key: cord-0996489-mq0wrr49 authors: Ye, Fei; Rupakheti, Dipesh; Huang, Lin; T, Nishanth; Kumar Mk, Satheesh; Li, Lin; Kt, Valsaraj; Hu, Jianlin title: Integrated process analysis retrieval of changes in ground-level ozone and fine particulate matter during the COVID-19 outbreak in the coastal city of Kannur, India() date: 2022-05-16 journal: Environ Pollut DOI: 10.1016/j.envpol.2022.119468 sha: 10d3c93a2b4056f28c3c93096182624c8bc293e2 doc_id: 996489 cord_uid: mq0wrr49 The Community Multi-Scale Air Quality (CMAQ) model was applied to evaluate the air quality in the coastal city of Kannur, India, during the 2020 COVID-19 lockdown. From the Pre1 (March 1–24, 2020) period to the Lock (March 25–April 19, 2020) and Tri (April 20–May 9, 2020) periods, the Kerala state government gradually imposed a strict lockdown policy. Both the simulations and observations showed a decline in the PM(2.5) concentrations and an enhancement in the O(3) concentrations during the Lock and Tri periods compared with that in the Pre1 period. Integrated process rate (IPR) analysis was employed to isolate the contributions of the individual atmospheric processes. The results revealed that the vertical transport from the upper layers dominated the surface O(3) formation, comprising 89.4%, 83.1%, and 88.9% of the O(3) sources during the Pre1, Lock, and Tri periods, respectively. Photochemistry contributed negatively to the O(3) concentrations at the surface layer. Compared with the Pre1 period, the O(3) enhancement during the Lock period was primarily attributable to the lower negative contribution of photochemistry and the lower O(3) removal rate by horizontal transport. During the Tri period, a slower consumption of O(3) by gas-phase chemistry and a stronger vertical import from the upper layers to the surface accounted for the increase in O(3). Emission and aerosol processes constituted the major positive contributions to the net surface PM(2.5), accounting for a total of 48.7%, 38.4%, and 42.5% of PM(2.5) sources during the Pre1, Lock, and Tri periods, respectively. The decreases in the PM(2.5) concentrations during the Lock and Tri periods were primarily explained by the weaker PM(2.5) production from emission and aerosol processes. The increased vertical transport rate of PM(2.5) from the surface layer to the upper layers was also a reason for the decrease in the PM(2.5) during the Lock periods. (Resmi et al., 2020). To clarify the reasons for the increasing trends in the O3 concentrations and the 151 descending trends in the other air pollutants during the lockdown and triple-lockdown" period, the 152 integrated process rate (IPR) analysis embedded in the CMAQ is employed to evaluate the 153 contributions of individual atmospheric processes, such as gas-phase chemistry, dry deposition, 154 cloud processes, aerosol processes, emissions, vertical transport, and horizontal transport. The CMAQ model version 5.2 was applied with the SAPRC07 gas-phase photochemical 189 mechanism (Carter, 2010) and AERO6 aerosol reaction mechanism (Binkowski and Roselle, 2003) Where N represents the total number of the data; Pi is the i th predicted value; Oi is the i th 231 observed value. where N, Pi, and Oi are consistent with that mentioned earlier. considering all the simulated data during the different periods (indicated by All in Table 1 ), the NMB 312 value was slightly overestimated but also within the benchmark. In general, the PM2.5 concentration 313 simulations reasonably compared with the observed concentrations during the predicted periods, 314 except during the Pre1 period. Although the NMB value during Pre1 exceeded the benchmark, the 315 NMB and NME values during the other periods all satisfied the benchmarks, and the NMB values 316 of all the model predictions for PM2.5 from March 1 to May 17 were lower than the benchmark. 317 Note that the PM2.5 concentrations during three periods were underpredicted as indicated by the 318 negative NMB values. The modeled PM10 concentrations were also underestimated during all the 319 periods. The statistical values of the NMB and NME for the other model species were nearly less 320 than one, except for NO2. Uncertainties in the observational data, meteorology, emission inventory, 321 and the scaling factors could be possible reasons for the discrepancy between the model predictions 322 and observations. Overall, the model provided comparable simulations for trace pollutants over 323 Kannur during the COVID-19 outbreak. 