ON THE RELATIONSHIP BETWEEN PM2.5 CONCENTRATIONS AND REGIONAL-SCALE PRIMARY PM2.5 EMISSIONS FOR THE UNITED KINGDOM Dick Derwent Co-Chairman EMEP/TFMM This works was supported by the United Kingdom Department for Environment, Food and Rural Affairs. UNIFIED EMEP MODEL EMEP Task Force on Measurement and Modelling TFMM has reviewed the Unified EMEP model and recommended that: • it is currently not suitable for modelling PM10 and PM2.5 because of its coverage of PM components • it gives a reasonable description of secondary inorganic PM and primary PM components of PM2.5 EVALUATION OF PM HEALTH EFFECTS IN INTEGRATED ASSESSMENT MODELLING PM health effects are evaluated in the integrated assessment modelling carried out for EU CAFÉ and NECD programmes and for the revision of the UN ECE Gothenburg Protocol process. Attention is restricted to PM2.5 and then only to secondary inorganic and primary PM2.5 components Unified EMEP model delivers regional scale concentrations of secondary inorganic and primary PM2.5 components on a 50 km x 50 km grid How are urban PM-driven health effects to be evaluated? CONCEPT OF URBAN PM INCREMENT Urban PM2.5 increment provides the bridge between the regional scale 50 km x 50 km PM2.5 component concentrations provided by the Unified EMEP model and the urban–scale PM concentrations required by RAINS. Urban PM2.5 increment = Urban PM2.5 concentration - Regional scale PM2.5 background concentration Urban PM2.5 concentration = Urban PM2.5 increment + Regional scale PM2.5 background concentration Regional-scale background is the PM concentration that would be observed if all local PM sources were switched off. URBAN PM2.5 INCREMENT • Useful concept in policy evaluations • Entirely theoretical concept defined by modelling • No observational basis to the concept • Dependent on an understanding of scale issues • What are urban and local scales? • At what scale does the urban scale hand over to the regional scale? DETERMINING URBAN PM2.5 INCREMENTS FOR ALL THE MAIN POPULATION CENTRES OF THE UNITED KINGDOM Employed a site-specific PM source attribution model to estimate the following PM2.5 components at 132 sites: • regional secondary PM2.5 concentrations • regional primary PM2.5 concentrations • local primary PM2.5 concentrations The approach uses an empirical model, incorporating some observations and some atmospheric dispersion modelling. Details are available in Stedman et al. (2006). LOCAL PM2.5 COMPONENTS • area sources using 1 km x 1 km PM2.5 emission inventory from NAEI and ADMS model • point sources using ADMS model REGIONAL PM2.5 COMPONENTS • sulphate, nitrate and ammonium on 20 km x 20 km scale maps based on observations with addition of ‘bound water’ • secondary organic PM2.5 calculated using HARM/ELMO model for anthropogenic and biogenic components • primary PM2.5 from point sources using ADMS model • primary PM2.5 from area sources using ADMS model • primary PM2.5 from distant sources using TRACK model OTHER PM2.5 COMPONENTS • Sea salts: PM2.5 0.2 µg m-3 across UK • Residual: PM2.5 3.0 µg m-3 across UK Residual PM components include all non-modelled contributions to ambient PM2.5 concentrations, particularly wind-blown dusts, agricultural activities and construction and demolition. MODEL VALIDATION PM10 PM2.5 RELATIONSHIP BETWEEN LOCAL PM2.5 COMPONENT AND 50 KM X 50 KM GRID SQUARE PM2.5 EMISSIONS Location Regional Regional Local Total Emission backgnd primary primary µg m-3 µg m-3 µg m-3 µg m-3 Belfast Clara St 5.9 0.7 8.4 15.0 2159 Bottesford 10.3 1.5 0.8 12.6 1381 Glazebury 8.3 1.2 3.1 12.7 2037 London Bexley 10.8 1.8 3.6 16.1 2506 London Eltham 10.8 1.8 3.8 16.3 2506 London Hillingdon 10.7 1.6 4.8 17.1 1536 Manchester South 8.6 1.3 3.6 13.6 2037 Redcar 8.0 1.0 4.6 13.5 1231 Walsall Willenhall 9.2 1.3 5.8 16.3 1779 THE RELATIONSHIP BETWEEN LOCAL PRIMARY PM2.5 COMPONENT AND REGIONAL SCALE PRIMARY PM2.5 EMISSION DENSITY FROM EMEP 14 12 y = 5.5619x Local primary PM 2.5, ug/m3 R2 = -0.1351 10 8 y = 2.2712x + 2.8461 6 R2 = 0.0824 4 2 0 0.0 0.5 1.0 1.5 2.0 PM2.5 emission, tonne/km 2/yr URBAN PM2.5 INCREMENTS USED IN RAINS THE RELATIONSHIP BETWEEN LOCAL PRIMARY PM2.5 COMPONENT AND REGIONAL SCALE PRIMARY PM2.5 EMISSION DENSITY FROM EMEP 14 12 y = 5.5619x Local primary PM 2.5, ug/m3 R2 = -0.1351 10 8 y = 2.2712x + 2.8461 6 R2 = 0.0824 4 Values used 2 in RAINS 0 0.0 0.5 1.0 1.5 2.0 PM2.5 emission, tonne/km 2/yr ALTERNATIVE APPROACHES FOR RAINS • Constant local primary PM2.5 increment of 4.4 ± 3 µg m-3 • Constant ratio of local primary PM2.5 increment to regional PM2.5 emission density of 6.9 ± 5 µg m-3 per tonne km-2 yr-1 • Population-weighted local primary PM2.5 increment of 2.9 µg m-3 CONCLUSIONS It is likely that the urban PM2.5 increments have been grossly underestimated in the United Kingdom in the integrated assessment modelling performed for the EU CAFÉ programme and in UN ECE Gothenburg Protocol reviews. PM-driven health impacts in the UK have been underestimated and so have the benefits of emission controls. Alternative approaches have been described relating urban PM2.5 increments to regional 50 km x 50 km PM2.5 emissions in the UK.