September 1997

 

 

PAMS Data Analysis: An Investigation of Local Meteorological Effects on Ozone During the OTAG 1995 Episode and the Weekday/Weekend Differences in the Northeast Corridor

 

 

 

This Report is Prepared For

 

United States Environmental Protection Agency

Office of Air Quality, Planning, and Standards

Research Triangle Park, NC 27711

 

 

 

by

 

Science Applications International Corporation

615 Oberlin Rd.

Raleigh, NC 27605

 

 

 

Under

 

EPA Contract No. 68-D3-0030

Work Assignment No. III-105

 

 

FOREWORD

 

This research project was sponsored by the Office of Air Quality Planning and Standards of the United States Environmental Protection Agency under Contract Number 68-D3-0030, Work Assignment Number III-105. Dr. Dave Guinnup was EPA technical monitor. Funding of this project does not necessarily signify that the results reflect the views of the Environmental Protection Agency. Dr. Fred M. Vukovich was Project Leader. He was assisted by Dr. Robert Wayland. This project benefited greatly by discussion between Dr. Vukovich and Professor Harvey Jeffries, School of Public Health, University of North Carolina, Chapel Hill, North Carolina.

 

TABLE OF CONTENTS

 

Foreword ;i

List of Figures ;ii

List of Tables ;v

Abstract ;vi

1.0 Introduction ;1

2.0 Analysis Procedure ;2

3.0 The 1995 OTAG Episode ;3

3.1 The General Meteorology ;3

3.2 Air Quality and Local Meteorology ;5

4.0 The OTAG Exceedance Cases ;25

5.0 Weekday/Weekend Effect ;30

6.0 Summary and Discussion ;30

7.0 The Study Findings ;40

8.0 Recommendations for Future Studies ;42

References ;43

 

LIST OF FIGURES

 

 

Figure 1 Geographical location of the PAMS stations used in this analysis. 2

Figure 2 Daily positions of the high pressure systems that influenced the eastern United States during the 7-18 July 1995 OTAG episode. 4

Figure 3 Surface analysis for 15 July 1995 at 1200Z showing frontal position, position of the high pressure center, and the surface wind direction. Dots indicating station position without lines indicating wind direction are locations with zero winds. 6

Figure 4 Mean diurnal profiles for ozone and NOx for the average non-exceedance and exceedance days in the June-July 1995 period for Washington DC.. 7

Figure 5 Departures for ozone and NOx from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for Washington DC. The average non-exceedance profiles were for the period June-July 1995. 8

Figure 6 Departures for various meteorological parameters from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for Washington DC. The average non-exceedance profiles were for the period June-July 1995. 9

Figure 7 Departures for ozone and NOx from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the upwind Baltimore site. The average non-exceedance profiles were for the period June-July 1995. 10

Figure 8 Departures for various meteorological parameters from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the upwind Baltimore site. The average non-exceedance profiles were for the period June-July 1995. 12

Figure 9 Departures for ozone from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the Baltimore site. The average non-exceedance profiles were for the period June-July 1995. 14

Figure 10 Departures for various meteorological parameters from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the Baltimore site. The average non-exceedance profiles were for the period June-July 1995. 15

Figure 11 Departures for ozone and NOx from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the downwind Baltimore site. The average non-exceedance profiles were for the period June-July 1995. 16

Figure 12 Departures for various meteorological parameters from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the downwind Baltimore site. The average non-exceedance profiles were for the period June-July 1995. 17

Figure 13 Departures for ozone, NOx, CO, and SO2 from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the Philadelphia site. The average non-exceedance profiles were for the period June-July 1995. 19

Figure 14 Departures for ozone and NOx from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the New York site. The average non-exceedance profiles were for the period June-July 1995. 20

Figure 15 Departures for various meteorological parameters from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the New York site. The average non-exceedance profiles were for the period June-July 1995. 22

Figure 16 Diurnal profiles of ozone and NOx on 14 and 15 July 1995 and on the average non-exceedance day for the June-July 1995 period at the Washington DC site. 23

Figure 17 Diurnal profiles for the meteorological parameters on 14 and 15 July 1995 and on the average non-exceedance day for the June-July 1995 period at the Washington DC site. 24

Figure 18 Diurnal profiles of ozone and NOx on 14 and 15 July 1995 and on the average non-exceedance day for the June-July 1995 period at the upwind Baltimore site. 25

Figure 19 Diurnal profiles for the meteorological parameters on 14 and 15 July 1995 and on the average non-exceedance day for the June-July 1995 period at the upwind Baltimore site.26

Figure 20 Diurnal profiles of ozone and NOx on 14 and 15 July 1995 and on the average non-exceedance day for the June-July 1995 period at the downwind Baltimore site. 27

Figure 21 Diurnal profiles for the meteorological parameters on 14 and 15 July 1995 and on the average non-exceedance day for the June-July 1995 period at the downwind Baltimore site. 28

Figure 22 The mean diurnal profiles of ozone and NOx on a weekday versus a weekend for the period June-July 1995 at the Washington DC site. 31

Figure 23 The mean diurnal profiles of ozone and NOx on a weekday versus a weekend for the period June-July 1995 at the Philadelphia site. 32

Figure 24 The mean diurnal profiles of ozone and NOx on a weekday versus a weekend for the period June-July 1995 at the New York site. 33

Figure 25 The MM5 wind field for 15 July 1995 at 1200 Z. Wind speeds were as large as 12.0 m/s over the ocean (black area) and as small as 1.0 m/s over southern Louisiana. Over the remainder of the area wind speeds varied from 3.0 to 6.0 m/s. 36

LIST OF TABLES

Table 1 Gas Chemistry and Meteorological Data Availability at the Selected PAMS Stations in the Northeast Corridor.. 3

Table 2 PAMS Meteorology Used in this Study and the Availability of That Meteorology in the June-July 1995 Period.. 3

Table 3 The Mean DMOC for Ozone and the Mean Sunrise-to-Sunset Value of NOx and NMHC for the Non-Exceedance Days in June-July 1995.. 7

Table 4 The DMOC and the Mean Values of the NOx Concentration and Meteorological Parameters Between Sunrise and Sunset on 14 and 15 July 1995. 29

Table 5 Summary of the Weekday/Weekend Effect at Washington DC, Philadelphia, and New York. The Ozone Concentrations are Diurnal Maxima and the NOx and the NMHC Concentrations are Averages from Sunrise to Sunset. RO is the Ratio of the Sunrise to Sunset Integrated Ozone on the Weekend to That on the Weekday. RN and RHC are the the Ratios of the Sunrise to Sunset Integrated the NOx and NMHC, respectively, on the Weekday to That on the Weekend.. 34

Table 6 The Mean DMOC for Ozone and the Mean Sunrise-to-Sunset Value of NOx and NMHC for All Exceedance Days in June-July 1995. 37

 

ABSTRACT

 

