United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/4-81-011
March 1981
Air
Philadelphia Oxidant
Data  Enhancement Study
Analysis and Interpretation
of Measured  Data

-------
                                EPA-450/4-81-011
Philadelphia  Oxidant Data
    Enhancement Study
Analysis and Interpretation

       of  Measured  Data

                  by

    Douglas AT lard, Michael Chan, & Chris  Marlia
         AeroVironment,  Incorporated
             145 Vista  Avenue
         Pasadena, California  91107

                  and

             Dr. Edgar  Stephens
     Statewide Air Pollution Research Center
      University of California at Riverside
         Riverside, California 92521


           Contract No.  68-02-3332
         EPA Project Officer: Norman Possiel
                Prepared for


      U.S. ENVIRONMENTAL PROTECTION AGENCY
         Office of Air, Noise and Radiation
      Office of Air Quality Planning and Standards
      Research Triangle Park, North Carolina 27711


                March 1981

-------
This report was  furnished  to  the  U.S. Environmental
Protection Agency  by AeroVironment Inc., Pasadena,
California, in  fulfillment  of Contract No. 68-02-3332.
The contents of this  report are  reproduced  herein  as
received  from  AeroVironment  Inc.   The   opinions,
findings, and conclusions  expressed are  those of the
authors   and   not   necessarily    those    of   the
U.S. Environmental  Protection Agency.   Mention  of
company or product names is not  to be considered as  an
endorsement  by  the  U.S. Environmental  Protection
Agency.

-------
                                     SUMMARY
     The objectives of this study were twofold:

     1.    To obtain air quality and meteorological data of sufficient extent and quality
           to apply the Systems Application, Incorporated (SAI) photochemical model to
           the Philadelphia area.

     2.    To  recommend days suitable for model validation through  analysis of field
           data.

     Sixteen  surface air quality  sites,  eighteen  surface  meteorological sites, and three
upper-air meteorological sites contributed data to the study during  the summer  of 1979.
In addition, helicopter-based measurements  of air quality and meteorology and measure-
ments of hydrocarbon species were performed.

     Based upon  analysis  of all data, the following days were recommended for model
validation:

                July 12                August 4              August 10
                July 13                August 5              August 22
                July 16                August 6
                July 19                August 7

-------
                               ACKNOWLEDGEMENTS
      Many people at AeroVironment Inc., besides the authors, contributed significantly to
this study. Other primary participants and their areas of responsibility are:

      Mr.  Robert Baxter, Field Manager
      Mr.  Don Gonzales, Data Processing
      Ms.  Diane Barker, Report Editing
      Mrs. Darlene Asamura, Technical Typing
      Ms.  Gloria Best, Technical Illustrating
      Mr.  Timothy Press, Data Analysis
      Mr.  Howard Hammeren, Field Calibration
      Mr.  David Pankrantz, Senior Field  Technician
      Mr.  Steven Fisher, Field Technician
      Mr.  David Emery, Field Technician
      Mr.  Stan Krzywonos, Field Technician

      A number of other organizations  were also involved in this  study.  The U.S. EPA's
Environmental Monitoring Systems Laboratory, Las Vegas,  Nevada (EMSL-LV), supplied an
instrumented helicopter.   Mr.  Charles  Fitzsimmons was  responsible for  EMSL-LV field
operations.  The helicopter measurements were made by Northrop Services' Environmental
Science Center under the supervision  of Mr. Calvin Hancock.  Mr. Barry  Martin, EPA
Environmental  Monitoring and Support  Laboratory,  Research  Triangle  Park,  North
Carolina,  and his staff gave invaluable assistance with instrumentation  problems.  Finally,
the contribution of Mr. Norm Possiel, EPA Project  Officer, to  the technical direction of
the project is certainly one of the most appreciated.
                                         u

-------
                               TABLE OF CONTENTS
     SUMMARY                                                                   i
     ACKNOWLEDGEMENTS                                                      li
     LIST OF FIGURES                                                           iii
     LIST OF TABLES                                                            ix

1.    INTRODUCTION                                                             1

2.    FIELD MEASUREMENT PROGRAM                                            2

     2.1   Objectives of the Field Program                                          2
     2.2   Program Design                                                         2
     2.3   Surface Measurements at Fixed Sites                                      3
           2.3.1    Special Study Sites                                              3
           2.3,2    Existing Air Quality/Meteorology Monitoring Stations              8
           2.3.3    Supplemental Meteorological Sites                               14
           2.3.14    Special PAN/Nitric  Acid/NOx Monitoring                        14
     2.4   Aircraft Measurements of Air Quality and Meteorology                    14
     2.5   Balloon Measurements of Upper Air Meteorology                          19
           2.3,1    Radiosonde Releases                                           19
           2.5.2    Pibal Releases                                                 19
     2.6   Hydrocarbon Species Measurements                                      22
     2.7   Quality Assurance                                                      22
           2.7,1    Surface Monitoring Program                                     24
           2.7,2    Audits                                                        26
     2.8   Data Reduction and Compilation Procedures                              27

3.    OZONE AND PRECURSOR CONCENTRATIONS IN THE STUDY AREA           29

     3.1   Data Summaries                                                        29
           3.1.1    Representativeness  of the 1979 Summer Months                   29
           3.1.2    Average and Maximum Levels of Pollutants                       32
           3.1,3    Ozone Wind Roses                                              36
     3.2   Hydrocarbon Species Analysis                                           42
           3.2.1    Data Summary                                                 42
           3.2.2    Acetylene/n-Butane Ratios                                     46
           3.2.3    Hydrocarbon Data Validity                                      47
           3.2.4    Limitation of Analytical Scheme                                48

4.    CASE STUDY DAYS                                                         53

     4.1   July 12, 1979  (Thursday)                                                 61
           4.1.1    Synoptic Meteorology                                          61
           4.1.2    Mesoscale Meteorology                                         61
           4.1.3    Precursor  Patterns                                             65
           4.1.4    Ozone Patterns                                                65
           4.1.5    NO2 Patterns                                                  69
           4.1.6    Concentrations  Aloft                                           69

-------
                          TABLE OF CONTENTS
4.2   July 13,  1979 (Friday)                                                    72
      4.2.1    Synoptic Meteorology                                           72
      4.2.2    Mesoscale Meteorology                                          72
      4.2.3    Precursor Patterns                                              75
      4.2.4    Ozone Patterns                                                 75
      4.2.5    NO2 Patterns                                                   80
      4.2.6    Concentrations Aloft                                            80
4.3   July 16,  1979 (Monday)                                                   83
      4.3.1    Synoptic Meteorology                                           83
      4.3.2    Mesoscale Meteorology                                          83
      4.3.3    Precursor Patterns                                              86
      4.3.4    Ozone Patterns                                                 86
      4.3.5    NO  Patterns                                                   89
      4.3.6    Concentrations Aloft                                            89
4.4   July 19,  1979 (Thursday)                                                 92
      4.4.1    Synoptic Meteorology                                           92
      4.4.2    Mesoscale Meteorology                                          95
      4.4.3    Precursor Patterns                                              95
      4.4.4    Ozone Patterns                                                 98
      4.4.5    NO- Patterns                                                  101
      4.4.6    Concentrations Aloft                                           101
      4.4.7    Hydrocarbon Species Data                                      101
4.5   August 4, 5, 6, 7, 1979 (Saturday, Sunday, Monday, Tuesday)               104
      4.5.1    Synoptic Meteorology                                          104
      4.5.2    Mesoscale Meteorology                                         110
      4.5.3    Precursor Patterns                                             115
      4.5.4    Ozone Patterns                                                121
      4.5.5    NO- Patterns                                                  133
      4.5.6    Concentrations Aloft                                           133
      4.5.7    Hydrocarbon Species Data                                      141
4.6   August 10, 1979 (Friday)                                                144
      4.6.1    Synoptic Meteorology                                          144
      4.6.2    Mesoscale Meteorology                                         144
      4.6.3    Precursor Patterns                                             147
      4.6.4    Ozone Patterns                                                147
      4.6.5    NO- Patterns                                                  151
      4.6.6    Concentrations Aloft                                           151
      4.6.7    Hydrocarbon Species                                           154
4.7   August 22, 1979 (Wednesday)                                            157
      4.7.1    Synoptic Meteorology                                          157
      4.7.2    Mesoscale Meteorology                                         160
      4.7.3    Precursor Patterns                                             160
      4.7.4    Ozone Patterns                                                160
      4.7.5    NO- Patterns                                                  163
      4.7.6    Concentrations Aloft                                           163
                                    IV

-------
                              TABLE OF CONTENTS
5.    CONCLUSIONS                                                          167

     5.1   Morning Ozone Levels Aloft                                          167
     5.2   Ozone Levels in the Philadelphia Urban Plume                          169
     5.3   Mesoscale Transport                                                 171
     5A   Synoptic Transport                                                  173
     5.5   Days Recommended for Model Verification                             173

6.    BIBLIOGRAPHY                                                         175

-------
                                 LIST OF FIGURES


Number                  Description

   1          Locations of monitoring sites set up specifically for the study           5

   2          Exterior view of the monitoring station at Site 5                        9

   3          Existing air quality/meteorological monitoring stations                 10
              used in the study

   4          Locations of supplemental meteorological monitoring sites             15
              used in the study

   5          EPA-LAS Vegas instrumented UH-1H helicopter                       17

   6          Locations of helicopter soundings and sampling sites                    20

   7          Locations of ground-level sampling sites                              23

   8          Ozone wind rose for Summit Bridge, DE                               37

   9          Ozone wind rose for Downingtown, PA                                38

   10         Ozone wind rose for Lumberton, NJ                                  39

   11         Ozone wind rose for Robbinsville, N3                                  40

   12         Ozone wind rose for Van Hiseville, N3                                 41

   13         Total ion chromatograph from ambient air sample                     51

   14         Ozone profile over Summit Bridge, DE at 0535 EST on                 58
              August 10, 1979

   15         Synoptic situation, 0700 EST, July 12, 1979                            62

   16         Synoptic trajectory of  air parcels arriving at Philadelphia              63
              at 1300 EST, July 12, 1979

   17         Streamline analysis depicting surf ace flow across the study             64
              area at 0700 EST on July 12, 1979

   18         Isopleths of ozone at 1300, 1400, and 1500 EST, duly  12, 1979           67

   19         Surface trajectory for  various air parcels located within the            68
              study area on duly 12,  1979
                                          VI

-------
                                LIST OF FIGURES



20         Diurnal profiles of ozone at upwind sites on July 12, 1979              71

21         Synoptic situation, 0700 EST, July 13, 1979                            73

22         Synoptic trajectory of air parcels arriving at Philadelphia              74
           at 1300 EST, July 13, 1979

23         Isopleths of ozone at 0900, 1100, 1200, and 1300 EST,                  77
           July 13, 1979

24         Surface trajectory for various air parcels located within the            79
           study area on July 13, 1979

25         Diurnal profiles of ozone at upwind sites and Chester, PA on           81
           July 13, 1979

26         Synoptic situation, 0700 EST, July 16, 1979                            84

27         Synoptic trajectory of air parcels arriving at Philadelphia              85
           at 1300 EST, July 16, 1979

28         Isopleths of ozone at 1200, 1300, and 1400 EST, July 16,  1979           88

29         Surface trajectory for various air parcels located within the            90
           study area on July 16, 1979

30         Diurnal profiles of ozone at upwind stations on July 13,  1979           91

31         Synoptic situation, 0700 EST, July 19, 1979                            93

32         Synoptic trajectory of air parcels arriving at Philadelphia              94
           at 1300 EST, July 19, 1979

33         Diurnal profiles of PAN at upwind sites and Van  Hiseville, NJ           97
           on July 19, 1979 and July average

34         Isopieths of ozone at 1400, 1500, and 1600 EST, July 19,  1979           99

35         Surface trajectory for various air parcels located within the           100
           study area on July 19, 1979

36         Diurnal profiles of ozone and NO at upwind stations on                102
           July 19, 1979

37a        Synoptic situation, 0700 EST, August 4, 1979                         106

37b        Synoptic situation, 0700 EST, August 5, 1979                         107

37c        Synoptic situation, 0700 EST, August 6, 1979

-------
                                LIST OF FIGURES
37d        Synoptic situation, 0700 EST, August 7, 1979                          109

38a        Synoptic trajectory of air parcels arriving at Philadelphia             111
           at 1300  EST, August 4, 1979

38b        Synoptic trajectory of air parcels arriving at Philadelphia             112
           at 1300  EST, August 5, 1979

38c        Synoptic trajectory of air parcels arriving at Philadelphia             113
           at 1300  EST, August 6, 1979

38d        Synoptic trajectory of air parcels arriving at Philadelphia             114
           at 1300  EST, August 7, 1979

39         Isopleths of ozone at 1400,  1500, 1600, and 1700 EST,                  122
           August 4, 1979

40         Isopleths of ozone at 1400,  1500, 1600, and 1700 EST,                  123
           August 5, 1979

41         Isopleths of ozone at 1400,  1500, 1600, and 1700 EST,                  124
           August 6, 1979

42         Isopleths of ozone at 1400,  1500, 1600, and 1700 EST,                  125
           August 7, 1979

43         Streamline analysis depicting surface flow across the study            126
           area at  1500 EST on August 4, 1979

44         Surface  trajectory for various air parcels located within the           128
           study area on August 4,  1979

45         Surface  trajectory for various air parcels located within the           129
           study area on August 5,  1979

46         Surface  trajectory for various air parcels located within the           131
           study area on August 6,  1979

47         Surface  trajectory for various air parcels located within the           132
           study area on August 7,  1979

48         Ozone concentrations reported during helicopter transects             136
           during mid-afternoon flight on August 4,  1979

49         Vertical profiles of O-j, NOx,  and Temperature obtained               137
           by the instrumented helicopter over West Chester, PA at
           0758 EST on August 5, 1979
                                       vm

-------
                               LIST OF FIGURES
50         Ozone concentrations reported during helicopter transect              138
           from downtown Philadelphia to iMedford, NJ on August 5, 1979
           at an altitude of approximately 550 m MSL

51         Vertical profiles of O,, NO , and Temperature obtained               140
           by the instrumented helicopter over Pottstown, PA at 0628 EST
           on August 7, 1979

52         Synoptic situation, 0700 EST, August 10, 1979                        145

53         Synoptic trajectory of air parcels arriving at Philadelphia              146
           at 1300 EST, August 10, 1979

54         Isopleths of ozone at 1300, 1400, 1500, and 1600 EST,                 149
           August 10, 1979

55         Surface trajectory for various air parcels located within the           150
           study area on August 10, 1979

56         Vertical profiles of O,, iNO , and Temperature obtained               152
           by the instrumented helicopter over Summit Bridge, DE at
           0535 EST on August 10, 1979

57         Ozone concentrations reported during helicopter transect from         153
           Brandywine, PA to Philadelphia on August 10, 1979 at an
           altitude of approximately 850 m MSL

58         Diurnal profiles of ozone at Summit Bridge, DE and Downingtown,      155
           PA on August 10,  1979

59         Synoptic situation, 0700 EST, August 22, 1979                        158

60         Synoptic trajectory of air parcels arriving at Philadelphia              159
           at 1300 EST, August 22, 1979

61         Isopleths of ozone at 1400  and 1500 EST, August 22, 1979              162

62         Surface trajectory for various air parcels located within the           164
           study area on August 22, 1979

63         Diurnal profiles of ozone at Downingtown, PA and Summit Bridge,      166
           DE on August 22,  1979
                                       IX

-------
                                  LIST OF TABLES


Number                   Description


   1          Responsible  organizations and periods of operation of fieid               4
              monitoring program components

   2          Meteorological and air quality parameters monitored at the              6
              five sites set up specifically for the study

   3          Monitoring instruments with their lower detetection limits,              7
              data precision, and lower limits of data validity

   14          Meteorology and air quality parameters monitored at                   11
              existing stations

   5          Site locations and operating agencies of the existing air                12
              quality/meteorology stations

   6          Locations of existing air quality/meteorology monitoring                13
              sites with respect to Philadelphia urban core

   7          Locations of supplemental meteorological  monitoring sites              16
              with respect to Philadelphia urban core

   8          Parameters measured  by  helicopter  and instruments used                18

   9          Summary of  aircraft sampling missions actually flown                  21

   10         Comparison  of meteorological parameters recorded during each         30
              summer month of 1979 with corresponding long-term means

   11         Summary of  high ozone days from 3uly 2 through September             31
              18, 1979 for  the study year  and the previous four years
              at Bristol, Pennsylvania

   12         Summary of  ozone data collected during the summer  of 1978            33
              and 1979 at the same location

   13         Peak and average ozone  precursor concentrations for the  study          34
              period at all air quality monitoring sites

   14         Ozone frequency distribution for the study period at several             35
              sites in and around Philadelphia

   15         Species breakdown of  Franklin Institute samples                       44

-------
                                LIST OF TABLES
32         6:00 to 9:00 a.m. LDT averages of NO  and NMHC and the
           ratio of NMHC to NO for rural and urban sites on August 10
                               /C

33         6:00 to 9:00 a.m. LDT averages of NO  and NMHC and the            161
           ratio of NMHC to NO for rural and urban sites on August 22
                               \

34         Summary of morning ozone concentrations observed generally         168
           upwind on case study days

35         Locations of peak ozone concentrations with respect to               170
           Philadelphia

36         Source areas for peak ozone levels and areas impacted by             172
           emissions from Philadelphia's central business district as
           determined by mesoscale surface trajectories
                                       xu

-------
                                LIST OF TABLES
 16         Species breakdown of industrial site samples                           45

 17         Distribution of hydrocarbon species in simultaneous sample             49
           pairs, taken at Lancaster and Downingtown, Pennsylvania,
           respectively

 18         Urban mixing heights for the case study days as determined             55
           from radiosonde released from downtown Philadelphia

 19         Time when rural mixing  heights first exceeded 1000 m as               57
           determined by AV Acoustic Sounder at Summit Bridge, Delaware

 20         Peak O, and NO,  concentrations, wind data,  and helicopter             60
           data availability Tior case study days for the entire study area

 21         6:00 to 9:00 a.m. LOT averages of NO  and NMHC and the              66
           ratio of NMHC to NO for rural and urban sites on 3uly 12

 22         6:00 to 9:00 a.m. LOT averages of NO  and NMHC and the              76
           ratio of NMHC to NO for rural and urban sites on July 13

 23         6:00 to 9:00 a.m. LOT averages of NO  and NMHC and the              87
           ratio of NMHC to NO for rural and urban sites on July 16
                               .X

 24         6:00 to 9:00 a.m. LOT averages of NO  and NMHC and the              96
           ratio of NMHC to NO for rural and urban sites on July 19

 25         Maximum  ozone and NO- concentrations observed in the study         105
           area during the period August 4 through August 7,  1979

 26         Vertical extent of thermal roots as  determined from acoustic          116
           sounder data at Summit Bridge, DE

 27         6:00 to 9:00 a.m. LDT averages of NO  and NMHC and the             117
           ratio of NMHC to  NO  for rural and urban sites on August 4
                               A

 28         6:00 to 9:00 a.m. LDT averages of NO  and NMHC and the             118
           ratio of NMHC to  NO  for rural and urban sites on August 5
                               X

 29         6:00 to 9:00 a.m. LDT averages of NO  and NMHC and the             119
           ratio of NMHC to  NO  for rural and urban sites on August 6

 30         6:00 to 9:00 a.m. LDT averages of NO  and NMHC and the             120
           ratio of NMHC to  NO  for rural and urban sites on August 7
                               A
31         Average mixed layer ozone concentrations aloft on August 4,          134
           4, 5, 6 and 7, 1979
                                       XI

-------
                                 1. INTRODUCTION

      The Clean Air Act Amendments of 1977 require  states to submit State Implementa-
tion Plans (SIPs) that demonstrate the attainment of the National Ambient Air Quality
Standard (NAAQS) for ozone (and certain other pollutants) by 1982.  If such attainment is
shown to  be impossible using reasonably available control  measures, an extension to  1987
can be granted.  To  assist the states  in developing their SIPs, the U.S. Environmental
Protection Agency (EPA)  is conducting a program to provide technical  information and
guidance on available procedures for estimating control requirements needed to attain the
NAAQS for ozone.

