Fine Particle Composition Study, Logan, Utah

Fine Particle Composition Study, Logan, Utah, January-March, 2004

Our first major air pollution study in Logan was during Winter 2004. The aerosol mass spectrometer sampled from January-March at two locations in Logan. During the winter of 2004, Logan made some national headlines due to the high particle concentrations that occured here during severe winter inversions. A summary of some of the data obtained by various researchers can be found on Randy Martin's website . During the months of January and February, PM_2.5 concentrations exceeded the 24-hour NAAQS standard of 65 ug m^-3 a total of 17 times. From early January through mid-March, the AMS sampled ambient particles in Logan at two locations.

Utah State University:

During January, 2004, the AMS sampled from the third floor of the Maeser building (Chemistry) at Utah State University. At right is a picture of the AMS in the laboratory. USU is located on the east bench and is somewhat higher in altitude than downtown Logan. Because we were nearer to the top of the inversion, we detected slightly lower mass concentrations than did the state's samplers in Logan during this time. However, our data immediately confirmed the state's filter data that the major contributor to PM2.5 in Logan is ammonium nitrate. Nitrate (blue color below) was the major component every single day, making up ~50% of the particle mass loadings.

Downtown Logan:

At the beginning of February, 2004, we finally got power installed to an auxilliary shed (gray shed below) located in downtown Logan. (125 W. Center) This is the site where the Utah Department of Air Quality monitors all pollutants in Logan (yellow shed below). The DAQ posts data from all of its sites online; you can find real-time air quality data for the last two days at Logan . Being able to sample at this site was especially helpful, since we could now compare our data to the co-located state data. We found very good qualitative agreement between our data and the real-time TEOM monitor that the UDAQ has at the downtown sampling site.

A comparison of AMS data (black) and TEOM data (blue) in downtown Logan shows a correlation in the time trends of >90%. The AMS generally detected ~75% of the total mass concentration that the TEOM did. There are 2 reasons for this: (1) The TEOM is a true PM2.5 sampler, while the AMS does not efficiently collect the largest particles (~>1 um). (2) The AMS relies on heating the particles on an oven to detect them, and is not capable of detecting refractory components because it cannot vaporize them. This incudes elemental (black) carbon and most metals. As you will see below, we believe that in Logan, explanation 2 is the major reason that the AMS sees lower mass loadings than the TEOM.

The composition of particles at the downtown sampling site was quite comparable to the particles we sampled at USU during January. Nitrate was again the major component, representing ~50% of the total mass concentration detected by the AMS. You can see the large buildup of particle concentrations over several day periods. These periods correlate to inversion conditions in the Cache Valley. Unlike many other urban locations, PM2.5 shows no weekday/weekend correlation in particle concentrations. Mass loadings can be either high or low during either the weekday or weekend. The major force behind whether air pollution gets bad or not is the presence or absence of an temperature inversion.

The organic fraction made up 20-30% of the total mass and we found that we could split the organic fraction up into several different particle types, shown in the plot below.

Two of the organic particle types appear to represent primary (directly emitted) sources: vehicle emissions (probably diesel exhaust) and wood-burning smoke. The plot at right shows our marker for vehicle exhaust (m/z 57, black) which generally shows small spikes of particles in the morning and evening (rush-hour). A second primary particle type that is observed is wood-smoke. The yellow plot in the lower graph shows a peak ( m/z 60) that can be indicative of wood-smoke (though it is not a unique tracer and is not perfect for quantifying the effects of wood-smoke.) But this trend shows small spikes of particles, often in the evening.

There also appears to be at least two processes contributing to organic carbon buildup during the inversion. We believe these are secondary particles (formed in the atmosphere through reactions.) The blue and red plots at right represent these two processes. In general, both show rapid buildup in concentrations during inversion conditions. The blue line represents m/z 43, a peak that represents oxidated organic compounds in the atmosphere, and while it builds up continuously during the inversion, it tends to have its highest loadings during the day, in
mid-afternoon. This type of particle is commonly observed during sampling campaigns conducted with the AMS, and is thought to be indicative of photochemical particle formation. On the other hand, the red plot shows m/z 58. This peaks also builds up during the inversion, however if you look at the detailed time trend, you find that it almost always reaches it's highest values in the middle of the night. We have not identified what this peak means chemically, though we have some hypotheses. However, as far as we can tell in talking to other researchers who utilize the AMS, no one else has observed this phenomenon before, with this particular peak. It appears to be fairly unique to the atmospheric chemistry in Logan.

Another interesting phenomenon in Logan are the size distributions of the particles. Below is a representative size distribution detected by the AMS for a 30-minute period. Next to it is an image plot which shows the size distribution changing over time. These plots tell us a couple things. First, the particles in Logan are small. Our distribution shows nitrate (blue histogram) peaking at ~350 nm particles and other components (organic, green; sulfate, red) at ~500 nm. In looking at this data, it seemed unusual that we could account for so much of the particle mass when the AMS doesn't detect particles larger than ~1 um. However, Randy Martin , our colleague in engineering put out some filter samplers to look at this. His data indeed confirmed what the AMS saw. Approximately 90% of the PM10 is PM2.5, and approximately 85% of the PM2.5 is PM1. Almost all of the particle mass is below 1 um in size. The image plot tells us is that during an inversion, the number of particles in the atmosphere is building up, not just the mass. On the image plot, the y-axis is particle size, x-axis is time, and color axis is particle concentration. It could be possible that during inversion conditions, extra pollution simply condenses onto the existing particles, increasing the particle mass concentration, but keeping the particle numbers relatively constant. However, our data indicates that during the inversion in Logan, the number of particles grows. For a given size of particle, we see many more particles as the inversion continues. Our particle number concentrations approximately quadrupled during the severe inversion from ~ January 10-17, 2004.

We are nearing completion on a couple publications from this data set. The first will give an overview of the overall particle chemistry in Logan, while the second will focus on the organic aerosol component which looks rather complex. Thanks go out to the Utah Department of Air Quality (especially Cheryl Heyning, Brock LeBaron, and Dave McNeill) for providing the shed for us to sample in on Center Street, the City of Logan for allowing us to put another shed there, and to Utah State University for support of the study.

This page was last updated: June 30, 2005
Please send your comments and suggestions