Long-range transported pollution from the Middle East and its impact on carbonaceous aerosol sources over Cyprus
Abstract. The geographical origin and source apportionment of submicron carbonaceous aerosols (organic aerosols, OA, and black carbon, BC) have been investigated here for the first time by means of high time resolution measurements at an urban background site of Nicosia, the capital city of Cyprus, in the Eastern Mediterranean. This study covers a half-year period, encompassing both the cold and warm periods with continuous observations of the physical and chemical properties of PM1 performed with an Aerosol Chemical Speciation monitor (ACSM), an Aethalometer, accompanied by a suite of various ancillary off and on-line measurements. Carbonaceous aerosols were dominant during both seasons (cold and warm periods), with a respective contribution of 57 % and 48 % to PM1, respectively, and exhibited recurrent intense night-time peaks (>20–30 µg m-3) during the cold period associated with local domestic heating. Findings of this study show that high concentrations of sulfate (close to 3 µg m-3) were continuously recorded, standing among the highest ever reported for Europe and originating from the Middle East region.
Source apportionment of the OA and BC fractions was performed using the Positive Matrix Factorization (PMF) approach and the combination of two models (aethalometer model and multilinear regression), respectively. Our study revealed elevated hydrocarbon-like organic aerosol (HOA) concentrations in Nicosia (among the highest reported for a European urban background site), originating from a mixture of local and regional fossil-fuel combustion sources. Although air masses from the Middle East had a low occurrence and were observed mostly during the cold period, they were shown to strongly affect the mean concentrations levels of BC and OA in Nicosia during both seasons. Overall, the present study brings to our attention the need to further characterize primary and secondary carbonaceous aerosols in the Middle East; an undersampled region characterized by continuously increasing fossil fuel (oil and gas) emissions and extreme environmental conditions, which can contribute to photochemical aging.
Aliki Christodoulou et al.
Aliki Christodoulou et al.
Long-range transported pollution from the Middle East and its impact on carbonaceous aerosol sources over Cyprus https://zenodo.org/record/7186341
Aliki Christodoulou et al.
Viewed (geographical distribution)
This study looks at a unique dataset gathered by advanced aerosol composition measurement techniques from a region not as well studied by online aerosol chemistry and composition. The study adds to the wealth of knowledge of aerosol pollution and long-range transport, especially owing to the unique impacts of sulfate on Cyprus and impact of Power Generation Plants. Several established scientific methods are used to establish pollution sources impacting the site and the authors do a nice job of using multiple techniques to evaluate and come to some conclusions. Generally, a nice paper, although with little novel methods, applied to a new and distinctive dataset. General comments are below with detailed comments following.
Based on the back trajectory cluster analysis alone, it is not convincing that there are enough air masses advecting over the Middle East to draw the conclusions the title suggest. The authors may consider adding in a graph which shows all 72-hr air mass back trajectories during the campaign (the authors use 120 hrs but this is excessive giving the uncertainty associated with HYSPLIT) at 6-12 hr intervals. Perhaps this would give a better idea of the potential influence of the Middle East. Please also include the Log10(n+1) trajectory footprint graphs from Zefir for the PSCF in Figure S17.
While the authors use NWR and PSCF productively to understand the potential sources of pollutants impacting the Nicosia site, I wonder that there is no mention of shipping emission impact given the high levels of sulfate measured. The sulphate burden over the Mediterranean has been shown to be higher in summer than that over Europe owing to shipping activities (Marmer & Langmann, 2005). While source apportionment techniques have shown a range of effects from ship related sources, from OA fractions of 4.5% in the Mediteranean (Chazeau et al. 2021) to 25% of PM2.5 in Hong Kong (Yau et al. 2013), the ship related burden does not seem to be even considered in the text. The Mediterranean will not become a sulfur emission control area (SECA) (marine fuels S content below 0.1%) until 2025 (COP 2021 under MARPOL (Annex VI), decision 10 June 2022), and has only been under the 0.5% S fuel content since Jan 2020 (IMO-2021, Low Sulfur Regulation), which is after this data 2018-2019 was collected. Therefore, it is strongly suggested that the impact of marine shipping be discussed within the main text as a possible source or reasons given why it is not being considered.
Marmer, E., and Langmann, B.: Impact of ship emissions on the Mediterranean summertime pollution and climate: A regional model study, Atmospheric Environment, 39, 4659-4669, https://doi.org/10.1016/j.atmosenv.2005.04.014, 2005.
