0708EUSEvent
From Datafedwiki
Backlink to Events
{{{EventTitle}}}
 Type: {{{EventType}}} | Location:Eastern U.S. | Dates: 2007-08-01 - 2007-08-12 | Lat: , | Lon: , |
Contents |
[edit] August 2007 Eastern US Air Pollution Event Preliminary Summary
[edit] Air Pollution Event Analysis
The analysis and of air pollution episodes ranges from cursary evaluation of available evidence to detailed study of causal factors and processes. The event analysis can also be performed, before, during or after the event itself.
- Forecasting. This is primarly a modeling process, projecting the event several days in advance. It is not consisered here.
- Near-Real Time Analysis. This analysis is conducted ind the days/weeks after the event using the best available observations, models and prior knowedge.
- Post Analysis. This detailed analysis is conducted months or years after event. It incorporates all the available data and models, including the 'slow' chemical, satellite and dignostic model results.
[edit] Near-Real Time Regional PM2.5 Pattern based on AIRNOW
This current effort is focused on the second, Near-Real Time Analysis of air pollutant events. It includes the following steps:
- The first step is to establish magnitude of the concentration pattern. If the values are near over the NAAQS. The concentration ranges, locations and time periods above the NAAQS need special attention.
- Next, the scope of 'event' or 'episode' is identified and the spatio-temporal pattern of the event is established.
- The third step is to begin the evaluation of the possible contributiong causes of the event.
The existence of an air pollution event is established through the analysis of the real-time AIRNOW PM2.5/Ozone monitoring network data. A convenient initial entassessment can be performet by spatio-temporal aggregation of the monitoring data.
1. Spatially Aggregated Time Series. The individual station data for each hour is averaged over a geographic region to yield the pattern of regional time series. The time range of events can be accessed by inspection of the regional time series (See EUS average below).
2. Time-Averaged Spatial Pattern. The data are averaged over the desired time period to yield the time averged spatial pattern. This analysis yield the spatial extent of the event.
3. Space-Time Iteration. Since neither the space or temporal pattern of the event is knwn, it is necessary to iterate the above analyses. Initally, the 1. Spatially Aggregated Time Series provides the general time range of the event. For that period, the 2. Time-Averaged Spatial Pattern can prvide the areas where the concentrtions were most severe. Next, step 1 can be repeated for the impacted regions, etc...
If the regional average concentration is near or above the NAAQS (35 ug/m3 for PM2.5), then some stations will be in violation of the standard.
For this analysis, the period aggregated time series is June - September 2007. Initially, only PM2.5 is considered over the Eastern US.
[edit] Eastern US Average
The average PM2.5 stations over the entire EUS shows a steady concentration increase from 10 ug/m3 on July 20 to about 30-35 ug/m3 by Aug 4-9. This is accumulation is indicative of a stagnation episode. The main period of the episode is August 1-12.
[edit] TN-KY Average
The PM2.5 event was most pronounced in the TN-KY multistate region. The peak regional concentrtion exceeded 35 ug/m3 for several days.
[edit] GA-AL Average
Further south over GA-AL, the conentrations also peked on August 5-6. Hoerver, there was also a secondary peak in the period Aug 12-20.
The above cursary analysis establishes with reasonable certainty that during the Aug 1-12, 2007 air pollution event, the the PM2.5 concentrations have exceeded the NAAQS for PM2.5 over numerous stations over the Eastern US. Therefore, better understanding of those exceedances can:
- aid effective mitigation of those vioaltions
- provide valid evidence for exceptional event waivers
- help future forecasting or charactarisation of such events
[edit] Meteorology Based on Surface Observations
Meteorological variables are key factors that influence the evolution of air pollution episodes. The meteorological data are derived from the surface meteorological network. The meteorological data console includes the maps for visibility (FBext), temperature (T), Dew point (TD), Sea-Level Pressure (SLP), Wind Speed (SPD), and Wind Vector. For general reference the meteorological console also includes the color maps from the MODIS sensors on Terra (10AM) and Aqua (1PM) satellites. The variable of particular interest is the wind speed since it is an indicator of air mass ventilation (inverse of stagnation)
2007-08-01 |
2007-08-02 |
2007-08-03 |
2007-08-04 |
2007-08-05 |
2007-08-06 |
The images above are snapshots of the meteorological console for six days, August 1-6, 2007. During the period August 1-4 there is a persistent stagnation over much of the Eastern U.S., south of the Ohio River. Over the same period there was a rise in the maximum surface temperature which exceeded 100 degrees over parts of the Southeast.
