Seasonal Pattern of Ozone over the OTAG Region

Rudolf B. Husar

Center for Air Pollution Impact and Trend Analysis (CAPITA)

Washington University

St. Louis, MO 63130-4899

February 16, 1997

Introduction

Ozone is a strongly seasonal pollutant, with the warm months exhibiting higher concentrations than the cool winter months. The ozone seasonality is imposed by the seasonality of the tropospheric background ozone, the temporal pattern of precursor emissions (NOx, VOC), the systematic seasonal shifts of transport meteorology and by the seasonality of photochemical oxidation and removal processes.

The shape of the seasonal cycle at any given location is determined by the combined influence of the above factors. As a result, in some areas the ozone signal is virtually constant over all seasons, while at other locations it varies by up to factor of four. In some areas ozone has a single summer or spring peak, while at other monitoring sites show a distinctly bimodal seasonal pattern.

The main significance of ozone seasonality in AQ management is that ozone control measures can be confined to the "ozone season". According to EPA, the ozone season is defined as April-September. Ozone monitoring by the states, and ozone reduction measures, if any, are mandated only during this ozone season. Detailed ozone precursor emission inventories are also confined to the warm ozone season.

This report was prepared to support the deliberations of the OTAG Air Quality Analysis workgroup. The specific purpose of this report is to examine ozone seasonality over the OTAG region with particular reference to issues considered relevant to the OTAG process.

OTAG Mission and Goals

The mission of OTAG is to identify control strategies and implementation options for the reduction of regional ozone over the eastern US The operational goals of OTAG are stated as (1) A general reduction in ozone and ozone precursors aloft throughout the OTAG region and (2) a reduction of ozone and ozone precursors at the boundaries of nonattainment areas.

Policy-Relevant and Scientific Results

The results of this work confirm that April-September is the proper time window for the ozone season. About 95% of the OTAG-wide exceedances occur during this warm part of the year. This holds for both the 120 ppb and the proposed 80 ppb standard.

The results of this ozone seasonality may help understanding some broader aspects of the ozone problem including ozone source identification, quantification of ozone-relevant processes, de-seasonalization of the ozone signal and improvement in UAMV model boundary conditions.

Ozone source identification. Each ozone source type has a characteristic seasonal pattern. For example, tropospheric background ozone tends to peak in the spring season, while anthropogenic ozone shows a strong summer peak.

Process signatures. The seasonal cycle may also reveal the roles of specific processes that influence ozone. For example, the spring peak of ozone may be indicative of stratospheric ozone intrusion into the troposphere. A July dip in the ozone concentration in some areas may be due to intense removal in the moist "monsoon" air over the Gulf of Mexico.

De-seasonalization of other signal components. The strong seasonality of ozone signal tends to obscure the long-term, weekly, synoptic, and diurnal ozone pattern. In order to evaluate the other, non-seasonal temporal pattern, it is beneficial to de-seasonalize the ozone signal. Devising a proper de-seasonalization procedure requires a deeper understanding of the seasonal pattern.

Data UAM-V model boundary conditions. The seasonality of ozone data may also aid the understanding of the data-model discrepancies, particularly at the OTAG boundaries For example, in the Gulf of Mexico subregion, the observed ozone concentrations are depressed in July, compared to April-May, and August-September. The assumed UAM-V ozone background concentrations over the Gulf states appear to be higher then the observed values. A possible explanation for the deviation may be found in the unusual bimodal seasonal ozone cycle over the Gulf states.

Data Sources and Processing

Data Sources and Quality Control

The ozone data used in this report were collected from multiple sources:
  • Data Set
  • Supplying Organization
  • Years
  • AIRS
  • EPA
  • 1991-1995
  • CASTNet
  • EPA
  • 1991-1995
  • EMEFS
  • Eulerian Model Evaluation and Field Study
  • 1988
  • SCION
  • Southern Oxidant Study
  • 1993, 1995
  • LADCO
  • Lake Michigan Air Directors Consortium
  • 1991 (88, 93, 95)
  • GEORGIA
  • State of Georgia
  • 1988, 91, 93, 95
  • NORTH CAROLINA
  • State of North Carolina
  • Data from each network were extracted and combined into a single integrated data set. The details of the data sources and quality control procedures are discussed in the report "Preparation of Ozone Files for Data Analysis."

