNY.

The mean HONO concentrations before CNY was higher than those reported in foreign cities (Acker et al., 2006; Crilley et al., 2016; Lee et al., 2016; Levy et al., 2014; Michoud et al., 2014; Shon et al., 2007) and several Chinese cities such as Guangzhou(Hu et al., 2002), Hongkong (Xu et al., 2015), Ji'nan (Li et al., 2018) and Xi'an (Huang et al., 2017) as summarized in Table 1 , while it was close to these previous observations performed in Beijing (Zhang et al., 2019e) and Shanghai (Cui et al., 2018). Even at noontime (11:00–13:00), the mean HONO concentration was 1.13 ± 1.21 ppbv throughout the observation, and 1.83 ± 1.52, 0.86 ± 0.51, 0.56 ± 0.46 ppbv before, during and after CNY, respectively. Before CNY, 66.7% of the noontime HONO concentrations were higher than 1.00 ppbv, which was the highest level compared with that ever reported in the urban atmospheres. The high HONO concentration observed in this work indicates the intensive HONO sources and potentially strong atmospheric oxidizing capacity in Shijiazhuang. The lockdown during & after CNY overlapping COVID-19 reduced ~55% of HONO concentration, which is ranked the top three pollutants mostly affected by the lockdown along with NO and NO₂ (Fig. 2E). The reduction ratio was ~31% after the improvement of diffusion ability of air masses was subtracted.

Table 1.

HONO concentrations measured in urban environment.

Observation duration	Location	Instrument	cHONO (ppbv) (median)	References
2001.5–6	Rome, Italy	DOAS	<2	(Acker et al., 2006)
2009.4–5	Houston, USA	ID-CIMS	<1.5(0.19)	(Levy et al., 2014)
2009.7, 2010.1–2	Paris, France	LOPAP	0.01–0.5, 0.01–1.7	(Michoud et al., 2014)
2014.6	York, UK	LOPAP	0.33–1.15	(Crilley et al., 2016)
2014.7–8	London, UK	LOPAP	0.2–1.8	(Lee et al., 2016)
2003.1–2	Kathmandu, Nepal	DOAS	0.15–7.45(1.7)	(Yu et al., 2009)
2004.5–2004.6	Soul, South Korea	WD-IC	1.8 ± 0.4(1.3)	(Shon et al., 2007)
2000.7, 11	Guangzhou, China	GAC	1.0–2.7	(Hu et al., 2002)
2007.8	Beijing, China	WD-IC	1.45 ± 0.58(1.47)	(Spataro et al., 2013)
2011.5, 11	Hongkong, China	LOPAP	0.35 ± 0.30, 0.93 ± 0.67	(Xu et al., 2015)
2014.2–3	Beijing, China	LOPAP	0.28–3.24	(Hou et al., 2016)
2015.7–8	Xi'an, China	LOPAP	1.12 ± 0.97	(Huang et al., 2017)
2006.8–9	Beijing, China	LOPAP, GAC	0.034–3.69	(Yang et al., 2014)
2015.9–10, 2016.1, 4–5, 6–7	Beijing, China	AIM-IC	2.27 ± 1.82; 1.05 ± 89;1.05 ± 0.95; 1.38 ± 0.90	(Wang et al., 2017)
2016.5	Shanghai, China	LOPAP	0.48–5.84(2.31)	(Cui et al., 2018)
2016.12	Beijing, China	LOPAP	3.5 ± 2.7	(Zhang et al., 2019e)
2015.9–2016.8	Jinan, China	LOPAP	1.15 ± 1.07	(Li et al., 2018)
2019.12.15–2020.1.22	Shijiazhuang, China	MARGA	2.43 ± 1.08(2.10)	This work
2019.1.23–2020.2.2			1.53 ± 1.16(1.20)	This work
2019.2.3–2020.2.22			0.97 ± 0.76(0.75)	This work

Fig. 3 shows the diurnal curves of HONO, NO, NO₂, PM_2.5, HONO/NO₂ ratio, CO, O₃ and PBL height on clean days (c _PM2.5 < 50 μg m⁻³) and pollution days (c _PM2.5 ≥ 50 μg m⁻³ and RH < 90%). The black and red lines are the diurnal curves before CNY and during & after CNY, respectively. The shadows indicate the standard errors (±1σ). Overall, the nighttime concentrations of HONO, NO, NO₂, HONO/NO₂ ratio and CO were higher than the corresponding daytime values regardless of the pollution level and the period, while O₃ showed an opposite trend due to increase in UV light after sunrise. PM_2.5 showed relative flat diurnal curves on pollution days although its nighttime concentrations were slightly higher than the daytime concentrations on clean days, in particular, during & after CNY. The PBL height also showed relative flat diurnal curves before CNY, while the slightly elevated PBL heights in daytime were discernable during & after CNY.