      Photochemical  models, particularly the Systems  Applications,  Incorporated (SAI),
Airshed Model  (Reynolds et al., 1978) can be used to quantify control  requirements.  The
EPA  plans to  verify  this  model  for  several  urban  areas  which  typify  different
photochemical  and meteorological regimes.  When this model is verified,  the EPA will be
in a position to establish the range of the model's applicability and to issue guidance on its
application.

      One  of the  urban areas under study  is Philadelphia, Pennsylvania.   Considerable
information on  the  transport  of  ozone and  precursors into the Philadelphia  area  was
obtained  in a  study  conducted in  1978 (Chan  et al.,  1979), even though the primary
emphasis  of that study was to evaluate and demonstrate different approaches to measure
such transport.  To enhance the data base for verifying  the Airshed Model, another study
was conducted  in Philadelphia  in the summer of  1979. Besides gathering data for model
verification,  another  objective of  that study  was  to  increase our knowledge of the
photochemical  and meteorological processes which lead  to peak ozone and NO., concen-
trations in Philadelphia.

      This  report documents the data collection  efforts  of the  1979 study, characterizes
the air quality  and meteorology during the monitoring  period, investigates periods of  high
ozone  and NO-, with modeling  potential,  and  provides recommendations  for  model
verification.

-------
                        2.  FIELD MEASUREMENT PROGRAM

      Data were  collected from  July 2, 1979 through September 18,  1979.   These data
included  continuous  surface  air quality  measurements  at  fifteen  locations,  surface
meteorological measurements at seventeen locations, aircraft measurement of air quality
and meteorological parameters in the vertical, upper air meteorology at three locations,
and hydrocarbon species measurements.

2.1   OBJECTIVES OF THE FIELD PROGRAM

      The primary objective of  the field program was to obtain air quality and meteoro-
logical  data of sufficient extent  and quality to apply the Systems Application, Incor-
porated (SAI), photochemical model to the Philadelphia area.  A secondary objective was
to collect data which would increase  the  EPA's knowledge  of the photochemical and
meteorological processes which lead to peak ozone and NO- concentrations in  a portion of
the Northeast Corridor.  Philadelphia was selected because it is a large city within the
Northeast Corridor capable of  generating high  ozone levels  while remaining subject to
significant inter-urban ozone/precursor transport.

2.2   PROGRAM  DESIGN

      To achieve  the  objectives of the  field program,  measurements were taken in the
vicinity of Philadelphia, Pennsylvania, to assess:

          the levels of ozone and precursors transported into  the study area;
          the levels of precursors produced from emissions within the study area;
          the highest levels of  NO- and ozone downwind of the Philadelphia urban area.

     The field monitoring  program had four components:

     1.    Surface measurements of air  quality and meteorology at fixed  sites;
     2.    Aircraft measurements of air quality and meteorology;

-------
     3.    Balloon measurements of upper air meteorology;
     '4.    Hydrocarbon species measurements.

     Table 1 shows the periods during which  each  of these field components contributed
data to the study and the organizations responsible for each component's operation.

2.3  SURFACE iMEASUREMENTS AT FIXED  SITES

     The fixed surface station component consisted of:

     1.    Five stations set up specifically for the study (special study sites);
     2.    Eleven existing air quality meteorological monitoring stations;
     3.    Supplemental meteorological sites;
     4.    Special PAN/nitric acid/NO  monitoring.
                                    J\

2.3.1       Special Study Sites

     Figure 1 shows the  locations of the five stations set up specifically for the study.
Table 2 lists the parameters monitored at each  of these AeroVironment-operated sites.
Table 3 lists the instruments used, their precision, lower detection limit, and lower limit
of data  validity.  All  monitoring equipment  at  these  sites,  with the exception of the
acoustic sounder used to  measure  mixing height  at Site 1, was  provided  by the EPA's
Environmental  Monitoring and Support  Laboratory in  Research  Triangle  Park,  North
Carolina.

     Sites 1  and 2 were upwind of  Philadelphia under prevailing flow which was from the
south-southwest to west.  Site 1 was near  Summit Bridge, Delaware, approximately 60 km
southwest  of the Philadelphia  urban core.  Site  2  was  near Downingtown,  Pennsylvania,
approximately 50 km west of Philadelphia. Site 2 was at the same position  as one  of the
sites in  the  1978 Philadelphia Oxidant  Study,  thus providing some year-to-year data
comparison.

     Sites 1 and 2 were  to monitor ozone and precursors transported into the city, both
along the surface and aloft, when  the prevailing flow was southwesterly.  Surface stations

-------
Table 1.    RESPONSIBLE ORGANIZATIONS AND PERIODS OF OPERATION OF FIELD MONITORING
          PROGRAM COMPONENTS.
Program
Component
Dates
Responsible Organization
Surface Measurements of Air Quality and Meteorological Data at Fixed Sites
Special study sites
Existing sites

Supplemental
meteorological sites
Special PAN/nitric acid
NO monitoring
A
July 2-September 18
Duly 2-September 18

July 2-September 18
July 15-August 22
AeroVironment Inc. /U.S. Environmental Protection
Agency
Philadelphia Department of Health (Air Management
Services)
Pennsylvania Department of Environmental Resources
New Jersey Department of Environmental Protection
National Weather Service
Federal Aviation Administration
U.S. Air Force, U.S. Navy
Washington State University
Battelle Columbus Laboratories
Aircraft Measurements of Air Quality and Meteorological Data
Bell UH-1H helicopter
July 18-August 16
Northrop Services/U.S. Environmental Protection
Agency, Environmental Monitoring Systems
Laboratory, Las Vegas
Balloon Measurements of Upper Air Meteorological Data
Pibal
Radiosonde
3uly 16-August 16
July 18-August 26
U.S. Environmental Protection Agency, Environmental
Monitoring Systems Laboratory, Las Vegas
Beukers Laboratories, Inc.
Hydrocarbon Species Measurements
July 15-August 15
Washington State University

-------
              ,J _>—-"r— ' -«^I-.f _ _.lr I-I  .f


              PENNSYLVANIA
              _  "IT " 7V^.,T">-V V^ N  5..LO-B*
FIGURE 1.  Locations of monitoring sites set up specifically for the study.

-------
Table 2.   METEOROLOGICAL AND AIR QUALITY PARAMETERS MONITORED AT THE FIVE
         SITES SET UP SPECIFICALLY FOR THE STUDY.
Component
Ozone
Nitrogen Dioxide
Oxides of Nitrogen
Nitric Oxide
Non-Methane Hydrocarbon
Total Hydrocarbon
Methane
Carbon Monoxide
Wind Speed
Wind Direction
Temperature
Solar Radiation
Mixing Height
Site
1
Summit
Bridge, DE
X
X
X
X
X
X
X

X
X
X
X
X
2
Downing-
town, PA
X
X
X
X
X
X
X

X
X



3
Lumber-
ton, NJ
X
X
X
X
X
X
X
X
X
X
X
X

4
Robbins-
viile, NJ
X
X
X
X




X
X
X


5
Van Hise-
ville, N3
X
X
X
X




X
X




-------
Table 3.   MONITORING INSTRUMENTS WITH THEIR LOWER DETECTION LIMITS, DATA
         PRECISION, AND LOWER LIMITS OF DATA VALIDITY.
Parameter
Instrument
Instrument Lower
Detection
Limit
Data
Precision
Lower Limit
of Data
Validity
Air Quality
°3
NO
N02
N0x
THC
CH^
NMHC
CO
Bendix 8002
Bendix 8101
Bendix 8101
Bendix 8101
Beckman 6800
Beckman 6800
Beckman 6800
Bendix 8501
0.001 ppm
0.005 ppm
0.005 ppm
0.005 ppm
0.100 ppm
0.100 ppm
0.200 ppm
0.500 ppm
0.005 ppm
0.010 ppm
0.010 ppm
0.010 ppm
0.100 ppm
0.100 ppm
0.200 ppm
0.500 ppm
0.005 ppm
0.010 ppm
0.010 ppm
0.010 pprn
0.100 ppm
0.100 ppm
0 . 200 pprn
0.500 pprn
Meteorology
Wind Speed
Wind Direction
Temperature
Solar Radiation
Mixing Height
Climatronics
Climatronics
Climatronics
Eppley (Model 2)
AeroVironment 300C
0.2 m/s
N/A
N/A
0.05
cal/cm /min
30 m
0.2 m/s
5°
0.5° C
0.05
2
cal/cm /min
10 m
0.2 m/s
N/A
N/A
0.05
2
cal/cm /min
30 rn

-------
can  provide  an  indication of transport  aloft, since ozone and  precursors trapped  aloft
overnight are brought to the ground through turbulent vertical mixing after the breakup of
the nocturnal radiation inversion.   Placing the  acoustic  sounder  at Site 1 provided the
time of inversion breakup.

      Sites 3, 4, and 5 were  downwind of Philadelphia under  prevailing flow.  Site 3 was
near  Lumberton, New Jersey, approximately  35 km  east-northeast of  the  Philadelphia
urban core.  Site ^ was near Robbinsville, New Jersey, approximately 60 km  northeast of
Philadelphia.  Site 5 was near Van Hiseville, New  Jersey,  approximately 75 km east-
northeast of Philadelphia. These sites were established to monitor  peak ozone and NG>2
concentrations resulting  from ozone  and precursors transported  into Philadelphia and
precursors  emitted  in  the  urban  core  under southwesterly flow.   Figure 2 shows an
exterior view of  the monitoring station  at  Site 5, which was  identical to the monitoring
stations at Sites 1, 2, 3, and 4.

2.3.2      Existing Air Quality/Meteorology Monitoring Stations

      Figure 3 shows the locations of the eleven existing air quality/meteorology monitor-
ing  stations which  contributed  data to  the study, and  Table k  lists the  parameters
monitored  at each.   The state and local agencies operating these sites  are named in
Table 5.   The  operating agencies themselves  supplied  most  of  the  air  quality  and
meteorological instruments.  The EPA's Environmental Monitoring and Support Laboratory
in Research Triangle  Park,  North  Carolina,  supplied some  additional  instrumentation.
Information on the accuracy, precision, and ranges of the instruments  can  be obtained
from the agencies themselves.

      The EPA selected  the stations  to be included in the  study  based primarily  upon
location and instrumentation available.  Sites 6,  11,  and 15 were included as  upwind
stations for south to southwest flow and  monitored transport  of ozone and precursors into
the study area under these flows.  Sites 9, 12,  13, and 14 were urban sites for monitoring
both urban  generated precursors and urban ozone  peaks  or,  in the case of Site 9, urban
wind flow.   Site  10 provided a near-urban monitor.  Sites 7, 8, and  16 served as downwind
ozone  and/or NO2  stations under southwesterly flow.  Table 6 gives the  distances and
bearings from the urban core for these stations.

-------
FIGURE 2.  Exterior view of the monitoring station at Site

-------
                 PENNSYLVANIA
FIGURE 3.     Existing air quaiity/meteoroiogicai monitoring stations used in the study.
                                 10

-------
Table 1.   METEOROLOGY AND AIR QUALITY PARAMETERS MONITORED AT EXISTING STATIONS.
Component
Ozone
Nitrogen Dioxide
Oxides of Nitrogen
Nitric Oxide
Non-Methane
Hydrocarbon
Carbon Monoxide
Wind Speed
Wind Direction
Temperature
Site
6
X
X


X


X
X
X
7
X
X


X


X
X
X
8
X
X

X






9







X
X

10
X
X

X


X



11
X





X



12
X
X

X
X

X



13
X
X

X
X

X


X
14
X
X

X
X

X



15
X
X





X
X

16
X










-------
Table 5.   SITE LOCATIONS AND OPERATING AGENCIES OF THE EXISTING
          AIR QUALITY/METEOROLOGY STATIONS.
  Site
Number
                   Location
        Operating
         Agency
      Chester, Pennsylvania
      Bristol, Pennsylvania
      North Philadelphia Airport
      Philadelphia, Pennsylvania
      Allegheny
      Philadelphia, Pennsylvania
10    Camden, New Jersey
11    Ancora, New Jersey
12    South Broad and Spruce Streets
      Philadelphia, Pennsylvania
13    Franklin Institute
      Philadelphia, Pennsylvania
      Air Management Services Laboratory
      Philadelphia, Pennsylvania
      Claymont, Delaware
16    New Brunswick, New Jersey
Pennsylvania Department of
Environmental Resources

Pennsylvania Department of
Environmental Resources

Air Management Services
(Philadelphia Department
of Health)

Air Management Services
(Philadelphia Department
of Health)

New Jersey Department of
Environmental Protection

New Jersey Department of
Environmental Protection

Air Management Services
(Philadelphia Department
of Health)

Air Management Services
(Philadelphia Department
of Health)

Air Management Services
(Philadelphia Department
of Health)

Delaware Department of
Natural Resources and
Environmental Control

New Jersey Department of
Environmental Protection
                                 12

-------
Table 6.   LOCATIONS OF EXISTING AIR QUALITY/METEOROLOGY
         MONITORING SITES WITH RESPECT TO PHILADELPHIA
         URBAN CORE.
Site
Number
6
7
8
9
10
11
12
13
14
15
16
Location
Chester, Pennsylvania
Bristol, Pennsylvania
North Philadelphia Airport
Philadelphia, Pennsylvania
Allegheny, Pennsylvania
Camden, New Jersey
Ancora, New Jersey
South Broad and Spruce Streets,
Philadelphia, Pennsylvania
Franklin Institute,
Philadelphia, Pennsylvania
Air Management Services Lab,
Philadelphia, Pennsylvania
Claymont, Delaware
New Brunswick, New Jersey
Position with Respect to
Philadelphia Urban Core
Distance
(km)
20.3
30.5
16.0
6.2
f.8
38.9
CBD*
2.5
9.0
28.1
85.0
Bearing
Direction
WSW
NE
NNE
NE
E
SE
CBD*
NNW
N
WSW
NE
*CBD = central business district
                               13

-------
 2.3.3      Supplemental Meteorological Sites

      Data  were  also obtained from eight existing meteorological  monitoring stations
 operated by the National Weather Service, the Federal Aviation Administration, the U.S.
 Air Force,  and the  U.S. Navy.  Figure 4 shows the locations of these stations.  All sites
 recorded one-minute averages of wind  speed, wind  direction,  and temperature once an
 hour.  These data were included to better define the study area wind field.  Table  7 gives
 the distance and bearing from the urban  core for these stations.

 2.3.4      Special PAN/Nitric Acid/NOx Monitoring

      Special  PAN,  nitric acid, and  oxides  of  nitrogen  monitoring was  conducted  by
 Washington State University and  Batteile Columbus Laboratories  during July 15 through
 August 22,  1979, at  Downingtown, PA (Site 2), and Van Hiseville, NJ (Site 5). Washington
 State  University monitored NO, NO  and PAN at Downingtown, and Batteile monitored
                                   X
 nitric acid at  Downingtown,  and  NO,  NO  ,  PAN,  and nitric acid  at  Van  Hiseville.
 Information  regarding  measurement  methods  can  be obtained  directly  from  these
 organizations.

 2.4   AIRCRAFT MEASUREMENTS  OF AIR QUALITY AND METEOROLOGY

      The primary objective of the aircraft  measurements was  to obtain vertical profiles
 and cross-sections of ozone and precursors  to quantify pollutant transport aloft. That is,
 from these  profiles  and cross-sections, we  could find the levels of ozone and precursors
 transported into  the  city in layers aloft  over  various  periods during  days of high
 photochemical  activity.   Such  measurements would also  provide boundary  and   initial
 conditions for verifying of the SAI model.

     Aircraft measurements were taken by the EPA's Environmental Monitoring Systems
 Laboratory in Las Vegas,  and  Northrop Services,  Inc., between July 18 and  August 16,
 1979, using  a Bell UH-1H  helicopter (Figure 5).  Parameters measured  included O.,, NO,
 NO^, visible light scattering,  temperature, and dew point versus location  and time.   In
addition, hydrocarbon  grab samples were taken for species analysis.   The hydrocarbon
sampling  will  be  discussed in  more  detail  later  in  this  chapter.   Table 8   lists the

-------
                 U > i • w-mi '-w^r.'ITit*^, ,_'Lim- \ ;*
                 PENNSYLVANIA
FIGURE 4.  Locations of supplemental meteorological monitoring sites used in the study.
                                 15

-------
Table 7.   LOCATIONS OF SUPPLEMENTAL METEOROLOGICAL MONITORING
         SITES WITH RESPECT TO PHILADELPHIA URBAN CORE.
Site
Number
8
17
13
19
20
21
22
23
Location
North Philadelphia Airport, PA
Philadelphia International Airport, PA
McGuire Air Force Base, NJ
Willow Grove Naval Air Station, PA
Lakehurst Naval Air Station, NJ
Trenton-Mercer County Airport, NJ
Millville Airport, NJ
Greater Wilmington Airport, DE
Position with Respect to
Philadelphia Urban Core
Distance
(km)
16.0
9.1
49.2
28.6
75.8
48.2
64.3
47.6
Bearing
Direction
NNE
SW
E
N
E
NE
S
SW
                           16

-------
FIGURE 5.  EPA-Las Vegas instrumented UH-1H helicopter.
                       17

-------
Table 8.   PARAMETERS MEASURED BY HELICOPTER AND INSTRUMENTS
          USED.
      Parameter
                                          Instrument
Hydrocarbons


Ozone

NO/NO 2/NO

Light Scattering


Temperature/Dew Point

Static Pressure (altitude)

Position
                           Grab samples using evacuated stainless steel
                           canisters

                           REM 612 B chemiluminescent analyzer

                           Monitor Labs 8440 chemiiuminescent analyzer

                           Meteorology Research Inc. Model  1550
                           nephelometer

                           EG&G 137-C1

                           National Semiconductor LX3702A

                           Two Collins DME  W
                                18

-------
instruments  used.  The  accuracy,  precision, and  ranges of these instruments have been
previously documented (Hancock, 1980).

      The aircraft  (helicopter) measurement program consisted  of  10 days of intensive
vertical soundings and transects upwind, over, and  downwind of the study city.  The choice
of these 10  days was based on forecasts predicting meteorological conditions favorable for
significant photochemical activity.

      On  each measurement day, three  flights were generally  made:   one in the early
morning (takeoff at 0500 EST), one at mid-morning (takeoff at 0900 EST),  and one in the
early afternoon (takeoff  at 1300 EST).

      Each flight consisted of vertical soundings  over three surface monitoring  stations
from as close to the ground surface as possible up to 2 km above ground level (AGL), and a
horizontal traverse between the stations at the elevation where  maximum ozone levels
were observed. Figure 6 shows the locations of these stations. Spirals were performed at
two upwind and one downwind sites during the  early morning; one upwind, one downtown,
and one downwind site during the late morning;  and one downtown and two downwind sites
during the  afternoon.   The  specific sites over  which  sampling was  performed varied
depending upon the wind direction.   The  exact flight  paths taken on each sampling day
have been documented elsewhere (Hancock, 1980).

      The mission  of  each  day was  to  obtain information  about ozone  and precursor
concentrations aloft being transported into the study  area, concentrations present aloft
near  the  central business  district, and  peak ozone concentrations  downwind.  Table 9
summarizes the aircraft sampling missions actually flown.