    The first examination of average daily maximum ozone maps has revealed anomalous ozone "holes" and peaks at unexpected locations. For those sites the hourly and daily maximum ozone values were re-examined for possible inconsistencies. Sudden systematic changes in the ozone concentrations, as well as major deviation from neighboring sites were the main clues for anomalous behavior. As a result of this quality control process, 6 out of 709 monitoring sites were discarded. The remaining data were used in all the subsequent computations exactly as submitted by the networks.

    Data Processing Procedures

    The data processing was conducted in the following major steps below:

    1. Data from individual networks were quality controlled and formatted uniformly.
    2. The hourly ozone data from all the networks were combined into a single database.
    3. The daily maximum (1-hour average) ozone was extracted from the hourly data.
    4. For each monitoring station the average, percentiles and exeedances of daily maximum ozone was computed.
    5. The results for all stations were contoured and plotted on maps and for easy presentation.

    The seasonal ozone concentration charts for individual sites were obtained by plotting the daily 1-hour maximum ozone for a year. On the same seasonal chart , the 60 day running average concentration as well as the 10th, 50th, and 90th percentiles within the 60-day window were superimposed. The main purpose of the high and low percentile lines is to indicate the variations of ozone concentrations. The use of the non-parametric statistic (percentiles) for the distribution also allows the detection of background concentrations as well as skewed distribution functions with long tails.

    The Framework for Ozone Pattern Analysis

    The air quality analysis below, makes extensive use of the concepts of pattern and pattern analysis. Pattern analysis is a structured approach to the organization and presentation of air quality data.

    The ozone analysis over the OTAG region is also based on the following physical considerations..

  • 1. The ozone concentration at a given location is composed of contributions from global tropospheric background ozone, the regional ozone from the superimposed emissions within the OTAG region , and from urban ozone that is in excess of the tropospheric and regional background.
  • 2. The spatial pattern can be examined on global, regional (e.g. the OTAG region) or on urban scales, with distinct pattern at each scale. The global-scale of tropospheric ozone constitutes the boundary condition for the regional ozone. Similarly, the regional ozone is the boundary condition of the urban ozone in a hierarchical relationship. In fact, for purposes of this analysis, regional ozone is defined as the ozone concentration at the boundaries of urban/industrial areas.
  • 3. The temporal ozone signal at any location may be decomposed into several temporal scales: secular (1-100 years), yearly, weekly, synoptic (3-5 days), and daily time scale. Each scale exhibits a distinct temporal pattern. The seasonal scale is superimposed on the secular trend, the weekly cycle on the seasonal cycle, the diurnal cycle is superimposed on the weekly cycle, etc.
  • 4. The spatial and temporal scales are linked due to transport at, say at 3 to 5 m/s average transport speed. The corresponding transport distance for a week is on the order of 2-3000 km, for 4 days (synoptic scale) it is 1-2000 km and for one day is 250-400 km. For purposes of OTAG, the most relevant scale is the regional/synoptic scale up to 1,000 km in size and up to about two days of residence time.
  • Seasonal Maps of Ozone

    Ozone exhibits a dynamic, spatially and temporally varying seasonal pattern. The main spatial and seasonal features are examined through daily maximum ozone maps averaged over each of the four seasons, winter (December, January and February), spring (March, April, May), summer (June, July, August), fall (September, October, November), as shown in Figures 1 a-d. For sake of completeness, the average daily maximum ozone maps are also presented for each month, Figure 2 a-l. However, monthly maps will not be discussed in detail.

    The winter map in Figure 1a shows generally low ozone concentrations, ranging between about 20-45 ppb. The main spatial feature is the constant ozone concentration throughout the periphery of the OTAG region (~30 ppb), and a distinct "ozone hole", in the midsection stretching from Nebraska through Illinois, Ohio, to the eastern Washington-Boston corridor. The lowest winter ozone appears to be in the urban and industrial areas, north of the Ohio and Mississippi Rivers. There is no evidence of the winter time ozone hole over the southern half of OTAG. In fact, the southern metropolitan areas, such as Atlanta, Houston, and Dallas, have winter concentrations that are indistinguishable from their rural surrounding. It is also worth noting, that the winter ozone concentration at the high elevation sites of southern Appalachia have higher concentrations (>40 ppb) than the adjacent low lying areas. This suggests that the high elevation sites are within the free tropospheric air.

