Elsevier

Atmospheric Research

Volume 266, March 2022, 105951
Atmospheric Research

A case study of heavy PM2.5 secondary formation by N2O5 nocturnal chemistry in Seoul, Korea in January 2018: Model performance and error analysis

https://doi.org/10.1016/j.atmosres.2021.105951Get rights and content

Highlights

  • N2O5 heterogeneous chemistry was investigated by intervening the uptake coefficient (γN2O5) with WRF-CMAQ.

  • Nighttime nitrate production rate driven by N2O5 chemistrywas analyzed to be 0.8 ±0.4 μg/m3/hr in IPR analysis.

  • Simulations with different uptake coefficients (γN2O5) indicate that improving γN2O5 can contribute to better predicting fine nitrate by 10%.

  • Error ranges from both PBL schemes and uptake process of N2O5 showed approximately ~10% relative to the nighttime-average PM2.5.

Abstract

Heterogeneous hydrolysis of dinitrogen pentoxide (N2O5) plays an important role in nighttime nitrate (NO3) formation in urban areas, and sometimes influences the occurrence of heavy PM2.5 pollution the next day in the Seoul Metropolitan Area (SMA), Korea. Here, we discuss the heavy PM2.5 wintertime episode of January 13–15, 2018, which was mainly induced by nighttime N2O5 heterogeneous reaction in the SMA. In our case, we confirmed that nighttime N2O5 hydrolysis is the most critical factor in the rapid formation of aerosol nitrate at high levels during the night, which prevailed in the morning of the next day. Our Integrated Process Rate (IPR) analysis showed that nighttime nitrate production in the episode was almost solely attributable to N2O5 chemistry, with hourly mean production rates of 0.8 ± 0.4 μg/m3 per hour in SMA, which is comparable to the daytime nitrate photochemical production rates of 0.9 ± 0.5 μg/m3 per hour. We also carried out a series of assessment of N2O5–driven nitrate formation sensitivity, and relevant errors were quantified by applying different N2O5 uptake coefficients in the WRF-CMAQ model. The potential errors of nighttime-average nitrate concentrations induced by N2O5 uptake process were assessed from a linear perspective for the planetary boundary layer (PBL) variances caused by four different PBL parameterization schemes: YSU, ACM2, MYJ, and QNSE. The potential error ranges by N2O5 uptake process were analyzed to be 2.3 to 3.5 μg/m3 (~10% relative to the nighttime-average), while biases of PBL simulations from 4 parameterization schemes were 2.3 ± 1.0 μg/m3, showing similar ranges in our episode. Although N2O5–driven heavy PM2.5 episodes do not occur often in SMA, our findings suggest the importance of N2O5 chemistry in vigorous wintertime nitrate formation and operational prediction errors of such PM2.5 episodes, under the premise of enhanced PBL simulation capabilities.

Introduction

Heavy pollution of PM2.5 (Particulate Matter with the aerodynamic diameter of less than 2.5 μm), particularly secondary PM2.5 is one of the most urgent societal issues in Northeast Asia, and national measures to improve PM2.5 air quality have been implemented in South Korea (Kim et al., 2017a; Kim et al., 2017b). The South Korean government has established a new standard crisis management manual for PM2.5, comprising four different levels (attention, caution, alert, and serious) of alert standards and an associated response system in the event of heavy PM2.5, defined as daily average concentrations consecutively exceeding 50 μg/m3 on multiple days. Because atmospheric processes relevant to heavy PM2.5 are highly complex in urban areas (due to meteorological/chemical uncertainties, urban morphology, etc.), a basic understanding of secondary particle formation has become more important (Lee et al., 2019) and is still a challenge for reliable operational PM2.5 air quality forecasting.

