Client: Mr. Charles Pietarinen NJ Department of Environment

1. A Voyage of Discovery Using Statistics: An Assessment of NJ Mercury Air Pollution Levels Preliminary Analysis by: Fawn Hornsby1, Charles Rogers2, & Sarah Thornton3 1,3North Carolina State University 2University of Texas at El Paso Team VIGRE Client: Mr. Charles Pietarinen NJ Department of Environment Faculty Mentor: Dr. William Hunt Graduate Assistant: Mr. Andrew Moore Tuesday, July 25, 2006 @ North Carolina State University

2. Overview • Background • NJDEP Project • Dealing with Instrument Peculiarities • Previous Analysis • Objectives • Data Sources • Assessment I (Overall) • Assessment II (Transport)

3. Background

4. NJDEP Mercury Project • The NJDEP established mercury air monitoring sites in Elizabeth and New Brunswick, NJ to better understand this problem • Complex and extensive mercury data sets were collected using the Tekran Continuous Mercury Analyzer • Elemental, particulate, and reactive gas mercury were measured at each city during 2004 and 2005.

5. Dealing with Instrument Peculiarities • Because of a peculiarity in the way the instrument operates, the mercury levels were measured beginning at random start times after midnight, so that no two measurements were collected during the same time period • Each measurement was assigned to one of 24 hourly time blocks based on the midpoint of the observed hour • For correlation analysis, observations in New Brunswick and Elizabeth were matched by date and were examined for one to up to 24 hours out of phase.

6. Previous Analysis • Found Diurnal Patterns and compared with them with other forms of air pollutants. • Checked for yearly patterns • Utilized meteorological variables to find locations of possible sources. • Checked for regional effects between New Brunswick and Elizabeth.

7. ~Finding Diurnal Patterns~ Hourly Patterns Throughout the Day Elemental Mercury Box Plot Without outliers: Particulate Box Plot Without outliers: Reactive Gas Mercury Box Plot Without outliers:

8. Daily Averages of Elemental Mercury at New Brunswick versus Elizabeth ELEMENTAL MERCURY WITH Outliers: Correlation: r = 0.1769 R-Squared Value: r^2 = 0.0313 WITHOUT Outliers: Correlation: r = 0.174 R-Squared Value: r^2 = 0.0306 PARTICULATE MERCURY WITH Outliers: Correlation: r = 0.571 R-Squared Value: r^2 = 0.3262 WITHOUT FARTHEST Outlier: Correlation: r = 0.6912 R-Squared Value: r^2 = 0.4778 REACTIVE GAS MERCURY WITH Outliers: Correlation: r = 0.514 R-Squared Value: r^2 = 0.2719 Daily Max Hour: Correlation: r = 0.37661 R-Squared Value: r^2 = 0.1418

9. Source Graphs For Elizabeth • Particulate Mercury: • Highest concentrations from 180-230 and 240-270 degrees • Elemental Mercury: • Highest concentrations from 140-170, and 190-220 degrees • Reactive Gas Mercury: • Highest concentrations coming from the directions of: 0-20, 120-160, and 220-280 degrees

10. Source Graphs For New Brunswick • Particulate Mercury: • Highest concentrations from 70-100 and 240-290 degrees • Elemental Mercury: • Highest concentrations from 40-90 degrees • Reactive Gas Mercury: • Highest concentrations coming from the directions of: 60-100, and 200-240 degrees

11. Objectives Analysis I: • Re-examine diurnal patterns for the year as well as all seasons. • Examine the day of the week effect • Explore the relationship between the three phases of mercury, and ozone, temperature, fine particulate matter, and precipitation with emphasis on seasonal variations. Analysis II: • Analyze the mercury data in order to determine if pollution transport effects mercury levels in New Brunswick and Elizabeth • Focus on mercury pollution transport to answer a number of questions such as: • Is mercury air pollution regional in nature? • Do sources affect air monitoring sites in both Elizabeth and New Brunswick simultaneously or are they out of phase by an hour or more?

12. Data Sources • Wind speed & direction, temperature, and precipitation were examined along with ozone and fine particulate matter (from USEPA). • Ozone data are measured at two locations: Rutgers University as well as Bayonne, New Jersey • Temperature data was measured at NWS Cooperative Observing Station in New Brunswick and also measured at the Newark Airport in Newark, New Jersey.

13. Analysis I Another Look at Diurnal Patterns • All Diurnal Patterns use a 50% completeness criteria. • High values above 2 standard deviations were removed in order to better see the distributions regarding the seasonal diurnal patterns. • The Winter months consisted of December, January, & February for 2004 and 2005. • The Summer months consisted of June, July, & August for 2004 and 2005.