    The ozone levels in the spring time are higher than the winter values, averaging 40-55 ppb throughout the OTAG region. In general, the spring season is characterized by spatially uniform ozone concentrations. As in the winter months, the southern half of the OTAG region is somewhat higher, 45-55 ppb, than the northern half. Also, there is an ozone depression of 5-10 ppb across the mid-section of OTAG, from Nebraska to the Washington-Boston corridor.

    It is well known that the highest ozone concentrations occur during the summer season, June, July, and August. The average daily maximum ozone concentrations are particularly high over the mid-section of OTAG, from Missouri to the Mid-Atlantic East Coast. The periphery of OTAG has remained low at 30-40 ppb even in the midst of the "ozone season". Consequently, there is a clear ozone bulge over the industrial Midwest and the mid-Atlantic coastal states exceeding the tropospheric background values by 30-40 ppb. Higher ozone concentrations of limited geographic extent are also evident for the Atlanta, Chattanooga-Knoxville, TN corridor. Elevated summer average ozone concentrations (> 50 ppb can) also be observed over Dallas-Ft. Worth - Tulsa-Oklahoma, as well as over Memphis, TN, St. Louis, MO, and over western Michigan, downwind of Chicago, IL.

    The high elevation sites of southern Appalachia show lower values (~45 ppb) than their surrounding low lying areas. It is evident, that these high elevation monitoring sites are above the boundary layer ozone, at least some of the time during the summer season. On the other hand, the high elevation monitoring sites at Whiteface Mountain, upstate New York, show higher values than its low lying neighborhood. Evidently, at the 1,600 meter Whiteface Mountain the ozone concentration is higher than the ozone in the surrounding areas, yielding a clue about the vertical structure. The discussion of the interesting question of ozone vertical structure and its relationship to long range transport is beyond the scope of this report.

    The fall season, Figure 1d, is similar in many ways to the spring season, except that concentrations range between 35-50 ppb, i.e. about 5 ppb lower than the spring values. The southern half of the OTAG region clearly shows about 10 ppb higher concentrations (>50 ppb) than the northern half. Concentrations above 55 ppb are only recorded over Houston, TX, Dallas-Ft. Worth, TX, Shreveport, LS, as well over the high elevation sites of the southern Appalachian Mountains. It is evident, that during the fall season the ozone concentrations aloft (about 1,500-2,000m) are higher than inside the boundary layer.

    Given the strong spatial and seasonal variability of ozone it is instructive to examine the difference between the summer and winter concentrations as a measure of seasonal modulation of ozone. The difference between July and January average daily maximum ozone concentration is depicted in Figure 3. Over the four corners of OTAG (northern Minnesota, upstate Maine, southern coast of Texas, and southern Florida) the July concentrations are only 5-10 ppb higher than the January values. On the other hand, the Washington-New York corridor has July values 40 ppb in excess of the January values. It is also remarkable that throughout the Ohio River Valley the July-January difference exceeds 40 ppb. Additional high July-January differences can be observed downwind of Chicago, IL, St. Louis, MO Memphis, TN, Charlotte, NC. Less pronounced summer winter differences can also be observed over Atlanta, GA, Dallas-Ft. Worth, TX and Kansas City, MO. The large July-January difference over the industrial Midwest (Ohio River Valley) and Mid-Atlantic states is due to the distinct ozone bulge in the summer and an ozone depression during the winter months. The shape of the July-January ozone difference over the northern half of the OTAG region resembles the NOx emission fields.

    The higher elevation sites over the southern Appalachian Mountains show only a modest July-January difference of about 10 ppb, indicating that these monitoring stations report the seasonally stable ozone pattern in the free troposphere, above the polluted boundary layer.

    The spatial pattern of seasonally averaged daily maximum ozone concentration can only reveal the gross features of the ozone spatial and temporal dynamics. In order to pursue the seasonal behavior in more detail, the daily and seasonal concentrations are examined at individual sites in the Appendix A "Ozone Seasonality at Individual Sites".

    Average Seasonal Pattern over the OTAG Region

    This section discusses the seasonality of ozone averaged over the entire OTAG region. The purpose of this aggregated view is to examine the shape of the aggregated average seasonal cycle. It is recognized that such averaging over the entire OTAG region obscures the significant subregional differences in ozone seasonality that is unique to tropospheric background on the edges of OTAG, urban concentration pattern, regionally characteristic sites, high elevation sites, etc.

