ABSTRACT

INTRODUCTION

In this paper, we report a case study of an episode of high ozone concentrations in synoptic-scale hazy air masses over the eastern half of the United States. In recent years, ozone concentrations well in excess of the ambient air quality standard (0.08 ppm) have been observed in rural areas at distances of several hundred kilometers from known sources of ozone precursor gases (Coffee and Stasiuk, 1975). Rural ozone may be a product of atmospheric reactions involving anthropogenic (e.g. White et al., 1976) as well as biogenic emissions, and may also be due to the downward mixing of stratospheric ozone, and the role of synoptic meteorology, are not quantitatively established. High rural ozone concentrations, however, have been observed most frequently inside large slow-moving high pressure cells in midwestern and eastern U.S. (Ripperton and Worth, 1973).

The occurrence of large hazy air masses containing sulfate concentrations in excess of 20 g/m3 have also been reported recently (Hall et al., 1973; Husar et al., 1976). The present work was initiated to examine whether synoptic scale hazy air masses with high sulfate levels also contain high levels of ozone. It was hoped that certain relationships could be found between these two noxious secondary pollutants which would shed some light on their origins.

The method of analysis is based principally on inspection of contour plots of sulfates, ozone and ground visibility data over the eastern half of the U.S. Air parcel trajectory analyses, surface wind data and local measurements of the pollutants in St. Louis are also consulted. The spatial and temporal density of the data base for national sulfate distribution is sparser than that for ozone. However, hourly surface visibility observations reported routinely from several hundred weather stations have previously been used as an effective surrogate for sulfate data during air pollution episodes (Husar et al., 1976).

DATA BASE FOR THE ANALYSIS

The Following three sources of data have been utilized in this work.

1. U.S. Weather Service Network for hourly observations (Service A). Values of ground level visual range and other weather parameters are recorded every hour at several hundred stations distributed over the U.S. The data from about 300 of these stations (Figure 1a) are available on magnetic tapes supplied by the National Climatic Center, NOAA. The high spatial density of this network permits the meaningful use of computer contour plotting techniques. Using the visibility data, the spatial extent, the temporal evolution and the transport of hazy 'blobs' (according to the Random House Dictionary, Random House, New York (1966) "blob"=an object, esp. a large one, having no distinct shape or definition". In the context of this report, "blob" is used synonymously with "hazy air mass") may be followed for several days by inspection of chronological visibility contour maps (Husar et al., 1976). Noon visibilities were chosen to minimize the effect of early morning high humidities. The noon relative humidity for inland stations ranged between 50 and 80%.
2. The National Air Sampling Network (EPA) for sulfate. Hi-volume samples collected over 24-hour periods by this network are routinely analyzed for sulfate. The samples are collected every 12 days, simultaneously at all sampling locations. For the present analysis, we were able to obtain sulfate data for about sixty NASN locations in the eastern half of the U.S. The spatial distribution of these stations is shown in Figure 1b. Isopleths of sulfate concentrations are plotted manually from the available data at 12-day intervals.
3. The SAROAD database (EPA) for ozone and total oxidant. Ozone and oxidant concentrations are recorded regularly at aerometric monitoring stations reporting to the U.S. EPA. Hourly average values are obtained from EPA's SOROAD data bank for 89 sites located in the eastern half of the U.S. (Figure 1c). Such data obtained in St. Louis, MO are shown in Figure 6. In the upper part of the figure, the hourly average ozone concentrations are shown at station #7 of the St. Louis city-county monitoring network. The data exhibit the typical diurnal pattern consisting of near-zero readings overnight and early morning, rising to peak in the early afternoon followed by a drop in the evening hours. Although it is not clearly documented in the literature, it is reasonable to assume that the low overnight concentrations are the direct consequence of physical and chemical removal processes that tend to scavenge ozone from the atmospheric surface layer. It has been shown (Coffee and Stasiuk, 1975) that the ozone concentrations above the surface layer does not exhibit this diurnal pattern. They also gave evidence that ozone measurements near the surface, obtained during the convectively active noon hour, are representative of the ozone concentrations in the air mass in the absence of positively or negatively interfering local sources. Thus the ozone concentration of the mixed layer air mass may be estimated from the envelope of the daily peak concentrations. For each station, the diurnal ozone pattern was inspected and the daily peak concentration was chosen as the ozone level of the air mass. Subsequently, contour plots were prepared for each day of the episode based on these peak observations.

RESULTS

The relationship between ozone and haze is examined through a case study of a large hazy air mass ('blob') that resided over the eastern U.S. for an estimated two weeks. Successive contour maps of noon visibility (Figure 2) are plotted for every second day from June 25 through July5, 1975. Inspection of the sequence of maps reveals that multi-state regions are covered with a haze layer in which the noon visibility is less than 6 miles. From long range trajectory calculations and surface wind information, it was determined that the air mass of June 25, 1975, within which the visibility was less than 6 miles was of maritime origin in the Gulf of Mexico. This air mass had been transported in a northerly flow across Louisiana, Arkansas, Illinois, and Indiana. Between June 25 and 27, 1975, a NNE trajectory prevailed in the southern states, but relative stagnation prevailed in the Great Lake region. During this stagnation, the air mass became increasingly hazy. Thereafter, an easterly flow developed causing the hazy blob to drift slowly westward, passing over St. Louis, MO on June 28-29, and continuing across Missouri and Kansas (June 30), and then up into Iowa and Minnesota (July1 and 2). The surface wind pattern at noon of June 30 is shown in Figure 3. Observe the clockwise circulation in the Great Lakes region. By July3, the blob had circled over Lake Michigan continuing its residence over the high pollutant emission density regions of the Northeast. In the next two days, however, a cold Canadian front advanced southwards at a relatively rapid pace, sweeping the hazy air mass ahead of it. St. Louis once again experienced substantially reduced visibility on July3-4 as the hazy air mass passed over it. By July 5, visibility deterioration to less than 4 miles were experienced in Atlanta, GA, Birmingham, AL, and Tallahassee, FL. Air trajectory analysis confirmed this southward motion of the blob.