Fig. 3.

Diurnal curves of (A) and (B) HONO, (C) and (D) NO, (E) and (F) NO₂, (G) and (H) PM_2.5, (I) and (J) HONO/NO₂ ratio, (K) and (L) CO, (M) and (N) O₃, (O) and (P) PBL height on clean days (c_PM2.5 < 50 μg m⁻³) and pollution days (c_PM2.5 ≥ 50 μg m⁻³ and RH < 90%). The black lines are the diurnal curves before CNY and the red ones are during & after CNY. The shadows indicate standard errors (±1σ). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

The diurnal variation of HONO before CNY was similar to NO, while this similarity disappeared during & after CNY. Their similar nighttime profiles suggest that vehicle emissions may pose a significant effect on the measured HONO level (Li et al., 2018) before CNY, while other processes such as secondary conversion should become prominent during & after CNY. This was consistent with the fact that high concentration of HONO before CNY was highly correlated with NO concentration (Fig. 1B), while it was well correlated with the HONO/NO₂ ratio during the whole observation (Fig. S4D). The HONO/NO₂ ratio also showed similar diurnal curve of HONO on pollution days, which implies secondary conversion of NO₂ should play an important role in the observed HONO levels. Besides the first peak of HONO/NO₂ in the morning, which is determined by the complex effect among evolution of boundary layer, traffic emission and the photolysis loss of HONO, a second peak of HONO/NO₂ at around noontime (13:00–14:00) during & after CNY was observable on both clean and pollution days. This implies that additional HONO sources may be related to the solar radiation intensity (Lee et al., 2016; Li et al., 2018).

As shown in Fig. 3, diurnal curves of HONO, NO and NO₂ during & after CNY were significantly below those before CNY on both clean days and pollution days. It did so for CO only on pollution days. The diurnal curves of O₃ during & after CNY were significantly above those before CNY. These results further confirmed the reduction of HONO, NO, NO₂ and CO, but the increase of O₃ during & after CNY overlapping COVID-19 lockdown. However, the diurnal variations of PM_2.5 during & after CNY were only slightly below that before CNY. This might be related to the weak increase in PBL height during & after CNY compared with that before CNY (Fig. 3O and P). It should be noted that the mean nighttime and daytime HONO/NO₂ ratio were 0.095 ± 0.051 and 0.075 ± 0.053, respectively, throughout the observation. These values are in the range of the reported HONO/NO₂ in literatures (Elshorbany et al., 2009; Huang et al., 2017; Li et al., 2018; Liu et al., 2014; Tong et al., 2015), but at a high level end. Interestingly, the HONO/NO₂ ratio on both clean days and polluted days during & after CNY were slightly higher than the counterpart. This implies that chemical conversion from NO₂ to HONO should be more effective during & after CNY compared with that before CNY. These results also mean that the relative contribution of HONO sources should have changed in different periods.

3.3. Influence of COVID-19 lockdown on HONO sources on pollution days

Fig. 4 compared the budget of HONO before CNY with those during & after CNY on pollution days (with PM_2.5 concentration > 50 μg m⁻³ and RH <90%). Table S1 summarized the mean intensities of these sources and sinks. Overall, the nighttime sources were comparable with the sinks in different periods, while significant underestimation of the sources presented in the daytime. Besides the possible unknown HONO sources, a possible reason might be related to the variation of PBL height. When we calculating the HONO sources, heterogeneous conversion from NO₂ to HONO and emissions from vehicle and soil were dependent on PBL height according to Eqs. (2), (7). If these sources were more sensitive to surface HONO concentration (Fig. S4B) than NO₂, the increase in PBL height in daytime should underestimate their contribution to HONO sources. On the other hand, OH concentration was estimated based on J _O1D and J _NO2. Although the calculated OH concentrations were comparable with that observed in Beijing (Tan et al., 2018), the uncertainty of OH concentration might also contribute the underestimation of HONO sources. In addition, light-enhanced heterogeneous reactions on both aerosol (Liu et al., 2014) and ground surfaces (Liu et al., 2019) have not been considered in this work. This might lead to the observed underestimation. Actually, the contribution of photolysis of nitrate became prominent at noontime, in particular, during & after CNY. This decreased the fraction of the unknown HONO sources. In the following section, we will mainly concentrate on the nighttime HONO sources. However, the unknown sources will also be taken into consideration when we calculating the relative contributions if the total sinks are over the total sources.