324 The O3 concentrations at the surface layer during the Lock and Tri periods exceeded the O3 387 concentration during the Pre1 period. Compared with the Pre1 period, O3 enhancement during the 388 Lock period was primarily attributed to the lower negative contribution of the photochemistry 389 processes rather than O3 titration by higher NOx emissions as during the Pre1 period. The lower O3 390 removal rate by horizontal transport also accounted for a considerable portion of the O3 increase 391 during the Lock period. For the Tri period, gas-phase chemistry also exhibited a slower consumption 392 of O3 as compared to the Pre1 period, during which more O3 was removed due to titration by higher 393 NOx emissions. Another possible reason for the higher O3 levels during the Tri period was that the 394 upper boundary layer had a stronger vertical transport import of surface O3 than that during the Pre1 395 From the perspective of the entire planetary boundary layer (Fig. 5b) , the interpretation of 397 concentration changes during the COVID-19 outbreak was slightly different. The O3 enhancement 398 during the Lock and Tri periods were both related to the increased O3 vertical transport. For the Tri 399 period, a weaker photochemical removal still made sense for the O3 increase, while the contributions 400 of photochemistry during the Pre1 and Lock periods were comparable. It was concluded that 401 contributions of the individual atmospheric processes varied in the upper layers above the ground. 402 The mean hourly O3 change rates due to various atmospheric processes for layers 1 to 10, as well is predominantly attributed to strong gas-phase photochemistry, and O3 import at the surface layer 416 primarily originates from vertical transport. 417 periods. Additionally, the O3 peak appeared 1-2 h later than that during the lockdown. It is worthy 444 to note that the delay in O3 concentrations was generally in good agreement with the delay in the 445 vertical transport contributions, and this further confirmed the dominant influence of vertical 446 transport on the surface O3 input over Kannur. During the Pre1 period, the photochemical titration 447 by high NOx emissions was stronger, and the photochemical O3 generation was weaker than those 448 during the other two periods, making the biggest average negative effect of photochemistry appear 449 during the Pre1 period relative to the Lock and Tri periods. 450 Photochemistry made a small contribution to the surface O3 formation; nevertheless, it showed 451 a greater contribution to O3 generation in the entire boundary layer (as shown in Fig. 6b) and reasons for O3 changes during the lockdown were compared between this study and previous 471 studies in Table 2 . The average O3 increase (23.3%) for Lock and Tri days compared to Pre1 days 472 was slightly lower than 29% stated by Kumari and Toshniwal (2020), lower than 37% reported by 473 during the three periods were not as much as expected. Photochemistry during all three periods had 514 a net negative contribution to PM2.5, and these negative effects became obvious with tightened 515 restrictive measures. Considering the weak effects of cloud processes on PM2.5, we do not discuss 516 these two processes in this paper. It is worth noting that the PM2.5 concentrations during both the 517 Lock and Tri periods were lower than that during the Pre1 period, and this was primarily explained 518 by the weaker PM2.5 production from emission and aerosol processes. The increased PM2.5 vertical 519 transport rate from the surface level to the upper level was also a reason for the PM2.5 decrease 520 during the Lock period. 521 As Fig. 8b shows, from the perspective of the entire planetary boundary layer, the aerosol 522 process, vertical transport, and emissions were the major contributors to PM2.5. The net effects of 523 the vertical and horizontal transports within the boundary layer were opposite to that at the surface 524 layer, indicating a complicated distribution of transport effects in the upper layers. Therefore, the 525 hourly PM2.5 change rates due to the various atmospheric processes for layers 1 to 10 and the 526 evolution of the PM2.5 vertical profiles during the three periods are displayed in Fig. S3 to evaluate 527 the vertical distributions. As Fig. S3 displays, the PM2.5 emissions only existed within the first three 528 layers, and this was related to the height of the emission source. In the first three layers, the 529 horizontal transport and aerosol processes were the other two important sources of PM2.5, while 530 vertical transport was the major sink for removing the near-ground PM2.5. Dry deposition only 531 occurs at the first layer and serves as another sink for PM2.5. The production rate of PM2.5 via the 532 aerosol process decreases as the vertical layer rises. Vertical transport contributes positively to the 533 upper layers and negatively to the lower layers, while horizontal transport has the opposite effect 534 that vertical transport does. This results in a vertical export and a horizontal import at the surface 535 layer. This may have been related to the local air circulation existing over Kannur. The PM2.5 536 concentration had the highest value at the surface layer and decreased as the vertical layer increased 537 for layers 1 to 6, and this could have been attributed to the contribution of primary emissions and 538 the aerosol process. 539 respectively. Although the PM2.5 trends during the three periods were similar, the magnitudes of the 557 PM2.5 changes and the contributions to the net PM2.5 differed during the three periods. The primary 558 difference is that the two peaks of PM2.5 during the Pre1 period (88.5 μg/m 3 at 0800 LST; 60.5 μg/m 3 559 at 1200 LST) were higher than that during the other two periods. This result was primarily attributed 560 to the higher positive effects of the aerosol process and the higher PM2.5 emissions during the Pre1 561 period. 