    Seasonality of OTAG-Average Ozone Concentrations

    The OTAG average ozone concentration for all available sites averaged over 1986-1995 is shown in Figure 4. The averaging was performed for each day of the year by aggregating 10-years of data for days 1 through 365. In the winter season (December, January and February) the concentration is lowest, ranging between 25 ppb in December and 40 ppb on the end of February. Throughout the spring season (March, April, May) the concentration rises from 40 to 60 ppb, averaging about 50 ppb. Throughout the summer months (June, July, and August) the concentration is stable at about 60 ppb. In the fall season (September, October, November) the concentration drops steeply from 60 to 30 ppb, with an average of about 45 ppb. Evidently, the OTAG average ozone pattern is not symmetric seasonally, the spring values are about 10% (5ppb) higher than the fall concentrations.

    Seasonality of Ozone Exceedances in the OTAG Region

    From the point of view of air quality management the critical regulatory parameter is the number of exceedances above the ambient concentration standard. Since the average ozone concentration shows a pronounced seasonality, it is expected that the exceedances themselves will be seasonal.

    For the purposes of OTAG-scale exceedance analysis, the number of stations that exceed the national standard was counted up for the OTAG region and divided by the total number of stations that reported valid ozone data for that day. This measure is the fraction of the monitoring sites that exceed a standard threshold value. It is recognized that this averaging over OTAG is biased toward those geographic areas that have high density of monitoring stations. Hence, the exceedances in urban industrial areas are weighed more than remote locations that have low station densities. In effect, weighing the exceedances in this manner, applies a geographic relevancy factor that EPA and the states have employed when laying out the national ozone network.

    The daily exceedances over OTAG for a particular year 1995 is shown in Figure 5. The seasonal chart contains exceedance fractions for 80, 100, and 120 ppb threshold concentrations. The data show that there is a pronounced seasonality of exceedances with a strong summer peak, and virtually no exceedances in the winter. The other major feature of the exceedance seasonality is the large day to day variation. Thirdly, as expected, the fraction of the stations that show the exceedances over 80 ppb is well above those exceeding 100 or 120 ppb. It is beyond the scope of this report to examine the day to day fluctuation of OTAG-wide exceedances. Rather, attention will be focused on the shape of the seasonal pattern for 80, 100, and 120 ppb threshold concentrations. The particular question of concern is whether a significant fraction of ozone exceedances occur outside April-September ozone season.

    The relative exceedances for the 80, 100, and 120 ppb threshold concentrations are depicted in Figure 6. The magnitude of the exceedances for a specific threshold were normalized to the year total to facilitate relative comparison among the three threshold levels. Figure 6 shows that at 120 ppb threshold concentration, seasonal distribution function extends between mid-May and mid-September, with a strong peak in mid-July. The seasonal distribution function at 80 ppb threshold is broader, extending from beginning of April through end of September, and a rather flat between June and August. It is evident, that the processes that are responsible from creating the 120 ppb exceedances differ somewhat from the causal factors of the 80 ppb exceedances.

    The exceedance distribution functions clearly show that a small fraction of the exceedances occur in the cool months, i.e. outside April-September season (Figure 6). The fraction of the exceedances that occur within the April-September ozone season is shown in Figure 7, as a function of the ozone threshold concentration. At 80 ppb threshold about 94% of the exceedances are within the ozone season and for the higher thresholds this fraction is constant at about 95%. Therefore, the data show that there is no significant difference in the coverage of the ozone exceedances by the ozone season for the 80 and 120 ppb threshold. This is somewhat counterintuitive result, since it would be expected that the broader seasonal distribution function for the 80 ppb threshold would result in higher fraction of exceedances outside the April-September limits.

    Summary of Ozone Seasonality over the OTAG Region

    Spatial analysis of ozone seasonality reveals that stable tropospheric background ozone concentrations exist at the OTAG boundaries throughout the year. Within the OTAG region, the winter maps show an "ozone hole' compared to the tropospheric background stretching over the industrial Midwest and the Mid-Atlantic states. The same region exhibits a summertime ozone bulge of 70-80 ppb average ozone which is about twice the tropospheric background. A map of the difference between the summer and winter concentrations shows a pattern that closely resembles the NOx emission field over the industrial Midwest and Mid-Atlantic states. The southern half of the OTAG region shows less seasonal variation.