Chazeau, B., Temime-Roussel, B., Gille, G., Mesbah, B., D'Anna, B., Wortham, H., and Marchand, N.: Measurement report: Fourteen months of real-time characterisation of the submicronic aerosol and its atmospheric dynamics at the Marseille–Longchamp supersite, Atmos. Chem. Phys., 21, 7293-7319, 10.5194/acp-21-7293-2021, 2021.
Yau, P. S., Lee, S. C., Cheng, Y., Huang, Y., Lai, S. C., and Xu, X. H.: Contribution of ship emissions to the fine particulate in the community near an international port in Hong Kong, Atmospheric Research, 124, 61-72, https://doi.org/10.1016/j.atmosres.2012.12.009, 2013.
Title - Title reorder suggestion as LRT is not really impacting the sources but rather a large mixed source itself. ‘Carbonaceous aerosol over Cyprus impacted by long-range transported pollution from the Middle East’
Line 73 – insert of, ‘in terms of PM’.
Line 86 – consider cool rather than mild
Line 97 – use of ca. reconsider throughout text. Circa is not always appropriate (see line 166), and the term ‘about’ or ‘approximately’ is modern usage. Historically, it is most commonly used in reference to a date that is not accurately known.
Line 99 – c.a. or ca. please harmonise, should be ca. (see comment for line 97)
Line 127-138 – (i) SOP by the Cost Action COLOSSAL, that should be referenced. (ii) It is assumed that the Q-ACSM operated here had a PM1 lens as cyclone cutoff was 1.3 um, but it should be stated. If you used CDCE, it is a standard vaporiser, but this should also be stated. (iii) Mass concentrations are not calculated via the CDCE, but rather corrected by it. Concentrations are calculated based on signal intensity and Fragmentation Table. (iv) Please also include RIE and RF values used. (v) It is stated RH was below 30%, how was it monitored?
Line 139 – should read ‘…were conducted using an 7-wavelength aethalometer (AE33, Magee Scientific, USA) a…’
Line 139 -146 – For clarity, what tape was used for the AE33, was it the post 2017 tape (no. 8060) that had non-linearity issues with short wavelengths (https://mageesci.com/tape/Magee_Scientific_Filter_Aethalometer_AE_Tape_Replacement_discussion.pdf )? Not an issue for BC6 but an issue for BB fraction. What was the MAC used?
Using an AAE of 1 and 2 for FF and WB is the default settings. However, using an AAE of 2 for WB is risky in coastal sites unless WB is the only residential solid fuel burning source. For example if other fuel types such as turf or peat are used the AAE could be quite different due to the combined effects of combustion efficiency and fuel moisture content while the AAE can also be generally affected by photochemical aging and the mixing state of black carbon (Garg et al. 2016). Additionally the AAE value can be highly sensitive to small changes at coastal sites. If you apply the Zotter at al. 2017 values for FF and biomass burning (BB rather than WB) of 0.9 and 1.68 rather than 1 and 2, it can lead to a doubling of the BB percentage (Spohn 2021).
Can the authors please elaborate on the known fuel sources and why the default values were used rather than optimising the alpha values?
Garg, S., Chandra, B. P., Sinha, V., Sarda-Esteve, R., Gros, V., and Sinha, B.: Limitation of the Use of the Absorption Angstrom Exponent for Source Apportionment of Equivalent Black Carbon: a Case Study from the North West Indo-Gangetic Plain, Environmental Science & Technology, 50, 814-824, 10.1021/acs.est.5b03868, 2016.
Zotter, P., Herich, H., Gysel, M., El-Haddad, I., Zhang, Y., MoÄnik, G., Hüglin, C., Baltensperger, U., Szidat, S., and Prévôt, A. S. H.: Evaluation of the absorption Ångström exponents for traffic and wood burning in the Aethalometer-based source apportionment using radiocarbon measurements of ambient aerosol, Atmos. Chem. Phys., 17, 4229-4249, 10.5194/acp-17-4229-2017, 2017.
Spohn, T. K.: A study of black carbon and related measurements from Ireland's atmospheric composition and climate change network, Thesis (Ph.D.) : NUI Galway, Galway
Line 152 – 154 – Why convert volume concentrations to mass concentrations by assuming variable density? Why not convert ACSM measurements to volume as you know the bulk mass fraction of species? You are adding 1 extra unnecessary approximation. Please try by volume and see if there are any differences.
