Simulation of the cooking organic aerosol concentration variability in an urban area
Introduction
Exposure to PM2.5 (particles with a diameter lower than 2.5 μm) can lead to serious respiratory and cardiovascular problems (Du et al., 2016; Xing et al., 2016). Organic aerosol (OA) often represents more than 50% of PM2.5 in urban areas (Kanakidou et al., 2005) and is a major component of atmospheric pollution (Zhang et al., 2007; Jimenez et al., 2009). OA can be emitted directly by a source or can be formed in-situ by condensation of low volatility products of the oxidation of organic vapors (Seinfeld and Pandis, 2006).
Cooking operations can be an important fine PM source for urban areas. In Los Angeles in the 1980s, approximately 20% of the primary organic aerosol was due to meat cooking (Hildemann et al., 1991). Cooking emissions contributed 19% to the total OA in Manchester during winter (Allan et al., 2010) and 17% in Barcelona (Mohr et al., 2012). In New York City, COA accounted for 16% of the total OA during summertime (Sun et al., 2011). In London, COA represented 22 and 30% of the total OA for two winter periods with different temperatures (Allan et al., 2010). Similar behavior was observed in Paris with cooking contributing 18 and 20% to the total OA during winter and summer time respectively (Crippa et al., 2013a, 2013b). In Athens, the contribution of COA to the total OA during summer and winter 2012 was 17 and 16% respectively (Kostenidou et al., 2015; Florou et al., 2017). In China, the contribution of cooking emissions is similar to that in European countries. During the 2008 Beijing Olympic Games 24% of the total OA was due to cooking (Huang et al., 2010). In Lanzhou, northwest China, during summertime COA contributes 24% to the total OA (Xu et al., 2014).
Meat charbroiling is one of the most important COA sources in many cities (Hildemann et al., 1991; Kostenidou et al., 2015). Hildemann et al. (1991) measured aerosol emissions up to 40 g kg−1 of meat cooked, during charbroiling of regular meat (with a fat percentage of 21%). During frying of regular meat, the measured aerosol emissions were a lot lower at 1 g kg−1 of meat cooked. Kostenidou et al. (2015) have shown that in Greek cities COA emissions are mainly due to the charbroiling of meat. The average COA contribution to the total OA in Patras was 10–15% depending on the season (Kostenidou et al., 2015; Florou et al., 2017). COA concentrations in Patras peaked between 14:00 and 15:00 LT and during the night (22:00–23:00 LT). Similar daily behavior has been observed in Paris, Athens, London, Manchester and Barcelona (Allan et al., 2010; Mohr et al., 2012; Crippa et al., 2013a; Kostenidou et al., 2015) which indicates the importance of cooking emissions during lunch and dinner periods.
COA can partially evaporate in the atmosphere after its emission and can also react with O3, OH, etc. Significant changes in COA composition during photooxidation in combination with small changes in mass have been reported by Kaltsonoudis et al. (2017). Heterogeneous reactions seem to be the most probable reason for these changes. Louvaris et al. (2017) determined the volatility distribution of COA during meat charbroiling combining thermodenuder and isothermal dilution measurements.
There have been relatively few modeling efforts of the spatial distribution of COA in urban areas mainly due to the lack of suitable emission inventories. Fountoukis et al. (2016) applied the PMCAMx model over Paris assuming COA emissions of 5.3 tons d−1 for summer and 5.1 tons d−1 for winter. These emissions were estimated based on the measured concentrations. They also assumed that 50% of the COA was emitted during lunch time (12:00–14:00 LT), while 20% during dinner time (20:00–22:00 LT). The emissions were distributed in space based on the population density. Ots et al. (2016) applied the EMEP4UK model (a regional version of the EMEP MSC-W model) over Britain for a year assuming emissions of 7.4 Gg COA. These cooking emissions were based on measurements of COA at two sites in London during the same period. The temporal variation of these emissions was based on measurements at a hot spot and peaked during lunch (12:00–14:00 LT) and dinner times (18:00–21:00 LT). The spatial distribution was also based on population density.