This study examined the behavior of ozone and its precursors during the 1995 OTAG episode (i.e., 7-18 July 1995) at sites in the northeast corridor using PAMS data. Four urban sites (i.e., Washington DC, Baltimore, Philadelphia, and New York) and two non-urban sites (i.e., sites upwind and down wind of Baltimore) were used in the study. The departures of ozone, of its precursors, and of various meteorological parameters from their average values on non-exceedance days in the June-July 1995 period were examined during the 1995 OTAG episode period. The differences between ozone and its precursors on a weekday/workday and a weekend day during the June-July 1995 period were also examined. We found that during the OTAG 1995 episode and at the PAMS sites examined, when exceedances took place, temperatures and solar radiation were consistently higher than those found on the average non-exceedance day. The wind speed and direction were highly variable from site to site, producing no temporal or spatial consistency in the flow pattern. The wind field was diffusive (i.e., governed by turbulence and not the synoptic pressure gradient), and not conducive of transport, particularly long-distance transport. Secondary mid day maximums in the NOx distribution developed when a local wind minimum developed. Since the presence of the secondary maximums in the NOx distribution will undoubtedly affect the local chemistry, it was concluded that the local wind variations were affecting the local chemistry.

At certain urban PAMS sites examined (i.e., Washington DC, Philadelphia, and New York), weekdays in the June-July 1995 period had, on average, 60 percent more NOx and 38 percent more NMHC between sunrise and sunset than those found on a wekend day, but there was 11 percent more ozone on the weekend days during that period. When meteorological conditions were conducive for episode development and the meteorological conditions on a weekday/workday and on a neighboring weekend day (e.g., a Friday and the following Saturday) were reasonably similar, the DMOC on the weekend day was from 30 to 50 ppb higher and the integrated sunrise to sunset ozone was, on average, 22 percent higher than that on the weekday/workday. In the case examined, the integrated NOx concentration between sunrise and sunset was, on average, about 60 percent higher on the weekday.

1.0 Introduction

States which contain areas that are classified as serious, severe, or extreme ozone non-attainment regions have established a network of photochemical assessment monitoring stations (PAMS) in those regions to meet certain federal regulation. PAMS stations usually measure gas chemistry data (i.e., ozone, oxides of nitrogen, and VOCs) and meteorology data (i.e., wind speed and direction, air temperature, humidity, pressure, precipitation, and solar radiation) to support studies characterizing ozone processes during the so-called "ozone season." Stations are generally located upwind, within, and downwind of an urban complex. PAMS stations have been developed in cities varying in population from 500, 000 to greater than 2 million. Stations are located in states as far west as California and as far east as Maine. The PAMS database provides the potential to examine the temporal behavior of and the processes involved with ozone and its precursors at stations across the US that are urban, urban influenced, and regional in character, particularly in the eastern half of the continental United States where considerable problems with non-attainment have been noted. These analyses will be particularly useful in determining information on ozone processes for control strategies and on the nature of the controls needed. Furthermore, these data can be examined and subsequently processed in a manner that can be very useful to those individuals involved with model evaluation.

The objective of this research project is to evaluate the temporal behavior of ozone and its precursors at selected PAMS sites in the eastern half of the continental United States. This study focused on the period (i.e., June and July 1995) surrounding the 1995 Ozone Transport Assessment Group (OTAG) episode (i.e., 7-18 July ) primarily because the PAMS data set was most complete for this period. The study examined the temporal behavior of ozone at PAMS stations in the so-called "northeast corridor" (i.e., Washington DC, Baltimore, Philadelphia, and New York).

2.0 Analysis Procedure

PAMS gas chemistry and meteorology data were analyzed for the period June and July 1995, which includes the 1995 OTAG episode period (i.e., 7-18 July), for the following stations in the northeast corridor: Washington DC; an upwind station for Baltimore (Nota Bene: this station is also a downwind station for Washington DC); Baltimore; a downwind station for Baltimore; Philadelphia; and New York. Figure 1 shows the location of these stations and Table 1 provides the data availability at each of these stations. The mean diurnal profiles of ozone, NOx, VOCs, and certain meteorological parameters were developed at each station for the ozone non-exceedance days and the ozone exceedance days (i.e., an exceedance was said to occur when the one-hour average ozone concentration was 120 ppb) in the June-July 1995 period and for the OTAG 1995 episode period (7-18 July). The meteorological parameters and their availability is given in Table 2. It should be noted wind direction in degrees is the direction from which the wind originated (i.e., a wind direction of 180_ means a wind from the south or air moving from the south to the north).

Figure 1 Geographical location of the PAMS stations used in this analysis.

 

 

Table 1 Gas Chemistry and Meteorological Data Availability at the Selected PAMS Stations in the Northeast Corridor.

 

Location

 

Ozone

 

NOx

 

VOC

 

Meteorology

Washington DC

Hourly

Hourly

Hourly

Hourly

Upwind Baltimore

Hourly

Hourly

3-Hourly

Hourly

Baltimore

Hourly

*

*

Hourly

Downwind Baltimore

Hourly

Hourly

3-Hourly

Hourly

Philadelphia

Hourly

Hourly

3-Hourly

*

New York

Hourly

Hourly

Hourly

Hourly

*-- Missing data

The mean diurnal profiles of ozone, NOx, and VOCs, and of the meteorological parameters for the non-exceedance days in the June-July 1995 period were subtracted from diurnal data during the OTAG 1995 episode period to create a time series of departure data that could be used to characterize the gas chemistry and meteorology at the selected stations in the northeast corridor during that period.

Table 2 PAMS Meteorology Used in this Study and the Availability of That Meteorology in the June-July 1995 Period.

 

 

Location

 

Solar Radiation

 

 

Temperature

 

Wind Direction

 

 

Wind Speed

 

 

Washington DC

*

Hourly

Hourly

Hourly

 

Upwind Baltimore

Hourly

Hourly

Hourly

Hourly

 

Baltimore

Hourly

Hourly

Hourly

Hourly

 

Downwind Baltimore

Hourly

Hourly

Hourly

Hourly

 

Philadelphia

*

*

*

*

 

New York

*

Hourly

Hourly

Hourly

 

*-- Missing data

3.0 The 1995 OTAG Episode

3.1 The General Meteorology

During the 1995 OTAG Episode, the eastern United States was basically influenced by the movement of two high pressure systems (HPS) into the region. The first HPS moved out of central Canada on 7 July into the Midwest, and then moved eastward across the east coast region into the Atlantic Ocean by 10 July (

2). This HPS was followed by a weak front/low pressure trough that moved through the region which, in turn, was followed by the second HPS which first appeared on the western fringe of the eastern half of the United States on 9 July. This HPS moved basically in an eastward direction until it reached the east coast on 13 July. It stalled momentarily, and then moved southwestward, becoming quasi stationary on 15-16 July over Alabama. Finally, it moved into the Gulf of Mexico on 18 July. A quasi-stationary front was found in the Great Lakes region extending eastward to the coast, which influenced the meteorology in that region through periodic north-south oscillations of the front. On 16 July, a cold front began to move through the region and was approaching the east coast by 18 July as the HPS to the south began to move into the Gulf of Mexico.