As themain components of air quality forecasting system, meteorology and chemistry are two major factors in determining the PM2.5 concentrations. These two processes are interrelated and sometimes act in a compensatory direction in the comprehensive regional air quality model structure, and thus, numerical sensitivity simulations for reliable predictions are prerequisite. Numerous numerical sensitivity results documented by the National Institute of Environmental Research (NIER) based on failed PM2.5 forecasts have shown that the cause of the forecast uncertainty originating from meteorological variables over the 5 years (2015–2019) could be attributed to the wind speed overpredictions (of more than 50%), as well as other factors, such as planetary boundary layer height, wind direction and temperature (NIER, 2018). In particular, wind speed overprediction has frequently occurred in long-range transport cases, and it also becomes important in in urban-scale air quality predictions for high concentrations in stagnant atmospheric conditions in the Seoul Metropolitan Area (SMA), Korea (Park and Kim, 1999; Kim and Ghim, 2002; Park et al., 2004).

The planetary boundary layer (PBL) refers to the lower atmospheric layer, with a depth that is generally less than 2 km, in which human activities take place. PBL has been reported to play an important role in the vertical distribution of air pollutants (Stull, 1988). Previous studies have shown the differences and inconsistencies originating from the PBL parameterization scheme used in modeling studies (Madala et al., 2015; Kim et al., 2015; Banks and Baldasano, 2016; Mohan and Gupta, 2018; Sarkar et al., 2019; Yang et al., 2021). Moreover, the importance and improvement of PBL simulations by turbulence have been highlighted for air pollution modeling in numerous previous studies (Madala et al., 2015; Kim et al., 2015; Mohan and Gupta, 2018; Yang et al., 2021). For example, Kim et al. (2015) studied the sensitivities of the vertical dispersion of pollutants to different PBL schemes using offline meteorology (WRF) and chemistry-transport (Polair3D/Polyphemus) models, and showed that they influence the PM vertical distributions, not only because they influence vertical mixing (PBL height and eddy diffusion coefficient), but also the horizontal wind fields and humidity. Mohan and Gupta (2018) found that PBL parameterization schemes can also have a significant impact on exploring the physical mechanism of the pollution process and predicting the dynamic variations of pollutants, while PBL-PM2.5 coupled studies were also performed to examine PBL sensitivity (Su et al., 2018; Lee et al., 2019a; Lee et al., 2019b; Li et al., 2021).

On the other hand, the analysis of errors originating from chemical mechanisms is also important. For example, secondary nitrate (NO3) is an important chemical component of PM2.5, and is recognized to be one of the most highly uncertain factors in predicting PM2.5. In the SMA, Korea, the mean concentrations of nitrate have generally been higher than those of sulfate (SO42−) in recent years (NIER, 2017; Jo et al., 2020; Kim et al., 2021), while secondary inorganic aerosol (SIA) species are becoming dominant in PM2.5 (Pathak et al., 2009; Khan et al., 2010; Squizzato et al., 2012; Shin et al., 2016; Seo et al., 2017). Recently, PM2.5 forecasting has also explored considerable uncertainties in the nighttime heterogeneous dinitrogen pentoxide (N2O5) chemistry (Prabhakar et al., 2017).

Nevertheless, the error quantifications of general biases induced by chemical processes (i.e., secondary organic/inorganic formation process) are yet to be assessed and remain highly uncertain in PM2.5 forecasting. In this context, sensitivity analysis of inorganic species, together with observational studies to validate the simulation, are needed to improve the nitrate formation mechanism over Northeast Asia for the improvement of PM2.5 predictions. This is because nitrate is a major component in urban areas, and inaccurate representation of nitrate PM2.5 formation chemistry directly results in severely failed PM2.5 forecasts.

Nitrate aerosols are formed mainly by two atmospheric pathways: (1) the reaction of OH with NO2 in the daytime and (2) the N2O5 heterogeneous hydrolysis during nighttime (Finlayson-Pitts and Pitts Jr., 1997; Ravishankara, 1997; Finlayson-Pitts and Pitts Jr., 2000; Brown et al., 2006a; Brown et al., 2006b). The product of these two (1) and (2) reactions, HNO3, is a limiting reagent and/or will thermodynamically partition to the aerosol phase, depending on the ammonia or other inorganic gaseous species (Franchin et al., 2018; Ibikunle et al., 2020; Nenes et al., 2020).