    The main features of the ozone seasonality are illustrated in Figure 8. It shows the characteristic concentration pattern for the tropospheric background, regional elevated ozone (such as over the industrial Midwest), and the average excess concentration within major urban areas. The main feature of this semi-quantitative illustration is the substantial magnitude of the tropospheric background ozone and its spring peak. Secondly, it conveys the magnitude of the regional and urban perturbations which cause an ozone bulge in the summer and a depression in the winter compared to the tropospheric background.

    The seasonal chart of the ozone concentration averaged over the entire OTAG region shows a somewhat asymmetric pattern. The lowest values of 25 ppb occur in December, steadily rising to about 60 ppb by June, and remain at that level throughout June, July, and August. In September, there is a sharp drop from 60 to 40 ppb, such that the fall concentrations are about 5 ppb lower than the springtime values.

    The seasonality of the ozone exceedances show strong peak in July for the 120 ppb threshold concentration. However, for the 80 ppb threshold, the exceedances are roughly constant through June, July, and August. Also, at the 80 ppb threshold, significant fraction of the exceedances occur in April and May. Interestingly, overall the fraction of exceedances that occur within April-September season is about 95% irrespective whether the threshold is set at 80 or 120 ppb.

    The examination of the ozone seasonality for individual sites reveals a dynamic signal, that may serve as a rich resource for further analysis. Possible topics for detailed analysis may include the quantification of tropospheric background throughout the OTAG domain and its comparison to the man-induced ozone levels. The description and the quantification of the synoptic ozone variations for tropospheric, regional and urban ozone, separate from the seasonal pattern would also deserve further exploration. Finally, the approach outlined herein could be useful in re-evaluating the ozone boundary conditions for OTAG scale ozone modeling efforts.

    Appendix A: Ozone Seasonality at Individual Sites

    This Appendix contains the seasonal ozone concentration pattern for individual monitoring sites. It is "work in progress". For some sites and subregions the seasonal charts are provided with annotations, for others only the figures are given.

    The seasonal pattern of ozone is imbedded in the long-term ozone signal which also contains the synoptic scale and the diurnal scale temporal variations. When all these temporal scales are left in the unfiltered ozone signal it is difficult to discern the behavior at any of these scales.

    The strong diurnal ozone cycle can be eliminated from the total signal by considering only daily maximum ozone concentrations. The rationale of this diurnal filter is that the daily maximum ozone is believed to represent the columnar ozone concentration. The discussion of the diurnal ozone pattern and its relationship to the vertical ozone structure is beyond the scope of this report.

    Separating the seasonal and the synoptic ozone pattern is more difficult. A traditional procedure involves the use of a 30-day low-pass filters on the temporal signal which eliminates the "stochastic" component. The available data indicate that both the mean and the stochastic ozone components are different for the tropospheric background, regional, and urban ozone. Our inability to separate these mean and stochastic component for tropospheric, regional, and urban ozone. prompted us to present the seasonal and synoptic ozone pattern without separating the signal components.

    The ozone seasonality may be characterized in several ways, including:

    1. Shape of the ozone seasonal cycle (e. g. unimodal, bimodal, sinusoidal).
    2. Amplitudes of the seasonal modulation (e.g. factor of 1.3-3.5).
    3. Month of the peak ozone (April-September).

    In the Figures that follow, the seasonal pattern at individual monitoring sites is displayed on seasonal time charts. The presented data are the daily maximum ozone for a particular year, usually for 1991. Other years are shown when 1991 data were not available. Many of the monitoring sites were selected from the CASTNet network, because of its complete seasonal coverage. Most AIRS sites have data only for April-September ozone season.

    The seasonal ozone charts for individual sites also show the running average concentration with a 60 day moving window (solid line). Finally, the seasonal charts also contain the 10th 50th, and 90th percentiles of ozone within the 60-day moving window. The purpose of showing these percentiles is to quantify the day to day variation of the daily ozone concentration.

    Ozone Concentration West of the OTAG Region

    The seasonal pattern of the regional scale ozone concentration just west of the OTAG region is illustrated in the charts for Glacier National Park, MT, Chiricahua, AZ, and Ghothic, CO. The seasonal pattern at each site has similar characteristics. The shape of the seasonal cycle is unimodal and the peak concentration is consistently in May. The ratio of maximum to minimum concentration is about 1.3, and the day to day variations of the ozone signal compared to the 60-day average are on the order of 10%. Also, the variation is symmetric around the mean which is characteristic of stochastic "noise" superimposed on the mean signal. The only deviation between the three sites is that the northern site at Glacier National Park has about 15 ppb lower average concentration compared to the Colorado and New Mexico sites.