 

Figure 2 Daily positions of the high pressure systems that influenced the eastern United States during the 7-18 July 1995 OTAG episode. The open circles represent the first HPS, and the blackened circles, the second HPS

 

The second HPS that moved into the region produced the meteorology that was conducive to episode development. Between 12 July and 15 July that HPS dominated the meteorology of most of the eastern United States. At that time, the winds at the surface were weak, but it was not until 15 July that the flow at the surface as well as aloft lost any semblance of being organized (i.e., transport on the 15 July was, according to the wind field from the surface to the 500-mb level over the a major portion of the eastern United States, probably not an issue). On 12 July through 14 July, anticyclonic flow was evident aloft and, to some extent, at the surface. Those characteristics vanished, to a large extent on 15 July. As the cold front approached on 16 July, the HPS was forced out into the Gulf of Mexico and the flow field in the region was dominated by pre-frontal conditions.

Figure 3 provides the surface flow field at about 1200 Z on the 15 July. The wind speeds varied from 0 to 2 m/s in the region dominated by the HPS. There is no semblance of organized flow except near the front located in the northern part of the region. Though exceedances relative to the one-hour ozone standard were found on days other than 15 July at some of the six PAMS stations used in this study during the 7-18 July 1995 OTAG episode period, exceedances were found at all the stations except New York on 15 July. New York was in the area that was influenced by the front found in the northern part of the region which may have influenced the chemistry in that region.

 

3.2 Air Quality and Local Meteorology

For this study, the mean diurnal profiles of ozone, NOx, and VOCs, and of the meteorological parameters for the non-exceedance days in the June-July 1995 period were subtracted from diurnal data during the OTAG 1995 episode period to create a time series of departure data. Figure 4 shows the mean diurnal profiles of ozone and NOx for the average non-exceedance and exceedance days in the June-July 1995 period for Washington DC. These data generally show the kind of ozone and NOx departures that will be created in an episode situation; that is, ozone is significantly greater and NOx is ussually smaller during the episode than that found on a non-exceedance day.

Washington DC: The high ozone (i.e., positive ozone departures or ozone in excess of that found, on average, on days when the ozone concentration did not exceed the 1-hour standard [i.e., non-exceedances days]) began to build up on 10 July (hour 84), climaxing with departures in excess of 80 ppb on 15 July(Figure 5). (Nota Bene: the hour in the parentheses that follows a data in this section corresponds to the noon hour on that day; that is hour 84 is noon on 10 July) It should be noted that PAMS data for this site on 17-18 July were missing. Since the average diurnal maximum ozone concentration (DMOC) on non-exceedance days for Washington DC (Table 3) was approximately 69 ppb and departures in excess of 50 ppb were required to reach the standard, the departure data in Figure 5 indicate that the ozone concentrations exceeded the standard on 14 (hour 180) and 15 (hour 204) July. Prior to 10 July (hour 84), the NOx data indicated instances of major buildups of NOx at night and the early morning hours. After 10 July (hour 84), the NOx concentration was, on average, below the values normally found on non-exceedance days.

 

Figure 3 Surface analysis for 15 July 1995 at 1200Z showing frontal position, position of the high pressure center, and the surface wind direction. Dots indicating station position without lines indicating wind direction are locations with zero winds.

Figure 4 Mean diurnal profiles for ozone and NOx for the average non-exceedance and exceedance days in the June-July 1995 period for Washington DC.

 

 

Table 3 The Mean DMOC for Ozone and the Mean Sunrise-to-Sunset Value of NOx and NMHC for the Non-Exceedance Days in June-July 1995. Sunrise/Sunset means the average between 6:00 AM and 6:00 PM local time.

 

 

 

 

 

Location

 

 

 

DMOC

(ppb)

 

Sunrise/

Sunset

NOx

(ppb)

 

Sunrise/

Sunset

NMHC

(ppb)

Washington DC

69

21

180

Upwind Baltimore

71

14

117

Baltimore

61

25

*

Downwind Baltimore

69

10

92

Philadelphia

60

31

112

New York

54

40

235

 

 

 

Figure 5 Departures for ozone and NOx from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for Washington DC. The average non-exceedance profiles were for the period June-July 1995.

 

About 24 hours after large positive ozone departures began to develop, large positive temperature departures developed (Figure 6). The temperature departures reached a peak on 15 July (hour 204) but were also very large on 14 July (hour 180). The mean wind direction at Washington DC for the non-exceedance days was from the south southwest. However, during the 1995 OTAG episode, the winds were from the west or northwest most of the time with brief periods of south and southeast winds.

Figure 6 Departures for various meteorological parameters from their mean diurnal profiles on non-profiles were for the period June-July 1995.exceedance days during the 7-18 July 1995 OTAG episode for Washington DC. The average non-exceedance profiles were for the period June-July 1995.

 

The large positive wind speed departures around 8 (hour 36) and 9 (hour 60) July are associated with the passage of the front or trough line that moved through the region episode period was highly variable that followed the first HPS. The large positive departure around the evening of 15 (hour 204) July is believed to be associated with a storm front stretching from New York to the Florida panhandle that passed through the region. The wind speeds on the 14 (hour 180) and 15 (hour 204) July were as much as 1.0 m/s below that normally found on a non-exceedance day.

Figure 7 Departures for ozone and NOx from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the upwind Baltimore site. The average non-exceedance profiles were for the period June-July 1995.

Upwind of Baltimore (or Downwind of Washington): The high ozone also began to build up around 10 July (hour 84) at this site. The largest positive departure which was in excess of 100 ppb, was reached on 15 July (hour 204) (Figure 7). In this case, the average DMOC on non-exceedance days for this site was approximately 71 ppb (Table 3) and positive departures must be around 50 ppb for an episode to be established on any given day during this period. The departure data in Figure 7 indicate that the ozone concentrations exceeded the one-hour standard on 15 (hour 204) and 17 (hour 252) July. The DMOC on 15 July (hour 204) was 174 ppb. The NOx data indicated instances of buildups of NOx at night and the early morning hours prior to 15 July (hour 204). For the most part, NOx concentrations days on the days when exceedances occurred at this site were below the values normally found on non-exceedance.

As in the case of the Washington DC site, large positive temperature departures began to develop at this site about 24 hours after the large positive ozone departures appeared (Figure 8). The temperature departures reached a peak on 15 July (hour 204) and decreased significantly thereafter, reaching values on 17 July (hour 252) about half that on 15 July (hour 204). Solar radiation data were available at this site and these data indicate that relatively clear skies characterize 15 (hour 204) and 17 (hour 252) July and the magnitude of the solar radiation on those two days exceeded, for the most part, that normally found on non--exceedance days (i.e., positive departures characterized both those days). [Nota Bene: Large negative departures such as those found on 7 (hour 12) and 16 (hour 228) July are generally indicative of cloud effects whereas persistent positive departures are indicative of the lack of cloud interference.] The largest positive departures were found on 8 (hour 36) and 9 (hour 60) July or around the period immediately after the front or trough line passed through the region. Though clouds did not predominate at this site on 15 (hour 204) and 17 (hour 252) July, some other atmospheric factor, possibly suspended particulates or water vapor or both which have been observed to increase in concentration during ozone episodes (Vukovich, 1979), must have influenced the solar radiation measurements since the positive departures on those two days were as much as 40 % smaller than those on 8 (hour 36) and 9 (hour 60) July.