2.5   BALLOON MEASUREMENTS OF UPPER AIR METEOROLOGY

2.5.1       Radiosonde Releases

      Beukers Laboratories, Inc.  was responsible  for  releasing five radiosondes  per day
from July 18 through August 26, 1979, from 315 East Chelten Avenue, Philadelphia (about
10 miles  west-southwest of  North Philadelphia Airport).  These five  releases provided
                                         19

-------
FIGURE 6.  Locations of helicopter sounding and sampling sites.
                           20

-------
Table 9.   SUMMARY OF AIRCRAFT SAMPLING MISSIONS ACTUALLY
          FLOWN.
Date
duly 23
duly 25
August 4
August 5
August 6
August 7
August 10
August 14
August 15
August 16
Flight Time
(EST)
1044-1243
1*32-1658
0532-0737
0854-1052
1314-1530
0919-1124
1308-1536
0738-0938
1305-1530
0517-0741
0907-1057
1316-1525
0517-0723
0844-1050
1301-1507
0519-0724
0832-1045
1258-1527
0522-0723
0851-1043
1300-1530
0519-0732
0848-1126
1317-1515
0510-0715
0835-0927
Pattern*
Southwesterly
Southwesterly
Westerly
Westerly
Westerly
Westerly
Westerly
Westerly
Westerly
Northwesterly
Northwesterly
Northwesterly
Northwesterly
Northwesterly
Northwesterly
Southwesterly
Northwesterly
Northwesterly
Southwesterly
Southwesterly
Southwesterly
Northwesterly
Northwesterly
Northwesterly
Northwesterly
Northwesterly
*The exact sounding locations for each flight pattern varied.
                                 21

-------
vertical profiles of  wind speed, wind direction, temperature, and dew point for the hours
0500, 0700, 0900, 1200,  and  1500 EST.  The  0700  and 1200 EST releases provided data at
"mandatory" levels (surface, 1000 mb, 850 mb, 700 mb, and 500 mb), while the 0500, 0900,
and  1500 EST  releases provided data at other significant levels (based upon  temperature
lapse rate).

2.5.2      Pibal Releases

      The EPA  Environmental Monitoring  Systems  Laboratory  in Las  Vegas,  Nevada
(EMSL-Las Vegas), was responsible for  releasing and  tracking  approximately 12  pilot
balloons per  weekday from  July 16 through August  16, 1979, from Robbinsville Airport,
iNew  Jersey (Site 4),  and from Greater Wilmington  Airport, Delaware (Site 23).  These
releases provided  vertical  profiles of  wind speed  and  direction for 0500 EST  through
1600 EST.

2.6   HYDROCARBON SPECIES MEASUREMENTS

      This component of the program,  which was the responsibility of Washington State
University, provided information on the hydrocarbon  species as they were transported into
and emitted within the urban area.

      There  were three  parts  to  the  sampling program:   (1) integrated samples were
collected at ground level; (2) grab samples  were  collected at ground level; and (3)  grab
samples were  collected  aloft.  Samples  were collected  between July 15  and August 15,
1979.

      Figure 7 shows the ground level  sampling locations.  One-hour integrated samples
were taken at  the Franklin Institute (Site  13) and at a site near  an industrial area in south
Philadelphia along the Delaware River.  Eighty-five of these samples were collected from
0600 to 0700  EST and 0800 to 0900 EST.  On selected days, a surface grab sample was
collected at  each  of two roving  sites, the location of  which varied from  day  to  day.
Fourteen of this type of  samples were collected.  Four grab samples  were collected by the
EPA-EMSL helicopter during each flight day for a total  of  forty.  Nine samples were
collected for intercomparison purposes. A grand total of 148 samples were analyzed.
                                         22

-------
FI = Franklin Institute
15 = Industrial Site
Rl-6  = Roving Sites
               FIGURE 7.   Locations of ground-levei sampling sites.
                                         23

-------
     All samples were analyzed within 24 hours of collection.  The gas  chromatography
procedures used  to analyze these  samples  are documented  elsewhere  (Westberg  and
Sweany,  1980).

2.7  QUALITY ASSURANCE

     This section summarizes  the procedures used to assure data quality for AeroViron-
ment's portion of the field program,  and specifically describes the calibration methods  and
frequency, station check methods and frequency, and audit and interlaboratory tests.

2.7.1      Surface Monitoring  Program

     The equipment listed  in Table 3, except  for the  acoustic  radar,  was tested  and
furnished to AeroVironment  by the EPA.   Both EPA and AV extensively bench-tested all
equipment to assure proper operation.   To meet the  manufacturers' specifications,  the
equipment and recording  devices were  installed  in environmentally controlled shelters at
each of  the five  stations. Additional air conditioners were installed at all sites to insure
that stations were operating  within the  specified temperature range.

o    Calibration of Air Quality Instruments

     AV personnel  performed multipoint calibrations of all air  quality  instruments  at
each station  at least  monthly throughout the  measurement program.   In addition, an
instrument was calibrated whenever any of  the following conditions occurred:  (a)  the
control limit for the span check as specified in the Station Check List Log was exceeded;
(b) after repair of a malfunctioning analyzer; (c) after replacement  of major components
of an  analyzer;  and (d)  when  the  audit results  exceeded  the limits  established  (see
discussion  of  audits, Section 2.7.2).  Zero  plus a minimum of four calibration points
equally spaced over the analyzer  range were used to  generate a calibration  curve.   A
"master"  calibrator was used  for  all  of the  multipoint  calibrations  to  ensure data
comparability. The purpose of  the multipoint calibrations was twofold:  (a)  to check  the
instrument linearity; and (b) to assign  the values of  the on-site calibration sources used
for daily station checks so that  they  would be  traceable.

-------
     Dynamic calibration of the NO-NC^-NO   and  (X analyzers was  performed using a
Monitor Labs  calibrator.   For  NO  and NO  channels, a National  Bureau of  Standards
Standard Reference Material (NBS SRM) gas cylinder containing approximately 50 ppm
NO in N2 was diluted to ambient levels for the calibration. For the NO? channel, AV  used
gas  phase  titration (GPT)  of  NO and  O3  prescribed  by  the EPA  (40 CFR, Part 50,
Appendix F).

     For  the  O^  analyzers, the calibrator's O, output was determined  on  site by  the
ultraviolet (UV) photometry method (M Federal Register 8221-8233,  February  8, 1978).  A
modified  Dasibi  Model  1003-AH ozone monitor  served as a transfer standard.   Gas
cylinders containing approximately 100 ppm NO in N9 were used for the NO/NO  spans.
                                                L.                       A.

     To calibrate  the THC/ChL instruments,  two compressed gas  cylinders containing
zero ppm and approximately 8 ppm of  CI-L in an ultrapure air were used. Two  compressed
gas cylinders  containing zero ppm  and  approximately  50 ppm of  CO in an ultrapure air
with about 350 ppm  CO.,  were  used to calibrate  the  CO analyzer.   In  both  cases,
multipoint calibrations  were performed  using an AeroVironment Model BB-100 dilution
system. All gas cylinders were  NBS-traceable.

     The  NO  flow rate  and dilution  system  of  the  calibrator and the flow rate for  the
AeroVironment Model BB-100 dilution system were calibrated with a  bubble flowmeter at
the beginning,  mid-way through, and at the end of the measurement program.

o    Calibration of Meteorological  Equipment

     Calibration checks of  the meteorological equipment were done on site.  The wind
cups were  turned  to make sure that  the bearings were  in normal condition.   The wind
vanes were aligned to the  true north, true south, and true  north again to calibrate  the
recorder output.  The local  magnetic  north declination of +10° was  accounted  for.  The
temperature output from  each station  was checked  using  an NBS-traceable mercury
thermometer.
                                         25

-------
o    Station Checks

     AV personnel checked stations daily using the Station Check List Log prepared for
the project. Their primary concerns were to ensure that the analyzers and the supporting
equipment  were in proper working order.

     Zero  and span were checked for ail gaseous analyzers during the station check. For
the  NO/NO   and O,  instruments,  this  was  done  by  using  the on-site  Monitor Labs
Model  8500 calibrator.   For  the   hydrocarbon analyzer, a  zero  cylinder  containing
hydrocarbon-free  air and  a span  cylinder  containing  approximately 2 ppm of CH^  in
uitrapure air  were used.  A similar procedure for  CO  was used, with the span cylinder
containing  approximately 50 ppm CO in uitrapure air.

     The zero and span values for each instrument were used to determine control  limits.
The control limits were a useful tool to detect early instrument problems.

     Hydrocarbon  analyzers  were  precision-checked   three   times   a  week   using
EPA-furnished gas cylinders containing known  concentrations of hydrocarbon.  The  results
of these checks were recorded and forwarded to the EPA.

     The station check personnel reported any instrument problems  to the Field Manager
on  the same  day so  that  corrective action  could  be  initiated  as  soon  as  possible.   In
addition, the Field  Manager provided a weekly instrument status report to  AV's Program
Manager in Pasadena, California, to update him  on  the field program status and to  enable
him to provide direction when needed.

2.7.2      Audits

     Quality assurance verification of the O^, NO-NO2-NOx, THC-CH^,  and CO systems
used for calibrating was done  at Research Triangle Park, North Carolina, at the start  of
the measurement program. The agreements for all systems were within 5%.

     During the monitoring phase, the project was further audited  by Research Triangle
Institute (RTI) under contract to the EPA.  Each of the five surface monitoring stations
were audited twice. RTI also audited other study participants.
                                          26

-------
     The acceptable criteria for ail audits were set at + \5% agreement for the slope of
regression analysis of the audit data.  Any audit results indicating that these limits were
exceeded resulted in prompt  recalibration or repair/recalibration of the instrument(s) in
question.

2.3  DATA REDUCTION AND COMPILATION PROCEDURES

     The continuous air quality and meteorological  data AV collected on chart recorders
were reduced at  AV's main office in  Pasadena, CA.  The data were digitized, tabulated,
and spot-checked for obvious instrument malfunctions and unusual data points.  Site logs,
field personnel, and project scientists were  consulted  concerning the validity of  unusual
data points.   Calibration  factors  were applied  to  the data,  as appropriate,  using  the
monthly multi-point calibration data  supplemented by the daily  zero  and span check
values.

     The air  quality and  meteorological  data collected  by federal,  state,  and  local
agencies were  provided to AV by EPA in tabulated  form.   These data were merged with
data collected  by AV to form a data set with consistent units and time reference (Eastern
Standard Time -- EST).  AV prepared data volumes containing hourly averages of all these
data for all the sites.

     AV also obtained hard copy printouts of supplemental  meteorological data collected
by the  National  Weather Service, Federal Aviation Administration,  U.S. Navy, and U.S.
Air Force. AV entered these data into the computer, error  checked them, and produced a
magnetic tape.

     Special PAN/nitric acid/NO  data collected by Washington University and Battelle
Columbus Labs were provided to AV  by EPA in hard  copy and magnetic  tape format.
These  data were entered in a form  consistent with  the  rest  of  the  data base,  error
checked, and a magnetic tape produced.

     In  addition, EPA  provided AV  with data recorded by the EPA EMSL-Las Vegas
aircraft.  These  data were  sent in  hard copy and  magnetic  tape format  along with hard
copies  of pollutant profiles measured during soundings.  These data were used during  the
analysis.
                                         27

-------
     Hard  copy printouts  of  pibal data  from Trenton, New  Jersey, and  Wilmington,
Delaware, collected by EPA EMSL-Las Vegas were  provided  by EPA for  use  in  the
analysis.  Upper air radiosonde data collected by Beukers Labs were provided to AV by
EPA on magnetic tape for use in the analysis.

     Finally, hydrocarbon  species data collected by  Washington  State University were
provided to AV by EPA in hard copy format for  use in the analysis.
                                        28

-------
       3. OZONE AND PRECURSOR CONCENTRATIONS IN THE STUDY AREA

3.1   DATA SUMMARIES

     This section  summarizes the data collected during the 1979 field program. It also
briefly compares data collected during the study with  data collected during recent  years
and, when  possible,  with  long-term means  to  determine how representative  the  study
period was with respect to characteristic ozone concentrations.

3.1.1       Representativeness of the 1979 Summer Months

     Table 10 compares some meteorological parameters recorded in each of the summer
months of 1979 with the corresponding long-term means at the Philadelphia International
Airport.  The average maximum  monthly temperatures were lower than the long-term
mean during July,  August, and September of  1979, whereas the average minimum monthly
temperature was higher than the long-term mean during each of those months. Moreover,
the mean monthly temperature (which  is the  average of  the  minimum and  maximum
temperatures) indicates that July was slightly cooler and August and  September slightly
warmer  than  the  long-term  mean.  Another  useful statistic  for  relating long-term
meteorological data to ozone  climatology is  the number of days  in each month  with
temperatures greater than or equal to 90° F- From Table 10, we can see that even though
the mean temperature for July 1979 was lower than the long-term mean, July 1979  had a
normal  number  of  days  during  which  the  temperature  reached  or exceeded  90 F.
August 1979  had  two days more and  September  1979  had two  days less  than  the
corresponding long-term  means.   Also, July  had less precipitation and August  and
September  had more  precipitation than  normal, and each of the three  months had less
sunshine  than normal.

     Table 11 summarizes the number  of high ozone days and the  number of episode
periods (three or more consecutive high ozone days) from July 2 through September  18 for
the study year and the previous four years  at  Bristol, Pennsylvania.  At this  site, peak
ozone concentrations were generally  lower in  the study  year than  in the  four previous
                                         29

-------
Table 10.  COMPARISON OF METEOROLOGICAL PARAMETERS RECORDED DURING EACH SUMMER MONTH
         OF 1979 WITH CORRESPONDING LONG-TERM MEANS.
Meteorological
Parameter
Average Maximum
Monthly Temperature
Average Minimum
Monthly Temperature
Mean Monthly
Temperature ( C)
Number of Days with
Maximum Temperature
(>_ 90° F)
Total Precipitation (cm)
Average Station
Pressure (mb)
Average Wind Speed
(rn/sec)
Percent of Possible
Sunshine
Number of
Years in
Long-Term
Average
30
30
30
17
30
4
36
34

July
1979
29.3
19.7
24.6
7
10.0
1,015.9
3.3
46

Long-Term
30.4
19.3
24.9
7
10.4
1,014.4
3.6
63

August
1979
28.8
19.5
24.2
7
15.1
1,015.9
3.6
52

Long-Term
29.3
18.2
23.8
5
10.4
1,016.9
3.5
63

September
1979
24.8
15.7
20.3
0
12.4
1,017.3
3.9
52

Long-Term
25.8
14.3
20.1
2
7.7
1,017.1
3.7
60


-------
 Table 11.   SUMMARY OF HIGH OZONE DAYS (greater than 0.100 ppm and 0.120 ppm) FROM
           JULY 2 THROUGH SEPTEMBER IS,  1979, FOR THE STUDY YEAR AND THE PREVIOUS
           FOUR YEARS AT BRISTOL, PENNSYLVANIA (the national ambient air quality
           standard for ozone is 0.120 ppm).
Year
1979
1978
1977
1976
1975
Number of Days
with Peak O
(ppm)
>0.100
14
21
23
1*
24
>0.120
4
13
11
6
10
Number
of
Episode*
Cases
0(0)
3(5)
3(4)
0(0)
1(4)
Total
Number of
Observations
(maximum
possible 79)
71
63
68
45
67
Percent of Days
with O3
(ppm)
>0.100
19.7
33.3
33.8
31.1
35.8
>0.120
5.6
20.6
16.2
13.3
14.9
*An episode is defined as three or more consecutive days with O^ measurements greater than
 0.100 ppm.  Numbers in parentheses indicate the number of days in the longest episode.

-------
years.  For  the study period,  less than 20% of the daily maximum concentrations were
greater than 0.100 ppm, whereas during each of the previous four years more than 30% of
the days exceeded this  value.   A similar  reduction  in  the  number  of days exceeding
0.120 ppm was noted.

      While  one might infer from the  above discussion that maximum ozone levels were
lower in 1979, conditions at Bristol, Pennsylvania may  not be indicative of  the  entire
study area.   A field program similar to  the  one for this study was performed  in  the
summer of 1978  to evaluate measurement  approaches for studying ozone and precursor
transport into an  urban area (Chan et al.,  1979).  Four of the sites contributing data to
that study were also used in this study.  Table 12 summarizes data collected at these sites
during the two summers. From this table, we can see that ozone levels were generally not
substantially different between the two years.

3.1.2      Average and Maximum Levels of Pollutants

      Table  13 gives peak and  average daily maximum ozone  concentrations as well  as
peak and average morning (0500 through 0800  EST)  precursor concentrations observed at
all  monitoring locations during the study period.  Site locations were  shown in Figures 3
and 14.  Ozone levels were  similarly high  even at some stations located predominantly
upwind. Precursor concentrations were generally quite low except at downtown locations.

      Table  14 is  a frequency distribution of ozone levels observed at various sites in and
around Philadelphia.  The greatest frequency  of  concentrations  in excess  of  0.12 ppm
occurred at  outlying stations.  Downingtown, PA (Site  2), to the west of Philadelphia, and
Lumberton,  NO (Site 3), to the  east, each had relatively  high frequencies of  observations
(>.85%) greater than 0.12 ppm  compared to downtown  Philadelphia (<.15%).  This pattern
is also true  for frequency of ozone values  greater than 0.08 ppm.  Sites 1 through 5 had
concentrations greater than 0.08 ppm for more than 3.5% of the observations, whereas in
downtown Philadelphia (Site 12) the frequency  is less than 1.0%.  The  high frequency of
low ozone  concentrations  at  Site 12  was  expected   due   to  the  higher levels  of
ozone-scavenging NO in urban areas.  Surprisingly, the frequency of high ozone levels at
outlying sites is somewhat independent of  the direction from  Philadelphia.  Sites  to the
west or southwest have roughly the same frequency of concentration  levels  in excess of
                                         32

-------
Table 12.  SUMMARY OF OZONE DATA COLLECTED DURING THE SUMMER OF 1978 AND 1979 AT THE SAME
         LOCATION.
Location
Downingtown,
Pennsylvania
Ancora ,
New Jersey
South Broad and
Spruce Streets
(Philadelphia)
Franklin Institute
(Philadelphia)
Site
Number
2
11
12

; 13
Peak
1978
0.1*2
0.142
0.180

0.250
1979
0.146
0.147
0.140

0.150
Number of Days
with Peak O, (ppm)
>0.100
1978
11
15
7

6
1979
11
15
4

4
>0.120
1978
3
4
2

5
1979
6
6
1

1
Daytime Monthly Averages
Jul
1978
—
0.049
0.028

0.036
Jul
1979
0.057
0.051
0.030

0.031
Aug
1978
0.055
0.052
0.024

0.030
Aug
1979
0.044
0.047
0.020

0.029
Sep
1978
0.041
0.034
0.018

0.024
Sep
1979
0.042
0.039
0.013

0.020

-------
TABLE 13.   PEAK AND AVERAGE OZONE PRECURSOR CONCENTRATIONS (in ppm) FOR THE STUDY
           PERIOD AT ALL AIR QUALITY MONITORING SITES.
Site
No.
1
2
3
b
5
6
7
8
10
11
12
13
11
15
16
Location
Summit Bridge, DE
Downingtown, PA
Lurnbei ton, N3
Robbinsville, N3
Van Iliseville, N3
Chester, PA
Bristol, PA
North Philadelphia
Airport, PA
Camden, N3
Ancora, NJ
South broad & Spruce Sts.
Philadelphia, PA
Franklin Institute,
Philadelphia, PA
Air Management Services
Lab, Philadelphia, PA
Claymont, DE
New Brunswick, N3
Ozone
Maximum
Hourly
Average
0.137
0.157
O.I'|6
0.111
0.161
0.183
0.135
0.160
0.161
0.117
0.1*0
0.150
0.160
0.170
0.105
Mean Daily
Maximum
3ul
0.080
0.087
0.079
--
0.078
0.077
0.073
0.077
0.078
0.079
0.057
0.061
0.071
0.070
—
Aug
0.072
0.068
0.078
0.072
0.072
0.071
0.080
0.079
0.076
0.078
0.011
0.058
0.059
0.062
0.056
Sep
0.058
0.061
0.052
0.052
0.051
0.061
0.060
0.053
0.055
0.060
0.026
O.O'll
0.056
0.033
0.038
Oxides oj Nitrogen
Maximum
Hourly
Average
0.063
0.083
0.13I
0.207
0.082
-
-
-
-
—
-
—
—
-
—
Morning Average
(0500-0800 EST)
Jul
0.010
0.019
0.025
-
0.012
-
-
0.012
0.068
—
0.069
0.068
0.055
-
—
Aug
0.006
0.013
0.022
0.028
0.007
-
--
0.031
0.052
-
0.059
0.060
0.057
-
--
Sep
0.012
0.022
0.021
0.018
0.010
-
-
0.037
0.080
—
0.035
0.067
0.059
~
—
Non-Methane Hydrocarbons
Maximum
Hourly
Average
1-0
1.1
1.5
--
-
6.6
1.7
—
—
—
2.9
1.7
3.1
-
—
Morning Average
(0500-0800 EST)
3ul
0.1
O.I
0.1
-
-
1.1
0.5
—
—
—
0.6
0.8
0.6
-
—
Aug
0.2
O.I
0.1
-
-
1.0
—
-
—
—
0.3
0.3
0.1
-
-
Sep
1 0.3
0. 1
0.1
-
-
1.3
—
—
—
—
0.3
0.3
0.5
-
—

-------
Table 14.   OZONE FREQUENCY DISTRIBUTION FOR THE STUDY PERIOD AT SEVERAL SITES IN AND
          AROUND PHILADELPHIA.
Site Name
Summit, DE
Downingtown, PA
Lumberton, NJ
Robbinsville, NJ
Van Hiseville, N3
South Broad Street, PA
Total Hours Observed with Specified Ozone Concentration (ppm)
Site
1
2
3
4
5
12
.000-.039
990
994
1,194
807
1,111
1,430
.040-.079
541
689
480
202
487
224
.080-.099
97
81
70
38
69
11
.100-.120
15
27
30
9
19
9
>.120
8
17
16
4
8
2
Total
1,651
1,808
1,790
1,060
1,694
1,676

-------
0.08 ppm as do sites to  the  east  and  northeast.  This suggests the importance of ozone
transport within the Northeast Corridor.