Links to the Meteorolgoical Consoles:
070801 - Console PDF
| 070802 - Console PDF
| 070803 - Console PDF
| 070804 - Console PDF
| 070805 - Console PDF
| 070806 - Console PDF
| 070807 - Console PDF
| 070808 - Console PDF
| 070809 - Console PDF
| 070810 - Console PDF
| 070811 - Console PDF
| 070812 - Console PDF
[edit] Fire Activity and Smoke
The fire activity during the pollutions event is accessed using data from the MODIS Rapid Response System Global Fire Maps. The images below represent the accumulated fire pixels over 10-day periods.
06/30/07 - 07/09/07 (2007181-2007190) |
07/10/07 - 07/19/07 (2007191-2007200) |
07/20/07 - 07/29/07 (2007201-2007210) |
07/30/07 - 08/08/07 (2007211-2007220) |
08/09/07 - 08/18/07 (2007221-2007230) |
The images above show the fire pixel maps before,during, and after the August 07 air pollution event. These observations indicate that during the 50-day period June 30- August 18 the fire activity throughout the Eastern U.S. was rather modest. Also the pattern of fire pixels during July 30-August 8 was similar to other 10 day periods before and after the event.
For comparison the figure below shows an intense biomass burning period, showing the April 1-10, 2007 grass burning in Kansas.
The evidence from MODIS satellite-based fire pixels indicates that the fire activity during the August 1-12 period was modest and distant from the main PM2.5 hotspots (KY, TN,GA, AL) over the eastern U.S.
[edit] Sulfate Based on Model Simulations
The overall pattern of PM concentrations and related variables are shown in the consoles linked below. The maps on each console include chemical transport simulations from two global models, GOCART and NAAPS. The simulated sulfate concentrations indicate that the highest the highest model levels coincide spatially with the highest PM2.5 concentrations. However, there is considerable deviation between the sulfate patterns simulated by the two models.
This section will be extended further. Links to the Pollution Consoles: 070801 - Console PDF | 070802 - Console PDF | 070803 - Console PDF | 070804 - Console PDF | 070805 - Console PDF | 070806 - Console PDF | 070807 - Console PDF | 070808 - Console PDF | 070809 - ConsolePDF | 070810 - ConsolePDF | 070811 - Console PDF | 070812 - ConsolePDF
[edit] Event Summary
The images above are snapshots of the meteorological console for six days, August 1-6, 2007. During the period August 1-4 there is a persistent stagnation over much of the Eastern U.S., south of the Ohio River. Over the same period there was a rise in the maximum surface temperature which exceeded 100 degrees over parts of the Southeast.
Since the PM concentration was rising during this stagnation period it is evident, that the concentration increase had to result from the accumulation of local emissions rather than caused by long-range transport. The main pollutant emissions in the Kentucky, Tennessee region is from sulfur oxides therefore it is possible that the concentration rise was due to sulfurous sources. This is coroborated by the high sulfate concentrations simulated by the chemical transport models.
It is doubtful that biomass smoke was the primary cause of the PM event. During the August 1-4 period the fire activity was relatively low and steady, mostly over the Kansas-Arkansas border.
Given the available evidence, it is here suggested that the overwhelming fraction of the high PM over the KY, TN, GA, AL region is contributed by anthropogenic sulfate. However, it is conceivable that at some stations even small amount of smoke could raise the ambient concentration from below to above the NAAQS of 35ug/m3.
[edit] User-Contributed, Harvested Resources
[edit] Automatic Harvesting of Florida Smoke-relavent Resources
Blogs with tags Florida + Smoke posted between 2007-05-07 to 2007-05-12.
- Google Blog Search:
- Flickr
[edit] Additional Human Filtering of Automatically Harvested Resources
By scanning the harvested resources (blogs and videos) deamed relevant to event analysis were tagged in del.icio.us. This added an additional human selection beyond automatic harvesting.
- Del.icio.us
- YouTube:
[edit] Data Resources
[edit] DataFed-Google Earth Mashup
| End Date | 12 August 2007 + |
| EventDescription | {{{EventDescription}}} |
| EventTitle | {{{EventTitle}}} |
| EventType | [Oops! “{{{EventType}}}” is not in the list of possible values (Dust, Smoke, Pollution) for this attribute] |
| Lat max | [Oops! “” is no integer number] |
| Lat min | [Oops! “” is no integer number] |
| Location | Eastern U.S. + |
| Lon max | [Oops! “” is no integer number] |
| Lon min | [Oops! “” is no integer number] |
| Start Date | 1 August 2007 + |