Figure 8 Departures for various meteorological parameters from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the upwind Baltimore site. The average non-exceedance profiles were for the period June-July 1995.

The mean wind direction at this site for the non-exceedance days was also from the south southwest. However, like that at the Washington DC site, the winds were from the west or northwest during most of the 1995 OTAG episode period. There was a brief period of east and southeast winds around 16 (hour 228) and 17 (hour 252) July. The wind speed was again highly variable. The large positive wind speed departures are associated with the passage of the front or trough line that moved through the region early in the period and with a storm front that stretched from New York to the Florida panhandle that passed through the region later in the period. It is interesting to note that there were dramatic reductions in the wind speed on 14 (hour 180), 15 (hour 204), and 17 (hour 252) July. The wind speeds on those days were more than 2.0 m/s below that normally found on a non-exceedance day. This magnitude of wind speed reduction did not occur at the Washington DC site nor at any of the other sites used in this study during the period.

Baltimore: The ozone variations during the OTAG 1995 period at this site were similar to those found at the previous two sites. The high ozone began to build up around 10 July at this site. The largest positive departure which was about 80 ppb, was reached on 15 July (hour 204) (Figure 9). In this case, the average DMOC on non-exceedance days was approximately 61 ppb (Table 3) and positive departures must be at least 60 ppb for an episode to be established. According to the departure data in Figure 9, the ozone concentrations exceeded the one-hour standard on 12 (hour 132) and 15 (hour 204) July. Unfortunately, good quality NOx data were not available from this site during this period.

Like the previous sites, large positive temperature departures began to develop at this site about 24 hours after the large positive ozone departures appeared (Figure 10), reaching a peak on 15 July(hour 204). The temperature departures decreased significantly after 15 July (hour 204) but remained positive over the remainder of the period . The solar radiation indicated that relatively clear skies on 15 July (hour 204) (N.B., solar radiation data were missing on 12 July [hour 132]). The largest positive departures were found on 10 July (hour 84) which is soon after the front or trough line passed through the region. The magnitude of the solar radiation departure on 15 July (hour 204) was about 35 % smaller than that on 10 July (hour 84).

 

Figure 9 Departures for ozone from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the Baltimore site. The average non-exceedance profiles were for the period June-July 1995.

 

The mean wind direction at this site for the non-exceedance days varied through the day. It was generally from the east in the morning and at night and from the southeast during most of the daylight period. However during the 1995 OTAG episode period and unlike the previous sites, winds from the north or northeast persisted. The wind speed was again variable during the entire period. Two of the large positive wind speed departures that are indicated in the data, are associated with the passage of the front or trough line that moved through the region early in the period and with a storm front that stretched from New York to the Florida panhandle that passed through the region later in the period. Negative wind speed departures as large as 1.0 m/s characterized the two episode days.

Downwind Baltimore: Again, the high ozone began to build up around 10 July at this site. The largest positive departure, which was in excess of 100 ppb, was reached on 15 July (hour 204) (Figure 11). The average DMOC on non-exceedance days for this site was approximately 69 ppb (Table 3) so that positive departures must be around 50 ppb for an episode to be established. Episodes occurred at this site on 14 (hour 180) and 15 (hour 204) July and the DMOC on 15 July (hour 204) was 179 ppb. The NOx data indicated two large buildups of NOx at night and the early morning hours just prior to the two episodes, but the NOx concentration during the two episodes was below the values normally found on non-exceedance days.

Figure 10 Departures for various meteorological parameters from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the Baltimore site. The average non-exceedance profiles were for the period June-July 1995.

 

The temperature departure data at this site indicated a slight change in the temperature pattern that took place prior to and during the episodes (Figure 12). The large positive temperature departures began to develop about 48 hours, not 24 hours, after the large positive ozone departures began to develop. However, the peak positive departures were still reached on 15 July (hour 204) though the temperature departures on 14 (hour 180) and 15 (hour 204) July were not significantly different. The temperature departures decreased again after 15 July (hour 204) but remained positive over the remainder of the period. The solar radiation data indicated that there relatively clear skies on 14 (hour 180) and 15 (hour 204) July. The largest positive departures were found on 9 July (hour 60) or soon after the front or trough line passed through the region. The magnitude of the solar radiation departure on 14(hour 180) and 15 (hour 204) July was only slightly less than that on 9 (hour 60) July.

 

Figure 11 Departures for ozone and NOx from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the downwind Baltimore site. The average non-exceedance profiles were for the period June-July 1995.

 

Figure 12 Departures for various meteorological parameters from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the downwind Baltimore site. The average non-exceedance profiles were for the period June-July 1995.

The mean wind direction at this site for the non-exceedance days was, for the most part, from the south southwest. However during the 1995 OTAG episode period, winds were from the north or northwest with brief periods of east winds and northeast winds. The wind speed was variable during the entire period and the variability was similar to that found at the Baltimore site. Two of the large positive wind speed departures that are indicated in the data, were again associated with the passage of the front or trough line that moved through the region early in the period and with a storm front that stretched from New York to the Florida panhandle that passed through the region later in the period. Negative wind speed departures as large as 1.0 m/s characterized the two episode days.

Philadelphia: As in the previous cases, the high ozone began to build up around 10 July (hour 844) at this site. The largest positive departure, which was greater than 60 ppb, was reached on 15 July (hour 204) (Figure 13). The average DMOC on non-exceedance days for this site was approximately 60 ppb (Table 3) so that positive departures must be around 60 ppb for an episode to be established during this period and this occurred only on 15 July (hour 204). The NOx data indicated that a number of large buildups of NOx occurred at night and the early morning hours just prior to the episode on 15 July (hour 204), but the NOx concentrations during the later stages of the ozone buildup and during the episode were below the values normally found on non-exceedance days.

 

Figure 13 Departures for ozone, NOx, CO, and SO2 from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the Philadelphia site. The average non-exceedance profiles were for the period June-July 1995.

 

 

Meteorological data were not available for this site during the OTAG 1995 episode period, but this site did provide CO and SO2 data which can be useful in identifying sources for the NOx (Figure 13). The CO and SO2 departures had essentially the same correlation with the NOx departures during the OTAG 1995 period (i.e., for CO vs. NOx, r = 0.50 and for SO2 vs. NOx, r = 0.51), but there were short periods when the correlation was higher. For example, the nighttime peak associated with the large double peak in the NOx departures found on the night of 9/10 July (hour 60-84) appears to be well correlated with increased CO at that time, whereas the early morning peak appears to be associated more with increased SO2. Furthermore, it appears that the increased NOx departures on 13-14 July (hour 156-180) are associated more with increased SO2 than with increased CO.