3.1.3      Ozone Wind Roses

      Ozone wind roses for Sites  1 through 5, shown in Figures 8 through 12, give some
information concerning the  general source areas  of these high concentrations.   These
depict  simultaneously  the  frequency distribution  of  ozone  concentrations and wind
direction.   The wind  directions  associated with  high  ozone  concentrations  are easily
identified from these  figures.   Some  data are not included in  Figures 8 through 10 since
coincident wind direction and ozone data were not available. It  is not expected that these
data would significantly alter the wind direction frequency distribution.

      Site 1 is southwest of  the Philadelphia urban  area.   However, at  this site ozone
concentrations greater than 0.12 ppm generally occurred  with winds from  the west  or
west of southwest  and  concentrations greater  than 0.10  ppm generally  occurred with
winds from the northwest, west or southwest. This indicates that high ozone levels at this
site are associated with surface  transport from  source  regions (such  as Baltimore  or
Washington D.C.) other than  Philadelphia.

      Site 2 is west of the Philadelphia area.  During the field study, ozone concentrations
greater than 0.12 ppm at this site generally occurred with winds from the east, southeast
or south which suggests  that the  Philadelphia urban  plume affects this site.   This differs
from the case during the summer of 1978 when winds from  the  southeast were only
associated with ozone concentrations  lower than 0.08 ppm  (Chan et al.,  1979). However,
winds from the southeast were  much less  frequent  during the summer of 1978 than during
the summer of 1979 and,  as can be seen from Figure 9, the ozone concentrations with the
greatest frequency  associated with southeast winds in 1979 were less than 0.08 ppm.

      Site 3 is east  of the Philadelphia urban area.  From Figure 10 we can see that ozone
concentrations higher than 0.12 ppm at this site occur with winds ranging from the west
to south to east.   Only  winds  from the north or northwest are not associated with high
ozone levels at this site.
                                          36

-------
              (~n .030 - .099 ppm


                  .100 - .120 ppm


                  >.i2o
t
N
0%                     5%                      10%

           Relative Frequency of Occurrence




       FIGURE 8.  Ozone wind rose for Summit Bridge, DE (Site 1).
                                     37

-------
         ^.000 - .039 ppm
         B .040 - .079 ppm
         n~l .0X0 - .099 ppm
         I—1 .100 - .120 ppm
             >.120 ppm
I	i
                                                                       t
                                                                       N
0%                     5%
          Relative Frequency of Occurrence
                                              10%
            FIGURE 9.  Ozone wind rose for Downingtown, PA (Site 2).
                                   3S

-------
    HI.OOO - .039 ppm
    1=^.040 - .079 ppm
    FT! .080 - .099 ppm
    I—| .100 - .120 ppm
    mm >.i2o
Relative Frequency of Occurrence
   !    l     I     I     I     I
                                                                       t
                                                                       N
o%
                                   10%
                FIGURE 10.  Ozone wind rose for Lumberton, NJ (Site 3).
                                     39

-------
            - .039 ppm
    f=) .040 - .079 ppm
    FT! .OSO - .099 ppm
    I—I .100 - .120 ppm
    irnro >.i2o
0%
           Relative Frequency of Occurrence
5%
                                                  t
                                                  N
10%
             FIGURE 11:  Ozone wind rose for Robbinsville, NJ (Site

-------
0%
    BB-000 - .039 ppm
    gE3.0.120 ppm
          Relative Frequency of Occurrence
5%
                                                  t
                                                  N
10%
              FIGURE 12.  Ozone wind rose for Van Hiseville, NJ (Site 5).

-------
     Sites 4 and  5, northeast of Philadelphia, record high ozone levels with winds from
the southwest  indicating the effects of the Philadelphia urban plume.  Moreover, as  with
Site 3, Site 5 also measures high ozone levels with winds from the southeast.  This finding
is  unexpected  since flow from this direction is usually associated with high pressure off
the coast to  the  northeast.  Sites 4 and  5 also  show some ozone  levels  higher  than
0.10 ppm with winds from  the northwest, indicating some long-range transport.

3.2  HYDROCARBON SPECIES ANALYSIS

     The 148 samples analyzed between mid-July and mid-August by Washington State
University (WSU) fall into  several different categories.

      1.    Over half (85)  of the samples were taken in the morning either at the Franklin
           Institute site (Site 13) near center city  or at the "industrial site"  in south
           Philadelphia (see Figure 7 for sampling locations).  Of these,  half were taken
           at  each site  between 0500 and 0600 EST and the other half between 0700 and
           0800 EST.  This forms the most uniform data set in the study and includes  data
           from six of the ten case study days to be discussed in Chapter 4.

     2.    Five afternoon samples were taken at  the industrial site between 1300 and
           1400 EST.

     3.    A set  of 35 samples was taken by EPA/Las Vegas at various locations.  Fifteen
           of  these were taken on five of the case study days.

     4.    A  group of fourteen  roving samples was taken  at various points in the
           Philadelphia  area.

     5.    Another set of nine samples was taken for intercomparison purposes at various
           locations.

3.2.1       Data Summary

     Most of  the samples were taken  between duly 17 and August 15,  1979,  during a
consistent sampling program at the Franklin Institute (Site 13) and at  a site  in south

-------
Philadelphia near the  industrial  area along the  Delaware  River.   Franklin  Institute  is
within a few blocks of the City Hall  which is  in the center of Philadelphia.  Integrated
samples  were taken from 0500 to 0600 EST and from 0700  to 0800  EST at each site for
about 20 days during the heart of the study.  The distribution by types (%) along with the
total identified  (known) species and the sum, including unidentified peaks for this block of
data, are given in Tables 15 and 16.   The  averages for  each of the  two  sites and  two
collection times are also given.  The most striking feature of these averages is that there
is so little difference  between the four sets of data.  Paraffins make up about two-thirds
of  the  total,  aromatics  about one  fourth,  oiefins (omitting ethylene) about 6%  and
acetylene 2%.  This distribution is valid at both sites at both hours even though individual
samples  depart from this average significantly.

      It  is  also  noteworthy  that  the  average  total  NMHC at the Franklin  Institute  site
(Site 13) for the 0700 to 0800 EST samples  was slightly higher than  for any of the other
three sets of data.  This provides no support for the idea  that industrial emissions are the
major source of hydrocarbon pollution in the  Philadelphia area.  If industrial emissions
were the major source of hydrocarbons, the industrial site would be expected to  show the
higher hydrocarbon concentrations.   The higher value  at 0700 to 0900 EST at Franklin
Institute may reasonably be ascribed  to the higher traffic levels at this location at the
later hours.

      Some  departures from the average  type  distribution can be explained by reviewing
the  species distribution.  Before citing  examples of this, it is  well to note  that these
anomalies normally affect only individual samples, not samples taken at different sites on
the same day or at different times at the same day.  The  early Franklin Institute sample
of 3uly 20,  1979, showed high paraffins and low aromatics, while the sample taken later at
Franklin Institute and the  early  sample at the industrial site showed just the reverse.
Review  of  the  species data reveals  very  high concentrations of butane  and pentanes,
suggesting that  this sample contained hydrocarbons from  a nearby gasoline vapor source.
This sample also contained large concentrations of light oiefins, especially propene, which
are more difficult to  explain.  On July 24,  1979,  all four samples contained the highest
levels of NMHC in the entire set.  The industrial site samples were nearly average  in type
composition while the  Franklin Institute samples were  both high  in  paraffins  and low in
aromatics.   Again,  referral  to  the  species  distribution reveals   high levels of  light

-------
Table 15.  SPECIES BREAKDOWN OF FRANKLIN INSTITUTE SAMPLES.


Date
July 17
July 18
July 19**
July 20
July 23*
July 2
-------
  Table 16.  SPECIES BREAKDOWN OF INDUSTRIAL SITE SAMPLES.


Date
July 17
July 18
July 19*-
July 20
July 23*
July 21
July 25*
July 26
July 27
July 30
July 31
August 1
August 2
August 3
August f««
August 5'»
August 6'-
August 7«»
August S
August 9
August 10*-
August If*
Augusi 15*
Average/
Standard
Deviation
0500 to 0600 E.ST
(percent)
Acetylene
3
2
2
2
3
0
1
0
—
2
2
2
2
2
2
2
3
2

2
0
1
3
Oleljn
5
6
5
1
9
9
7
8
—
2
:
S
2
5
6
3
7
i*
5
5
3
IS
5

1.9/1 6/3.1

Aromatic
13
25
32
31
27
20
25
36
_
35
35
IS
31
25
19
27
21
27
10
31
19
12
20

26/9

Paraffin
79
68
60
61
61
71
66
55
_
60
60
72
65
69
73
68
71
66
81
61
18
69
73

66/8

(ppeC)
Known
1031
131
376
160
392
1119
350
S60
—
2SS
288
285
189
157
2S2
255
167
273
817
336
663
310
189



Sum
1511
520
118
575
575
1700
193
1130
—
361
361
315
226
539
307
290
550
272
878
118
711
367
5E-S

603/387


n-Butane Ratio
0.277
0.156
0.367
0.28!
C.268
0.012
0.060
C.100
—
0.326
0.326
0.079
. 116
.226
. Ill
.207
.356
.113
.038
.228
.075
.358
0.236



0700 to OSOO £5T
July 17
July 18
July 19**
July 20
July 23*
July 21
July 25*
July 26
July 27
July 30
August 1
August 2
August 3
August 1**
Augusi 5*
August 6*»
Augusi 7*»
August 8
August 9
August 10*-
August 11*
Augusi 15*
Average/
Standard
Deviation
3
1
3
—
1
1
0
I
3
3
1
3
2
2
5
1
2
3
1
2.5/1

„
9
5
—
7
S
(4
10
I*
5
0
5
5
15
2
S
5
9
7
u
6.6/2.9

21
21
37
—
20
21
15
31
50
32
11
28
26
16
30
23
27
15
21
25
21
21.9/S.5

69
63
56
—
73
69
81
58
13
59
75
65
66
67
63
65
66
67
66
66
78
72
65.S/8

561
106
511
—
786
831
S21
108
<*76
330
818
217
568
136
230
588
330
181
387
257
375
302


809
558
630
—
1018
101"
9S8
572
666
385
975
278
700
532
255
691
399
507
152
306
119
316
592/251

0.210
0.523
0.179
—
0.073
0.131
0.030
0.138
0.132
0.389
0.067
0.010
0.195
0.131
0.112
0.383
0.127
0.155
C.278
0.102
0.377


* Helicopter data avauaole
»Case study oay

-------
paraffins.  So gasoline vapor is  suggested, although the propane is also quite high in both
Frankiin Institute samples.  The industrial site  samples on July 24,  1979, were unusually
high in olefins, even higher carbon number oiefins.  The two 2-hexenes totaled 39 ppbC in
the early sample. This source signature is not recognizable.  One-hexene was high also on
August  10, 1979, at the industrial site in the later sample, but benzene was unusually high
in the early sample that day (185 ppbC).

     The  type distribution,  indeed  the  individual  species  distributions, are  strongly
suggestive  of  gasoline composition.   It  would  be  very useful to  know  the  species
composition of the gasoline sold in  the Philadelphia area.  Even more valuable would be
information on the brand-to-brand variations in fuel  composition.  Lacking this specific
information, comparison  can be made only with the rather  meager  open literature on
gasoline composition.  One of the most complete analyses was reported by the Shell Oil
Company,  Wood  River Laboratory  (Sanders,   1968;  Maynard,  1969).   Two  fuels  were
analyzed with no claim to representativeness.  They each contained about 60% paraffin,
30% aromatics, and the balance olefin. This is consistent with the type distribution in the
Philadelphia study.  The  detailed breakdown by species also shows  similarities with the
fuel analysis.

     It is  evident from this data set that,  while the species distributions are similar  in
different areas and at different times, some samples show excesses of small groups of
hydrocarbons.  With more information on signatures  it might be possible to identify the
sources of these hydrocarbons.

3.2.2      Acetylene/n-Butane Ratios

     Acetylene is of  interest  in  studies  of   this  kind, not because  it  is  toxic or
photoreactive, but because it is almost exclusively derived  from automobile exhaust.  It is
not present in gasoline (or any other motor  fuel) but  is formed in the  engine.  Since it is
one  of  the  least  reactive of  atmospheric hydrocarbons, it serves  as  a good  tracer for
engine exhaust.

     Normal  butane (n-butane)  is  also  of  fairly  low   reactivity but  it  enters the
atmosphere from several sources: automobile exhaust, gasoline vapor, vaporized gasoline,

-------
as well as industrial sources.   The ratio  of  these two hydrocarbons then should  give a
measure of the contribution of automobile exhaust to any given sample. If each ambient
sample  represents exhaust emissions from  a large number of vehicles, and if the n-butane
is all due to exhaust, this ratio should  be approximately constant and equal to  that found
in average automobile exhaust. Mayrsohn et  al.  (1977) used a ratio of about 2 acetylene
to 1 n-butane for their  source reconciliation studies.   Since air samples  showed lower
values,  substantial contributions of n-butane  from other sources were indicated. In fact,
Mayrsohn et al.  attributed roughly one-half the total hydrocarbon to automobile exhaust
based on their source reconciliation procedure.

      The acetylene/n-butane ratios from the  present study are tabulated in Tables  15 and
16.  Most values are  much lower than the 2:1  ratio used by Mayrsohn et al.  for automobile
exhaust.  They are even  lower  than the approximate 1:1 ratio reported  by Stephens (1973)
for air  samples  in Riverside, California.   The ratios given in Tables 15 and 16 are  also
quite variable, which suggests  a nonuniform air mass with  varying contributions of other
sources compared to automobile exhaust.   No  distinctive pattern in these variations could
be discerned.  If local sources of gasoline vapor are responsible for much of the n-butane,
a recognizable pattern is not to be expected.

3.2.3      Hydrocarbon  Data Validity

      The analytical group  at Washington   State University  presented  some data on
precision, providing  in their report  (Westberg and Sweany, 1980) comparisons with  two
other laboratories on  one calibration  sample which contained  one  hydrocarbon (C.H,0;
 isomer not stated).  Agreement was fair.  They  also reported an exchange sample study
done with an  EPA field laboratory.   Complete  species breakdown is  not  given but the
totals and the breakdown by types were in fair agreement.

      Another possible comparison  is  between the sum of individual hydrocarbons indi-
cated by the  gas chromatograms  and the total  measured by an  NMHC  (non-methane
hydrocarbon) analyzer.  Data from  such  an analyzer (Beckman  Instruments Model 6800)
are available,  but the NMHC data are  given to only one significant figure  so this method
hardly provides a useful  test.  This instrument measures methane and total hydrocarbons
separately -- indeed, on separate samples — and then subtracts to determine NMHC.

-------
Since methane is not small compared to the total hydrocarbon, it is necessary to subtract
a large number from a slightly larger number.  This magnifies the error.

     Hydrocarbon levels in  ambient air can be quite variable, so it is necessary to  take
simultaneous samples if agreement is to be expected.   Two  such  replications  were
included in the data base; one  pair taken in Lancaster, Pennsylvania on August 1*,  1979
(time not stated), and a second pair taken in Downingtown, Pennsylvania, at 0700 to
0800 E5T on August 17, 1979.  The. data are  shown in Table 17.  It should be noted  that
these two cities are many miles from Philadelphia, yet hydrocarbon concentrations were
comparable  to those in urban  air.  Automobile exhaust is probably the  main source in
these cities.  Table 17 follows the format used for the main data table for the Franklin
Institute and industrial site  samples.  The type distribution gives the percent acetylenes,
olefins, aromatics, and  paraffins while the last two columns  give  total of  identified
compounds in  ppbC and the sum  which includes unidentified  compounds.   Ethylene is
omitted from these olefin totals for two reasons:  to keep the data presentation uniform,
and because the quite  high values are suspect although some sort of local source cannot be
ruled out. The two Lancaster samples each showed about 600 ppbC of hydrocarbon  with
quite similar type distributions.   Even though the two  Lancaster  samples appear  very
similar, some  large differences  between  individual  hydrocarbons  are  apparent.   For
example, in  Sample 2  the  benzene  and toluene were about 40% smaller than  in Sample 1
but  some of the  CQ  aromatics were larger.   The  type distributions are  reasonably
consistent even though the total hydrocarbon differed by a factor of  two between the two
Downingtown samples.  The second Downingtown sample contained only  half  as much
hydrocarbon as  the  first:   concentrations were lower across  the  board so the  type
distribution  remained  unchanged.  There  is no ready  explanation for the disagreement
between these two samples  taken at the same time and place.  Successive samples taken
the  same morning were  similar  in  composition  except for  much  lower  ethylene
concentrations (4 to 6  ppbC instead of  50 to 75 ppbC).

3.2.4      Limitations of Analytical Scheme

     The present  analytical  scheme had some limitations which should be recognized.

-------
                  Table 17.   DISTRIBUTION (%) OF HYDROCARBON SPECIES IN SIMULTANEOUS SAMPLE

                            PAIRS, TAKEN AT LANCASTER AND DOWNINGTOWN, PENNSYLVANIA,

                            RESPECTIVELY.
-P
\D
Sample
Number
Type Distribution (%)
Acetylene
Olefin
Aromatic
Paraffin
Total (ppbC)
Identified
Compounds
Including
Unidentified
Compounds
Lancaster, Pennsylvania (time not stated) August 14, 1979
1
2
5
5
11
8
23
25
61
62
487
447
592
552
Downingtown, Pennsylvania (0700-0800 EST) August 17, 1979
1
2
3
2
3
2
32
29
61
66
169
84
205
94

-------
1.    Data are limited to  hydrocarbons, although oxygenates must be present, some
     from atmospheric reactions and some from sources.