562 Table 2 However, there were discrepancies between the results of this study and other results in Table 2 A sensitive simulation was conducted to assess which factor (meteorology or emission) has a 582 more significant impact on air quality changes during the lockdown. In the sensitive simulation, the 583 emissions were not adjusted with the scaling factors (denoted as 'Base' in the following description), 584 and the meteorology was kept the same as in the Mod case. Therefore the difference between Base 585 and Mod was due to emission changes during the lockdown. The contributions of emissions 586 reductions were calculated by subtracting the value of Base with Mod in the corresponding period. 587 The absolute differences between Lock and Pre1, as well as between Tri and Pre1, include 588 contributions from both meteorology changes and emission reductions. Therefore, the contributions 589 of meteorology in Lock and Tri days can be estimated by subtracting the emission contribution from 590 the total absolute difference. This method was used in a previous study (Liu et al., 2020) Mod, the discrepancies between the Lock, Tri, and Pre1 periods, as well as between Mod and Base 597 were very clear. The O3 difference between Base and Mod during Lock was small and more 598 pronounced during the Tri period. As shown in Fig. S5 , emission reductions could influence CHEM 599 more significantly than meteorological variations. Since a slower consumption of O3 by 600 photochemistry was associated with surface O3 increase during the Lock and Tri periods as 601 previously described in section 3.2, it was concluded that the changes in CHEM caused by emission 602 reduction have a large impact on O3 increase during the lockdown. For PM2.5, the decreases in PM2.5 603 concentrations during the Lock and Tri periods were mainly explained by the weaker PM2.5 604 production from emission and aerosol processes as previously described in section 3.3. As Fig. S6 The second uncertainty is associated with the scaling factors (Table S1, The CMAQ model was applied to evaluate the air quality in the coastal city of Kannur, India, during 655 the 2020 COVID-19 lockdown. Both the simulations and observations showed a decline in the PM2.5 656 concentrations and an enhancement in the O3 concentrations during the Lock and Tri periods 657 compared with that in the Pre1 period. The IPR analysis was employed to isolate and quantify the 658 contributions of individual atmospheric processes to explore the causes of O3 and PM2.5 659 concentration changes. The results revealed that the surface O3 enhancements during the Lock and 660 Tri periods were primarily attributable to the weaker negative effect of photochemistry, which was 661 caused by a lesser NOx titration effect during the nighttime and more active photochemical 662 generation during the daytime. In contrast with previous studies, it can be concluded that the O3 663 photochemistry in Kannur was in a VOC-limited condition during the study period and that O3 664 titration with NO was reduced when the NOx emissions were reduced. During the Lock and Tri 665 periods, the surface PM2.5 decrease was primarily caused by the reduced emissions and the aerosol 666 process, and this represents a reduction in the primary source of PM2.5 and a reduction in the 667 secondary aerosol formation via gas-to-particle conversion. Horizontal transport and vertical 668 transport both had a significant effect on the changes in surface O3 and PM2.5. During the Lock 669 period, the surface O3 enhancement was also caused by the lower removal rate by horizontal 670 transport, and the PM2.5 decline was also caused by the higher removal rate by vertical transport. 671 During the Tri period, a stronger vertical import from the upper layer to the surface also accounted 672 for the O3 increase compared with the Pre1 period. This study demonstrates the contribution of 673 individual atmospheric processes and estimating the causes of quality changes due to lockdown 674 measures, and provides insights for regulatory authorities when considering the formulation of 675 The changes in the air quality of Wazirpur, Delhi due to the COVID-19 686 shutdown Rethinking of the adverse effects of NOx-control on the reduction of 688 methane and tropospheric ozone -Challenges toward a denitrified society Evaluation of relationship between meteorological 691 parameters and air pollutant concentrations during winter season in Elazığ Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2): Emergence, 694 history, basic and clinical aspects Surface Ozone and its Precursor 696 Gases Concentrations during COVID-19 Lockdown and Pre-Lockdown Periods in Hyderabad City, India. 