Figure 14 Departures for ozone and NOx from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the New York site. The average non-exceedance profiles were for the period June-July 1995.

New York: The ozone departure data for this site shows a significantly different pattern than at the previous sites. The high ozone did not begin to build up until 13 July (hour 156) at this site or about three days after the buildup had commenced at the other sites. The largest positive departure, which was almost 80 ppb, was reached on 14 July (hour 180) (Figure 14) not 15 July (hour 204) as in the previous cases. The average DMOC on non-exceedance days for this site was approximately 54 ppb (Table 3) so that positive departures must be around 70 ppb for an episode to be established during this period, and this occurred only on 14 July (hour 180). The NOx data indicated that a number of large buildups of NOx occurred at night and the early morning hours just prior to the episode on 14 July (hour 180), but the NOx concentrations during the ozone buildup as well as during the episode were below the values normally found on non-exceedance days.

There was also a significant change in the meteorology at this site compared to the previous sites. The temperature departure data (Figure 15) indicated that the large positive departures appeared on or about 13 July (hour 157), not 10 (hour 84) or 11 (hour 108) July like at the previous sites, or precisely when the ozone buildup began. The peak positive departures were still reached on 15 July (hour 204). The temperature departures decreased significantly after 15 July (hour 204) but remained positive for the remainder of the period. Solar radiation data were not available at this site during the OTAG 1995 episode period.

Both the wind speed and wind direction departure data at this site indicated that a strong 4-5 day oscillation affected their variability. The New York site is located in the northern portion of the northeast corridor and this region was affected by frontal passages and frontal oscillations more than the sites to the south. The mean wind direction at this site for the non-exceedance days was, for the most part, from the south. During the 1995 OTAG episode period, winds oscillated from the east and southeast to the north or northwest. The wind speed oscillated from values 1 to 2 m/s below the values on non-exceedance days to as much as 3 m/s greater which is significantly different than that found at the sites to the south. During the episode on 14 July (hour 180), the wind speeds were larger than those found on non-exceedance days by almost 3 m/s in some cases. Between 14 (hour 180) and 15 (hour 204) July, the wind shifted and wind speeds began to decrease, suggesting a front passed through the area which may explain why there was no exceedance at this site on 15 July (hour 204).

 

Figure 15 Departures for various meteorological parameters from their mean diurnal profiles on non-exceedance days during the 7-18 July 1995 OTAG episode for the New York site. The average non-exceedance profiles were for the period June-July 1995.

 

Figure 16 Diurnal profiles of ozone and NOx on 14 and 15 July 1995 and on the average non-exceedance day for the June-July 1995 period at the Washington DC site.

Figure 17 Diurnal profiles for the meteorological parameters on 14 and 15 July 1995 and on the average non-exceedance day for the June-July 1995 period at the Washington DC site.

Figure 18 Diurnal profiles of ozone and NOx on 14 and 15 July 1995 and on the average non-exceedance day for the June-July 1995 period at the upwind Baltimore site.

 

4.0 The OTAG Exceedance Cases

As indicated in the last section, a major ozone exceedance took place in the southern part of the of the northeast corridor on 15 July during the OTAG 1995 episode period at all the PAMS sites examined in the region except New York. At some of these PAMS stations, ozone exceedances also took place on the 14 July, but these were not as large as those on 15 July. The meteorology on 14 and 15 July did not appear to be significantly different, the region being under the influence of the same HPS on those two days. The chemistry and meteorology on those two days were examined to provide insight into those factors that may have caused the more significant episode on 15 July. The PAMS sites at Washington DC, upwind of Baltimore (or downwind of Washington), and downwind of Baltimore were used in this analysis since they had available high quality chemistry and meteorology data on these two days.

 

Figure 19 Diurnal profiles for the meteorological parameters on 14 and 15 July 1995 and on the average non-exceedance day for the June-July 1995 period at the upwind Baltimore site.

 

Figures 16, 18, and 20 provide the diurnal profiles for ozone and NOx and Figures 17, 19, and 21, the diurnal profiles for the meteorological parameters for 14 and 15 July at the three sites. Also included in the figures are the average diurnal profiles for each parameter for the non-exceedance days in June and July 1995. These data are provided for comparative purposes. The meteorological data at these sites indicate that the surface temperature was 6-9 C higher than that on the average non-exceedance day on both days, and though the temperature on 15 July was about 2 _C higher than that on 14 July, the important issue is that the temperature was substantially higher that that on the average non-exceedance day on both days (See Table 4).

Figure 20 Diurnal profiles of ozone and NOx on 14 and 15 July 1995 and on the average non-exceedance day for the June-July 1995 period at the downwind Baltimore site.

Figure 21 Diurnal profiles for the meteorological parameters on 14 and 15 July 1995 and on the average non-exceedance day for the June-July 1995 period at the downwind Baltimore site.

 

 

The winds were from the west on 14 July and from the west southwest on 15 July at the Washington DC and upwind Baltimore sites, and from the west northwest at the downwind Baltimore site both days. The solar radiation (at the sites where data were available) and the wind speed were not significantly different on these two days.

The DMOC (Table 4) was from 30 to 50 ppb higher and the integrated (i.e., the sum) ozone from sunrise to sunset, the period of primary production of ozone, was, on average, 22 percent higher on 15 July than on 14 July even though the meteorology on those two days were similar. The interesting characteristic on those two days was, however, the differences in the NOx concentrations. The integrated sunrise to sunset NOx concentration was, on average, about 60 percent higher on 14 July. The integrated NOx concentration ranged from 37 percent higher (upwind Baltimore) to 86 percent higher (downwind Baltimore). Furthermore, there was a large injection of NOx in the morning at between 5:00 AM and 7:00 AM on 14 July at all three sites similar to that found on non-exceedance days which was not evident on 15 July. The 14th July 1995 is a Friday and the 15th is a Saturday. The difference in the NOx concentration is most likely due to the absence on a Saturday of the rush hour and general automobile traffic that characterizes a weekday/workday like Friday. The difference in the NOx concentration may also explain the difference in the ozone concentration.

Table 4 The DMOC and the Mean Values of the Ozone and NOx Concentration and Meteorological Parameters Between Sunrise and Sunset on 14 and 15 July 1995.

 

Parameter/Place

 

Washington DC

 

Upwind

Baltimore

 

Downwind Baltimore

 

 

14 July

15 July

14 July

15 July

14 July

15 July

DMOC (ppb)

124

155

119

174

143

179

Average* Ozone (ppb)

88

105

90

111

89

109

 

Ratio(15 July O3/14 July O3) (%)

119

124

123

 

Average* Nox (ppb)

12

8

13

9

12

7

 

Ratio(14 July NOx/15 July Nox) (%)

157

137

186

 

Average* Solar Radiation (W/m2)

*

*

526

488

499

515

Average* Temperature (C)

31.5

33.8

31.3

33.1

28.5

31.1

Average* Wind Direction (degrees)

270

249

272

250

289

283

Average* Wind Speed (m/s)

3.2

3.1

1.2

0.8

2.1

1.8

*--Designates sunrise to sunset average.