Many previous studies have pointed out that nitrate formation via N2O5 heterogeneous hydrolysis is important in producing high PM2.5 concentrations, especially during winter, because of the longer nighttime length; thus, the N2O5 uptake coefficient (γN2O5) is an uncertain, but important parameter in N2O5 heterogeneous hydrolysis (Brown et al., 2006a; Baasandorj et al., 2017; Wang et al., 2018). Nevertheless, the extremely high PM2.5 (i.e., up to the alert or serious levels of predictions) induced purely by nighttime N2O5 formation in urban areas, did not frequently occur in SMA in Korea. However, the heavy PM2.5 wintertime episode of January 13–15, 2018, was found to be mainly induced by nighttime N2O5 heterogeneous reaction, inferring from the modeling-based estimation of reaction rates for HNO3 formation and the relevant measurement (NIER, 2018). This made it possible to assess quantitatively the operational PM2.5 predictions system capabilities and uncertainties on nitrate formation originating from the N2O5 uptake process in SMA.

In this study, we carried out a series of numerical simulations to (1) confirm the importance of the N2O5 heterogeneous hydrolysis process in the SMA during nighttime and (2) compare the uncertainty ranges originating from two factors: the N2O5 uptake process as an uncertain factor in the chemistry, and the parameterization of PBL height as an uncertain factor in meteorology. We first investigated PM2.5 concentrations and weather conditions and selected stagnant winter days for a case study to minimize the wind speed uncertainty in the SMA. N2O5 experiments were conducted under the same framework of Jo et al. (2019), and four PBL schemes (YSU, ACM2, MYJ, and QNSE) were employed to quantify the PBL bias, while numerical tests were carried out to evaluate the ranges in PBL errors in the WRF-CMAQ model.

Section snippets

Modeling system and domain

To conduct sensitivity analyses associated with air quality, we adopted the Weather Research and Forecast model (WRF, https://www.mmm.ucar.edu/weather-research-and-forecasting-model) and the United States Environmental Protection Agency's (US EPA's) Community Multi-scale Air Quality model (CMAQ v5.0.2, https://www.cmascenter.org/cmaq/). The WRF (v3.6.1) was used to provide input meteorological fields for the CMAQ (ver. 5.0.2), using the grid nudging technique as a data assimilation method (

Base case simulation

The simulated meteorological and chemical variables were compared with the measurements at the Bulgwang site, as shown in Fig. 3. For the meteorological variables, 2 m temperature, RH, 10 m wind speed, and PBL height were extracted at the nearest grid point to the Bulgwang site. In Fig. 3, the overall simulated meteorological variables generally showed similar temporal variations with the measurements, with the exception of the underestimated PBL height. The resulting index of agreement (IOA)

Summary and conclusions

This study analyzed the sensitivity of major meteorological and chemical uncertainty factors that greatly affect the accuracy of air quality models. The WRF-CMAQ was applied to simulate experiments on the PBL as a meteorology uncertainty factor and N2O5 uptake coefficient as a chemical uncertainty factor to identify the effect on the results of PM2.5 numerical simulation in winter. The study period covered January 13–15, 2018, during which the measured NO3 concentrations indicated that the

Author statement

H.-Y. Jo, G. Heo, and C.-H. Kim: Conceptualization, M. Lee, J-.A. Kim, and T. Lee; Data curation M.-S. Park, Y.-H. Lee and S.-W. Kim.; Formal analysis, H.-J. Lee. and Y.-J. Jo.; Investigation, G. Heo.; Methodology, H.-Y. Jo and H.-J. Lee; Visualization, H.-Y. Jo, G. Heo, and C.-H. Kim; Writing—original draft preparation, H.-Y. Jo and C.-H. Kim; Writing—Review and Editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2020R1A6A1A03044834 and 2020R1I1A1A01072998).

References (65)

  • S.-U. Park et al.

    A numerical model for the simulation of SO2 concentrations in the Kyongin region, Korea

    Atmos. Environ.