2.    The  ethylene  values are  all suspect.   The operators  recognized that many
     samples taken  near  monitoring  stations were contaminated  by the effluent
     from chemiluminescent ozone analyzers.   Other samples, not subject  to this
     fault, showed such high ethylene values that they are hardly  believable.  The
     data summaries omitted.ethylene from the olefin total for the samples subject
     to contamination samples but not for others. This introduced a bias into the
     data reports. If distributions are to be compared it is better to omit ethylene
     for all  samples  since it cannot be included for all samples.  The data have been
     recalculated on this basis.

3.    Qualitatively,  all  the hydrocarbons  can  be  attributed  to auto  exhaust.
     Figure 13 shows a  total  ion chromatograph used  to  identify hydrocarbon
     species seen in field samples.   While the chromatograph is not  for a sample
     from  the Philadelphia study, it is characteristic of  most  of  the samples
     analyzed and includes the  hydrocarbon species in the C<--C1? molecular  weight
     range observed in the Philadelphia study.  The only species  listed in Figure 13
     with a prominent industrial use is  styrene.  While information on production
     and use of styrene in the Philadelphia area is lacking, it would  be  most helpful.
     Butadiene 1, 3, another prominent industrial hydrocarbon, was  not measured.

4.    Isoprene,  2  methyl  butadiene 1, 3, is  shown  on the sample chromatogram
     (Figure 13) as a well-separated, although  very small, peak.  However,  it was
     not included in any of  the data reports.  This hydrocarbon is important because
     it is the parent of the  entire terpene series.  Isoprene itself, along with alpha
     pinene, beta pinene  and a host  of  other terpenes are emitted by  vegetation.
     As  olefinic hydrocarbons they  are extremely  reactive   but  this  must  be
     balanced against quite low  emission rates.   Because the actual  size of these
     emissions is not accurately known,  the role they  play in oxidant  enhancement
     is still controversial.  A  large  family of  terpenes, including alpha  and  beta
     pinene, have the  formula  C1QH16-  They  would probably  appear  in  the
     chromatogram  interspersed  among the Cg aromatics.  They  were apparently
     not identified on the chromatogram.

-------
                                                                          TOTAL ION   CHfiOMATOGRAM
                           INCREASING  Tine  4ND TEMPERATURE
     1.   2  Methyl   1  butene        23.
    -2.   2 - Methylbutane              2"*.
     3.   Halocarbon                   25.
     14.   1 - Pentene                   26.
     5.   3 - Methyl   1  butene        27.
     6.   n  Pentene                   2S.
     7.   Isoprene                      29.
     S.   Carbon disuifide              30.
     9.   t - 2 - Pentene               31.
    10.   c - 2 Pentene                 32.
    11.   2 - Methyl - 2 butene          33.
    12.   2,2 - Dimethylbutane          34.
    13.   Cyclopentene                 35.
    14.  (Cyclopentane                 36.
        V - Methyl   1  Pentene       37.
    15.   2,3 - Dimethylbutane          38.
    16.  /2  Methylpentane             39.
        \t  1 - Methyl - 2  pentene    40.
    17.   c  4   Methyl  2 - pentene    41.
    18.   3 - Methylpentane             42.
    19.  /2 - Methyl   1 - pentene       43.
        *• 1  Hexene                   44.
    20.   n - Hexane                   45.
    21.   t - 2 - Hexene                46.
    22.  ,2 -Methyl   2 - Pentene       47.
        v t - 3 - Methyl - 2 - Pentene    48.
c - 2   Hexene               49.
Methylcyclopentane
c - 3 - Methyl - 2 pentene     50.
2,2,3   Trimethylbutane       51.
1,1,1   Trichloroethane        52.
2,4   Dimethylpentane        53.
Benzene                     54.
1 - Methylcyciopentene       55.
Cyclohexane                 56.
2 - Methylhexane
2,3 - Dimethylpentane        57.
3 - Methylhexane             58.
Dimethylcyclopentane        59.
Dimethylcydopentane        60.
Dimethylcyclopentane        61.
2.2,3, - Tnmetnylpentane     62.
n - Heptane                  63.
Methylcyclohexane           64,
Tnmethylcyclopentane        65.
EthylcycJopentane            66.
2,5 - Dimethythexane         67.
2,4 - Dimethylhexane         68.
2,3.4 - Tnmethylpentane      69.
Toluene                     7C.
2,3 - Dimethylnexane         71.
2 - Methylheptane            72.
/ 3 - Ethyihexane
^ 3 - Methylheptane
 C  - 9 Alkane
 Dimethylcyclonexane
 n - Octane
 Ethylcy cione xane
 C   9 Alkane
 Ethylbenzene
/ p - Xylene
* m  - Xylene
 Styrene
 o - Xylene
 n - Nonane
 i - Propvibenzene
 n - Propvibenzene
 3   Etnvitoluene
 2   Ethyitolune
 1   Ethyltoluene
 1,3,5   Trimethylbenzene
 1.2,4   Trimethyibenzene
 1,2,3 Trimethylbenzene
 Methylstyrene
 1,3  Dietnylbenzene
 1,4 - Diethylbenzene
 C   10 Substandard oenzene
 C   10 Substandard benzene
FIGURE 13.   Total ion chromatograph from ambient air sample (source:  Westberg
                 and Sweany,  1980).

-------
5.    The  nature  of  hydrocarbon isomerism introduces a potential bias into
     species  summation.  Thus, there  is only  one 6-carbon aromatic  hydro-
     carbon (benzene), and 10 ppbC of this will produce an easily measurable
     chromatographic peak.  There  are seven 6-carbon saturates (counting
     cyciohexane and methyl cyclopentane) each of which produces a separate
     peak in the chromatogram. There are twenty-two 6-carbon oiefins each
     giving a separate  peak  in the chromatogram.   If 10 ppbC of 6-carbon
     oiefins were distributed equally  among twenty-two  peaks,  each would
     contribute a little less than one-half ppbC.  Many would be unmeasurabie
     among the larger  peaks.  This problem  becomes more acute at lower
     hydrocarbon  loading and  severe  for larger molecules.   This "isomer
     explosion" not  only  makes the average  concentration at each carbon
     number very small, but it also complicates the problem the analyst has in
     identifying peaks.  It means that the absence of higher oiefins from the
     analytical list is partly an  artifact.

6.    The sample  chromatogram in Figure  13 tapers off at molecular weights
     above  C,Q,   suggesting that  not too  many hydrocarbons  are being
     overlooked in this range.   It is significant that the chromatograms end  at
     the end of the gasoline range.

-------
                                4.  CASE STUDY DAYS

     One of the primary objectives of this study was to  recommend days for which data
could be used for  model application and  validation.  The case study days were chosen to
present a variety of situations in which ozone concentrations in  excess of 0.12 ppm were
observed.   Other  factors  considered  in selecting  the  case study days  included  NO2
concentrations, data availability (including EPA-EMSL helicopter data), and  wind speed
and wind direction at the surface and aloft during early morning (0400 to 0800 EST), late
morning (0800 to  1200 EST) and afternoon (1200  to 1800 EST) hours.  AV selected ten case
study days, with the concurrence of the  EPA Project Officer.  These days are: 3uly 12,
13, 16, and 19, and August 4, 5, 6, 7, 10, and 22.  August 4 through August 7 represents an
episode of high ozone levels.

     This section  discusses and compares the case study days, particularly with respect to
meteorology,  trajectories, transport of ozone and precursors into  and  within the study
area, pollutant sources, and concentrations aloft.

     Synoptic scale trajectories  referred to in the following discussions were  provided by
the EPA. An atmospheric transport and diffusion model was used to compute the latitude
and longitude of an air parcel at several  time intervals prior to its reaching Philadelphia.
The program computes trajectories by taking average winds in  the layers  of interest from
0000 GMT (1700 EST)  and   1200 GMT  (0700 EST)  wind  soundings  and,  if  available,
0600 GMT (0100 EST)  and  1800 GMT (1100 EST)  soundings.    Each  observation  was
weighted by the inverse  of the square of the distance between  the parcel  and the sounding
station.

     Mesoscale trajectories were derived using surface wind information from all  sites in
the study area for which data were available.  The interpolation scheme uses  weighted
averages of the wind components.  The weighting factors are inversely proportional to the
square  of the distance between the air parcel and the wind site. More weight was given to
wind observations  made  directly upwind  or downwind of the  air parcel  than  when made
removed to one side.  This feature is included to reflect the tendency of  winds to  change
more rapidly in cross-streamline directions than  in the along-streamline direction.

-------
      For  each case study day, mesoscale trajectories were constructed for the following
selected air parcels:   (1) an  air parcel near downtown Philadelphia at 0600 EST, (2)  a
control trajectory  approximately  10 miles  northwest  of  downtown  Philadelphia  at
0600 EST  (these were constructed to determine the sensitivity of the starting point on the
trajectory path), (3)  an  air parcel at an upwind site  at 0600 EST,  and (4) an air parcel
ending at  the site and time of the maximum ozone concentration measured during the day.
Other trajectories were constructed when appropriate.

      Rural and urban mixing heights  were derived for the case study days for which data
were  available.   Urban mixing heights  were computed from radiosondes released at
Beukers Laboratories  in  downtown  Philadelphia.   Detailed  temperature  profiles were
constructed at 0500, 0900, and 1500 EST for  the seven  case study days that  soundings
were  made. The three hours mentioned above are the only times that sufficient data were
available  to construct meaningful  profiles.  From these profiles the height of the base of
the first stable layer (i.e., where the variation of  temperature with height is less than the
dry adiabatic  lapse rate:  1° C per 100 m) was taken to be the height of the urban mixed
layer. The height of the mixed layer at hours other  than 0500, 0900, and 1500 was taken
to be the  height at which the temperature profile represented by the dry adiabatic lapse
rate (or moist adiabatic lapse rate,  when the air was near  saturation) intersected  the
detailed profile measured by the temperature sounding. The surface temperature for the
hour of interest and the most recent temperature sounding available (0500, 0900, or 1500)
were  used for  this determination.  In order  to minimize  the errors  involved in  extrapo-
lating the 0900 sounding to  1300 and  1400 in this way, the  mixing height determined for
1500 EST  was  used as an upper bound for mixing heights of earlier hours.

      The  heights of the mixed layer for 0500 through  1800 EST are presented in  Table 18
for the case study days.  On all days but one, the mixing height  was zero meters above
ground level at  0500 EST, signifying a stable layer started at the surface.  By 1500 EST
the mixed layer height was usually above  1000 m  MSL indicating significant mixing in the
lower atmosphere.  The mixing heights usually  reached  a  maximum between  1300 and
1600 EST.  At  that  time, surface temperatures usually began to  decrease, causing  a
ground-based inversion to form, thus inhibiting mixing from the ground level.  The time at
which surface  temperatures began to decrease after  1500  is indicated by a zero mixing
height in Table 18.  A ground-based inversion  began to form by 1800 EST on all of  the days
shown.

-------
Table 18.  URBAN MIXING HEIGHTS (m MSL) FOR THE CASE STUDY DAYS AS DETERMINED FROM RADIOSONDE
          RELEASED FROM DOWNTOWN PHILADELPHIA.
Date
July 19
August 4
August 5
August 6
August 7
August 10
August 22
Hour (EST)
0500
0
0
0
0
0
125
0
0600
187
0
0
0
0
221
0
0700
407
0
0
0
181
291
171
0800
407
0
147
96
250
291
197
0900
741
157
388
345
325
339
411
1000
741
157
424
591*
620
503*
900
1100
851
505
606
591
972
503
900
1200
991
606
606
666
1332
503
1080
1300
1572
698
1072
666
1332
1245
1567
1400
1572
698
1072
666
1332
1516
1567
1500
2314
930
1072
666
1332
1516
1707
1600
2314
1270
1287
1044
1439
0
0
1700
0
1270
1287
0
1439
0
0
1800
0
0
0
0
0
0
0
* Radiosonde "surface" temperature used instead of Franklin Institute.

-------
     Rural mixing height information was obtained from AV's acoustic sounder at Summit
Bridge, DE (Site 1).  From these data,  the  time at which surface ozone concentrations
were representative of  concentrations  aloft was determined in the following way:  The
surface-based inversion layer  usually begins eroding shortly after sunrise.  Heat trans-
ported upward from the surface acts to lift the inversion above the surface at the same
time that it  is eroded  from below.  When the bottom  of  this elevated inversion layer
reaches the height of the top of the original morning ground-based inversion, some of the
ozone trapped aloft is transported downward.  However, the bulk of the ozone reservoir
aloft  is  not mixed downward  until the inversion layer, now elevated, is  completely
destroyed. After the destruction of the inversion layer, mixing usually proceeds upward
to the  vertical limit of  the acoustic sounder (1000 m) in a matter of minutes. The time  at
which this thorough mixing within  the  mixed layer occurs is the time  at which  surface
ozone concentrations are most representative of concentrations aloft.  For the case study
days,  the times at  which the mixing  height first  exceeded  1000 m  M5L are  given  in
Table 19. From this table, we can see that, for the case study days, thorough mixing had
occurred  by  1100 EST and that the  average time of the rural mixing heights exceeding
1000 m is 1000 EST.

     A comparison of ozone concentrations at several stations before  and after the time
of uninhibited vertical  mixing revealed that concentrations  were generally  low  in the
study area before the breakup and became higher afterwards due to mixing of ozone from
aloft.  This is supported by the  observation that on days when the helicopter was flown,
the surface ozone  concentrations just  after  the inversion  breakup were similar  to the
concentrations measured aloft by the helicopter. For example, Figure 14 shows an ozone
profile taken  over  Summit  Bridge, DE  (Site 1) on  August 10 at  Q535  EST.  The  surface
concentration at this site was 0.026 ppm  at 0500 EST.  The  profile indicates concentra-
tions  higher  than  0.06 ppm  in  a  layer from  about 600 to  900 meters MSL.  Surface
concentrations at this site increased later as the higher concentrations aloft were mixed
down.  At 1000 EST —  the first hour when  the mixing height was above 1000  meters --
this site recorded a surface concentration of 0.079 ppm.

     Information on the locations of the major point sources of volatile organic emissions
within  the study area  was  provided by Philadelphia's Air  Management Services.  The
majority  of these  sources are along or close to the Delaware  River,  with  the largest

-------
Table 19.    TIME WHEN RURAL MIXING HEIGHTS FIRST EXCEEDED
           1000 m AS DETERMINED BY AV ACOUSTIC SOUNDER AT
           SUMMIT BRIDGE, DELAWARE (Site 1).
Date
July 19
August 14
August 5
August 6
August 7
August 10
August 22
Time (EST)
1000
0900
0900
0930
0945
1100
0915

-------
    2500r
     2000
     1500
   1)
   *•*

   1)
     1000
      500
            0.02       0.06       0.10       0.14       0.1S


                          Concentration (ppm)
FIGURE 14.  Ozone profile over Summit Bridge, DE (Site  1) at 0535 E5T

             on August 10,  1979.
                                58

-------
concentration of sources in Philadelphia  and Chester.  These are also areas with heavy
commuter  traffic.  These  sources were  considered  in  the  analysis of patterns of  high
ozone concentrations.

     The hydrocarbon species data set from the Franklin Institute/industrial site provides
data for six of  the ten  case study days.   Although  these  days  were  chosen because  of
elevated oxidant values, the hydrocarbon species data for these days do not stand  out  as
distinctive with regard to total concentrations or species distributions.

     A resultant  wind speed and wind direction vector was derived for each station in the
study area for each of the time periods (early morning, late morning, and  afternoon) on
each case  study day.  From these resultant vectors for each station, then, resultant wind
speed and wind  direction vectors were  approximated for the entire study area.   Table 20
gives,  along with  peak ozone (OJ and  nitrogen dioxide  (NO-) concentrations, the surface
resultant wind  directions  and  wind  speeds,  and average  wind  speeds and  prevailing
directions  aloft in the 500- to 1500-m layer, for each of these periods on the case study
days. It also notes whether or not helicopter data were available for each day.

-------
            Table 20.  PEAK O AND NO  CONCENTRATIONS, WIND DATA, AND HELICOPTER DATA AVAILABILITY
                     FOR CASE STUDYTJAYS FOR THE ENTIRE STUDY AREA.
Date
July 12
July 13
July 16
July 19
August 4
August 5
August 6
August 7
August 10
August 22
Peak
°3
(ppm)
.147
.183
.151
.157
.144
.135
.136
.129
.170
.140
Peak
NO2
(ppm)
.110
.170
.069
.070
.090
.090
.070
.080
.080
.100
Wind Direction
Resultant Surface
Time (EST)
04-08
SW
Var
E
N
NW
NW
NW
N
SW
SW
08-12
W
SE
E
NE
N
N
NW
NE
SW
Var
12-18
NW
Var
SE
SE
N
NW
NW
W
W
SW
Prevail-
ing
Aloft
—
—
NE-E
NE-E
N
N-W
NW
W
W
—
Wind Speed (m/s)
Resultant Surface
Time (EST)
04-08
1.5
1.0
1.0
1.5
1.0
1.5
1.5
1.5
2.5
1.5
08-12
1.5
1.3
1.8
2.5
2.0
2.0
2.0
1.5
3.5
2.0
12-18
3.0
2.0
2.5
1.5
1.0
2.0
2.0
1.5
3.0
2.0
Aver-
age
Aloft
--
—
2.0
4.0
4.0
4.0
3.0
2.0
8.0
—
Avail-
ability
of
Heli-
copter
Data
No
No
No
No
Yes
Yes
Yes
Yes
Yes
No
OS
O

-------
*.l   JULY  12, 1979 (THURSDAY)

     Important  aspects  of the July 12 case include  the  weak influence  of the Bermuda
high-pressure cell, indications of  significant ozone levels aloft, and high surface ozone
concentrations upwind of Philadelphia.   The highest ozone concentration observed this
day occurred downwind of the industrial areas  near Wilmington, DE, although high ozone
levels also occurred upwind of the  urban complex, indicating long-range transport.

     On July 12,  1979,  the peak  ozone  concentration  in the study area  was 0.147 ppm.
Two of  the  twelve operating stations reported ozone levels higher than 0.12 ppm and eight
stations reported ozone  levels higher than 0.10 ppm.  In  addition,  two stations reported
NO- levels higher  than 0.10 ppm.

4.1.1       Synoptic Meteorology

     The surface  synoptic situation at 0700 EST is shown in Figure  15.  Pressure gradients
were weak  over the study area and circulation was  weakly dominated by the Bermuda
high-pressure cell  near  the Bahamas.   A very moist low pressure  cell, the remnants of
Hurricane  Bob,  was  approximately 700 km  to the  southwest  and moving toward the
northeast.  Upper level wind flow was westerly and weak.

     The synoptic  trajectories of  air parcels in the 500- to 1000-m and 1500-  to 2000-m
MSL layers  arriving at Philadelphia at 1300 EST are shown in Figure 16. The paths are
generally from the west  and are entirely over land. The urban area of Pittsburgh is near
the path of  the 1500- to  2000-m layer parcel.

4.1.2       Mesoscale Meteorology

     The mesoscaie flow was weak and from the southwest  during the early morning.
Figure 17 shows  a  streamline  analysis  depicting  the flow  across  the  study  area at
0700 EST.  At 1100 EST  the flow was from the west in the southern part of the study area
but primarily from the northwest at sites north of Philadelphia. During the afternoon, at
1500 EST, most  sites were reporting winds from the northwest, with wind speeds at six
sites greater than  5 m/s.  Wind speeds were lighter after 1700 EST and primarily from the
west or northwest. No winds aloft data were available.
                                         61

-------
                                                                                                          70C
OX
K)
                                       "CURE I;,   synoptic situation, 0700 EST, 3uly 12, 1979.