 

5.0 Weekday/Weekend Effect

Further evidence of the weekday/weekend differences in the ozone concentrations can be seen in the comparison of that effect in Washington DC, Philadelphia, and New York (Figures 22, 23, and 24). The data in the figures were generated using the weekday/weekend PAMS data for June and July 1995. The results in the figures are summarized in Table 5. Though the DMOC was only a few ppb higher on the weekend, there was, on average, about 11 percent more ozone from sunrise to sunset on the weekend. On the other hand, the integrated NOx concentration and the integrated NMHC concentration were, on average, about 60 percent and 38 percent, respectively, higher during the weekday. Furthermore, as in the case of 14 and 15 July 1995, there were a large injections of NOx in the morning on the weekday at all these site, but not on the weekend. Large injections of NMHC were noted in the morning at Philadelphia and during most of the daytime period in New York, but were not evident on the weekend in either place. In Washington DC, the morning injection of NMHC was larger on the weekday, but there was also a morning injection of NMHC on the weekend that was only slightly smaller than that on the weekday.

6.0 Summary and Discussion

This study examined the behavior of ozone and its precursors during the 1995 OTAG episode (i.e., 7-18 July 1995) at sites in the northeast corridor using PAMS data. Four urban sites (i.e., Washington DC, Baltimore, Philadelphia, and New York) and two non-urban sites (i.e., a site upwind and down wind of Baltimore) were used in the study. In examining departures of ozone, of its precursors, and of various meteorological parameters from their average values on non-exceedance days in June-July 1995, we found that the ozone at the sites from Philadelphia southward began to build up around 10 July, immediately after a weak front or low pressure trough passed through the region and a large HPS moved in. Peak ozone concentrations were reached at these sites on 15 July. The level of the ozone departures decreased after 15 July as a cold front approached the region, but the departures generally remained positive through the end of the 1995 OTAG episode period (i.e., 18 July). At some of these sites, nighttime and early morning buildups of the NOx concentration were noted immediately before exceedances took place, but the only consistent pattern noted at all these sites was that the NOx departures were usually negative on the exceedances days.

Figure 22 The mean diurnal profiles of ozone and NOx on a weekday versus a weekend for the period June-July 1995 at the Washington DC site.

Figure 23 The mean diurnal profiles of ozone and NOx on a weekday versus a weekend for the period June-July 1995 at the Philadelphia site.

Figure 24 The mean diurnal profiles of ozone and NOx on a weekday versus a weekend for the period June-July 1995 at the New York site.

 

 

Table 5 Summary of the Weekday/Weekend Effect at Washington DC, Philadelphia, and New York. The Ozone Concentrations are Diurnal Maxima and the NOx and the NMHC Concentrations are Averages from Sunrise to Sunset. RO is the Ratio of the Sunrise to Sunset Integrated Ozone on the Weekend to That on the Weekday. RN and RHC are the the Ratios of the Sunrise to Sunset Integrated NOx and NMHC, respectively, on the Weekday to That on the Weekend.

 

 

 

Location

 

Weekday

DMOC

(ppb)

 

Weekend

DMOC

(ppb)

 

 

RO

(%)

 

Weekday

NOx

(ppb)

 

Weekend

NOx

(ppb)

 

 

RN

(%)

 

Weekday

NMHC

(ppb)

 

Weekend

NMHC

(ppb)

 

 

RHC (%)

Washington DC

68

71

112

23.4

14.5

161

183

159

115

Philadelphia

55

56

105

30.4

19.3

158

127

95

134

New York

52

57

117

43.9

27.9

157

268

162

165

 

At the New York site, the only site examined north of Philadelphia in the northeast corridor, the buildup of the ozone did not begin until 13 July or three days after it began at sites to the south. The peak ozone concentration was reached on 14 July at this site, and the departures decreased on 15 July becoming negative on 16 July around the time a front passed through the region. The NOx departures indicated that a number large buildups of NOx occurred at night and the early morning hours just prior to the episode on 14 July, but the NOx concentration during the ozone buildup as well as during the episode was below the values normally found on non-exceedance days.

At Philadelphia and the sites to the south, the temperature and solar radiation variations were consistent at all the sites. Large positive temperature departures began to build up 24 to 48 hours after the large ozone buildup began. The temperature departures reached a peak on 15 July, decreasing significantly thereafter, but remaining positive through the end of the period. The solar radiation departures were generally large and positive when the exceedances took place, indicating clear skies and significant amounts of solar energy available for the photo chemistry.

The wind speed and direction variations, on the other hand, were not consistent at these sites during this period. The diurnal variation of the wind direction was significantly different from site to site. Winds varied from the north and northwest with a brief period of south and southeast winds at one site, from west and northwest with brief periods east and southeast winds at another site, and from north and northeast at still another site. The diurnal variation of the wind speed was just as variable as that for the wind direction. On the same day, mean wind speeds varied from about 0.8 m/s to 3.0 m/s from site to site. At some sites, the wind speed was less than that found on non-exceedance days in the morning hours, reaching values nearly equivalent to than that found on non-exceedance days in the early afternoon. At other sites, the wind speed decreased significantly at around sunrise, remaining very low throughout most of the daylight period, and increasing again near sunset. The data suggest that these local features in the wind speed had significant effects on the chemistry. For example, in the case when the wind speed was very small through out most of the daylight period, the NOx concentration was noted to build up around noon time, presumably associated with the noon time lunch period. This buildup of the NOx concentration was not noted on non-exceedance days. It is suggested that the buildup of the NOx concentration must influence the local ozone chemistry.

At New York, there was also consistency in the temperature variations in that large positive temperature departures began to build up prior to and during the exceedance day. The large positive temperature departures first appeared on 13 July, not 10 or 11 July like the sites to the south. The peak positive departures were also reached on 15 July, and the temperature departures decreased significantly later on 15 July. The wind speed and wind direction departure data indicated a strong 4-5 day oscillation which affected their variability. This site is in the northern portion of the northeast corridor which was affected by frontal passages and frontal oscillations more than the southern portions. During this period, wind direction oscillated from the east and southeast to the north or northwest, and the wind speed,

from 1 to 2 m/s below the values on non-exceedance days to as much as 3 m/s greater. During the episode on 14 July, the wind speeds were, at times, larger than those found on non-exceedance days by almost 3 m/s. Around the 15 July, a front passed through the area which may explain why there was no exceedance of the 1-hour standard for ozone at this site on 15 July.

