August 1995 Forest Fire Impacts in New England and Atlantic Canada

Rich Poirot, VT DEC, Al Leston, CT DEP; Cliff Michaelsen, ME DEP

Abstract

Large forest fires occur less frequently in Northeastern North America than in more arid regions to the West. However, Northeastern fires do occur periodically and can have significant impacts on regional air quality for particulate matter and a variety of gaseous pollutants. Developing a better understanding of historical fire impacts in the Northeast may provide useful insights for identifying effects of future fire events and for developing future air quality management strategies for fine particles, regional haze and other associated pollutants.

During late August of the hot, dry summer of 1995, a number of forest fires broke out in the Northeastern US and Eastern Canada. In the Province of Quebec alone, fires ignited during the 12-day period of 8/12/95-8/20/95 consumed over 150,000 hectares. Maximum (known) PM-10 impacts approached 500 ug/m3 for a 1-hour (TEOM) PM-10 monitor in Presque Isle, ME on the early morning of Tuesday 8/22/95, at approximately the same hour that an unusually high gaseous mercury concentration was observed at Kejimkujik NP, NS. 24-hour PM-10 concentrations in the range of 100-150 ug/m3 were observed at Presque Isle on 8/22/95, with PM-2.5 concentrations in excess of 50 ug/m3 in St. John, NB and 40 ug/m3 in Halifax, NS. While concentrations appear to have declined on the following day (Wednesday, 8/23/95) unusually high PM-2.5 concentrations, composed predominantly of organic carbon compounds, were recorded at a number of IMPROVE sites which sampled that day. Several days earlier, unusually high nighttime isoprene, CO and NOx concentrations were observed at several PAMS sites in CT and MA - which also appear to be associated with forest fire(s) impacts.

Understanding the specific origin of the fire impacts is challenging due to the number of different fires burning concurrently, and also because of complex meteorological flow patterns resulting from the presence of hurricane Felix of the North Atlantic coast. Because there were a number of strong, clear "signals" of woodsmoke (in air that was otherwise relatively clean), and relatively detailed spatial and temporal information in the relevant monitoring data, additional analysis of this episode period may be helpful in:

- discerning detailed "fingerprints" of forest fire impacts in both speciated PM-2.5 and gaseous VOC data

- recognizing future forest fire influences during less "obvious" fire events

- clarifying discrepencies between emission inventories and measured isoprene for ozone model episodes

- evaluation of meteorological model performance under difficult flow conditions.

Figures and Animations

Figure 1. Locations of Major Quebec Forest Fires starting between 8/12/95 and 8/23/95

Figure 2. Hourly PM-10 at Presque Isle ME, and Gaseous Hg at Kejimkujik, NS: 8/16-26/95

Figure 3. NOAA HY-SPLIT back trajectories for Presque Isle for hour of maximum PM impact

Figure 4. 24-Hour PM-10 Mass Concentrations at Selected Maine Sites: 8/14-30, 1995

Figure 5. 24-Hour PM-2.5 Mass Concentrations at Selected NE Sites: 8/19-26, 1995

Figure 6. Regional PM-2.5 Composition on 8/23/95, with Summer '95 Time Series at Acadia National Park

Figure 7. PM-10 and PM-2.5 Composition at Acadia NP on 8/23/95

Figure 8. Isoprene at East Hartford, CT for 8/95, showing atypically high levels on night of 8/20

Figure 9. Diurnal Isoprene Cycles on 8/20/95 and averaged for August 1995 at selected sites

Figure 10. Diurnal Carbon Monoxide Cycles on 8/20/95 and averaged for August 1995 at selected sites

Figure 11. Diurnal NO₂ Cycles on 8/20/95 and averaged for August 1995 at selected sites

Figure 12. Carbon Monoxide at Hartford, CT and averaged for all CT and MA sites, 8/15-25, 1995

Figure 13. NOAA HY-SPLIT back trajectories for E. Hartford, CT for hour of maximum Isoprene impact

Animation 1. Visible Satellite images suggesting possible fire locations: 8/17-23, 1995 (view unfortunately cropped to exclude parts of eastern Canada)

Animation 2. CAPITA Monte Carlo Back Trajectories for Halifax, NS, Presque Isle, ME and E. Hartford, CT 8/17-23/95

Animation 3. Surface Wind Speed and Direction vectors (from AIRS met data) 8/17-23, 1995