-------
                            90°
                                              1900 EST°--o
FIGURE 16    Synoptic trajectory of air parcels arriving  at  Philadelphia at 1300 EST, July 12, 1979.  Solid line indicates
              parcel in the 500 to 1000-meter layer; dashed line is for a parcel in the 1500 to  2000-meter layer.  Intervals
              are six hours.

-------
FIGURE 17.  Streamline analysis depicting surface flow across the study
             area at 0700 EST on July 12, 1979. The numbers near the
             sites indicate wind speeds (meters per second).

-------
4.1.3       Precursor Patterns

     The highest levels  of morning precursors occurred downtown, a pattern followed on
all  of  the case  study  days;  however, concentrations at  Lumberton,  NJ  (Site 3), and
Bristol, PA (Site?), to  the east and  northeast, respectively, of the Philadelphia  urban
area,  were abnormally high.  The  6:00 to  9:00 a.m. LOT averages of NMHC and  NOX>
along with the ratio of  iNMHC to NOx are given  in Table 21 for rural and  urban  sites.
From  this table,  the downtown levels  on  July 12 were  about average when  compared to
other case study  days.  The morning levels at  rural sites indicate an interesting pattern.
Low concentrations of both pollutants  were measured at Summit Bridge, DE (Site 1), and
Downingtown, PA (Site 2),  which are to the southwest  and west, respectively, of the
Philadelphia urban area.  However, unusually high levels were measured at  Lumberton, N3
(Site 3), which is  to the east.  The 6:00 to 9:00 a.m. LOT average NMHC concentration of
1.0 ppm measured at Lumberton was  the highest recorded  at this  site during the field
program;  furthermore,  concentrations  of   NO   at this  site   increased to  0.080 ppm
(0.068 ppm NC^)  by 0900 E5T.  This pattern of morning precursors is most likely a result
of the airflow which  was light and either westerly or southwesterly during the nighttime
and early morning  hours (Figure 17).   This implies  transport of these pollutants  from
Philadelphia.

4-1.4       Ozone Patterns

     Considering the afternoon flow directions, the areas of highest  ozone  levels  shown in
Figure 18, are as expected.  The lowest  ozone concentrations  occurred near downtown
Philadelphia, probably due to NO scavenging. The highest concentrations  were measured
at locations southeast of Philadelphia,  with  Ancora, NJ (Site 11), recording the maximum
value of 0.147 ppm.  Concentrations in excess of 0.10 ppm  were  recorded  by  1100 EST at
Downingtown, PA  (Site 11), and  remained  above this  level at  Lumberton  and Ancora
through 1700 EST.

     To determine  possible  source  and  impact  areas,  mesoscale  trajectories  were
constructed for several air parcels of  interest  and are  shown in Figure 19.   A backward
trajectory was developed for an  air parcel ending at Ancora, NJ (Site 11),  at 1300 EST,
with the highest  recorded ozone  level  of the day, 0.147 ppm.  This  air parcel followed  a

-------
Table 21.    6:00 to 9:00 a.m. LOT (0500-0800 EST) AVERAGES OF NO
           and NMHC (ppm) AND THE RATIO OF NMHC TO NO FOR
           RURAL AND URBAN SITES ON JULY 12.
Site
No.
1
2
3
5
Rural Sites
Summit Bridge, DE
Downingtown, PA
Lumberton, NJ
Van Hiseville, N3
6:00 to 9:00 a.m. LDT Averages
N0x
0.00
0.02
0.0*
0.01
NMHC
0.0
0.0
1.0
—
NMHC/NO
Ratio X
—
1.5:1
25.8:1
—
Urban Sites
6
7
8
10
12
13
1*
Chester, PA
Bristol, PA
North Philadelphia Airport, Pfi
Camden, NJ
So. Broad & Spruce Streets,
Philadelphia, PA
Franklin Institute,
Philadelphia, PA
AMS Lab, Philadelphia, PA
—
—
0.05
0.12
0.07
0.07
0.10
1.1
1.*
—
—
0.6
—
1.0
—
—
—
—
8.1:1
—
10.0:1
                              66

-------
            1300 EST
1400 EST
              1500 E5T
FIGURE IS.  Isopleths of ozone at 1300, 1400, and 1500 EST, July 12, 1979.
                                  67

-------
FIGURE 19.   Surface trajectory for  various air parcels located within the study area on
              July 12, 1979.  Solid line  indicates  a trajectory  beginning at downtown
              Philadelphia at 0600  EST; dashed  lines  are  for  parcels  ending  at  the
              indicated site and  time. Dots show the positions of the parcels at  one hour
              intervals.
                                       68

-------
trajectory somewhat south of the Philadelphia source area but was close to the highly
industrialized area between Philadelphia and Wilmington, Delaware, at 0600 E5T.

      Mesoscale trajectories are  also shown for air parcels with  high ozone levels at sites
upwind  of  Philadelphia.  Air parcels  arriving at Sites 1 and 2 at 1400 EST and 1500 EST,
respectively,  contained  ozone concentrations in excess of 0.10 ppm.   These air  parcels
apparently traversed no major source areas during the  five to six  hours before reaching
the end of the trajectory,  indicating  the possibility of longer-range transport, particularly
in light of the long over-land synoptic  trajectory previously shown in Figure 15.   The
evidence of transport aloft is discussed below.

      An air  parcel  beginning at  downtown Philadelphia at 0600  EST was located between
Lumberton, NJ  (Site 3),  and Van Hiseville, NJ (Site 5), at  1500 EST.  These stations had
high O, concentrations of  0.124 and 0.105  ppm, respectively, at this time.

4.1.5      NO-, Patterns

      Peak NO_ concentrations  occurred  near the urban area,  very late in the evening.
Conversion of O, into NO^ by scavenging of NO (relatively plentiful in  the urban area) is
the probable cause.  For example, the Franklin Institute (Site 13) NO2 peak of 0.090 ppm,
which occurred at 2100  EST,  followed  an  ozone concentration trend of 0.070 to 0.020 to
0.000 ppm from 1900 EST to 2100 EST.

4.1.6      Concentrations Aloft

      Although no helicopter data were available for July 12, pollutant concentrations at
upwind  surface  stations  at the estimated  time of morning  inversion  breakup  were
reviewed  to  get some idea of ozone and  precursor transport aloft. No acoustic sounder
data were available to determine the exact time of  inversion breakup, so 1000 EST was
 assumed,  based upon the  fact that all but one of the case study days for  which acoustic
sounder data were available experienced the first hour of uninhibited vertical mixing at
1000 EST.
                                          69

-------
     The diurnal profiles of ozone for the upwind sites at Summit Bridge, DE (Site 1) and
Downingtown, PA  (Site 2), for  July  12 are shown in Figure 20.   A rapid morning ozone
build-up, characteristic of mixing from aloft, is evident.  At 1000 EST, Summit Bridge and
Downingtown  recorded ozone concentrations of 0.068 ppm  and  0.075 ppm, respectively,
suggesting significant  ozone transport aloft.  NO and NMHC concentrations were near
zero at this time suggesting no significant precursor transport.  Transport of such levels
of ozone helps explain the high afternoon ozone concentrations  recorded at Sites  1 and 2
(0.109  and  0.114 ppm, respectively)  despite  their positions  predominantly upwind  of
Philadelphia.
                                         70

-------
                                           Downingtown, PA
                                               (Site 2)
                                         Summit Bridge, DE
                                              (Site i)
0.00
        0200
0600
1000         1*00
    Time (E5T)
1800
2200
     FIGURE  20.    Diurnal profiles of ozone at upwind sites on July 12, 1979.
                                    71

-------
4.2  JULY  13, 1979 (FRIDAY)

     Important aspects of the July 13 case include the highest ozone and NO2 concentra-
tions recorded during the field program, variable flow, and indications of a reservoir  of
elevated ozone concentrations aloft.  Peak concentrations were due  to the influence  of
the urban plume, coupled with near stagnation and ozone mixed down from aloft.

     The peak O, and NO2 concentrations recorded during the field  program (0.183 and
0.170 ppm,  respectively) occurred on  July  13.  Ten of the  thirteen operating stations
reported O, levels higher  than 0.12 ppm  and four  stations recorded NO2 concentrations
higher than  0.10 ppm. July 13 followed another case study day (July  12) during which high
NO_ and O-, levels were observed.

4.2.1       Synoptic Meteorology

     The surface  synoptic pattern  is depicted in Figure 21.   The Bermuda high-pressure
cell, about 500 km  east of Florida, exerted a weak influence over  the study area.  The
remnants of Hurricane  Bob  were 700 km west-southwest  of the study  area,  moving
eastward.  These two features were reflected  in  the  upper-level flow which  was very
weak and primarily northwesterly over the study area.

     The trajectories of air parcels traveling in the 500- to 1000-m  and  1500- to 2000-m
layer toward Philadelphia (arrival time 1300 EST) are shown in Figure 22. The  paths are
very similar to the  synoptic trajectories for July 12, which is  reasonable, considering the
stability of the weather systems.  The path brings the 500- to 1000-m parcel near the high
emission density  area of Washington, D.C.-Baltimore  less than 36 hours upstream and
provides a continual over-land trajectory with slight anticyclonic curvature.

4.2.2       Mesoscale Meteorology

     Early  morning wind flow was variable and very weak.   Late morning southeasterly
wind flows were primarily weak; however, a great deal  of variability was evident.  In the
afternoon, there was a significant spatial variation in the wind  speed, with some stations
reporting calm winds and some stations reporting winds  in excess of 5 m/s; however, most
stations reported wind directions with a southerly component.
                                         72

-------
                                                                           70'
90
                                        80°                75°





            FIGURE 21.   Synoptic situation, 0700 EST, July 13, 1979.

-------
FIGURE 22.
Synoptic trajectory of air  parcels arriving at Philadelphia at  1300 E$T, July 13, 1979.  Solid line indicates
parcel in the 500 to 1000-meter layer; dashed line is for a parcel in the 1500 to 2000-meter layer. Intervals
are six hours.

-------
     The weak and  variable  wind flow  at  the  surface, coupled  with  relatively little
movement  of  air parcels on the synoptic  scale (Figure 21) between 0100 and  1300 EST,
indicates that July 13 was a case of near stagnation.

4-2.3       Precursor Patterns

     Urban precursor levels were elevated while rural areas had characteristically low
concentrations. Maximum morning NO and NMHC concentrations at urban stations were
in excess of  0.10 ppm and 1.0 ppm,  respectively,  with 3.5 ppm of NMHC recorded  at
Chester  (Site  6) at 0600 EST.  The  6:00 to 9:00 a.m.  LOT averages of NMHC  and NOx,
along with the ratio of  NMHC to NO  are given in  Table 22 for  rural  and urban sites.
                                     X
Elevated levels of both  pollutants were recorded at  the urban  sites,  whereas levels  at
outlying  rural stations  were  generally very low.  Chester, PA (Site 6),  measured the
highest average NMHC level (2.6 ppm) and also reported relatively high  levels (0.040  to
0.056 ppm) of  NO2 from  6:00 to 9:00 a.m. LOT.

     The elevated precursor levels measured at  these urban and  near-urban  sites were
above normal  and generally above levels observed  on other case study days.  This, coupled
with the near-stagnant weather  situation  on this  day,  helps to explain the  unusually high
ozone concentrations which occurred along the fringe of the urban area.

4.2.4       Ozone Patterns

     Isopieths of ozone  concentrations at 0900,  1100,  1200,  and 1300 EST are shown  in
Figure 23.    By  0900 EST, a  concentration  of  0.101 ppm  was   already  recorded   at
Ancora, NJ  (Site 11),  southeast  of Philadelphia,  while concentrations  were  less than
0.04 ppm downtown.   After this, highest  ozone concentrations  were centered near the
urban area.   At  1100 EST, ozone concentrations in excess of 0.12 ppm  were  evident  at
Chester,  PA (Site 6), Camden, NJ (Site 10), and  Ancora, NJ  (Site  11), the first  two  of
which  are  near urban sites.   Thus,  high  levels  were widespread very  early, although
downtown sites were still low.  Downingtown, PA (Site  2) and Lumberton, NJ (Site 3), west
and east  of Philadelphia, respectively, were recording  ozone levels in excess of 0.10 ppm.
The following  hour (1200 EST),  the highest  level recorded during  the program,  0.183 ppm,
occurred  at   Chester.   At  1300 EST,Camden, NJ   (Site 10),   had the  highest   ozone
                                         75

-------
Table 22.    6:00 to 9:00 a.m. LOT (0500-0800 EST) AVERAGES OF NO
           and NMHC (ppm) AND THE RATIO OF NMHC TO NO  FOR
           RURAL AND URBAN SITES ON JULY 13.
Site
No.
1
2
3
5
Rural Sites
Summit Bridge, DE
Downingtown, PA
Lumberton, NJ
Van Hiseville, NJ
6:00 to 9:00 a.m. LOT Averages
N0x
0.00
0.0*
0.03
0.02
NMHC
0.1
0.2
0.3
—
NMHC/NO
Ratio
0.0:0
5.8:1
11.0:1
—
Urban Sites
6
10
12
13
1*
Chester, PA
Camden, NJ
So. Broad & Spruce Streets,
Philadelphia, PA
Franklin Institute,
Philadelphia, PA
AMS Lab, Philadelphia, PA
—
0.17
0.17
0.15
0.12
2.6
—
0.9
—
1.2
—
—
5.5:1
—
10.2:1
                             76

-------
              0900 EST
              1200 EST
1300 EST
FIGURE 23.   Isopleths of ozone at 0900, 1100, 1200, and 1300 EST, July 13, 1979.
                                  77

-------
concentration (0.161 ppm),  but the  entire metropolitan  area of Philadelphia also had
relatively high ozone levels (greater  than 0.110 ppm).  Moreover, high ozone levels were
measured at Sites 2, 3, 8, and 11 as well.  Ozone levels persisted in excess of 0.12 ppm
until late afternoon at some locations.

      Surface mesoscale trajectories were developed to help determine the origin of some
of the observed high ozone concentrations.  These trajectories are shown in  Figure 24.
The  flow was weak  and  variable with a tendency towards southerly and  southeasterly
winds, particularly in the late morning and afternoon.

      An air  parcel  beginning  at  Philadelphia  at 0600 EST  was about  20 km  east  of
Downington, PA (Site 2), at 1300 EST.  Unfortunately, this is an area  for which no data
were  available;  however, this parcel is between Downingtown, PA (Site 2), and the
AMS Lab, Philadelphia (Site 14).  These  two  sites reported ozone  levels  of  0.123 and
0.140 ppm,  respectively, at this time, and we  might expect even higher  levels along the
path of the trajectory.

      Chester, PA (Site 6),  as discussed  above,  had the highest  ozone  concentration
measured on this day and  recorded a value of 0.183 ppm  at 1200 EST.  A parcel ending at
this site  and time followed a trajectory from the southeast and started (0600 EST)  in  an
area of expected low precursor concentrations 20 km south-south west of Philadelphia's
central business district.

      Air  parcels  with high ozone  levels  at  1400 EST  at  the  AMS Lab,  Philadelphia
(Site 14), and 1500 EST at Lumberton, N3 (Site 3), began in the eastern portion of the
study area.  In fact, the air parcel ending at the AMS Lab at 1400 EST  began very near to
Lumberton at 0600 EST.  The NMHC  and NO   concentrations at Lumberton at that time
were  both zero, indicating another high ozone parcel originating in areas of low precursor
levels.

      A re-examination of  the ozone isopleths  shown  in Figure 23  indicates that the
surface mesoscale trajectories do not tell the whole story.   The center  of  high  ozone
concentrations does not  move westward  across  the study as the mesoscale trajectories
imply, but  rather shows the spreading out tendency  characteristic  of  a stagnation day.
The contribution of ozone  transport aloft to this process is  discussed below.
                                         78

-------
FIGURE 2^.   Surface trajectory for various air parcels located within the study area on
              July 13, 1979.   Solid line  indicates  a  trajectory  beginning at downtown
              Philadelphia at  0600  E5T; dashed  lines  are for  parcels  ending  at the
              indicated site and time. Dots show the positions oi the parcels at one hour
              intervals.
                                     79

-------
      A comparison of the mesoscale trajectories with the synoptic trajectories shown in
Figure 22 implies some wind shear aloft, since the air parcels  approach from  different
quadrants.  However, the variability  of the surface winds and the fact that the synoptic
trajectories may be  based solely upon  0700 EST  and 1700 EST winds  aloft make  this
implication speculative.

it.2.5      NO2 Patterns

      NO? concentrations were generally higher on this day at all sites than on other case
study days. Peak concentrations  in excess of  0.10 ppm (including the highest value of the
entire program, 0.17 ppm) were recorded at  urban and  near-urban sites in  the late
morning.   South Broad Street (Site 12) recorded concentrations in  excess of 0.10 ppm  all
day.  These concentrations were  apparently the result of  NO reactions with the very high
ozone levels observed even in the urban areas.

4.2.6      Concentrations Aloft

      From the  above discussion, it  is evident  that areas both  upwind and downwind of
Philadelphia experienced  high  ozone levels.   The site  which  experienced the  highest
observed  levels — Chester, PA  (Site 6) — was not  directly downwind as determined  by
surface flow.   Thus,  in  light  of the high  levels  observed  at the  upwind  stations  at
Ancora, NJ (0.141 ppm), and Lumberton, NJ  (0.125 ppm), possible explanations  are that
there was some contribution of long-range  transport and/or, since high  levels were
observed  the  previous  day,  there was  some carryover  or buildup.   Unfortunately,  no
helicopter data  are available for that day to assess ozone concentrations aloft prior to
inversion  breakup.  However, a  review of  concentrations at upwind sites (taken to be to
the east, in this  case) during the first  hour after thorough mixing can provide an indication
of ozone  levels aloft.  Based upon acoustic sounder  data obtained for other case study
days (no such data are available for July 13), 1000 EST  is a good approximation  of this
first hour of mixing, and  the concentrations shown in Figure 25 are consistent with this.
Ozone  concentrations  of  0.097 ppm   at   Lumberton, NJ   (Site 3),   0.064 ppm   at
Van Hiseville, NJ  (Site 5), and   0.123 ppm  at  Ancora, NJ (Site 11),  were  observed  at
1000 EST.  At the time of  mixing, these ozone levels  were some of the highest  observed
during the program, implying a substantial ozone reservoir aloft on July 13.  The  effect of
                                         80

-------
   0.20
   0.18



   0.16



7 0-14
 5.
 0.

I 0.12
*j
 Ifl
 5 0.10
 u
 5
U
 
-------
mixing of ozone aloft  down to  the  surface  is further illustrated in Figure 25, showing
diurnal profiles  of  ozone at Chester and  several upwind sites.  Very  rapid  increases  in
ozone levels are evident after mixing begins.
                                         82

-------
^•3   JULY 16, 1979 (MONDAY)

     Important  features of the July 16 case include southeasterly flow, confinement of
high ozone concentrations to areas downwind of  Philadelphia,  and an unusual over-the-
water synoptic air  parcel trajectory.  High ozone levels were due primarily to emissions
from the greater Philadelphia urban area.

     The peak  ozone level on  July  16 was 0.151 ppm.  Three of the twelve operating
stations recorded levels higher than 0.12 ppm.  All other sites recorded 0.10 ppm or less.
NC>2 levels were also fairly low.  The peak NO2 concentration was 0.069 ppm with three
stations reporting concentrations in excess of 0.05 ppm.

4.3.1       Synoptic Meteorology

     The synoptic  situation is shown  in Figure 26. A weak, slow-moving cold front was to
the northwest, while the wind flow was dominated by a sluggish low pressure center to  the
southeast, bringing primarily east to  southeast flow  over the study area. Upper level flow
was extremely weak with little pressure  gradient  evident.   Also, thundershowers were
present in the study area in the afternoon.