Figure 25 The MM5 wind field for 15 July 1995 at 1200 Z. Wind speeds were as large as 12.0 m/s over the ocean (black area) and as small as 1.0 m/s over southern Louisiana. Over the remainder of the area wind speeds varied from 3.0 to 6.0 m/s.

The meteorological data indicated that there was consistency in the surface temperatures and solar radiation during the episode days; that is, the temperatures and the solar radiation were higher than that found on non-exceedance days at all sites. However, the wind speed and direction were highly variable from site to site, producing no temporal or spatial consistency in the flow pattern, particularly on exceedance days. The wind field was diffusive (i.e., dominated by turbulence and not the large-scale pressure gradient), and not conducive of transport, particularly long-distance transport. This result is highly inconsistent with the kinds of wind fields produced by mesoscale prediction models (e.g., MM5) and used by air quality models. These wind fields usually demonstrate very consistent flow patterns with well defined anticyclonic flow (i.e., dominated by the large-scale pressure gradient and not by turbulence) in such episodes (Compare Figure 25 to Figure 3). Since the data indicate that the local winds will influence the local chemistry, the results not only suggest that the wind fields used by the air quality models are questionable in terms of transport in episode situations, but the subsequent chemistry is also suspect.

Besides examining aspects of the meteorology and chemistry of the 1995 OTAG episode, the study also examined weekday/weekend differences in ozone and its precursors. Exceedances occurred both on 14 and 15 July at most of the PAMS sites examined in the southern part of the northeast corridor. The DMOC on 15 July at those sites was from 30 to 50 ppb higher than on 14 July and the integrated sunrise to sunset ozone on 15 July (i.e., during the period of primary production of ozone) was, on average, 22 percent higher even though the meteorology on those two days was very similar. The temperature was about 2 _C higher on 15 July but that difference was smaller than the temperature departures relative to the average non-exceedance day (i.e., about + 8.0 _C, on average) that existed on these two days. The solar radiation, the wind speed, and wind direction were not significantly different on the two days. The integrated NOx concentration between sunrise to sunset, on the other hand, was, on average, about 60 percent higher on 14 July. Furthermore, there was a large injection of NOx in the morning between 5:00 AM and 7:00 AM on 14 July at these sites which was not evident on 15 July. The 14th July 1995 is a Friday and the 15th July is a Saturday. The difference in the NOx concentration is most likely due to the absence on a Saturday of the rush hour and general automobile traffic that characterizes a weekday/workday (i.e., Friday).

 

Table 6 The Mean DMOC for Ozone and the Mean Sunrise-to-Sunset Value of NOx and NMHC for All Exceedance Days in June-July 1995.

 

 

 

 

Location

 

 

 

DMOC

(ppb)

 

Sunrise/

Sunset

NOx

(ppb)

 

Sunrise/

Sunset

NMHC

(ppb)

Washington DC

132

11

122

Upwind Baltimore

141

11

*

Baltimore

120

*

*

Downwind Baltimore

130

15

128

Philadelphia

120

27

138

New York

121

40

247

*--missing data

These same characteristics were noted in our comparison of the weekday/weekend differences in the ozone concentrations at Washington DC, Philadelphia, and New York. The integrated sunrise to sunset ozone concentration was about 11 percent higher on the weekend. On the other hand, the integrated NOx concentration and the integrated NMHC concentration were, on average, about 60 percent and 38 percent, respectively, higher during the weekday. Furthermore, there was a large injection of NOx in the morning at all the sites on the weekday, but not on the weekend. Large injections of NMHC were noted in the morning at Philadelphia and during most of the daytime period in New York, but were not evident on the weekend in either place. In Washington DC, the morning injection of NMHC was larger on the weekday, but there was also a morning injection of NMHC on the weekend that was only slightly smaller than that on the weekday. The notion that lower values of NOx and NMHC may lead to higher values of ozone in this region can also be seen in comparing the average DMOC with the average sunrise to sunset NOx and NMHC concentrations on non-exceedance days (Table 3) and on exceedance days (Table 6). Some of the highest values of the DMOC are associated with the lowest values of NOx and NMHC.

Ozone production is a complex process (Seinfeld, 1989; Hough, 1988; Dodge, 1989). On a weekday/workday when the morning rush hour traffic produces the large morning injections of NOx and when there are generally more NOx emissions during the daylight hours because there is normally more automobile traffic during that period, ozone levels are reduced by the presence of fresh NO since NO reacts rapidly with ozone, removing ozone from the system. NO makes up a significant portion of the NOx emissions in urban regions. The fresh NO must be removed from the system before sizeable ozone production can develop. Furthermore, around mid-day when the solar radiation is most intense due to the high solar zenith angles, the excited atomic oxygen atom, O(1D), can be produced by photodissociation of ozone when sufficient UV solar radiation is available. The excited atomic oxygen atom subsequently reacts with water vapor to produce two OH radicals and the OH radical is used to produce the peroxy radicals through various reactions. The peroxy radical then reacts with NO, removing NO from the system and producing NO2. The photodissociation of NO2 produces O(3P) which reacts with oxygen to produce ozone. The production of O(1D) drives the system to produce very large concentrations of ozone when the system initially has a large ozone concentrations (i.e., ozone begets ozone when NO is also present).

For the weekend day, or any day (e.g., holidays) that has lower NOx (i.e., lower fresh NO) concentrations than that normally found on a weekday in most large cites, large ozone concentrations can be developed earlier in the day than on a weekday. Therefore, the mid-day ozone production through the O(1D) process may be more intense on the weekend than on the weekday (i.e., under the same general meteorological conditions and with the limitations on the number of hours of sufficient amounts of solar energy for ozone photodissociation on a given day, net mid-day ozone production should be larger on a weekend day than on a weekday/workday). Overall, there should be more ozone produced from sunrise to sunset on a weekend day than on a weekday/workday because there is less NO available to remove ozone and potentially more mid-day ozone production through the photodissociation of ozone process. Meteorology will perturb the system, so estimates of absolute effects can only to attained when meteorological condition are similar on a weekday/workday and a weekend day.

The weekend/weekday comparisons for Washington DC, Philadelphia, and New York indicated that, on average, ozone production began earlier in the morning on the weekend day than on the weekday/workday due presumably to the lower morning NOx concentrations. The DMOC was only a few ppb higher on the weekend day suggesting that there was no dramatic mid-day production of ozone, on average. In the episode case study, when the same HPS influenced the region and the meteorology was essentially the same over the region on a weekday/workday (i.e., Friday, 14 July) and the neighboring weekend day (i.e., Saturday, 15 July), ozone production began earlier in the morning on the weekend day than on the weekday/workday due again to the lower morning NOx on the weekend day. However in this case, the data also indicated that there were mid-day periods of intense ozone production on the weekend day at all sites studied and the DMOC was 30 to 50 ppb higher suggesting that the ozone production process involving the photodissociation of ozone may have been driving the system. Because ozone production started earlier in the morning on the weekend day, the ozone concentration reached higher values at mid-day when the photodissociation takes place. The presence of higher concentrations of ozone at this time suggest that a more significant effect of the photodissociation would take place. It should be noted, however, that secondary maxima for NOx developed during on the weekend day (i.e., 15 July) around mid-day when the intense ozone production occurred which was not found on the weekday/workday. The secondary maxima for NOx, in some cases but not all cases, were associated with local wind variations (i.e., low wind speeds). The role of the secondary maxima in NOx on the intense ozone production is not clear.

