Diesel. Cooking. Wood-burning.

High resolution modelling of particulate matter air quality in the UK with a focus on carbonaceous aerosol

Transcript

1. High resolution modelling of particulate matter air quality in the UK with a focus on carbonaceous aerosol: diesel, cooking and wood-burning R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, Edinburgh Riinu Ots - data Scientist Data Manager at Clinical Surgery - University of Edinburgh PhD in Environmental Chemistry - University of Edinburgh BSc in Physics - University of Tartu, Estonia
2. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Talk outline • Introduction • Model - EMEP4UK • Diesel: Ots et al.: Simulating secondary organic aerosol from missing diesel-related intermediate-volatility organic compound emissions during the Clean Air for London (ClearfLo) campaign, Atmos. Chem. Phys., 16, 13773-13789, doi:10.5194/ acp-16-13773-2016, 2016. • Cooking: Ots et al.:Model simulations of cooking organic aerosol (COA) over the UK using estimates of emissions based on measurements at two sites in London, Atmos. Chem. Phys., 16, 6453-6473, doi:10.5194/acp-16-6453-2016, 2016. • Wood-burning: Ots et al.: A model investigation of carbonaceous aerosol from residential solid fuel burning with different assumptions for the spatial distribution of emissions, In prep. for Atmos. Chem. Phys., 2017. • Conclusions
3. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Particulate matter (PM) [= aerosol]* * Conceptually, the two terms are different: aerosols are two-phase systems, consisting of the particles and the gas in which they are suspended, whereas PM only refers to the condensed (and dispersed) phase. In this research field, however, the terms are used interchangeably. Introduction The impacts of atmospheric PM include adverse health effects, climate forcing, chemical processing, deposition, and effects on visibility. PM2.5 annual average concentrations: WHO air quality guideline: 10 μg m−3 UK and Europe year 2020 target value: 25 μg m−3 Scotland year 2020 target value: 12 μg m−3 PM2.5 daily average concentrations: WHO air quality guideline: no more than 3 days >25 μg m−3
4. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh PM composition and sources - complicated*. Organic aerosol -OA ~30% of total PM. Introduction SOA (secondary OA) OA (organic aerosol) POA (primary OA) Hydrocarbon-like OA from fossil fuel combustion, mainly vehicular Cooking OA from charbroiling and frying Biomass Burning OA, residential or natural (forest fires) Coal Combustion OA SV-OOA LV-OOA Solid Fuel OA, combined factor of BBOA and CCOA HOA COA BBOA SFOA OOA OOA 1|2 Oxygenated OA, used if OOA factors are not distinguishable OOA type 1 and type 2; used if factors are not identifiable CCOA Semi-Volatile oxygenated OA Low-Volatility oxygenated OA POA can evolve into SOA through atmospheric ageing *Anthpopogenic and natural, primary (emitted as particulates) and secondary (formed in the atmosphere from gases or semi-volatile components). + Elemental carbon (EC) also known as black carbon or soot
5. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Organic aerosol (OA) ~30% of total PM Introduction SOA (secondary OA) OA (organic aerosol) POA (primary OA) Hydrocarbon-like OA from fossil fuel combustion, mainly vehicular Cooking OA from charbroiling and frying Biomass Burning OA, residential or natural (forest fires) Coal Combustion OA SV-OOA LV-OOA Solid Fuel OA, combined factor of BBOA and CCOA HOA COA BBOA SFOA OOA OOA 1|2 Oxygenated OA, used if OOA factors are not distinguishable OOA type 1 and type 2; used if factors are not identifiable CCOA Semi-Volatile oxygenated OA Low-Volatility oxygenated OA POA can evolve into SOA through atmospheric ageing 0 1 2 3 4 Obs. Modelled Annual average OA concentration, µg/m3 COA SOA SFOA HOA Annual average AMS- PMF speciated OA concentration at an urban background site in London North Kensington. Note that green represents different components on the two figures, other colours are unified.
6. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Air pollution modelling 5 km x 5 km (x 21 vertical layers, lowest=40 m) UK nested into a 50 km x 50 km European domain Model - EMEP4UK
7. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Air pollution modelling Model - EMEP4UK NMB = −32%, NMGE = 36%, r = 0.78, COE = 0.35 (a) NOx NMB = −1%, NMGE = 24%, r = 0.79, COE = 0.41 (b) O3 NMB = 6%, NMGE = 41%, r = 0.73, COE = 0.32 (c) SO 4 2− NMB = −12%, NMGE = 50%, r = 0.65, COE = 0.32 (d) NH 4 + NMB = −23%, NMGE = 67%, r = 0.57, COE = 0.34 (e) NO 3 − 0 100 200 300 0 25 50 75 100 0.0 2.5 5.0 7.5 10.0 0 5 10 0 10 20 30 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Daily concentration, µg/m3 Measured Base Daily averaged measurements, London North Kensington 2012: measured inorganic pollutants (gaseous and particulate) are captured well by the model.
8. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Air pollution modelling Model - EMEP4UK Daily averaged measurements, London North Kensington 2012: measured organic aerosol is not well captured by the model. NMB = −54%, NMGE = 56%, r = 0.53, COE = 0.11 (a) HOA NMB = −71%, NMGE = 72%, r = 0.72, COE = 0.01 (b) SFOA 0 1 2 3 4 5 6 0 1 2 3 4 5 6 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Daily concentration, µg/m3 Measured Base 0 1 2 3 4 Obs. Modelled Annual average OA concentration, µg/m3 COA SOA SFOA HOA
9. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Unincluded diesel emissions - OA vs VOCs (i.e. aerosol vs gases) Diesel SOA VOCs 107 SOA yields SOA ageing SOA in 5 volatility bins 1 µg/m3 non-volatile volatile VOC - Volatile organic compound The partitioning calculations on this Figure include a non-volatile POA component with a concentration of 3 μg m−3 SNAP10 SNAP9 SNAP8 SNAP7 SNAP6 SNAP5 SNAP4 SNAP3 SNAP2 SNAP1 0 10 20 30 40 UK national total emissions, Gg/year SFOA HOA EC Other/Mineral PM SNAP 10 SNAP 9 SNAP 8 SNAP 7 SNAP 6 SNAP 5 SNAP 4 SNAP 3 SNAP 2 SNAP 1 0 100 200 300 UK national total emissions, Gg/year Other/Mix VOC emission Gasoline−VOCs emission Diesel−VOCs emission Added diesel−IVOCs (SNAP sectors are different pollution sources ranging from power plants to agriculture.) RANGE PM VOCs
10. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Unincluded diesel emissions - OA vs VOCs (i.e. aerosol vs gases) Diesel SOA 3.2. Methods Comparison of diesel and gasoline NMVOCs in the UK National Atmospheric nventory (NAEI) with the urban background ambient concentrations measured learfLo winter Intensive Observation Period in London. NAEI 2012 Measurements a Diesel-(I)VOCs 8 Gg yr 1 107 µg m 3 Gasoline-VOCs 31 Gg yr 1 33 µg m 3 Diesel/Gasoline 0.26 3.2 a Dunmore et al. (2015) 9 8 7 6 5 4 3 2 1 Other/Mix VOC emission Gasoline−VOCs emission Diesel−VOCs emission Added diesel−IVOCs SNAP 10 SNAP 9 SNAP 8 SNAP 7 SNAP 6 SNAP 5 SNAP 4 SNAP 3 SNAP 2 SNAP 1 0 100 200 300 UK national total emissions, Gg/year Other/Mix VOC emission Gasoline−VOCs emission Diesel−VOCs emission Added diesel−IVOCs A.3. Increasing all NMVOCs from SNAP7 by 3.3 times - femis.dat file rcemis:C15H32 = C15H32 ; * Yields Presto 2010 * These yields were from C15H32 (M 212.41), so scaled for our C (12) and non_C (1) to ,! get moles; assuming an OM/OC of 1.7 2.07e-11 OH + C15H32 = 0.69 ASOC_ng100 + 5.77 NON_C_ASOA_ng100 + 1.11 ,! ASOC_ug1 + 9.31 NON_C_ASOA_ug1 + 6.40 ASOC_ug10 + 53.79 NON_C_ASOA_ug10 + 4.69 ,! ASOC_ug1e2 + 39.36 NON_C_ASOA_ug1e2 ;
11. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Unincluded diesel emissions - Results Diesel SOA (a) Year (b) Spring (c) Summer (d) Autumn (e) Winter 0.0 0.5 1.0 1.5 2.0 2.5 Obs. Base addDiesel Obs. Base addDiesel Obs. Base addDiesel Obs. Base addDiesel Obs. Base addDiesel Concentration, µg/m3 ASOA BSOA Background OA SOA
12. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Cooking emissions - charbroiling, frying, deep-frying Cooking OA (not emissions from the wood or fuel used for cooking, but from the food or cooking- oil) Workday Residential
13. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Estimating cooking emissions Cooking OA r = 0.99 (a) r = 0.93 (b) 0.0 2.5 5.0 7.5 0 3 6 9 12 15 18 21 0 3 6 9 12 15 18 21 Hour of day Measured Modelled (c) (d) 0.0 1.0 2.0 3.0 4.0 Mon Tue Wed Thu Fri Sat Sun Mon Tue Wed Thu Fri Sat Sun Day of Week Measured Modelled Marylebone Road North Kensington Mean COA, μg/m3 COA emission for the UK (gridded to workday population density). Model normalized mean biases of COA concentrations at the Lon- don Marylebone Road and North Kensington sites are shown for total UK emissions of 2, 8, and 7.4 Gg. A total emission of 7.4 Gg was chosen and is used in the rest of the simulations presented in this work. Site Measured Modelled (NMB) 2 Gg 8 Gg 7.4 Gg North Kensington 0.8 µgm 3 70 % +18 % +8 % Marylebone Road 2.2 µgm 3 75 % 2 % 4 % to population density data was aggregated appropriately to the coarser model resolution during input data preparation. 2.4 Annual total emitted COA Based on sensitivity tests (Table 1), the annual total COA emissions for the UK applied to the model was set to 7.4 Gg. (The spatial distribution applied to these emissions is ex- plained in the previous section, the temporal variation is explained in the following section.) This is a 9 % addi- tion to the UK national total PM2.5 emissions for the year 2012 (82 Gg; NAEI, 2013). This emission corresponds to about 320 mgperson 1 day 1 (for a population of 63 mil- lion), which is 4 times higher than estimated by Fountoukis et al. (2016) for France. This difference might be explained Figure 2. Average temporal profiles of COA conc two sites in central London in 2012: (a) diurna Marylebone Road site, (b) diurnal profile at the ton site, (c) day-of-week profile at the Marylebon (d) day-of-week profile at the North Kensington tamp of (a) and (b) is at the beginning of the hour. standard deviations for each mean value. Annual average cooking OA (COA) Diurnal and weekly profiles 0.0 —0.1————> 0.9 μg m−3 10% addition to currently reported PM2.5 emissions
14. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Comparison with independent measurements - Manchester 2007 • • • • • • • • • • • • • • • y = −0.06 + 0.52 ⋅ x, r = 0.86 (c) 0 1 2 3 4 5 6 0 1 2 3 4 5 6 Measured daily COA, µg/m3 Modelled daily COA, µg/m3 NMB = −50%, NMGE = 57%, r = 0.63, COE = 0.12 (a) 0.0 2.5 5.0 7.5 10.0 Jan 24 Jan 26 Jan 28 Jan 30 Feb 01 Feb 03 Feb 05 Feb 07 Hourly COA concentration, µg/m3 Measured Modelled r = 0.80 (b) 0 2 4 6 0 3 6 9 12 15 18 21 Hour of day Mean COA, µg/m3 Measured Modelled Workday - Manchester Cooking OA
15. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Evaluation of cooking OA concentrations for other major cities Cooking OA Leeds Manchester Birmingham
16. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Solid fuel OA - coal and wood burning (in smoke control areas!) wood/coal OA Since the Great Smog of 1952 in London, several legislative interventions have substantially reduced the use of solid fuels in residential heating. For example, most of London is a smoke control area whereby the solid fuel burning is prohibited, unless undertaken in approved wood burners. This control is only applied on appliances with a chimney, incidental sources such as bonfires or barbecues are allowed. However, it appears that these measures are no longer effective. Based on Fuller et al.: New Directions: Time to tackle urban wood burning? Atmos. Env., 68, 295-296, 2013. http://dx.doi.org/10.1016/j.atmosenv. 2012.11.045 Image source: Wikimedia Commons
17. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Offical and tested solid fuel OA distributions wood/coal OA SNAP10 SNAP9 SNAP8 SNAP7 SNAP6 SNAP5 SNAP4 SNAP3 SNAP2 SNAP1 0 10 20 30 40 UK national total emissions, Gg/year SFOA HOA EC Other/Mineral PM Total PM2.5 emission in 2012: 82 Gg SNAP2 = 25 Gg (30%) Not including grey sources.
18. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Offical and tested solid fuel OA distributions wood/coal OA
19. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Model experiments of solid fuel OA emissions distributions wood/coal OA
20. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Residential solid fuel burning - Results wood/coal OA Base Base4x Redist combRedist North Kensington -69% -18% +33% -18% Marylebone Road -59% +14% +57% -1% Annual average normalised mean bias: (a) Detling (b) Harwell (c) North Kensington 1 2 3 4 0 3 6 9 12 15 18 21 0 3 6 9 12 15 18 21 0 3 6 9 12 15 18 21 Hour of day Mean SFOA, µg/m3 Measured Base 4xBase combRedist
21. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Elemental/black/soot carbon - measurements can’t agree either wood/coal OA spring (MAM) summer (JJA) autumn (SON) winter (DJF) 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 2.0 0.0 2.5 5.0 7.5 10.0 Harwell North Kensington Marylebone Road BC EC−T EC−R EC EC EC BC EC−T EC−R EC EC EC BC EC−T EC−R EC EC EC BC EC−T EC−R EC EC EC Mean concentration, µg/m3 Measured Base Base4x combRedist
22. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Elemental/black/soot carbon - Results wood/coal OA spring (MAM) summer (JJA) autumn (SON) winter (DJF) 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 0.00 0.25 0.50 0.75 0.0 0.5 1.0 1.5 2.0 2.5 BEL BIR CAR GLA LISB NORW STAB BC EC EC EC BC EC EC EC BC EC EC EC BC EC EC EC Mean concentration, µg/m3 Measured Base Base4x combRedist
23. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Elemental/black/soot carbon - Results wood/coal OA spring (MAM) summer (JJA) autumn (SON) winter (DJF) 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 BEL BIR CAR /m3 0.0 0.5 1.0 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 0.00 0.25 0.50 0.75 0.0 0.5 1.0 1.5 2.0 2.5 CAR GLA LISB NORW STAB Mean concentration, µg/m3 0.0 0.5 0.0 0.5 1.0 1.5 0.00 0.25 0.50 0.75 0.0 0.5 1.0 1.5 2.0 2.5 A LISB NORW STAB BC EC EC EC BC EC EC EC BC EC EC EC BC EC EC EC Mean conce Measured Base Base4x combRedist
24. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Results overview Conclusions 0 1 2 3 4 Obs. Base This work Annual average OA concentration, µg/m3 COA SOA SFOA HOA (a) SOA (Chapter 3) (b) COA (Chapter 4) (c) SFOA (Chapter 5) 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 Obs. Base This work Obs. Base This work Obs. Base This work Annual average concentration, µg/m3 Base - emissions as reported by the national emissions inventory This work - inclusion of missing species, spatial modifications to existing ones
25. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Results overview Conclusions ✤ Diesel emissions EVEN MORE complicated than thought ✤ Cooking - significant source in major urban areas; better filters? ✤ Wood and coal burning is ✤ undertaken in smoke control areas ✤ both discretionary and for heating purposes
26. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Acknowledgements PhD Supervisors Mathew Heal (School of Chemistry, UoE) Massimo Vieno (NERC CEH-Edinburgh) Stefan Reis (NERC CEH-Edinburgh) University of Manchester Dominique Young James Allan Hugh Coe University of York Rachel Dunmore Jacquiline Hamilton King’s College London David Green NERC CEH-Edinburgh Chiara Di Marco Eiko Nemitz Anais Detournay GeoSciences, UoE ian Mackenzie Georgia Institute of Technology lu Xu Sally Ng University of Gothenburg Robert Bergström David Simpson TNO: Department of Climate, Air and Sustainability (Netherlands) Jeroen J.P. Kuenen Collaborators: Aerodyne Research Inc. leah Williams Scott Herndon PhD examiners David Stevenson (GeoSciences, UoE) Sean Beevers (Kings’ College London)
27. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh Thank you for listening!
28. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh
29. R.Ots@ed.ac.uk 1-March 2017: Global Change Research Institute Seminar, University of Edinburgh The above table shows that, for 2012, NAEI reported 8 Gg of diesel- VOCs and 31 Gg of gasoline-VOCs. The measurements, however, showed that the concentration of diesel-(I)VOCs (IVOCs + VOCs) is 3.2 times the concentration of gasoline-VOCs. Based on this, diesel-(I)VOCs emission should be 31 Gg×3.2 = 99 Gg, and the updated total NMVOCs from SNAP7 therefore become 31 Gg + 99 Gg = 130 Gg. The new total (130 Gg) is 3.3 times higher than what is officially reported (39 Gg), thus the missing diesel emission is a 230% (2.3× the original, reported, amount) addition to this sector. Diesel SOA Emissions Inventory (NAEI) with the urban background ambient concentrations measured during the ClearfLo winter Intensive Observation Period in London. NAEI 2012 Measurements a Diesel-(I)VOCs 8 Gg yr 1 107 µg m 3 Gasoline-VOCs 31 Gg yr 1 33 µg m 3 Diesel/Gasoline 0.26 3.2 a Dunmore et al. (2015) SNAP 10 SNAP 9 SNAP 8 SNAP 7 SNAP 6 SNAP 5 SNAP 4 SNAP 3 SNAP 2 SNAP 1 0 100 200 300 UK national total emissions, Gg/year Other/Mix VOC emission Gasoline−VOCs emission Diesel−VOCs emission Added diesel−IVOCs