Atlanta Olympic Games Ozone levels in Air and Asthma Rates.

3 August 2004 (starting at the bottom of the page readers can see my previous postings exploring the data relevant to the claims linking reduced asthma to reduced ozone due to reduced traffic)
Claims have been given wide creedence on the internet that asthma was reduced in Atlanta during the Olympic Games of 19th July to 4 August 1996 due to reduced traffic lowering air pollution, see downloadable pdf report at; http://www.eere.energy.gov/cleancities/conference/2001/pdfs/Michael_Friedman.pdf
The above presentation is based on a paper originally published in 1999 in the Journal of the American Medical Association (JAMA)
Another view by University of Georgia Professor Emeritus Harold Brown pointing out weaknesses in the connection can be found at; http://www.gppf.org/article.asp?RT=19&p=pub/AirQuality/envurbanmyth040716.htm
Dr John Ray's weblog http://antigreen.blogspot.com/ on 19 July 2004 has drawn attention to this issue so I wanted to see what the data shows,  ideally I want to analyse hourly ozone, temperature and whatever other air monitoring elements available for say June to August 1996.  All my ozone data analysis experience tells me that big city summer ozone peaks will be broadly influenced by daytime max temperature and met factors such as wind; I would be sceptical that a slight decrease in vehicle emissions could dominate over these more important factors.
The Georgia Dept for Natural Resources runs air monitoring in Atlanta and their web site has this graphic full size at; http://www.air.dnr.state.ga.us/tmp/olympic_ozone/index.html
You can also click on a calendar and see hourly data for each day in the Olympic period.    There is no text to go with their graphic which is for exactly the Olympic period and I can only assume that readers will be left with the general impression that ozone levels were lowered during the Olympics.  Because it took a week or so to clear up with the Georgia DNR exactly what station data they had sent me I sought data from other sources that was relevant to the general problem and that is all posted from the bottom of the page upwards in order of posting.
Luckily I was able to find hourly ozone and other air data for  various SE USA stations (see map left)  from the web site of   Atmospheric Research Inc which has a useful public database at http://www.atmospheric-research.com/public/index.html
 but unfortunately no Atlanta data for 1996.  Their JST site is at Jefferson Steet in Atlanta.   Luckily,  1996 ozone data is available for Yorkville (YRK)  a rural site 80 kms NW of Atlanta and Centreville (CTR)  a rural site in Alabama

 3 August posting
The graphic below shows ozone and NOx (NO for Yorkville) in the Olympic summer of 1996 for South DeKalb station in Atlanta and rural Yorkville 80 kms NW.  As other data has shown(see 27 July below), the Olympic period was  typified by low ozone across a broad region and this also is the case for South DeKalb. Traffic emissions  for South DeKalb as represented by NOx are wildly erratic and show no steady relationship with ozone.  Yorkville ozone daily peaks are  often higher than South DeKalb yet Yorkville NO is negligible compared to NOx in South DeKalb, which reflects the very low traffic levels at Yorkville. Clearly ozone can often generate high daily peaks without traffic. Another graphic just showing the Olympic period and ten days before and after.
South DeKalb wind speed is usually higher when ozone is lower and vice versa.
It is clear that the politically correct notion that reduced Olympic traffic caused lower asthma rates has no scientific basis that can be detected in air monitoring data.  


27 July 2004   The graph below shows clearly that across a broad region of  Georgia and Alabama the Olympic period happened to be characterized by the lowest ozone levels of the 1996 summer.  So you would have to suspect that the lower ozone levels in Atlanta during the Olympics are also following this regional trend which is no doubt due to broad scale meteorological factors and nothing to do with traffic.  I am still waiting to get the 1996 summer hourly data from an Atlanta station from the Georgia DNR.


22 July, still awaiting the Atlanta June-August 1996 data covering the Olympic period.
To keep ourselves busy we decided to compare data from a central Atlanta station and a rural site 80 kms NW, Yorkville; taking ozone and nitrous oxide numbers for August 2001, a cooler year with lower ozone similar to 1996.
It is strikingly obvious that Atlanta and Yorkville have broadly similar daily ozone peaks, some days Yorkville even exceeds Atlanta !!!    This is despite much higher overall NO numbers in Atlanta, those car exhausts, with several very high peaks.
Note the very high NO peaks do not shift the ozone peaks much, so it is hard to see a strong relationship between NO  and ozone.  So where does this leave the much trumpeted claim that a slight reduction in traffic during the Olympics could have caused lower ozone levels ?
NO numbers for Yorkville are insignificant, just showing as the purple trace along the bottom of the graph
When will the EPA's start telling the truth about urban ozone ?




21 July.
An analysis of Yorkville data downloaded from the Atmospheric Research Inc web site shows a relationship between periods of low wind speed and high ozone in the graph below.  Note the periods of  higher  ozone below around June 15, 22, 29 and August 16 (then wind data gap).  And the Olympic period 19 July to 4 August had several higher wind speed days.

So it seems that the issue of what causes variations in ozone levels is more complex than just considering exhaust emissions.
Although we were sent some 1996 data from the Georgia air monitoring Department, the ozone numbers did not agree with those on their web site so we are checking further.

Posted 20 July 2004

On the Georgia Goverment air monitoring site I read a pdf Report for 2002 at http://www.air.dnr.state.ga.us/amp/  which had a graph of ozone exceedences for the years 1993 to 2002 and it is obvious that I could test out my idea of the influence of temperature on ozone levels by comparing Atlanta summer ozone peaks and temperature for the high ozone year 1999 and a low ozone year, 2001.
Luckily I was able to find hourly data for  Atlanta from the web site of  Atmospheric Research Inc which has a useful public database at http://www.atmospheric-research.com/public/index.html
for SE USA air monitoring  but unfortunately no Atlanta data for 1996.As yet I have not found those 1996 data for central Atlanta I require but while reading the
The graph below shows clearly the broad relationship between ozone peaks and  daytime temperatures, the warmer year 1999 is typified by higher ozone daily peaks.


Back to Ozone front page.

Back to WarwickHughes.com front page.