1. 1 Recent PM 2.5 Trends in Georgia André J. Butler Mercer University EVE 290L 14 April, 2008

2. 2 Background v What is PM 2.5 ? –Solid (or liquid) particles with aerodynamic diameters < 2.5  m  Why is the 2.5  m distinction used? –Fine particles: insufficient inertia to deposit in nasal passages (imbed deep within lungs) –Toxicity: particles of this size contain chemicals that may be toxic

3. 3 Background v What are the sources of PM 2.5 ? –Anthropogenic u Primary: IC engines, fireplaces, meat-cooking operations, industrial processes u Secondary: atmospheric conversions of SO 2 and NO x to SO 4 2- and NO 3 -, respectively –Biogenic u Primary: volcanoes, soil erosion, sea spray u Secondary: atmospheric conversion of organic gases to particulate organic carbon

4. 4 Background v What are the effects of PM 2.5 ? –Environmental impacts u Climate change (cooling) u Visibility degradation –Human health impacts u Morbidity (decreased lung function, increased respiratory hospitalizations, school absenteeism) u Mortality (cancer, cardiac arrest)

5. 5 Temporal Analysis of PM 2.5 Mass Data v Summary statistics

6. 6 Temporal Analysis of PM 2.5 Mass Data v Weekly-averaged and seasonal trends in PM 2.5 mass (a) Weekly-averaged trends in PM 2.5 mass (b) Seasonal variations in PM 2.5 mass

7. 7 PM 2.5 Mass Results v Temporal –Highest values in summer –Lowest values on Monday –Highest values at night and AM rush v Spatial –Overall, little spatial variation

8. 8 Temporal Analysis of SO 4 2- v Seasonal SO 4 2- variations and Spearman rank correlation coefficient matrix Sulfate R -values Summary statistics Sulfate R -values

9. 9 Temporal Analysis of NH 4 + v Seasonal NH 4 + variations and Spearman rank correlation coefficient matrix Ammonium R -values Summary statistics

10. 10 Temporal Analysis of NO 3 - v Seasonal NO 3 - variations and Spearman rank correlation coefficient matrix Nitrate R -values Summary statistics

11. 11 Major Ionic Species Results v SO 4 2- –Peak values in summer –Highest when wind from west (180-360  ) –Spatially homogeneous v NO 3 - –Peak values in winter –Spatially homogeneous v NH 4 + –Peak values in summer –Associated with SO 4 2- and NO 3 - –Spatially homogeneous

12. 12 Temporal Analysis of OC v Seasonal OC variations and Spearman rank correlation coefficient matrix OC R -values Summary statistics

13. 13 Temporal Analysis of EC v Seasonal EC variations and Spearman rank correlation coefficient matrix EC R -values Summary statistics

14. 14 Indicators of Secondary Aerosol Production v Total carbon (TC=OC+EC) to EC ratio

15. 15 Estimation of Secondary Aerosol Production v Method of Turpin and Huntzicker (1991) and Castro et al. (1999) OC sec = OC tot – OC pri OC pri = EC  (OC/EC) pri OC sec = secondary OC OC tot = total OC OC pri = primary OC (OC/EC) pri = primary ratio estimate

16. 16 OC/EC Results v Organic carbon –Peak values in summer –Lowest values on Monday –Significant secondary contribution (43-80%) –Spatially homogeneous v Elemental carbon –Peak values in summer (August anomaly) –Lowest values on weekends

17. 17 Predicting PM 2.5 Spatially-averaged model for Case 3: R 2 = 0.63

18. 18 Conclusions v Temporal variation in PM 2.5 concentrations greater than spatial  Mean 24-hour levels: 19.3 - 21.2  g/m 3 –Annual NAAQS: 15  g/m 3 (3 yrs data) v Peak levels in summer (mass and all species except NO 3 - ) –August 1999 was anomalous v Slight increase in PM 2.5 during work week v Late-night, early-morning diurnal peaks

19. 19 Conclusions, continued v Major ions accounted for ~40% of mass v OC largest single species (~31% of mass) –Secondary formation important (43-79%)