Despite the progress in understanding COA in urban areas a number of issues are yet to be addressed. The estimation of emissions and their distribution in space and time are still quite uncertain. It is not clear if there are COA hot spots in urban areas and how high the corresponding concentration levels are. Also, the evaluation of the few modeling efforts has been limited due to the sparse observations which may be far from the concentration hot spots.
In this work we estimate the COA emissions in a major Greek city (Patras) based on the individual locations of each restaurant and the corresponding emission factors from meat grilling. A high-resolution version of PMCAMx is applied over Patras in an effort to simulate the spatial and temporal distribution of COA in this area. COA is assumed to be semi-volatile and to also react forming secondary COA. The predictions of the model are evaluated against measurements from a low-cost PM2.5 sensor network in the city including measurements in the area with the highest restaurant density.
Section snippets
Model description
PMCAMx is a three-dimensional chemical transport model (CTM) that simulates the emissions, horizontal and vertical advection, vertical and horizontal turbulent dispersion, gas, aqueous and aerosol chemistry, aerosol dynamics and removal through wet and dry deposition by solving the pollutant continuity equation (ENVIRON, 2013). The model is the research version of CAMx (ENVIRON, 2013) and uses the corresponding approaches for the simulation of atmospheric transport. The modified SAPRC mechanism
Cooking organic aerosol concentration
The predicted concentrations of fresh COA in the area with the highest restaurant density, Trion Navarchon Square, reach hourly levels of 5–15 μg m−3 during most nights (Fig. 5a). The corresponding peaks during noon are much lower (1–2 μg m−3) because of the intense summertime mixing and dilution together with the assumed lower emission rates during lunchtime. The COA concentrations during August 25–26 and on August 30 were lower due to the strong winds that blew on those days. On the other
Conclusions
Fresh COA emissions of 150 ± 75 kg d−1 were estimated for the city of Patras which corresponds to 0.6 ± 0.3 g d−1 per person. An intermediate COA emission factor of 8 g kg−1 of meat cooked was used for the base case analysis based on the cooking method used, the type and the average fat content of the meat. Emissions were distributed in space based on the locations of restaurants and in time based on measured COA diurnal profiles. PMCAMx with ultra-high resolution (1 × 1 km2) predicts peak COA
CRediT authorship contribution statement
Evangelia Siouti: Software, Formal analysis, Writing – review & editing. Ksakousti Skyllakou: Software, Writing – review & editing. Ioannis Kioutsioukis: Software, Writing – review & editing. Giancarlo Ciarelli: Software, Writing – review & editing. Spyros N. Pandis: Conceptualization, Supervision, Writing – review & 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
We acknowledge the support of this work by the project “Smart Air Quality Monitoring (SmartAQM)", funded by Western Greece's Smart Specialisation Strategy (RIS3) 2014–2020, co-financed by Greece and the European Union and by the project “PANhellenic infrastructure for Atmospheric Composition and climatE change” (MIS 5021516) which is implemented under the Action “Reinforcement of the Research and Innovation Infrastructure”, funded by the Operational Programme “Competitiveness, Entrepreneurship
References (53)
- et al.
Optimizing model performance: variable size resolution in cloud chemistry modeling
Atmos. Environ.
(2001) - et al.
Development and application of an efficient moving sectional approach for the solution of the atmospheric aerosol condensation/evaporation equations
Atmos. Environ.
(2003) - et al.
Integrated approaches to modeling the organic and inorganic atmospheric aerosol components
Atmos. Environ.
(2003) - et al.
Low-cost sensors for measuring airborne particulate matter: field evaluation and calibration at a South-Eastern European site
Sci. Total Environ.
(2020) - et al.
Consumer-perceived quality in “traditional” food chains: the case of the Greek meat supply chain
Appetite
(2007) - et al.
Predictions for particle deposition on natural-waters
Atmos. Environ.
(1980) - et al.
Solution of the discrete coagulation equation
J. Colloid Interface Sci.
(1980) Parameterization of surface resistances to gaseous dry deposition in regional-scale numerical models
Atmos. Environ.
(1989)- et al.
Contributions from transports solid fuel burning, and cooking to primary organic aerosols in two UK cities
Atmos. Chem. Phys.
(2010) - et al.
Atmospheric degradation of volatile organic compounds
Chem. Rev.
(2003)