     The synoptic  trajectory (Figure 27) shows an air parcel  in the 500- to 1000-m layer
arriving from the southeast, resulting in an over-water trajectory, not normally conducive
to high ozone concentrations in  the  study area.  The air parcel  in the 1500- to 2000-m
layer approached more from  the southwest, although it also spent the last hours of its
journey over the ocean.

4.3.2       Mesoscale Meteorology

     Surface airflow over the study area  was characterized by an easterly component
throughout the  day.  During the morning hours, sites near the urban areas, as well as to
the west and southwest of Philadelphia, reported winds primarily from the east. However,
sites to the northeast showed a wider  variation:   some sites  reported winds from  the
southeast, while  others reported winds  from the northeast.  Wind speeds  remained less
than k m/s at all sites prior to 1200 EST. During the afternoon, winds were stronger  and
                                          83

-------
OO
-p-
                             1024  A CHICAGO* mm
                                          FIGURE 26.   Synoptic situation, 0700 EST, 3uly 16, 1979.

-------
Go
                                                                                                                     45'
                            90
                                                                                                              TO"
         FIGURE 27.   Synoptic  trajectory of air  parcels arriving at  Philadelphia at 1300 EST,  July  16, 1979.  Solid line  indicates
                       parcel in  the  500 to 1000-meter layer; dashed line is for a parcel in the 1500 to 2000-meter layer.  Intervals
                       are six hours.

-------
primarily from the southeast  over the entire study area.  By 1900 E5T, some sites were
reporting winds from  the north or northeast.  Winds aloft, as indicated by pibal  data  at
Wilmington, Delaware, and Trenton,  New Jersey, were primarily easterly or southeasterly
through 2000-m throughout the day.

4.3.3       Precursor Patterns

     Morning precursor concentrations were generally lower at  all  sites than those
observed on any other case study day, although no downtown NMHC data were available.
The 6:00 to 9:00 a.m. LOT averages  of NMHC and NOx, along with the ratio of NMHC  to
NO ,  are  given  in  Table 23  for rural and urban  sites.    The AMS Lab  (Site 14)  in
Philadelphia, within 9 km of the  central business district, reported 6:00 to 9:00 a.m. LOT
average concentrations of 0.2 ppm for NMHC and 0.03 ppm for NO .
                                                              A

4.3.4       Ozone Patterns

     Ozone  concentrations were low at all  sites during the  morning  hours except for
Downington, PA (Site 2), west of Philadelphia.  This  site reported ozone  concentrations
exceeding 0.08 ppm by 1100 EST. Through the afternoon, concentrations remained low  at
all sites upwind of the urban area, as shown in Figure 28,  which also shows isopleths of O,
concentrations at  1200, 1300 and  1400 EST.  (Data were obtained for this figure from sites
which did not participate in the study but which were downwind under this  unusual flow
regime.) This figure  shows  a  broad  area of elevated O, extending from the edge of the
urban and  industrial areas downwind of  Philadelphia.   Chester, PA  (Site 6), and Clay-
mont, DE  (Site 15),  although  not   directly  downwind  of  Philadelphia,  reported  daily
maximum ozone concentrations of 0.131 and  0.130 ppm, respectively, at 1400 EST. Since
wind speeds were generally less than 5 m/s throughout the day,  and since  ozone levels
aloft appear to be low on this day (as evidenced by the low daily maximum  concentrations
at the  upwind sites of Lumberton, NJ (Site 3), Ancora, N3 (Site 11), and Van  Hiseville, NJ
(Site 5)  of  0.058,  0.061, and  0.061  ppm,  respectively),  the  high  levels at  Chester  and
Claymont are likely due to industrial/automotive emissions within 20 km of the sites.  The
region of elevated ozone concentrations shown on Figure 28 persisted through 1700 EST.
                                          86

-------
Table 23.    6:00 to 9:00 a.m. LOT (0500-0800 EST) AVERAGES OF NO
            and NMHC (ppm) AND THE RATIO OF NMHC TO NO  FO&
            RURAL AND URBAN SITES ON JULY 16.
Site
No.
1
2
3
5
Rural Sites
Summit Bridge, DE
Downingtown, PA
Lumber ton, NJ
Van Hiseville, N3
6:00 to 9:00 a.m. LOT Averages
NOX
0.00
0.03
—
0.01
NMHC
0.0
0.1
0.1
—
NMHC/NO
Ratio X
—
4.3:1
—
—
Urban Sites
10
13
14
Camden, NJ
Franklin Institute,
Philadelphia, PA
AMS Lab, Philadelphia, PA
0.04
0.04
0.03
—
—
0.2
—
—
6.7:1
                                87

-------
            1200 EST
                                                  1300 EST
          1*00 EST
FIGURE 28.   Isopieths of ozone at 1200, 1300, and 1400 EST, July 16, 1979.
                              88

-------
     The  surface  trajectories  for  this   day  shown   in  Figure 29  indicate   that
Downingtown, PA  (Site 2), was  downwind  of Philadelphia.  An  air  parcel  beginning  at
0600 EST in downtown Philadelphia was slightly northwest of Downingtown, PA (Site  2), at
1500 EST.  (Site 2 recorded the highest study area ozone concentration at this time.)  The
parcel  followed an  eastward path  during the morning until  winds  from the  southeast
became dominant in  the afternoon, then the parcel traveled in a northwest-only  direction.
An air parcel ending  at Chester, PA  (Site 6),  at  1400  EST with  a concentration  of
0.131 ppm was 40 km southeast of Philadelphia at 0600 EST, although it passed over highly
industrialized areas prior to its arrival at Chester.

4.3.5      NO., Patterns

     July 16 experienced the lowest NCX concentrations  of any of the case study  days.
The  peak  concentration  of  0.069 ppm occurred at Chester,  PA  (Site 6),  at  2300 EST,
reflecting the late evening  decomposition  of ozone.  The other two sites  recording levels
greater  than  0.050  ppm  (Claymont, DE and Franklin Institute, PA) also  reported the
2300 EST peaks.

4.3.6      Concentrations Aloft

     Although no  pollutant measurements were  recorded above the surface level on this
day, upwind surface  concentrations recorded at 1000 EST indicate that ozone levels  aloft
were lower than on  the first two case study days discussed.  (No acoustic  sounder data
were available for July 16.   1000 EST was selected  as the first complete hour  of mixing
through  1000 m, based  upon acoustic sounder data  on  other case study  days.)  Observed
concentrations were  0.049 ppm for ozone, 0.000 ppm for NO, and 0.000 ppm for  NMHC  at
Lumberton, NJ (Site  3), and 0.061 ppm of ozone and 0.002 ppm  of NO at Van  Hiseviile, NJ
(Site 5).   Diurnal  profiles  of  ozone  at  Lumberton, NJ (Site 3),  and Van  Hiseviile, NJ
(Site 5),  in Figure  30 demonstrate the relatively low ozone levels observed all day upwind
of Philadelphia. This indication of lower levels  aloft is consistent with  the observation
that high ozone concentrations were much less widespread and  occurred  downwind  of
Philadelphia on July  16.
                                          89

-------
FIGURE 29.   Surface trajectory  for various air  parcels located within the study area on
              July 16, 1979.  Solid  line  indicates a  trajectory beginning  at  downtown
              Philadelphia  at 0600  E5T;  dashed  lines  are  for parcels ending  at the
              indicated site and time.  Dots show the positions of the parcels at one hour
              intervals.
                                       90

-------
  0.18f
  0.16-
s
30.12
  0.10
   O.OS
   0.04
   0.02
   0.00
                                         Van Hiseviile, NJ
                                         /    (Site 5)
        0100
0500
0900         1300
    Time (EST)
1700
2200
       FIGURE 30.  Diurnal profiles of ozone at upwind stations on July 16, 1979.
                                     91

-------
4.4  JULY 19, 1979 (THURSDAY)

     The important aspects  of the July 19  case  were southeasterly flow, the apparently
long residence  time over the New York City area of a low-level air parcel arriving  at
Philadelphia at 1300 EST, indications of high concentrations  of ozone  transported aloft,
and some NO transport at the surface.  Peak ozone concentrations resulted from both the
greater Philadelphia urban plume and transport aloft.

     The peak ozone  concentration  recorded on July 19  was 0.160 ppm.   Six of the
thirteen operating stations  reported  ozone concentrations  equal to  or  greater than
0.12 ppm and an additional four stations reported ozone concentrations equal to or greater
than 0.10 ppm.  The peak NO2 concentration was 0.070 ppm, with six  stations reporting
NC>2 concentrations equal to or greater than 0.05 ppm.

4.4.1       Synoptic Meteorology

     On July 19,  a broad, high-pressure cell dominated the northeastern United States,
including Philadelphia, as shown in Figure 31.  This  high pressure cell  was  to the west
bringing flow with an easterly component to the study area.  A stationary front, which had
been present for several days, was about 150 km  to the southeast. The upper level flow
was predominantly west-southwesterly.

     The 100- to  500-m and 1500-  to 2000-m layer synoptic  air parcel  trajectories
depicting the 96-hour paths of parcels arriving at Philadelphia at 1300 EST are shown  in
Figure 32.  The 100- to 500-m trajectory shows a very erratic path, as might be expected
considering  the long-term  presence of  the stationary front  mentioned  above.   The
trajectory  indicates a prolonged  residence time  in  the vicinity  of the New York City
metropolitan area.

     The 1500- to 2000-m trajectory is  quite different. It  is  a long over-land trajectory
traversing  the  Great  Lakes  area.  Since  the  maximum  mixing height on  July 19  in
downtown Philadelphia  was greater than 2000 m  MSL,  transport at both levels  may be
important.
                                           92

-------
    451
40'
               90
                        FIGURE 131.   Synoptic situation, 0700 EST, July 19, 1979.

-------
FIGURE 32.   Synoptic  trajectory of air parcels arriving at Philadelphia at 1300 EST, 3uly 19, 1979.  Solid line indicates
              parcel in the 500 to  1000-meter layer; dashed line is for a parcel in the 1500 to 2000-meter layer.  Intervals
              are six hours.

-------
4.4.2       Mesoscale Meteorology

     Surface  wind flow  veered from  northerly  to northeasterly to southeasterly from
early  morning to late afternoon.  Several sites (North Philadelphia Airport,  Philadelphia
International  Airport, Greater  Wilmington  Airport,  Delaware, Trenton-Mercer County
Airport,  New  Jersey, and Downingtown, Pennsylvania) reported wind speeds greater than
2.0 m/s  from  0500 EST  through  2100  EST,  indicating some significant  wind flow  was
prevalent all day.  Winds aloft  veered  from  northeasterly to easterly.  The overall wind
flow patterns  were very similar  to those on July 16 discussed previously.

      Mixing heights on this day were the highest of all of the  seven case study days for
which  data were available (see Table 18).  Rural acoustic sounder data show uninhibited
mixing (mixing through  1000 m)  at  1000 EST  with  strong  vertical  mixing generally
confined to lower layers  throughout  the  day.  Urban mixing heights, as determined from
radiosonde observations, were 741 m at 0900  EST and 2314 m at  1500 EST.

4.4.3       Precursor Patterns
      The 6:00 to 9:00 a.m. LOT NMHC and NO  averages along  with the NMHC to NO
                                              A                                     X
ratios are given  in  Table 24  for  rural  and  urban sites.   Urban  morning  precursor
concentrations were quite low and NO concentrations at outlying stations were even lower
(<0.025 ppm), with the exception of Van Hiseville, NJ (Site 5),  75 km east-northeast of
Philadelphia,  which recorded  a 6:00 to 9:00 a.m. LOT  NO  average  of  0.08 ppm, the
                                                          A
highest morning concentration observed at  this  site during  the program.   The high
concentration  of  NO  observed  at Van Hiseville  (Site 5)  and  the synoptic trajectory
discussed previously raise the possibility of surface precursor transport.  This is supported
by peroxyacetyi nitrate (PAN)  data at Van Hiseville on this day, presented in Figure 33.
Concentrations  of  this pollutant,  a product  of  the  photochemical  process,  reached
3.70 ppb at 1000 EST, whereas  the  July average for this  hour is 1.4 ppb.  Early  morning
flow at the surface was from the north or northeast in the northern section of the study
area, which suggests some transport from the  New York City metropolitan area.
                                          95

-------
Table 24.    6:00 to 9:00 a.m. LDT (0500-0800 E5T) AVERAGES OF NO
           and NMHC (ppm) AND THE RATIO OF NMHC TO NO  FOR
           RURAL AND URBAN SITES ON 3ULY 19.
Site
No.
1
2
3
5
Rural Sites
Summit Bridge, DE
Downingtown, PA
Lumberton, NJ
Van Hiseville, N3
6:00 to 9:00 a.m. LDT Averages
N0x
0.02
0.02
0.05
0.08
NMHC
0.3
0.0
0.4
—
NMHC/NO
Ratio X
16.5:1
—
8.6:1
-—
Urban Sites
6
7
8
10
12
13
14
Chester, PA
Bristol, PA
North Philadelphia Airport, PA
Camden, N3
So. Broad & Spruce Streets,
Philadelphia, PA
Franklin Institute,
Philadelphia, PA
AMS Lab, Philadelphia, PA
—
—
0.03
0.07
0.05
0.05
0.03
0.7
1.4
—
—
0.2
—
0.4
—
—
—
—
3.4:1
—
12.3:1
                               96

-------
^.20
0.20-
         0800
1000
  1200
Time (EST)
1400
1600
 FIGURE 33-  Diurnal profiles of PAN at Van Hiseviile, N3 (Site 5) on July
              19, 1979 and July average.
                                    97

-------
i*.i*.i*       Ozone Patterns

     Ozone isopleths  for  the study area  at  1400, 1500, and  1600 EST  are  shown  in
Figure 34.   As on July 16, a day in which the surface study area flow was  from the
southeast, a broad area of elevated ozone was located west and southwest of Philadelphia.
Summit  Bridge, DE  (Site 1),  Downingtown, PA  (Site 2),  Claymont, DE  (Site 15),  and
Chester, PA (Site 6),  all reported ozone  concentration of 0.120 or higher by 1400 EST.
The  highest ozone  concentration of the day  (0.160 ppm) was  found at Claymont, DE
(Site 15), southwest of Philadelphia, at  1600 EST.  During  the day, urban and near-urban
sites also reported high concentrations.   Sites  to the east and southeast reported lower
peak concentrations.

     Surface  trajectories for this day are shown on Figure 35. During the early morning
hours,  winds were from the north or northeast; by 1100 EST, the  flow was from the east,
and during the afternoon the flow was from  the southeast.  A  parcel starting at 0600 EST
in  downtown Philadelphia was about eight  miles  west  of  Downingtown, PA  (Site 2), at
1500 EST when a 0.157 ppm ozone concentration was recorded.  As discussed above, the
maximum ozone concentration recorded in the study area  during this  day was 0.160 ppm
at Claymont, DE (Site 15), at 1600 EST.  Figure 35 shows a parcel ending at this site and
time with a trajectory beginning far to the east making an anticyclonically-curved path to
arrive  at the  site from the  southeast.   The  0600 EST starting  point  was  over an area
closest to Lumberton, N3 (Site 3). Peak morning NO and NMHC  levels at Lumberton, N3
(Site 3),  were  0.024 ppm and 0.5 ppm, respectively,  higher than  normal for this site and
generally higher than other rural sites.

     Mesoscale surface trajectories were  also developed for two cases in  which high
ozone  levels were reported downtown and to the east.  The 0600 EST  position  of an air
parcel  arriving  at  the AMS Lab,  Philadelphia (Site 14),  at  1300  EST, with  an ozone
concentration  of  0.130 ppm, was near the near-urban station  of  Bristol, PA (Site 7). An
air  parcel arriving  at Lumberton, N3  (Site 3),  at  1100 EST with  a 0.104  ppm ozone
concentration  was 45 km  to  the northeast  at  0600  EST near the rural Robbinsville, NJ
(Site 4),  station.   Such elevated ozone  levels at a site upwind  of  Philadelphia  suggest
long-range ozone transport,  particularly in light of the  synoptic  trajectory  shown  in
Figure  32.
                                           98

-------
                1*00 EST
500 EST
            1600 EST
FIGURE 34.   Isopieths of ozone at 1*00, 1500, and 1600 EST, July 19, 1979.
                                  99

-------
                                                            .104  ^^
                                                             Site 3
                                                            1100 E5T
FIGURE 35.   Suriace trajectory for various air parcels located within the study area.on
              July 19, 1979.  Solid line  indicates a  trajectory  beginning at downtown
              Philadelphia  at 0600 EST; dashed  lines  are for  parcels  ending  at  the
              indicated site and  time.  Dots show the positions of the parcels  at one hour
              intervals.
                                        100

-------
4.*.5       NO2 Patterns

     The  peak  NO2 concentration  (0.070 ppm) occurred  at Claymont, DE  (Site 13),
southwest  of  Philadelphia.   This  and  other  sites recording  NG>2  values  in  excess of
0.05 ppm were primarily downtown and to the southwest (downwind), and nearly all peaked
in the late morning.  NC>2  peaks downwind of the urban area in the late morning are the
characteristic  result of the NO/O-j reaction  and preceed the  afternoon  ozone buildup.
Bristol, PA (Site 7), to the northeast, also recorded  NO?  concentration in  excess of
0.05 ppm.   Van Hiseville (Site 5), east-northeast of Philadelphia,  recorded 0.046 ppm at
0800 EST, the highest NCu concentrations recorded there during the field program, a
reflection of the high NO concentrations observed during the hours  preceding 0800 EST.

4.4.6      Concentrations Aloft

     The combination of  ozone transport aloft and  surface  precursor  transport is a
plausible explanation for the  high NO-  levels  observed at Van Hiseville, NJ (Site 5), and
the high ozone levels observed at Lumberton, N3 (Site 3), both upwind of Philadelphia.

     Data from direct measurement of pollutants aloft were not available.  Mixing  height
data collected at Summit  Bridge, DE  (Site 1),  indicated that the first full  hour of mixing
to 1000 m was  at 1000 EST. Ozone and NO pollutant measurements at the upwind sites of
Lumberton, NJ,  and Van Hiseville, NJ  (Sites 3  and  5),  at  this  time  were  0.082 and
0.083 ppm, and 0.005 and 0.003 ppm, respectively.  The diurnal  profiles of ozone and NO
at these two upwind locations presented in Figure 36 show these indications of transport.
The  rapid increase in ozone and concurrent rapid decrease in NO are characteristic of
ozone aloft being mixed down. This suggests  that  high levels  of ozone were transported
aloft and that  there was very little  transport of precursors aloft.  However, as discussed
above, the abnormally high 6:00 to  9:00 a.m. NO  concentrations observed  in the eastern
                                               .X
portion  of the study area implies surface precursor transport,

4.4.7      Hydrocarbon Species Data

     Both samples collected at Franklin Institute and the industrial site had NMHC  below
the overall average on this day, which  is consistent with the fact that mixing heights were
                                          101

-------
   o.os
 a.
 a.


 o
0.06
g 0.04

4)
U


U 0-02

O
z

   0.00
                                             NO
                             Van Hiseviile, N3
                            /     (Site 5)
                               Lumberton, PA
                                   (Site 3)
              0300
                         0700
1100
1300
1900
2300
   0.12
± 0.10,


o


2  0.08
4*
C

-------
fairly  high  in  the morning  hours.   The  species  breakdown  showed  differing  unusual
features:

      1.    The Franklin Institute samples,  in addition  to the  background  of automobile
           exhaust hydrocarbons, showed C7-Co paraffins in unusual amounts.  Methyl
           cyclohexane, which showed large amounts in several samples, was 70 ppbC in
           the 0500 to 0600 E5T sample and somewhat elevated at 0700 to 0800  EST.

