GA Smoke Evansville Impact
From Datafedwiki
Back to GA Smoke Event
Contents |
This page contains analysis of the May 2007, Georgia smoke impact. Specifically, the focus is on the area surrounding Evansville,IN on May 29, 2007.
[edit] A: Event Description - PM2.5 Exceedance (>15ug/m3 or > 35ug/m3 )
Station 180372001, DUBOIS, IN had FRM PM concentration of:
- May 30, 2007, 31.8 ug/m3
- May 29, 2007, 39.5 ug/m3
- May 28, 2007, 34.2 ug/m3
- May 27, 2007, 30.5 ug/m3
- May 26, 2007, 41.5 ug/m3
General description of the May 2007 GA smoke event is here. The FRM PM mass concentration at the station 180372001, DUBOIS, IN, was 30-42 ug/m3 during the period May 26-May 30, 2007.
These high PM concentrations could have contributed either to the daily NAAQS (35 ug/m3) or to the yearly average NAAQS (15 ug/m3).
[edit] B: Clear Causal Relationship between the Data and the Event
[edit] Backtrajectories, May 27-30, 2007 -Click to enlarge
Backtrajectories were constructed using the available PM2.5 FRM dataset. For the days May 27-30, there was a strong uniform flow toward Evansville, IN from the Atlantic and passing through Georgia.
The above sequence shows the transport pattern from the Georgia fire toward the receptor sites. The trajectories are colored blue prior and colored rainbow after passing through the source rectangle.
The forward transport trajectories indicate that the smoke from the GA fires followed a curved trajectory toward Indiana and Illinois.
The above sequence shows the transport pattern to and from a rectangular source area placed over the Ohio River Region. The source area represents potential sources of sulfate to the receptor site near Evansville, IN. The trajectories are colored blue prior and colored rainbow after passing through the source rectangle. These trajectories indicate that the inflow to the Ohio River Region is from the S-SW. The outflow is toward the NE.
Hence, the Ohio River Region is not a significant source of sulfate art the receptor site.
The above sequence shows the general transport pattern over the Eastern US during May 27-30.
[edit] AIRNOW PM2.5 Time Series
The official monitoring networks for PM2.5 FRM, EPA speciation and IMPROVE have only 3 or 6 day sampling intervals and provide only daily average values. The cause of the exceedances and violations during the non-sample days is difficult to establish and to quantify. The continuous air monitors for the AIRNOW network provide the means to "fill the gaps" in between the routine sampling days. The hourly AIRNOW maps also provide evidence for the movement, ie. transport direction of aerosol palls.
On May 29, 2007, several 2.5 AIRNOW PM2.5 monitoring stations in the vicinity of Evansville, IN show concentrations in the excess of 35 ug/m3. Prior and after this "event" the concentrations were about 20 ug/m3 at all three stations shown below. Note, that all the times for the AIRNOW data are given in UTC, ie. Eastern + 5 hours
Location and time series for AirNow Stations 181630012, 210590005, 212218001.
The above charts show the time series for three sites, including the receptor site under consideration (181630012), as well as an upstream site in SW Kentucky (212218001).
It is evident that on May 29, 2007 the above three sites all indicate a peak in excess of 40 ug/m3. It also shows that the peak at SW Kentucky occurred 6-8 hours prior to the peak near Evansville, IN (This is not very strong argument. The spatial maps are more compelling.)
[edit] AIRNOW PM2.5 Hourly Concentration Map Sequence May 28 18:00 - May 29 22:00
The AIRNOW PM2.5 maps show the concentration pattern in four-hour increments. (The times are in UTC). The first map at May 28, 1800 Eastern time, shows that the high PM concentration pall is located just SE of the receptor site. As the time progressed, the aerosol pall was drifting toward the North.
The AIRNOW data provides evidence that on May 28 and 29, the smoke pall was transported from the South and passed over the receptor area.
[edit] MODIS Satellite AOT, May 27, May 28, May 29
The satellite-derived aerosol optical thickness is patchy due to the extensive cloudiness. Nevertheless, the patchy satellite data show the smoke over the central part of the Eastern US. (This is only corroborating evidence.)
[edit] C: The Event is in Excess of the "Normal" Values
[edit] VIEWS
The chemical data from speciated aerosol samples provide the most significant line of evidence for quantifying this exceptional event and its impact at the receptor site. Chemical data for the particular receptor site are not available. Hence, the composition at that site has to be inferred from the speciated data at neighboring locations.
The above charts show the OC concentration at the IMPROVE sites BOND1, LIVO1, CADI1. The time series are overlayed for the three sites to show the spatial coherence of the OC-pall impacting all three sites. The OC levels on May 27, and May 30, ranged between 5-10 ug/m3, compared to background level of about 2 ug/m3.
Hence, there is an excess OC of 3 -7 ug/m3. Assuming an OC-mass scaling factor of 1.8 for smoke, this corresponds to 5.4-12.6 excess mass during the smoke passage.
It is evident that during May 27 and May 30 at these three IMPROVE sites the sulfate aerosol concentration was about normal for these locations and season.
[edit] D: The Exceedance or Violation would not Occur, But For the Exceptional Event
Based on the above evidence, the PM2.5 concentration peak on May 29, 2007 was contributed by the addition of the GA smoke spike which amounted to at least 5-12 ug/m3. This excess can be considered to be the exceptional event contribution.