Line 142 – C = 1.39 is a correction to C=1.57 from Drinovec et al. 2015.
Drinovec, L., MoÄnik, G., Zotter, P., Prévôt, A. S. H., Ruckstuhl, C., Coz, E., Rupakheti, M., Sciare, J., Müller, T., Wiedensohler, A., and Hansen, A. D. A.: The “dual-spot” Aethalometer: An improved measurement of aerosol black carbon with real-time loading compensation, Atmos. Meas. Tech., 8, 1965–1979, https://doi.org/10.5194/amt-8-1965-2015, 2015.
Line 174 – please report in asl as you did for the stations rather than above ground.
Line 201 – m/z used for the first time and not defined.
Line 216 – consider adding literature for BBOA
Lin, C., Ceburnis, D., Hellebust, S., Buckley, P., Wenger, J., Canonaco, F., Prévôt, A. S. H., Huang, R.-J., O’Dowd, C., and Ovadnevaite, J.: Characterization of Primary Organic Aerosol from Domestic Wood, Peat, and Coal Burning in Ireland, Environmental Science & Technology, 51, 10624-10632, 10.1021/acs.est.7b01926, 2017.
Trubetskaya, A., Lin, C., Ovadnevaite, J., Ceburnis, D., O’Dowd, C., Leahy, J. J., Monaghan, R. F. D., Johnson, R., Layden, P., and Smith, W.: Study of Emissions from Domestic Solid-Fuel Stove Combustion in Ireland, Energy & Fuels, 35, 4966-4978, 10.1021/acs.energyfuels.0c04148, 2021.
Line 226 – 46% variation is quite high, are you sure BBOA should be related to BCwb, as discussed wood burning may not be the main source of fuel for home heating. This may also change the AAE value for the non-ff BC.
Line 237 – S4 order – a should come before b
Line 236-239 – Why compare PM and not Volume from the SMPS? (see comment for line 152-154)
Line 246-249 – Why does the low ratio clearly denote a major contribution of long-chain hydrocarbon OA, is there the possibility that the Org RIE used is incorrect and the OA concentrations are a bit too low?
Line 279-280 – That is not clear from the figure. It seems the reverse, that in the warm period Cluster 5 originates over Middle East/Asia. Also, several clusters pass over Turkey. Please specifically name the clusters and refer to figure in the main text.
Figure 3 – no need for the 100th place decimal place on the pie charts. Are the data stacked onto each other in the time series, or overlaid? In other words, are the 6-hr average peaks greater than 50 ug/m3 or do the peaks show total NR-PM1?
Line 304-310 – Significant figures are incorrect. All averages are only as accurate as the standard deviation. E.g. 10.30 ± 7.92 should be 10 ± 8. Check significant figures in the numbers throughout the text (E.g. Table 1).
Line 317 – omit the word much. SO4 is 3 ± 2 ug/m3 which is within the range reported in other studies shown in Table 2.
Figure 4 – These are diurnal averages, the ranges should be shown. Left and right axis should be the same scale. Correct for all average plots.
Line 386-387 – First definition of MO-OOA. While MO-OOA is synonymous with low-volatile OOA, it should be first defined here as More Oxidised OOA. Same argument for the first mention of LO-OOA. Suggest changing the text to read more like, ‘… namely the low-volatile MO-OOA (More-Oxidized Oxygenated Organic Aerosol) and semi-volatile LO-OOA (Less-Oxidized Oxygenated Organic Aerosol)…’
Figure 10 – Shows PSCF 75th Percentile (should represent the probability of an air parcel to be responsible for measured concentrations at the receptor site above the 75th percentile) - please amend the figure text to reflect what 75th Percentile means in this figure
Line 659 – Figure S19 shows LV-OOA, should it show MO-OOA as the figure caption and text suggest? Rather than abbreviated as low-volatility
Line 702 – insert the word of, ‘… of the acquired data’
Figure S4 – a) should be BC AE33 rather than ACSM
Figure S6 – a) cluster 1 and 7 are difficult to differentiate – why has 7 clusters been chosen rather than another number? Is warm and cold switched in the text – refer here to comment about line 279-280.
Figure S9 – it would be good to insert the correlations graphs between the factors in warm and cold periods. E.g. HOA-1 (cold) vs HOA-1 (warm).
Figure S17 – Also include the Log10(n+1) trajectory footprint graphs from Zefir for the PSCF.
Figure S19 – rename MO-OOA rather than LV-OOA