1. Risks, challenges and mitigation actions in the APICE partners’ area: between the scientific findings and new governance models - Genoa M.C. Bove, P. Brotto,F. Cassola, E. Cuccia, D. Massabò, A. Mazzino, P. Prati Department of Physics – University of Genoa Final conference – Venice, 8 th November 2012

2. APICE scientific issues: the case of Genoa Main goal: to provide Authorities and Stakeholders with a reliable tool to study and forecast air quality: a “Chemical Transport Model, CTM” Methodology (shared with all the Partners): 1)“picture” of air quality (i.e. PM2.5) with a 1-year monitoring campaign  Source Apportionment (SA). 2)CTM assessment with updated emission data 3)Check of CTM vs. real-world measured data 4)Comparison of SA by CTM and monitoring campaign 2

3. Site1: C.So Firenze Site2: Multedo Site3: Bolzaneto PORT Monitoring campaign Intensive campaign (May-Oct 2011) after prevailing meteo conditons analysis

4. The PM2.5 level is almost the same in the three sites Main PM2.5 sources: at “regional” scale PM2.5 levels The correlation between PM2.5 time series is stronger for the sites much closer to the port F M B

5. PM2.5 average apportionment: Corso Firenze (14 ± 5) %

6. PM2.5 average apportionment: Multedo (12 ± 4) %

7. PM2.5 average apportionment: Bolzaneto (9 ± 3) %

8. PM2.5 apportionment at a glance Basically: ship emissions

9. Temporal behaviour of ship emissions Many ferries to the Islands

10. Meteorological preprocessor: WRF 3-domain configuration (10 km + 3.3 + 1.1 km) Simulations driven by NCEP GFS fields (0.5°) 24-hr-long simulations, hourly outputs, year 2011 10

11. 11 Chemical transport model: CAMx Maritime sector (harbour activities) Road transport Industry Non-industrial combustion plants Other sources (including natural emissions) Outer domain covering Western and Central Europe (10 km resolution) 2-way nesting procedure Inner domain – focus on local area 47x47 grid points 1.1 km resolution PM source apportionment approaches: Zero-out CAMx PSAT routine City area Harbor area Pollutants: NOx, SOx, CO, PM….

12. Emission data Large-scale anthropogenic emission data provided by AUTH (TNO data processed through the MOSESS code) Natural emissions obtained processing WRF outputs with the NEMO code (developed by AUTH) Updated (2010) harbour emission data calculated by Techne Srl (provider of Province of Genoa) according to CORINAIR Guidebook 2011 (no disaggregation for different harbour activities contribution available) Local gridded emission data provided by Liguria Region (reference year 2008): 1 km spatial resolution hourly temporal resolution SNAP sectors disaggregation 12

13. Model validation – comparison with observed data (PM 2.5 ) 13

14. Model validation – comparison with observed data (Sulfates) 14

15. Model validation – comparison with observed data (NO x ) 15

16. 16 CTM source apportionment results (zero-out) PM2.5 NOx Contribution restricted to the area around the harbour (expecially for PM2.5) Contribution of harbour activities (%) Summer 2011

17. 17 CTM source apportionment results (zero-out) PM2.5NOx Contribution of road transport (%) Summer 2011 Contribution to concentrations over the whole city

18. 18 Sources SA by measured data (PMF) SA by CTM (CAMx with PSAT) Maritime (13 ± 5) % coast (9 ± 3) % inland 9% coast 5% inland Industrial (30 ± 10) %20% Road Traffic (40 ± 15) %45% Residential combustionNot resolved5 %5 % Others (crustal, sea, etc. )(15 ± 5) %20% SA of PM2.5: June- August 2011 - Intercomparison ± ???

19. 19 Harbor activities contribution to PM 2.5 concentrations CTM vs Receptor models CTMReceptor Models Cso Firenze 11 %(14 ± 5) % Multedo 9 %(12 ± 4) % Bolzaneto 4 %(9 ± 3) %

20. 20 -20 % +2 % Future scenario analysis: PM2.5 Scenario 1 – 2020 without mitigation actions

21. 21 - 35 % - 5 % Future scenario analysis: PM2.5 Scenario 2 – 2020 with S % reduction in fuels

22. 22 - 40 % - 5 % Future scenario analysis: PM2.5 Scenario 3 – 2020 with S % reduction in fuels and cold ironing of container and ferries terminal

23. 23 Summary A quite complete picture of PM2.5 levels and sources for the year 2011 has been obtained thanks to a considerable experimental effort A CTM model has been implemented and put in operation: validation vs. measured data pretty good Source apportionment by real-world data + receptor model (PMF) and CTM (WRF+CAMx) in fair agreement Future scenarios according to stakeholders inputs and APICE methodology completed (reference year 2020)