Fig. 4.

HONO budget before CNY and during & after CNY on polluted days with PM_2.5 concentration > 50 μg m⁻³ and RH <90%.

In the night, heterogeneous reaction of NO₂ on ground surface was the largest HONO source, followed by heterogeneous reaction on aerosol surface, vehicle emission, homogeneous reaction between NO and OH and soil emission in the whole observation period (Fig. 4). Emission from soil was a minor HONO source during our observation due to the low temperature. E _soil varied from 0.00044 to 0.0091 ppbv h⁻¹, with a mean value of 0.0026 ± 0.0014 ppbv h⁻¹ throughout the observation. It was slightly higher during & after CNY (0.0031 ± 0.0016 ppbv h⁻¹) than that before CNY (0.0020 ± 0.0008 ppbv h⁻¹) because of the increase in temperature (Fig. 1C). The nighttime E _soil were 0.0020 ± 0.0007 ppbv h⁻¹ before CNY and 0.0030 ± 0.0015 ppbv h⁻¹ during & after CNY, respectively. The corresponding relative contributions to nighttime HONO sources were (0.27 ± 0.08) % and (0.83 ± 0.36) %. Direct emission from vehicle exhaust was an important HONO source. E _vehicle ranged from 0.015 to 0.45 ppbv h⁻¹, with a mean value of 0.099 ± 0.078 ppbv h⁻¹ throughout the observation. It decreased from 0.16 ± 0.081 ppbv h⁻¹ before CNY to 0.052 ± 0.022 ppbv h⁻¹ during & after CNY due to the reduction of vehicle emission. The corresponding nighttime values were 0.16 ± 0.078 and 0.051 ± 0.021 ppbv h⁻¹, which contributed 22.2 ± 6.7% (BCNY) and 14.7 ± 5.9% (CNY&ACNY) to HONO sources. These relative contributions were close to that in Ji'nan (12%–21%) (Li et al., 2018), while they were much lower than that in Beijing (~50%) (Liu et al., 2020; Meng et al., 2019; Zhang et al., 2019e). This can be well explained by the fact that the vehicle population in Shijiazhuang is much smaller than that in Beijing (http://www.stats.gov.cn). Because the NO concentrations were very high before CNY (Fig. 1), homogeneous reaction between NO and OH played an important role in HONO source in the night. The nighttime P _NO-OH was 0.076 ± 0.091 ppbv h⁻¹ before CNY, while it decreased to 0.0092 ± 0.014 ppbv h⁻¹ during & after CNY. Their corresponding fractions in the nighttime HONO sources were 10.8 ± 6.0% and 2.2 ± 1.1%. Compared with above sources, heterogeneous conversion of NO₂ on aerosol and ground surfaces were the most important nighttime HONO source in Shijiazhuang. This is consistent with the high HONO/NO₂ as shown in Fig. 3. The nighttime P _aerosol were 0.14 ± 0.08 ppbv h⁻¹ before CNY, which was only second to the P _ground (0.30 ± 0.12 ppbv h⁻¹). During & after CNY, they were 0.069 ± 0.055 ppbv h⁻¹ (P _aerosol) and 0.20 ± 0.12 ppbv h⁻¹ (P _ground), respectively. Thus, heterogeneous reaction of NO₂ on aerosol surface contributed 19.0 ± 3.9% to the nighttime HONO sources before CNY. It increased to 21.2 ± 5.7% during & after CNY due to the decrease of E _vehicle and P _NO-OH. The relative contribution of ground surface increased from 38.9 ± 7.8% before CNY to 48.0 ± 13.1% during & after CNY. Photolysis was the major loss of HONO in the daytime. Before CNY, it was 2.65 ± 2.77 ppbv h⁻¹, while it decreased to 1.59 ± 1.86 ppbv h⁻¹ during & after CNY. In the night, vertical and horizontal transport dominated the sinks of HONO, with a mean value of 0.60 ± 0.31 ppbv h⁻¹ before CNY and 0.26 ± 0.19 ppbv h⁻¹ during & after CNY.