Though the amount of data used for the weekday/weekend studies is limiting, the preliminary results from that study would suggest that the following inferences can be made. In Washington DC, Philadelphia, and New York, reductions in the NOx concentration by much as 60 percent and the VOC concentration by as much as 38 percent, on average, may have a localized disbenefit effect; that is, the DMOC and the integrated amount of ozone from sunrise to sunset may increase as a result of these levels of NOx and VOC reductions. It appears that the disbenefit effects associated with lower NOx and VOCs are greatest on high ozone or episode days (i.e., when meteorological conditions are conducive to producing high ozone). It should be noted that for the cases studied, increases in ozone on the weekend were found when the NOx concentration on the weekday was as much as 86 percent greater than that on the weekday. It is not clear from these results at what level of reductions for NOx and VOCs must be achieved at the sites investigated to lower the value of the DMOC and of the sunrise to sunset integrated amount of ozone.

Finally, the results from the weekday/weekend study suggest a unique methodology for evaluation of air quality (AQM) models. AQMs can be challanged to reproduce the weekday/weekend differences noted above. Since such differences simulate results that may be obtained from control strategy measures, the results of the model evaluation based on such a challange would provide a clear indication of the utility of the model as a tool for the development of control strategies.

7.0 The Study Findings

The principal findings of this study are as follows.

During the OTAG 1995 episode period and at the PAMS sites examined, when exceedance took place, temperatures and solar radiation levels were consistently higher than those found on the average non-exceedance day.

During the OTAG 1995 episode period, particularly on days when exceedances took place, and at the PAMS sites in Philadelphia and southward, the wind speed and direction were highly variable from site to site, producing no temporal or spatial consistency in the flow pattern. The weak pressure gradients associated with the HPS could not sustain an organized wind field. The wind field was diffusive, and not necessarily conducive of transport, particularly long-distance transport.

During the OTAG 1995 episode period and at the New York PAMS site, a 4-5 day oscillation of the wind speed and direction characterized the temporal variability of the wind. Imbedded within this periodic variation were high frequency variations of both the wind speed and direction so that there was no temporal consistency in the wind at higher frequencies than 4 or 5 days. There was at times spatial consistency of the wind field in the neighborhood of New York city due to the presence of a frontal boundary near the site. The pressure field associated with the front produced an organized wind field

During the OTAG 1995 episode period and at the PAMS sites examined, secondary maxima in the NOx distribution were found, generally at mid-day, when a local wind minimum developed. Since the presence of the secondary maximum in the NOx distribution will undoubtedly affect the local chemistry, the local wind variations were affecting the local chemistry.

At the PAMS sites examined and during the June-July 1995 period, weekdays had, on average, about 60 percent more NOx and 38 percent more NMHC from sunrise to sunset than that found on a weekend day, but there was 11 percent more total ozone on the weekend days during that period.

At the PAMS sites examined and during the July 1995 OTAG episode period, when meteorological conditions were conducive for episode development and the meteorological conditions on a weekday/workday and on a neighboring weekend day (e.g., a Friday and the following Saturday) were reasonably similar, the DMOC on the weekend day was from 30 to 50 ppb higher and the integrated sunrise to sunset ozone was, on average, 22 percent higher than that on the weekday/workday. In the case examined, the integrated NOx concentration between sunrise to sunset was, on average, about 60 percent higher on the weekday.

8.0 Recommendations for Future Studies

The PAMS data set provides a remarkable tool to study the behavior of ozone and its precursors. Further study using these data will provide significant information useful in understanding processes that govern the behavior of ozone and in planning control strategies for ozone. The following is a brief list of studies that may be useful for these purposes.

1) The present study of the weekday/weekend differences in the diurnal cycle for ozone and its precursors used limited amounts of data and was centered in the northeast corridor region. The database should be expanded to included data for the summer ( i.e., June, July, and August) in the period 1994-1997 as well as include other areas where PAMS data are available. The data should be stratified according to the size of the cities in one case and according to the level of the DMOC (e.g., 120 ppb, 80 ppb but < 120 ppb, and <80 ppb) in another.

2) The present study of the weekday/weekend differences in the diurnal cycle for ozone and its precursors in episode situations treated only one case study. Further study is necessary, providing a sufficient number of case studies to determine conclusively the nature of the difference. The case studies should emphasize weekday/weekend episodes that take place under the same meteorological conditions, preferably the same meteorological situation (i.e., a Friday and Saturday or a Sunday and Monday that are under the influence of the same HPS).

3) A study of the spectrum of VOCs on a weekday versus a weekend day should be performed to see if there is a difference in which specie(s) dominates at those times. These results should be used to examine weekday/weekend differences in the volatility and subsequent ozone production, if any.

4) Averages of the diurnal variation of ozone and its precursors should be developed for each day of the week (i.e., Sunday, Monday, Tuesday, etc.) during the summer period ( i.e., June, July, and August) for the period 1994-1997 to determine if differences other than the weekend-weekday/workday differences can be detected and the cause of those differences determined.

5) The present study of the characteristics of the meteorology in the northeast corridor in episode situations provided some interesting information about the temporal and spatial distribution of the wind speed and direction in that region and its potential effects on the local chemistry. This study should be expanded to include other regions where PAMS data are available.

6) The PAMS data should be used for an in-depth evaluation of OTAG 1995 modeling results. The present study showed discrepencies between the observed and modeled wind field and suggested that not only would there be resulting transport differences but that the local chemistry would be effected. The chemistry as well as aspects of the wind field can be examined through time series analysis of ozone, its precursors, and of the meteorology at selected points in the modeling domain using the PAMS data and the model data.

7) A limited summertime climatology (1994-1997) of ozone and its precursors (i.e., NOx and NMHC) should be developed for the various cities where PAMS data are available. The results should be compared with a climatology of various meteorological parameters to determine how the ozone precursors vary with meteorology (N.B., studies have shown that ozone varies decisively with meteorology). The results should be examined in terms of city size, cities that are NOx limited, and cities that are VOC limited.

 

 

 

References

Dodge, M.C., 1989: A comparison of three photochemical oxident mechanisms. J. Geophys. Res., 94, 5121-5136.

Hough, A.M, 1988: An intercomparison of mechanisms for the production of photochemical oxidents. J. Geophys. Res., 93, 3789-3812.

Seifeld, J.H.: Urban Pollution: state of science. Science, 243, 745-753.

Vukovich, F.M., 1979: A Note on Air Quality in High Pressure Systems, Atmospheric Environment, 13, 255-265.

 

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.