    These data indicate that a stable tropospheric background ozone exists over the remote western U.S. with a clear spring peak. The seasonality, as well as the magnitude of this tropospheric ozone signal is similar to observations at other remote sites. The most remarkable feature of the remote western ozone data is the small day to day variability. It is evident that such a stable ozone concentration in the presence of highly variable airmass characteristics can only be achieved if the sources of ozone sources are far, many thousands of kilometers away from the monitoring sites.

    Ozone Concentration in the Upper Midwest

    The ozone concentration pattern at the northwestern corner of OTAG is examined through a monitoring site at the Experimental Lake Area (E.L.A.) site in southwestern Ontario and at Perkinstown, WI. The concentrations at the E.L.A. site are similar to the Glacier National Park, MT site in that the winter concentrations are about 35 ppb, rising to 50 ppb in May, followed by a decline through November and December back to the 35 ppb level. Between October and May the day to day ozone variation is small (+15%), comparable to the variance at the Glacier National Park. The main difference between the two sites occurs in the warm season June-September, when the E.L.A. site exhibits ozone peaks up to 60-70 ppb, lasting for several days. These warm season ozone peaks could arise from the transport of ozone containing airmasses from the south, i.e. the OTAG region. However, the full airmass history analysis has not been performed to substantiate this possibility.

    The concentration pattern at Perkinstown, WI is rather similar to the E.L.A. site. The average ozone concentration, the shape of the seasonal cycle, as well as the relative variation of the ozone signal within and outside the ozone season is similar to the two sites. Evidently, the northwestern corner of the OTAG region is exposed to the tropospheric background ozone throughout the year except for multiday excursions in the warm season when concentrations can reach 70-80 ppb.

    Ozone Concentration in New England.

    The New England ozone pattern is examined through the monitoring sites at Ashland, ME, Woodstock, NH, and Cape Elizabeth, ME. The concentration pattern at the northeastern corner of the OTAG region, Ashland, ME is remarkably similar to the concentration at the northwestern corner E.L.A., ONT site. The average concentrations range between 35-50 ppb, with a slight (50/35=1.?) ratio. A major feature of both sites is the occurrence of short duration ozone excursions well over 80 ppb, in the warm season April-September. The consequence of these peaks is that the 90th percentile of the data is about 50% higher than the running average concentration. On the other hand, the 10th percentile of the ozone values is rather similar to the 10th percentile at E.L.A., ONT or the Glacier National Park, MT. Again, these ozone peaks are tentatively attributed to ozone laden airmasses that are transported to the monitoring region from within the OTAG region.

    The concentrations at Woodstock, NH retain some features of the Ashland, ME site. However, the concentration peaks during the warm season are more pronounced. Evidently, the influence of transported ozone is more pronounced than at the northern Maine site.

    Cape Elisabeth, ME site is on the Maine Atlantic coastline. Ozone data show that during the cold season, the ozone concentrations and their variations is virtually identical to tropospheric background concentrations. However, the warm season values deviate dramatically from the stable 30-40 ppb background concentrations. Sharp peaks well exceeding 80 ppb occur throughout the 6-month period, causing many exceedances of the ozone standards, regardless of which metric is used. It is evident that high ozone plumes are transported to Cape Elisabeth dozen or more times a season.

    The data from this site clearly illustrate the need to decompose both the mean and the stochastic ozone signal into tropospheric background, regional background, and "urban impact". The nature of the fluctuations suggests that Cape Elisabeth is either exposed to the tropospheric background, or to regional and urban scale plumes. Most of the variation in the warm season zone is attributed to the latter ozone types.

    The 90th and the 10th percentiles are qualitatively different at the coastal Maine sites. The 10th percentile has an April-May peak, at 35 ppb, followed by a decline to 30 ppb in the winter. On the other hand, the 90th percentile steadily rises from about 40 ppb in the winter to 90-100 ppb in July and August. The sharp summer peak is partially visible in the 50th percentile. The summertime ratio of the 90th/50th percentiles is 95/50=? while the ratio of 50/10 is 50/30= ? It is clear that the distribution function is strongly asymmetric with the long tail of high concentrations causing a lognormal, rather than a normal distribution function.

