Air Quality trend analyses under EMEP/TFMM and link to EEA work Augustin COLETTE (INERIS), Chair of the TFMM/CLRTAP TFMM National Experts, CCC, MSC-E, MSC-W, ETC/ACM 20th EIONET Air Quality Management and Assessment Workshop Overview TFMM Trend Assessment Methodology Measurement Modelling (Eurodelta) Link with EEA-ETC/ACM trend work Torseth et al., ACP, 2012 TFMM Trend Assessment 2014-2015 Spring 2014:TFMM Annual Meeting Initial trend assessment studies Discussion on methodologies Agreement on methodology & modelling setup Draft analyses & definition of writing teams Drafts results + writing report Review Publications Fall 2014:Trend Workshop Winter 2014-2015 Spring 2015 : TFMM Annual Meeting Fall 2015: EMEP Steering Body Winter 2015-2016 Spring 2016 : TFMM Annual Meeting Trend methodology for observations Specifications Time periods: 1990-2012 / 1990-2001 / 2002-2012 Data completeness: 75% of annual coverage and 75% of years covered Statistical analysis: Concentrations and wet deposition fluxes of Pb, Cd, Hg and B[a]P + other POPs: biexponential on the basis of monthly values; Ozone (avg & max 8hr), NO2, PM10, PM2.5, SO2, sulphate, nitrate and carbonaceous atmospheric concentration and sulfur and nitrogen precipitation chemistry: MannKendall and Sen-Theil slope on the basis of annual values & by season Process Centralised analysis (Centres) published on open TFMM wiki Supplemented by national expert knowledge of (i) data issues, (ii) need for filtering, (iii) non-linearity and changes of slopes in the record, (iv) additional compounds, (v) add or discard sites, (vi) changes in seasonal cycles, (vii) relation with local emission changes Trend methodology for models: EURODELTA 6 regional CTMs + EMEP/MSC-W Common setup Meteorology Emissions (GAINS/ECLIPSE) Boundary conditions ED3-Trends 3 tiers of experiments 1990/2000/2010 reference [7] Sensitivity to emissions [6] Sensitivity to boundary conditions [3] Full 20yr hindcast [2] EURODELTA-Trends Key Questions Model validation How do model compare with observations in 1990, 2000, and 2010 ? How do models capture the trend in observations ? Can we conclude on the uncertainties in emissions? European policy effectiveness What if no emission change occurred in Europe ? Did “potency” changed over the past 20yrs? Can we identify changes in SIA formation ? Non-European influence What if no emission changed beyond Europe ? What is the uncertainty related to boundary conditions ? Meteorological variability Does meteorological variability contribute to the AQ trend over the past 20 yrs? Impacts Link with effect community (health & ecosystems) What model dynamical evaluation means for IAM? O3: Average trend Composite of observations at 54 EMEP sites with valid observations over 1990-2012 Annual means were increasing in the 1990s Ozone peaks decreased since 2000 O3: Statistics over the EMEP network Relative Change (%) 1990-2000 2001-2012 Annual Mean 7.7 -5.3 JJA avg of daily Max -6.9 -12.9 SOMO35 10.1 -32.5 Fraction of stations with significant trend (%, pval<0.05) Substantial average changes, yet unsignificant trends short time period & interannual variability AOT40 -21.7 -42.0 EMEP&AIRBASE Complementarity AIRBASE: ~230 RU background stations 2002-2012 EMEP: ~50 RU background stations 1990-2012 EMEP / 1990-2001 (54) EMEP / 2002-2012 (54) AIRBASE / 2002-2012 (231) Consistent conclusion on the O3 trends in 2000’s when expanding to airbase coverage O3: Lower photochemical production Larger decreases at EMEP sites where O3 levels where higher Courtesy D. Derwent Negative trends for mean O3 in summer, even at urban sites ETC/ACM - EEA O3: EURODELTA : modelled exposure 6-RCTM median pop-weighted average SOMO35 2000 Relative Change 2010 vs 2000 2010 Std dev Sulphur & Nitrogen Compounds General decrease, some gaps between model&observations Ox. Sulphur > Ox. Nitrogen > Red. Nitrogen Particulate Matter Average Trend at EMEP sites 2002-2012 Model Observed PM10 -2.6% / yr 2.4% / yr PM2.5 -3.7 % / yr 3.0 % / yr Solid line: Measurements Dottet line: EMEP model Confirmed at AIRBASE sites, with slighly larger trends close to the sources ETC/ACM - EEA PM2.5 Modelled Exposure (EURODELTA) 1990 Average population weighted exposure by country 6-model median ensemble in 1990 and 2010 2010 EURODELTA3 PM2.5 Modelled Exposure (EURODELTA) Change in exposure between 1990 and 2010 Left: relative change in the EMEP model (%) Right: std dev. of the relative change in the 6-model ensemble EURODELTA3 BaP Trends: example of Germany Air concentrations (modelling) Emissions (official data) 160 3 Total reduction Growth from 2005 to 2012 1.2 68% 22% Emissions, t/y 1 0.8 0.6 0.4 0.2 Value Total reduction Growth from 2005 to 2012 140 120 80 60 40 Value Trend 20 Trend 0 77% 31% 100 8 DE1 2012 2010 2008 2006 2004 2002 2000 DE9 ng/l 6 4 2 4 2 0 Total reduction 57% Growth from 2005 to 2012 22% 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 ng/l 1998 8 Total reduction 56% Growth from 2005 to 2012 29% 6 Trend analysis of measurements at German sites 1996 1994 1992 2012 2010 2008 2006 2004 2002 2000 1998 1996 1994 1992 1990 0 1990 Air concentrations, ng/m 1.4 Heavy Metal: total reduction (1990-2012) Lead Wet dep. Air conc. Mercury Wet dep. Air conc. 100 100 80 80 80 60 40 20 0 -20 60 40 20 0 -20 Total reduction, % 100 Total reduction, % Total reduction, % Air conc. Cadmium 60 40 20 0 -20 Wet dep. Heavy Metals Trends HMs: Relative increase of the importance of non-EMEP and secondary sources Conclusion Main messages O3: PM & SIA Larger decrease in the 1990s Next steps Larger decrease in the 1990s HM & POPs increase in baseline in the 1990s decrease of the peaks in the 2000s Relative small fraction of sites with significant trends Full draft end of 2015 Publication Spring 2016 Ongoing Eurodelta analysis Collaboration with other initiatives EEA-ETC/ACM CLRTAP AR WGE Trend Report TOAR