Year-long Concurrent MAX-DOAS Observations of Nitrogen Dioxide and Formaldehyde at Pune: Understanding Diurnal and Seasonal Variation Drivers

ABSTRACT  
 
Year-long observations of nitrogen dioxide (NO2) and formaldehyde (HCHO) using the Multi-Axis Differential Absorption Spectroscopy (MAX-DOAS) technique are reported from Pune City, India. We studied the diurnal and seasonal variations, effect of biomass burning and the weekend effect on both species. NO2 diurnal profiles displayed a traffic induced peak at ~09:00 hrs. HCHO also showed a morning peak ~10:00 hrs due to production from oxidation of VOCs in the presence of solar radiation. Both NO2 and HCHO show the highest average concentrations during the winter (October, November, December, January and February -ONDJF), with mean mixing ratios of 2.0 ± 1.4 ppb and 3.0 ± 1.4 ppb, respectively. These observations suggest that a lower boundary layer (BL) height during ONDJF leads to higher concentrations of trace gases. During June, July, August, and September (JJAS), both trace gases show a minimum in their concentrations, with average mixing ratios for NO2 and HCHO being 0.9 ± 0.6 ppb and 1.1 ± 0.7 ppb, most likely due to removal by wet deposition. There was no significant difference in both the trace gases on weekdays and weekends. Using back-trajectory analysis, we conclude that air parcels coming from regions of biomass burning increased the concentrations in Pune. Emissions from nearby industrial areas of Bhosari and Pimpri-Chinchwad increased NO2 concentrations in Pune city. Finally, we compared the observations with previous reports over India and found that both HCHO and NO2 concentrations are lower in Pune compared to the other large cities in India.


A C C E P T E D M A N U S C R I P T
INTRODUCTION 32 Reactive trace gases make up less than 0.1% of the atmosphere but affect it in numerous 33 ways. Some trace gases act as greenhouse gas and contribute to climate change by changing previous reports has been presented in Table 2. 83 In this work, we study tropospheric HCHO and NO2 over a one-year period from January-

A C C E P T E D M A N U S C R I P T
6 The study was conducted in Pune City, India from 2 nd January 2018 to 8 th February 2019. 105 Pune is the 8 th most populous city in India and its population has grown ~35% to 5.05 million    For detailed discussion on errors from DOAS analysis please refer to Platt and Stutz (2008).

172
The mixing ratios of HCHO and NO2 were calculated using the respective trace gas and O4 to be associated to 'high fire' events. Grid points with fire counts less than median value were 186 associated with 'low fire' events.

187
Back-trajectory analysis 188 Back-trajectory analysis for air parcels reaching the observation site was carried out using hours.

195
The MAX-DOAS instrument was operational for the period between 2 nd January, 2018 and

A C C E P T E D M A N U S C R I P T
September (JJAS) and October to February (ONDJF) respectively. Henceforth in this paper, 201 the three seasons will be abbreviated as the initials of the months mentioned above. top panel in Fig. 3 shows the time series of O4 dSCDs. molec. 2 cm -5 (ranging from 5.8 x 10 42 molec. 2 cm -5 to 3.0 x 10 43 molec. 2 cm -5 ) and 3.9 x 10 -4 .

207
During JJAS and ONDJF, the average O4 dSCDs at the 1° elevation angle were 1.7 x 10 43 ±  respectively. NO2 mixing ratios were calculated using the O4 method. Fig. 4a shows the 221 mixing ratios of NO2 measured during the current study. The average NO2 mixing ratio and 222 detection limit during MAM were 1.6 ± 1.0 ppb (ranging from 0.4 to 7.8 ppb) and 0.1 ppb

234
The average RMS of residual structure at the 1° elevation angle for MAM, JJAS and ONDJF 235 were 4.8 x 10 -4 , 6.2 x 10 -4 and is 5.0 x 10 -4 respectively. The average HCHO mixing ratio and The diurnal profiles of O4 dSCDs showed a 'U' shape with two dSCDs maxima, early the 247 morning and late evening, during ONDJF. Fig. S1(a) shows a typical day from ONDJF (7 th and reach a maximum at 10:00 hrs in the morning (4.2 ± 1.4 ppb in ONDJF; 3.7 ± 1.5 ppb in gradually until the late afternoon during ONDJF and JJAS, reaching minimum values of 1.8 ± 299 0.7 ppb and 0.7 ± 0.3 ppb respectively. The main reason for this is that in the morning, due to 300 the presence of solar radiation, VOCs get oxidized to HCHO leading to an increase. Later in 301 the day, with increasing solar radiation the photo-dissociation of HCHO also starts increasing, 302 leading to a loss of HCHO. These two competing processes result in a decrease after a peak at 303 10:00 hrs in the morning.

304
The average diurnal mixing ratios are lowest during JJAS. Due to wet deposition during 305 monsoon, the HCHO is lower. The average HCHO mixing ratios during ONDJF were the 306 highest among all the three seasons, which as explained earlier is due to the low boundary 307 layer heights in winter compared to the summer months of MAM.

A C C E P T E D M A N U S C R I P T
18 spent at least 60% of its time in last 12 hours and 24 hours in any sector. The hypothesis is 398 that if an air parcel spends 60% or more of its time in a sector, the air parcel will carry 399 representative information from that sector compared to the other.

A C C E P T E D M A N U S C R I P T
During MAM the prevailing wind direction is north-westerly. For 12 hour back-trajectories, 423 40.3% of the total NO2 observations travel over sector 7, followed by sector 8 (18.5%) and 424 sector 6 (7.7%). For both the 24 hour (Fig. 8B) and 12 hour (Fig. 8E)  circulation. This is seen in the back-trajectory data, with about 42.8% of the air parcels 433 passing over sector 6, followed by sector 7 (23.7%) and sector 5 (6.0%) for the 24 hour back-434 trajectories. For the 24 hour back-trajectories (Fig. 8C), sector 7 shows highest median value.

435
For the 12 hour back-trajectories sector 8 shows highest median values (Fig. 8F). Although   Fig. 9 shows a box-whisker plot for the sector-wise analysis of HCHO. Fig. 9A

477
Previous observations of NO2 in India are presented in Table 1. We have not discussed  Table 1, the annually average NO2 mixing ratio in Pune is lower than the urban sites induced morning peak (at ~ 10:00 hrs) compared to the present study. Also biomass (crop

A C C E P T E D M A N U S C R I P T
22 residue) burning increases NO 2 mixing ratios in north-west India, whereas the impact is less 497 in Pune.

498
Compared to NO2, relatively fewer observations of HCHO have been made over India (Table   499 2). The annual average HCHO mixing ratio in Pune was lower than the megacity of Kolkata NO2 was found at ~9:00 hrs resulting from automobile emissions during peak traffic hours.

517
HCHO showed a morning peak at ~10:00 hrs, which is due to emissions and the 518 photochemical oxidation of VOCs. We found that HCHO/NO2 ratio remains in VOC limited 519 region for O3 formation during 07:00 -09:00 hrs but the ratio is in the border regime for rest