    Ozone Concentration over Michigan

    The monitoring sites over Michigan show qualitatively different pattern compared to

    Ozone Concentration over Florida

    Concentration measurements in Florida may represent the tropospheric ozone concentrations over the subtropical Atlantic. However, urban influences from Miami, FL, Tampa Bay, FL and other cities are also present.

    The selected monitoring sites were in southern Florida, near the southern Florida Everglades (AIRS 12025030)and the CASTNet site at Sumatra, FL in the Florida panhandle. The unique feature at these sites is the bimodal ozone peak, the first larger peak occurring in March-April (50 ppb), followed by a decline in June-July to 40 ppb, followed by another smaller peak in September and October. In southern Florida, the low concentrations (10th percentile) is rather stable, however, sharp concentration peaks lasting one or two days are superimposed on this background. The interpretation is that the sharp ozone peaks are caused by urban plumes from the Miami metropolitan area.

    Summary of Ozone Seasonality over the OTAG Region

    Spatial analysis of ozone seasonality reveals that stable tropospheric background ozone concentrations exist at the OTAG boundaries throughout the year. Within the OTAG region, the winter maps show an "ozone hole' compared to the tropospheric background stretching over the industrial Midwest and the Mid-Atlantic states. The same region exhibits a summertime ozone bulge of 70-80 ppb average ozone which is about twice the tropospheric background. A map of the difference between the summer and winter concentrations shows a pattern that closely resembles the NOx emission field over the industrial Midwest and Mid-Atlantic states. The southern half of the OTAG region shows less seasonal variation.

    The main features of the ozone seasonality are illustrated in Figure 8. It shows the characteristic concentration pattern for the tropospheric background, regional elevated ozone (such as over the industrial Midwest), and the average excess concentration within major urban areas. The main feature of this semi-quantitative illustration is the substantial magnitude of the tropospheric background ozone and its spring peak. Secondly, it conveys the magnitude of the regional and urban perturbations which cause an ozone bulge in the summer and a depression in the winter compared to the tropospheric background.

    The seasonal chart of the ozone concentration averaged over the entire OTAG region shows a somewhat asymmetric pattern. The lowest values of 25 ppb occur in December, steadily rising to about 60 ppb by June, and remain at that level throughout June, July, and August. In September, there is a sharp drop from 60 to 40 ppb, such that the fall concentrations are about 5 ppb lower than the springtime values.

    The seasonality of the ozone exceedances show strong peak in July for the 120 ppb threshold concentration. However, for the 80 ppb threshold, the exceedances are roughly constant through June, July, and August. Also, at the 80 ppb threshold, significant fraction of the exceedances occur in April and May. Interestingly, overall the fraction of exceedances that occur within April-September season is about 95% irrespective whether the threshold is set at 80 or 120 ppb.

    The examination of the ozone seasonality for individual sites reveals a dynamic signal, that may serve as a rich resource for further analysis. Possible topics for detailed analysis may include the quantification of tropospheric background throughout the OTAG domain and its comparison to the man-induced ozone levels. The description and the quantification of the synoptic ozone variations for tropospheric, regional and urban ozone, separate from the seasonal pattern would also deserve further exploration. Finally, the approach outlined herein could be useful in re-evaluating the ozone boundary conditions for OTAG scale ozone modeling efforts.

    Figure 1. Seasonal average ozone concentration over the OTAG region.
    Figure 1a
    Figure 1b
    Figure 1c
    Figure 1d

    Figure 2. Monthly average ozone concentration over the OTAG region
    Figure 2a
    Figure 2b
    Figure 2c
    Figure 2d
    Figure 2e
    Figure 2f
    Figure 2g
    Figure 2h
    Figure 2i
    Figure 2j
    Figure 2k
    Figure 2l

    Figure 3. Difference between July and January ozone concnetrations

    Figure 4. Seasonality of ozone over the OTAG region

    Figure 5. Relative yearly exceedences, 1986-95
    Figure 5a
    Figure 5b
    Figure 5c

    Figure 6. Relative seasonality of exceedences at 80, 100 and 120 ppb threshold

    Figure 7. Fraction of exceedences in the April-October ozone season

    Figure 8. Typical Ozone Seasonality at Tropospheric Backround, Regional and Urban Sites.


    Submit your comments, feedback, questions, and ideas pertaining this page. Your input will be automatically added to the existing annotations. In order to add a new comment, you must be registered with the OTAG/AQA Peoples Page.