    (1999)
  • X. Ren et al.

    Behavior of OH and HO2 in the winter atmosphere in New York City

    Atmos. Environ.

    (2006)
  • S. Squizzato et al.

    A procedure to assess localand long-range transport contributions to PM2.5 and secondary inorganic aerosol

    J. Aerosol Sci.

    (2012)
  • J. Yang et al.

    Evaluation and improvement study of the Planetary Boundary-Layer schemes during a high PM2.5 episode in a core city of BTH region, China

    Sci. Total Environ.

    (2021)
  • M. Baasandorj et al.

    Coupling between chemical and meteorological processes under persistent cold-air pool conditions: evolution of wintertime PM2.5 pollution events and N2O5 observations in Utah’s Salt Lake Valley

    Environ. Sci. Technol.

    (2017)
  • J.H. Bowden et al.

    Examining interior grid nudging techniques using two-way nesting in the WRF model for regional climate modeling

    J. Clim.

    (2012)
  • S.G. Brown et al.

    Wintertime vertical variations in particulate matter (PM) and precursor concentrations in the San Joaquin valley during the California regional coarse PM/Fine PM air quality study

    J. Air Waste Manage. Assoc.

    (2006)
  • S.S. Brown et al.

    Variability in nocturnal nitrogen oxide processing and its role in regional air quality

    Science

    (2006)
  • J.M Davis et al.

    Parameterization of N2O5 reaction probabilities on the surface of particles containing ammonium, sulfate and nitrate

    Atmos. Chem. Phys.

    (2008)
  • E. Dinar et al.

    CCN activity and hygroscopic growth of organic aerosols following reactive uptake of ammonia

    Environ.Sci.Technol.

    (2008)
  • C. Emery et al.

    Recommendations on statistics and benchmarks to assess photochemical model performance

    J. Air Waste Manage. Assoc.

    (2016)
  • B.J. Finlayson-Pitts et al.

    Tropospheric air pollution: ozone, airborne toxics, polycyclic aromatic hydrocarbons, and particles

    Science.

    (1997)
  • B.J. Finlayson-Pitts et al.

    Chemistry of the Upper and Lower Atmosphere: Theory, Experiments, and Applications

    (2000)
  • A. Franchin et al.

    Airborne and ground-based observations of ammonium-nitrate-dominated aerosols in a shallow boundary layer during intense winter pollution episodes in northern Utah

    Atmos. Chem. Phys.

    (2018)
  • C.C. Guenther

    Estimates of global terrestrial isoprene emissions using MEGAN (Model of Emissions of Gases and Aerosols from Nature)

    Atmos. Chem. Phys.

    (2006)
  • X.-M. Hu et al.

    Evaluation of three planetary boundary layer schemes in the WRF model

    J. Appl. Meteorol. Climatol.

    (2010)
  • I. Ibikunle et al.

    Fine particle pH and sensitivity to NH3 and HNO3 over summertime South Korea during KORUS-AQ

    Atmos. Chem. Phys. Discuss.

    (2020)
  • H. Jeon et al.

    The characteristics of PM2.5 concentration and chemical composition of Seoul metropolitan and inflow background area in Korea peninsula

    J. Korean Soc. Urban Environ.

    (2015)
  • H.-Y. Jo et al.

    Identification of long-range transported haze phenomena and their meteorological features over northeast Asia

    J. Appl. Meteorol. and Climatol.

    (2013)
  • S. Kim et al.

    Korea emissions inventory processing using the US EPA’s SMOKE system

    Asian J. Atmos. Environ.

    (2008)
  • S. Kim et al.

    The primary and recycling sources of OH during the NACHTT-2011 campaign

    J. Geophys. Res.

    (2014)
  • H.C. Kim et al.

    Regional contributions to particulate matter concentration in the Seoul metropolitan area, South Korea: seasonal variation and sensitivity to meteorology and emission inventory

    Atmos. Chem. Phys.

    (2017)
  • Cited by (0)

    View full text