14. Another Look At Diurnal Patterns: Elemental Mercury Summer Ozone Season Elizabeth, 2005 Yearly Diurnal Patterns in Elizabeth: Yearly Diurnal Patterns in New Brunswick:

15. Another Look At Diurnal Patterns: Particulate Mercury Summer Ozone Season Elizabeth, 2004 Yearly Diurnal Patterns in Elizabeth: Yearly Diurnal Patterns in New Brunswick:

16. Another Look At Diurnal Patterns: Reactive Gas Mercury Summer Ozone Season Elizabeth, 2004 Yearly Diurnal Patterns in Elizabeth: Yearly Diurnal Patterns in New Brunswick:

17. Elemental Mercury Values During the Winter Seasons New Brunswick: Elizabeth:

18. Elemental Mercury Values During the Summer Seasons New Brunswick: Elizabeth:

19. Particulate Mercury Values During the Winter Seasons New Brunswick: Elizabeth:

20. Particulate Mercury Values During the Summer Seasons New Brunswick: Elizabeth:

21. Reactive Gas Mercury Values During the Winter Seasons New Brunswick: Elizabeth:

22. Reactive Gas Mercury Values During the Summer Seasons New Brunswick: Elizabeth:

23. Elemental Hg Day of the Week Effect

24. Particulate Hg Day of the Week Effect

25. Reactive Gas Hg Day of the Week Effect

26. 2 Seasons Per Year Winter Months: October ~ March 2. Summer Months: April ~ September 4 Seasons Per Year 1. Spring: March ~ May 2. Summer: June ~ August 3. Fall: September ~ November 4. Winter: December ~ February Seasonal Variations In order to better see the seasonal patterns of ozone and temperature as it relates to each mercury phase, we divided the year into four seasons and two seasons.

27. Examining Temperatures versus Elemental Mercury in New Brunswick Scatter Plot: Box Plot:

28. Seasonal Variations of Elemental Mercury in New Brunswick

29. Ozone versus Elemental Mercury Elizabeth: New Brunswick:

30. Relationship Between Elemental Mercury, Ozone, & Temperature Elizabeth: New Brunswick:

31. Examining Temperatures versus Particulate Mercury in Elizabeth Scatter Plot: Box Plot:

32. Seasonal Variations of Particulate Mercury in Elizabeth

33. Examining Temperatures versus Particulate Mercury in New Brunswick Scatter Plot: Box Plot:

34. Seasonal Variations of Particulate Mercury inNew Brunswick

35. Ozone versus Particulate Mercury Elizabeth: New Brunswick:

36. Relationship Between Particulate Mercury, Ozone, & Temperature Elizabeth: New Brunswick:

37. Examining Temperatures versus Reactive Gas Mercury in Elizabeth Box Plot: Scatter Plot:

38. Seasonal Variations of Reactive Gas Mercury in Elizabeth

39. Examining Temperatures versus Reactive Gas Mercury in New Brunswick Box Plot: Scatter Plot:

40. Seasonal Variations of Reactive Gas Mercury in New Brunswick

41. Ozone versus Reactive Gas Mercury New Brunswick: Elizabeth:

42. Relationship Between Reactive Gas Mercury, Ozone, & Temperature Elizabeth: New Brunswick:

43. Mercury versus Precipitation ~Elizabeth~ ~New Brunswick~ ~Elemental~ ~Particulate~

44. Mercury versus Precipitation ~New Brunswick~ ~Elizabeth~ ~Reactive Gas~

45. Continuous PM fine Measurements versus the Particulate Mercury Daily Averages in New Brunswick Particulate Mercury Daily Average Particulate Mercury Daily Average Particulate Mercury Daily Average

46. Continuous PM fine Measurements versus the Particulate Mercury Daily Averages BY SEASON in New Brunswick Winter: October ~ March R-squared = 0.3408 R-value = 0.5838 Particulate Mercury Daily Average Summer: April ~ September R-squared = 0.0305 R-value = 0.1746 Particulate Mercury Daily Average

47. Continuous PM fine Measurements versus the Particulate Mercury Daily Averages in Elizabeth Particulate Mercury Daily Average Particulate Mercury Daily Average

48. Conclusions • For both cities, Elemental and RGM have higher concentrations in the Spring and Summer months while Particulate Mercury has the highest concentrations in the Winter and the lowest concentrations in the warmer months. • Temperature is positively correlated with Elemental and Reactive Gas Mercury while negatively correlated with Particulate. • Precipitation is negatively correlated with all three phases of Mercury. • Describe the relationship between the 3 phases of hg and particulate matter and ozone

49. Recommendations • When comparing data collected with the Tekran Continuous Analyzer, we recommend assigning each measurement to one of 24 hourly time blocks based on the midpoint of the observed hour • Co-locate mercury instruments with fine particulate, ozone, and meteorological variables in order to better understand the physical processes.

50. New Jersey’s Pervasive Ghost: An Assessment of Lethal Ambient Air Mercury Transport Preliminary Analysis by: Charles Rogers1, Fawn Hornsby2, & Sarah Thornton3 1University of Texas at El Paso 2,3North Carolina State University Team VIGRE Client: Mr. Charles Pietarinen NJ Department of Environment Faculty Mentor: Dr. William Hunt Graduate Assistant: Mr. Andrew Moore Tuesday, July 25, 2006 @ North Carolina State University