Except for E _soil (increased 50%), the reduction ratio of all other sources and sinks ranged from 33% to 88% in the night during & after CNY when compared with that before CNY as shown in Table S1. In the daytime, besides E _soil, P _nitrate and L _HONO-OH increased 18% and 67%, respectively, whereas other sources and sinks decreased from 40% to 68%. The increase of E _soil was related to increase in temperature, while it was related to increase in irradiation intensity for P _nitrate and increase in OH concentrations for L _HONO-OH. In the night, vehicle related sources (E _vehicle and P _NO-OH) were the mostly influenced sources, while heterogeneous conversion (in particular, on ground surface) was less affected by the reduction of anthropogenic activities during and after CNY overlapping COVID-19. Therefore, this led to an increased relative contribution of heterogeneous reaction on ground and aerosol surfaces in the night (Fig. 4). This is consistent with the good correlation between HONO concentrations and NO, NO₂ and PM_2.5 concentrations as shown in Fig. S7.

As pointed above, the k _het during & after CNY (0.017 ± 0.003 h⁻¹, P < 0.05) was significantly higher than that before CNY (0.012 ± 0.005 h⁻¹). We further derived the uptake coefficient of NO₂ (γ_NO2) on both aerosol and ground surfaces according to,

$$k_{\mathit{het}}=\frac{\mathit{S\omega }{\gamma }_{{\mathit{NO}}_2}}{4}$$ (12)

$$\omega =\sqrt{\frac{8\mathit{RT}}{\mathit{\pi M}}}$$ (13)

where, S is the total surface to volume ratio of ground and aerosols (S _g + S _a, m⁻¹), ω is the mean velocity of NO₂ molecules, R is the ideal gas constant, T is the temperature (K), M is the molecular weight of NO₂ (kg mol⁻¹). Fig. 5A compared the derived nighttime γ_NO2 before CNY with that during & after CNY. It was 6.43 ± 9.02 × 10⁻⁶ before CNY, which is significantly (P < 0.05) lower than 1.20 ± 1.76 × 10⁻⁵ during & after CNY. These values are larger than that derived in Ji'nan (1.4 ± 2.4 × 10⁻⁶) (Li et al., 2018) and the laboratory values (5 × 10⁻⁹ - 9.6 × 10⁻⁶) on different particles (Ndour et al., 2009; Underwood et al., 1999; Underwood et al., 2001). The elevated γ_NO2 means that reaction kinetic is favorable of heterogeneous conversion from NO₂ to HONO during & after CNY although both NO₂ and PM_2.5 concentrations decreased compared with before CNY. As pointed in Section 2.2, direct emission of HONO from soil was not considered when we calculating the k _het. This will lead to an additional uncertainty for the γ_NO2. However, this uncertainty should be small because the emission rate of HONO from vehicle (0.099 ± 0.078 ppbv h⁻¹) was as ~38 times as that from soil (0.0026 ± 0.0014 ppbv h⁻¹).

Fig. 5.

(A) The derived nighttime γ_NO2 before CNY and during & after CNY, the dependence of the γ_NO2 on (B) the temperature, (c) the ratio of AWC/PM_2.5 and the aerosol pH. The box plots in (B)–(D) are the distribution of nighttime temperature, AWC/PM_2.5 and aerosol pH before CNAY and during & after CNY.