During the above episode, sulfate data were obtained and plotted for June 23 and July 5, 1975 A close relationship has been observed between the spatial extent of the hazy air mass and high sulfate concentrations as shown in Figure 4. On June 23, the blob was located east of Lake Erie where sulfate concentrations over 300 g/m³ were obtained. On July 5, both the haziness and high sulfate levels were reported from the southeastern U.S. (Georgia and Alabama). The coincidence of hazy air masses and high sulfate concentrations on these two days confirms the utility of visibility observations as a qualitative surrogate for sulfates.

Contour plots of the daily maximum ozone concentrations are shown in Figure 5. Inspection of the corresponding contour plots for ozone and visibility (Figures 2 and 5) reveal that the geographical location of high ozone concentrations roughly corresponds to the areas of low visibility (and high sulfate). As may be anticipated, however, the correlation with haziness (low visibility) is much better for sulfate than for ozone.

Contours of daily maximum ozone concentration for June 25, 1975, show that an area of approximately 1000 square kilometers located halfway between the Gulf of Mexico and the Great Lakes, had ozone concentrations in excess of 0.08 ppm. The air parcel trajectory followed a northerly course during that day. It is worth noting that, on that day, the haziness (Figure 2) developed somewhat farther north (i.e. later along the trajectory) compared to the area of high ozone levels. The spatial extents of ozone and haze areas roughly coincide on June 27 as the air mass stagnates. In the following days of the episode, high ozone levels (>0.08 ppm) continue their presence in the same approximate region of the U.S. where the haziness predominates. As the hazy air mass moves to the south by July 5, elevated ozone levels are also observed in that region. At that time, ozone levels are comparatively suppressed immediately behind the front, but continue to be over 0.08 ppm farther to the north and west.

The possible relationship between hazy air masses and high ozone concentrations may also be studied by the analyses of long-term observations of visibility and ozone at fixed locations. The lower part of Figure 6 shows the three-month variation of the extinction coefficient (b_ext=3.92/visual range). The visibility data were recorded at the St. Louis airport. The passage of the hazy blob observed previously over St. Louis is clearly shown by the two peaks of June 28-29 and July 3-4. Comparison of peak ozone concentrations and visibility data for June, July and August, 1997 show that ozone levels above the standard roughly coincide with haziness corresponding to b_ext greater than 5.

DISCUSSION

The above data analysis provides some evidence that synoptic scale hazy air masses also contain elevated ozone concentrations. This evidence is based on only one case study, and is weaker that the observed correspondence between sulfate levels and haziness. A possible explanation for the simultaneous occurrence of high ozone-haze-sulfate levels may be given on simple meteorological grounds: in stagnant or slow moving air masses, precursor gases for ozone and light scattering aerosol are emitted and mixed, leading to an accumulation of these secondary pollutants due to the low ventilation coefficient. It is likely that this mechanism is the primary cause of synoptic scale air pollution phenomena. A more intriguing question is whether these two pollutants have a synergistic effect upon each other's development. Does the presence of high ozone concentrations promote the oxidation of SO₂ to sulfate.

Figure 1. Geographic distribution and density of network stations where data utilized in this paper were recorded a) National Weather Service Network for visibility; b) National Air Sampling Network for sulfate; c) EPA/SOROAD network for ozone.

Figure 2. Sequential contour plots of ground visibility at 1200 CST (b_ext=3.92/visibility range).

Figure 3. Surface wind direction pattern at weather stations for 1200 CST on June 30, 1975.

Figure 4. Comparison of contour plots of noon visibility and daily average sulfate concentration for 6/23 and 7/5/1975.

Figure 5. Sequential contour plots of daily peak ozone concentration.
Figure 6. Comparison of long-term observations of ozone and light scattering coefficient, b_ext=3.92/visibility, at an air monitoring station in St. Louis, MO.

ACKNOWLEDGEMENT

This research was supported by the U.S. Environmental Protection Agency, Environmental Sciences Research Laboratory, Research Triangle Park, NC. We wish to express our thanks to W.E. Wilson, Jr., N. Turcu, and the St. Louis County Health Department.

REFERENCES

1. Coffee, P.E. and W.N. Stasiuk. Evidence of atmospheric transport of ozone into urban areas, Environmental Science and Technology, 9(1), 59-62, 1975.
2. Hall, F.P. Jr., C.E. Duchon, L.E. Lee and R.R. Hagan. Long-range transport of air pollution: A case study, Monthly Weather Review 101, 404, 1970.
3. Husar, R.B., N.V. Gillani, J.D. Husar, C.C. Paley and P.N. Turcu. Long -range transport of pollutants observed through visibility contour maps and trajectory analysis Preprint 3^rd Symposium of Atmospheric Turbulence, Diffusion, and Air Quality, American Meteorological Society, Raleigh, NC, October, 1976.
4. Ripperton, L.A. and J.J.B. Worth. Interstate ozone studies, 2^nd Joint Conf. On Sensing of Environmental Pollutants, Washington, D.C., December, 1973.