697 Environmental Processes On modelling 699 growing menace of household emissions under COVID-19 in Indian metros Models-3 Community Multiscale Air Quality (CMAQ) model aerosol 702 component 1. Model description COVID-19 lockdown-induced changes in NO2 705 levels across India observed by multi-satellite and surface observations Review of the governing equations, computational algorithms, and other 708 components of the models-3 Community Multiscale Air Quality (CMAQ) modeling system Development of the SAPRC-07 chemical mechanism Impact of quarantine measures on chemical 712 compositions of PM2.5 during the COVID-19 epidemic in Characteristics of 714 ozone and particles in the near-surface atmosphere in the urban area of the Yangtze River Delta Estimates and 25-year trends of the 720 global burden of disease attributable to ambient air pollution: an analysis of data from the Global 721 Burden of Diseases Study Variation of trace gases in 723 Ozone pollution during the COVID-19 lockdown in the spring 2020 over Europe 726 analysed from satellite observations, in situ measurements and models Impact of COVID-19 lockdown on NO2 5 and PM10 concentrations and assessing air quality changes in Baghdad Global, continental, and national variation in PM2.5, O3, and NO2 concentrations during the 777 early 2020 COVID-19 lockdown One-year simulation of ozone and particulate matter in China 779 using WRF/CMAQ modeling system Attributable to Particulate Matter in China: Source Contributions and Responses to Reductions Source 784 contributions and regional transport of primary particulate matter in China Numerical simulation and process analysis of 786 typhoon-related ozone episodes in Hong Kong Enhanced secondary pollution offset reduction of primary emissions 790 during COVID-19 lockdown in China Seasonal characteristics of aerosols (PM2.5 and 792 PM10) and their source apportionment using PMF: A four year study over Delhi Process analysis of the impact of atmospheric 795 recirculation on consecutive high-O3 episodes over the Seoul Metropolitan Area in the Korean Peninsula Modeling the impact of COVID-19 on air quality in southern California: implications for future 799 control policies Assessing the real-time thermal performance of 801 reinforced cement concrete roof during summer-a study in the warm humid climate of Kerala The role of emission reductions and the meteorological 860 situation for air quality improvements during the COVID-19 lockdown period in central Europe Lessons for air quality and climate policy Changes in air quality during the lockdown in Barcelona (Spain) one month into the SARS-CoV-2 950 epidemic Photochemical Process Modeling and Analysis of Ozone Generation Modelling air quality during the EXPLORE-YRD campaign -Part I. Model 955 performance evaluation and impacts of meteorological inputs and grid resolutions Assessment of summertime O3 formation and the 958 O3-NOX-VOC sensitivity in Zhengzhou, China using an observation-based model Process analysis and sensitivity study of regional ozone formation over the Pearl River Delta, China, 962 during the PRIDE-PRD2004 campaign using the Community Multiscale Air Quality modeling system Premature Mortality Associated with 965 Exposure to Outdoor Black Carbon and Its Source Contributions in China Associations of daily 967 mortality with short-term exposure to PM2.5 and its constituents in 970 Measurement of the key parameters of VOC emissions from wooden furniture, and the impact of 971 temperature Impact of the COVID-19 lockdown 973 on roadside traffic-related air pollution in Process analysis of typical summertime ozone 975 episodes over the Beijing area Summertime ozone pollution in Sichuan Basin, China: Meteorological conditions, sources and process 978 analysis Trend analysis of surface ozone at suburban 980 Opposite impact of 982 emission reduction during the COVID-19 lockdown period on the surface concentrations of PM2.5 and 983 O3 in Wuhan Secondary organic aerosol from polycyclic aromatic hydrocarbons in Southeast 985 Estimating population exposure to 987 ambient polycyclic aromatic hydrocarbon in the United States -Part II: Source apportionment and 988 cancer risk assessment Impact of reduced anthropogenic 990 emissions during COVID-19 on air quality in India Tri: O3 (+20.7%) Less O3 titration changes between the Lock period and the Pre1 period, as well as the Tri period and the Pre1 period Contributions of the individual processes to the concentrations of O3 (a) at the surface layer 1034 and (b) in the planetary boundary layer during the three periods VTRA, and CONC denote O3 change by gas-phase chemistry, reduction in O3 by dry deposition, 1036 change in O3 by horizontal and vertical transportation, and the hourly O3 (in ppb) respectively. dry deposition, HTRA and VTRA denote PM2.5 change by horizontal and vertical 1051 transportation respectively, AERO denotes PM2.5 change by the aerosol process Diurnal variations for contributions of different processes to PM2.5 formation (a) at the 1056 surface layer and (b) in the planetary boundary layer during the three periods. Abbreviations used 1057 in this figure are the same as in Fig Photochemistry contributes negatively on O3 concentrations at the surface layer 2. Lower chemical consumption and stronger vertical import lead to O3 increase 3. Emission and aerosol processes are the major contributors to surface PM2