As shown in Fig. 5B, the γ_NO2 is well negatively correlated with the reciprocal of temperature (lnγ_NO2 = −2.46 × 10³/T-2.69, R = 0.92). This means the conversion from NO₂ to HONO requires to overcome an apparent activation energy (20.5 kJ mol⁻¹). Thus, the higher nighttime temperature (278.0 ± 4.1 K) during & after CNY was favorable of this reaction compared with that before CNY (272.7 ± 2.6 K). On the other hand, the γ_NO2 was negatively correlated with both the ratio of aerosol water content to PM_2.5 mass concentration (AWC/PM_2.5) and the aerosol acidity (pH), which were calculated using the ISORROPIA-II model with the gas phase and particle phase concentrations of the relevant species, the temperature and the RH as inputs in a forward mode under metastable conditions (Ding et al., 2019). The corresponding equation is γ_NO2 = 1.28 × 10⁻⁵-6.71 × 10⁻⁶ AWC/PM_2.5 (R = 0.71) and γ_NO2 = 1.79 × 10⁻⁵-1.24 × 10⁻⁶ pH (R = 0.57) (Fig. 5C and D). Both the nocturnal AWC/PM_2.5 ratio (0.31 ± 0.27) and aerosol pH (4.87 ± 0.89) during & after CNY were significantly lower than those values (0.35 ± 0.31 and 5.17 ± 1.19) before CNY. This can be explained by the higher SIA fraction (Fig. S3) and the lower RH during & after CNY. These results are well agreement with previous laboratory studies that water can competitively inhibit the uptake of NO₂ (Liu et al., 2015; Zhang et al., 2012b) and particle phase acids can promote HONO formation from heterogeneous reaction of NO₂ (Bao et al., 2018; Han et al., 2016). This means the variation of aerosol properties might be another reason lead to the elevated reactivity for heterogeneous conversion from NO₂ to HONO during & after CNY. Thus, the increase in the relative contribution of heterogeneous reactions on ground and aerosol surfaces can be explained by both the effective heterogeneous conversion from NO₂ to HONO and the reduction of direct emission from vehicle.

4. Conclusions

During & after CNY overlapping COVID-19 lockdown, great reduction of air pollutants emission was observed in Shijiazhuang. The reduction ratio of NO, NO₂, NO_x, SO₂, CO, CO₂, PM_2.5, OC, EC and nitrate concentrations were 86.4%, 60.5%, 72.2%, 20.6%, 42.2%, 47.1%, 29.8%, 23.1%, 41.0% and 23.9%, respectively, compared with those before CNY. At the same time, the diffusion ability of air mass increased obviously as evident by the increase in wind speed (25.5%) and PBL height (22.7%) during & after CNY. The lockdown during & after CNY overlapping COVID-19 might reduce ~62% of NO_x from traffic emissions after improvement of diffusion ability was taken into consideration.

The mean HONO concentration was 2.43 ± 1.08 ppbv before CNY, which was significantly higher than those during CNY (1.53 ± 1.16 ppbv) and after CNY (0.97 ± 0.76 ppbv). The lockdown during & after CNY overlapping COVID-19 reduced ~31% of ambient HONO compared with that before CNY after the improvement of diffusion ability was considered. Heterogeneous reaction of NO₂ on ground surface was the largest nocturnal HONO source, followed by heterogeneous reaction on aerosol surface, vehicle emission, homogeneous reaction between NO and OH and emission from soil in the observation period. Except for emission from soil, most of the HONO sources and sinks decreased from 33% to 88% in the night during & after CNY when compared with that before CNY. The relative importance of heterogeneous reaction of NO₂ on surfaces further increased during & after CNY because of both the decrease in vehicle emission and the increase in heterogeneous reaction kinetics compared with that before CNY. This work confirmed that reducing anthropogenic activities definitely reduced ambient HONO concentrations, while it nonlinearly responded to the reduction of anthropogenic activities.

CRediT authorship contribution statement

Yongchun Liu: Conceptualization, Methodology, Data curation, Writing - original draft. Shuangying Ni: Conceptualization, Methodology, Investigation, Writing - original draft. Tao Jiang: Investigation, Resources. Shubin Xing: Investigation, Resources, Writing - review & editing. Yusheng Zhang: Investigation, Data curation. Xiaolei Bao: Conceptualization, Methodology, Data curation, Investigation, Resources, Writing - original draft. Zeming Feng: Investigation, Data curation. Xiaolong Fan: Investigation, Data curation. Liang Zhang: Investigation, Data curation. Haibo Feng: Writing - review & editing, Funding acquisition.

Declaration of competing interest

The authors declare that they have no conflict of interest.

Editor: Jianmin Chen

Appendix A. Supplementary data

Supplementary material