      2.    The industrial site samples did  not have unusual amounts  of this hydrocarbon
           but showed unusually high concentrations of  iso-propylbenzene (cumene).  This
           is an industrially important hydrocarbon, so this might reflect a  chance spill of
           this hydrocarbon.
                                         103

-------
4.5   AUGUST 4, 5, 6, 7, 1979 (SATURDAY, SUNDAY, MONDAY, TUESDAY)

      These days represent the only  extended  ozone episode observed during the monitor-
ing period.  August 4  was interesting because it  was a weekend day with high observed
ozone.  The highest concentrations  appeared  to be  transported into the study  area  from
the north, and  a low-level surface trajectory brought  the parcel up along the Atlantic
seaboard.   August 5, another  weekend day (Sunday), had  the following  interesting
characteristics:  (1)  peak ozone  concentrations  directly downwind of Philadelphia;  (2)
indications  of  significant early morning ozone  concentrations  aloft; and (3)  a  long,
over-land synoptic  trajectory.   August 6 was  remarkably similar to August 5 with  similar
ozone patterns, indications of transport,  and air parcel trajectories. August 7 demonstra-
ted significant wind shear, very little early morning ozone aloft, and surface flow reversal
during the day.

      This  four-day  series represented an  extended  episode  period  in  which  ozone
concentrations in excess of 0.120  ppm on each day  were recorded within the study area.
Table 25  gives  the maximum  ozone and NO-  concentrations and the number of  sites
exceeding certain levels on the four days.  This  table shows that, although peak ozone
concentrations  were   fairly  high  on  each  of  the  four days, the  number of stations
experiencing high  ozone  concentrations  did  not  increase from   day  to  day during the
episode; furthermore, high ozone concentrations were less widespread than on most of the
other case study days.  Peak NG>2 concentrations were less than 0.10 ppm, as was the  case
for all but two of ten case study days examined.

4.5.1       Synoptic Meteorology

      Figures 37a through 37d show the surface synoptic progression for the four days.  A
high-pressure cell was centered just to the  southwest of the study area on  August 4 and
became stronger so that most of the eastern half of  the United States was dominated by a
high-pressure ceil during the days following. A surface front situated off the coast of the
northeastern  states on August  4 underwent frontolysis on August 5.   A weak cold front
just north  of the Great Lakes on August 5 moved through the study area on August 6.
                                         104

-------
Table 25.  MAXIMUM  OZONE AND NO? CONCENTRATIONS OBSERVED
         IN THE  STUDY  AREA DURING THE PERIOD  AUGUST *
         THROUGH  AUGUST 7, 1979.
Date
August 4
August 5
August 6
August 7
Ozone
Concen-
tration
(ppm)
.m
.135
.136
.129
Number of Sites >
0.12 ppm
3
2
3
3
0.10 ppm
6
5
5
5
NO2
Concen-
tration
(ppm)
.090
.090
.07*
.080
Number
of Sites
_>0.05 ppm
5
6
5
5
                             105

-------
o
ON
                                          FIGURE 37a.  Synoptic situation, 0700 EST, August 4, 1979.

-------
FIGURE 37b.   Synoptic situation, 0700 EST, August 5, 1979.

-------
                                                                                                     -70°
o
00
                45 °J
            40
                            90
                                                                                                                    45'
                                                                                                                       40"
                                        FIGURE 37c.  Synoptic situation, 0700 EST, August 6, 1979.

-------
FIGURE 37d.  Synoptic situation, 0700 EST, August 7, 1979.

-------
     It is interesting  to  note  that  a similar synoptic pattern  and sequence of weather
events was also associated  with  the only  extended episode situation observed  during  a
similar 1978 study in Philadelphia (Chan et al., 1979).

     The approach of air  parcels  to  the study area became more northerly as the episode
progressed.  Figures  38a through 38d show the synoptic trajectories of air parcels arriving
at Philadelphia at  1300 E5T on August 4 through 7, respectively.  On August 4, the 100- to
500-m and 1500- to  2000-m layer trajectories were quite different. The path followed by
the 100- to 500-m parcel  was erratic and included a long, over-land trajectory along the
Atlantic seaboard. The path followed by the 1500- to 2000-m layer parcel was  overland
and from the west. On the last three days of  the episode, the 100- to 500-m layer parcels
followed anticycionically-curved  paths approaching from the west, northwest, and then
north of Philadelphia.  The lower-level air parcel arriving on August 4 appeared to have
traversed the  high  emission  density areas  of Baltimore  and Washington, D.C., while
parcels arriving on August 5 and 6 did not directly traverse any major source area but
traversed areas near the major  urban centers of Cincinnati,  Cleveland, and the lower
Great Lakes.  The lower-level trajectory ending on August 7 approached through  the New
York City area.

4.5.2      Mesoscale Meteorology

     The general  surface study area flow was north  to  northwest  on August 4  and  was
predominantly from the northwest on August 5 and 6.  On  August 7 the flow was from the
northeast in the morning but from the southwest during the afternoon.

     Upper air wind  flow over the study area  was generally northerly on August 4 through
the morning of August  5, backing to  westerly  in the afternoon of August 5. Northwesterly
flow aloft prevailed on  August 6 and  westerly  on August 7.

     Urban  mixing heights  on August 4 through  6 were  the lowest of any of the seven
case study days for which  data were available (see Table 18).  Mixing heights on August 4
were particularly low.  The  0900  EST mixing height was  157 m  and urban mixing heights
remained under 1000  m most of the day.  The  maximum  urban mixing heights on August 4,
5,  6, and 7 were 1270 m,  1287 m, 1044 m and 1439 m, respectively. Rural  mixing height
                                         110

-------
                                                                                 75'
FIGURE 38a.  Synoptic  trajectory of air  parcels arriving at Philadelphia at 1300 EST, August 4, 1979.  Solid line indicates
              parcel in the 500 to  1000-meter layer; dashed line is for a parcel in the 1500 to 2000-meter layer. Intervals
              are six hours.

-------
N>
        FIGURE 38b.  Synoptic trajectory of air parcels arriving at Philadelphia at 1300 EST, August 5, 1979.  Solid line indicates
                      parcel in the  500 to 1000-meter layer; dashed line is for a parcel in the 1500 to 2000-meter layer. Intervals
                      are six hours.

-------
                 90
                                                           80C
75
FIGURE 38c.  Synoptic trajectory of air parcels arriving at Philadelphia at  1300 EST, August 6, 1979.  Solid line indicates
              parcel in the  500 to  1000-meter layer; dashed line is for a parcel in the  1500 to 2000-meter layer.  Intervals
              are six hours.

-------
FIGURE 38d. Synoptic trajectory of air parcels arriving at Philadelphia at  1300 EST, August 7, 1979.  Solid line indicates
              parcel in the  500 to 1000-meter layer; dashed line is for a parcel in the  1500 to 2000-meter layer.  Intervals
              are six hours.

-------
data, as  determined from the  Summit Bridge, DE  (Site i), acoustic  sounder  records
indicate  that rural mixing heights exceeded  1000 m on ail  days  during  the  episode  and
reached this level between 0900 and 1000 EST.  However, analysis of the  acoustic sounder
traces reveals that the intensity of mixing ranged from weak to very weak on  August 4, 5,
and  6.   The intensity  of  mixing within the mixed layer  is related to  "thermal roots"
appearing on acoustic sounder traces.  Thermal roots appear on acoustic sounder traces as
vertical  spikes that taper with height.  They are typical  of free convection conditions
(after  the breakup of the ground-based inversion) with light winds and strong radiational
heating (Baxter et ah, 1978).  The degree of mixing is a function of both  the darkness  and
vertical  extent of the thermal  roots.  The vertical  extent  of  the  thermal  roots  for
August 4, 5, 6, and  7 after the surface-based inversion was lifted is  given  in Table 26.
The  table shows that, while mixing occurred through the 1000 m level, it was not vigorous
and  thermal roots displayed  little  vertical extent, seldom exceeding 300 m.  Vertical
mixing on August 7 was somewhat more vigorous in comparison to the other case  study
days.

4.5.3      Precursor Patterns
      The 6:00 to 9:00 a.m. LOT  averages of NMHC and NO , along with the ratio  of
                                                           X
NHMC to NO  for rural and urban sites, are given in Tables 27  through 30  for August 4,
5, 6,  and 7,  respectively.   Precursor concentrations were generally lower on these days
than on  other case study days.  With the exception of Camden, NJ (Site 10), urban and
near-urban sites  recorded 6:00 to 9:00  a.m. LOT average NO  concentrations lower than
0.10 ppm on  each of these days.  Camden reached an NO  concentration of 0.10 ppm  on
August  6,  but measured lower  concentrations  on the other  days.   Franklin  Institute
(Site 13) recorded hourly-averaged NO  levels in excess of 0.10 ppm  later in the morning
                                    X
on August 6 and  7.  Much of the NMHC data  at the urban and near-urban sites is missing
for these days. The Air Management Service  Laboratory (Site 14) was the only urban site
that recorded  NMHC  on each of  these  days and  the 6:00 to 9:00 a.m. LOT average
concentrations at this  site were less than 0.50 on each day.  Whereas most sites  recorded
low precursor concentrations,  it  should be noted that Chester, PA, recorded a 6:00  to
9:00 a.m. LOT average NMHC  concentration  of 2.5 ppm on August 4. Precursor concen-
trations  at rural  sites were, as usual, low, although on August 4 Lumberton, N3 (Site  3),
experienced 6:00 to 9:00 a.m. LOT NO and NMHC concentrations  of 0.03 and 0.3 ppm,
                                         115

-------
Table 26.    VERTICAL EXTENT OF THERMAL ROOTS (meters) AS DETERMINED
           FROM ACOUSTIC SOUNDER DATA AT SUMMIT BRIDGE, DE (Site 1).
Date
August i+
August 5
August 6
August 7
Time (EST)
0900
0
180
220
280
1000
180
250
300
400
1100
180
270
300
WO
1200
200
270
320
360
1300
250
240
340
280
1400
240
210
290
300
1500
180
200
240
340
1600
100
140
160
300
                            116

-------
Table 27.    6:00 to 9:00 a.m. LOT (0500-0800 EST) AVERAGES OF NO
            and NMHC (ppm) AiND THE RATIO OF NMHC TO NO  FO&
            RURAL AND URBAN SITES ON  AUGUST t.        X
Site
No.
1
2
3
k
5
Rural Sites
Summit Bridge, DE
Downingtown, PA
Lumberton, NJ
Robbinsville, NJ
Van Hiseville, NJ
6:00 to 9:00 a.m. LOT Averages
N0x
0.01
0.02
0.03
0.02
0.01
NMHC
0.1
0.1
0.3
—
—
NMHC/NO
Ratio
7 . 0: 1
4.3:1
10.0:1
—
—
Urban Sites
6
8
10
12
13
1*
Chester, PA
Northeast Airport, PA
Camden, NJ
So. Broad & Spruce Streets,
Philadelphia, PA
Franklin Institute,
Philadelphia, PA
AMS Lab, Philadelphia, PA
—
0.01
0.05
0.07
0.06
0.0*
2.5
—
—
—
—
0.5
—
—
—
—
—
11.8:1
                                117

-------
Table 28.    6:00 to 9:00 a.m. LOT (0500-0800 EST) AVERAGES OF NO
           and NMHC (ppm) AND THE RATIO OF NMHC TO NO  FO&
           RURAL AND URBAN SITES ON AUGUST 5.
Site
No.
1
2
3
4
5
Rural Sites
Summit Bridge, DE
Downingtown, PA
Lumberton, NJ
Robbinsville, NJ
Van Hiseviile, NO
6:00 to 9:00 a.m. LDT Averages
N0x
0.00
0.01
0.03
0.00
0.00
NMHC
0.0
0.1
0.2
—
—
NMHC/NO
Ratio X
—
10.0:1
5.7:1
—
—
Urban Sites
10
12
13
14
Camden, NJ
So. Broad & Spruce Streets,
Philadelphia, PA
Franklin Institute,
Philadelphia, PA
AMS Lab, Philadelphia, PA
0.0*
0.03
0.03
0.03
_
—
—
0.3
__
—
—
10.0:1
                              118

-------
Table 29.     6:00 to 9:00 a.m. LOT (0500-0800 EST) AVERAGES OF NO
            and NMHC (ppm) AND THE RATIO OF NMHC TO NO  FOR
            RURAL AND URBAN SITES ON AUGUST 6.
Site
No.
1
2
3
14
5
Rural Sites
Summit Bridge, DE
Downingtown, PA
Lumberton, NJ
Robbinsville, NJ
Van Hiseville, NJ
6:00 to 9:00 a.m. LOT Averages
NOX
0.01
0.01
0.03
0.03
0.01
NMHC
0.0
0.1
0.3
—
—
NMHC/NO
Ratio X
3.0:1
7.0:1
5 . 5: 1
—
—
Urban Sites
10
12
13
1*
Camden, NJ
So. Broad & Spruce Streets,
Philadelphia, PA
Franklin Institute,
Philadelphia, PA
AMS Lab, Philadelphia, PA
0.10
0.05
0.06
0.04
—
—
—
0.4
—
—
—
10.0:1
                               119

-------
Table 30.    6:00 to 9:00 a.m. LOT (0500-0800 EST) AVERAGES OF NO
           and NMHC (ppm) AND THE RATIO OF NMHC TO NO  FOR
           RURAL AND URBAN SITES ON AUGUST 7.
Site
No.
1
2
3
4
5
Rural Sites
Summit Bridge, DE
Downingtown, PA
Lumber ton, NJ
Robbinsvilie, NJ
Van Hiseviile, NJ
6:00 to 9:00 a.m. LDT Averages
N0x
0.02
0.03
0.04
0.02
0.01
NMHC
—
0.0
0.2
—
—
NMHC/NO
Ratio x
—
1.0:1
4.2:1
—
—
Urban Sites
6
10
12
13
14
Chester, PA
Camden, NJ
So. Broad & Spruce Streets,
Philadelphia, PA
Franklin Institute,
Philadelphia, PA
AMS Lab, Philadelphia, PA
_
0.07
0.04
0.06
0.05
0.2
—
0.4
—
0.3
__
—
10.0:1
—
6.6:1
                              120

-------
respectively.   These were  almost as  high  as  the  concentrations measured  at  the
near-urban Air Management Service Laboratory (Site 14) on this day.

4.5.4       Ozone Patterns

     The areas of peak  afternoon  ozone on each of  the  four  study days  are shown in
Figures 39 through 42.   On August 4, high ozone levels  (greater than 0.10 ppm) were
recorded  at stations  to the southeast and northeast of Philadelphia.  On August 5, high
levels were measured at Summit Bridge,  DE (Site 1), and Ancora, NJ  (Site 11), southwest
and southeast of  Philadelphia, respectively. On August 6 the high levels were measured at
Lumberton, NJ (Site 3), east of Philadelphia, and Ancora, NJ.  On  August 7, the  highest
concentrations occurred to the east  and southwest.

     A comparison of the areas of  peak concentrations, as shown  in Figures 39 through
42, and the resultant wind directions listed earlier in Table 20, show a general  pattern of
peak ozone concentrations occurring downwind of  Philadelphia,  with  some  interesting
exceptions.

     August 4 is a  case  of  peak  ozone  concentrations  not  occurring   downwind  of
Philadelphia.  Robbinsville, NJ (Site  4), recorded the highest ozone level in the study area
(0.144 ppm) on August 4 despite its  location  northeast of  Philadelphia. At this  site, weak
northerly or  northwesterly flow occurred in the morning hours, whereas stronger north-
easterly flow  was prevalent by 1500  EST. Other sites in the northeast section of the study
area also reported winds  from  the  northeast  in the  afternoon,  whereas   winds  at
Downingtown, PA (Site 2), Chester, PA (Site 6), New Brunswick, NJ (Site 16), and Philadel-
phia International Airport (Site 17) reported winds from the  northwest.  An example of the
afternoon  surface  wind  flow on this day  is  presented  on  Figure 43, which shows  a
streamline analysis at 1500 EST.  This flow pattern set  up an area of convergence just
east of  Philadelphia.   This is  supported by National Weather Service observations  of
towering  cumulus clouds  to the southeast of Philadelphia  during  the afternoon,  and  by
pibal  data which  indicate  that   the   winds  were  northerly  or  northwesterly  over
Wilmington, DE and northerly  to northeasterly over Trenton, NJ throughout the  afternoon.
The convergent wind flow possibly resulted in a merging of  the New York and Philadelphia
plumes  just southeast of  Philadelphia,  which helps  explain the ozone  pattern indicated in
                                         121

-------
                1*00 EST
1300 E5T
                    1600 E5T
                                                      1700E5T
FIGURE 39.  Isopieths of ozone at 1*00, 1500, 1600, and 1700 EST, August 4, 1979.
                                  122

-------
              1400 EST
 1500 EST
            1600 EST
1700 EST
FIGURE ^0.  Isopieths of ozone at 1400, 1.500, 1600, and 1700 EST, August 5, 1979.
                                  123

-------
                [400 EST
1500 EST
               1600 EST
                                                    1700 EST
FIGURE 41.  Isopleths of ozone at 1400, 1500, 1600, and 1700 EST, August. 6, 1979.
                                    124

-------
                 1400  EST
1500 EST
             1600 EST
1700 EST
FIGURE 42.  Isopleths of ozone at 1400, 1300, 1600, and 1700 EST, August 7, 1979.
                                    125

-------
FIGURE 43.  Streamline analysis depicting the surface flow across the
             study area at 1500 EST on August 1, 1979. The numbers
             by the site indicate wind speeds (meters per second).
                           126

-------
Figure 39.    In  particular,  in  light  of  the  observed  low  morning  concentrations  in
Philadelphia  on  this day, it  appears that  high ozone formed   from emission sources  in
northern  New  Jersey  was  transported  across   central  New  Jersey  and  impacted
Robbinsville, NJ  (Site 4),  Bristol, PA  (Site 7),  North  Philadelphia  Airport  (Site 8), and
Lumberton, NJ (Site 3), in the  late afternoon.  Evidence of ozone transported  aloft  is
presented later.

      Mesoscale surface  trajectories  constructed   for selected air parcels on August 4
confirm the above theory. Air parcels traveled generally north to south and Philadelphia
emissions  were advected far to  the  south.   Figure 44  shows  trajectories ending  at
Robbinsville, NJ (Site  4), and North Philadelphia Airport (Site 8) at times of peak ozone
concentrations (1600 and 1900 EST, respectively).  Also shown is a trajectory beginning in
downtown  Philadelphia at 0600 EST.  The  air parcels arriving  at Robbinsville, NJ (Site 4),
and  North Philadelphia Airport  (Site 8) carried ozone concentrations  of  0.144 ppm and
0.130 ppm, respectively.  They originated near the precursor-rich areas of  northern New
Jersey. The  air parcel originating in downtown Philadelphia is advected southward into an
area for which no ozone data are  available, although Ancora, NJ (Site 11), southeast of
Philadelphia, reported a peak concentration of  0.102 ppm at  1700 EST.  Flow reversal at
the end point of this trajectory at about  1700 EST indicates that  the Philadelphia plume
did not continue to  the  south but remained  in the study area for possible contribution to
pollutant concentrations on August 5.

      The highest  ozone concentration recorded on August 5 (0.135  ppm) occurred to the
southeast of  Philadelphia at Ancora, NJ (Site 11),  at  1300 EST.   The surface mesoscale
trajectories  shown  in  Figure 45 show  that  this site was almost  directly downwind  of
Philadelphia.  However, as was noted earlier, there was no  indication  of high  precursor
concentration in Philadelphia on  this  day.   The  6:00  to 9:00 a.m. LOT  average NO
                                                                                    A
concentrations were less than 0.05 at all the urban  and near-urban sites  that reported this
pollutant on  this day (Table 28).  Therefore, the high ozone level measured at Ancora may
have resulted from long-range transport or carry-over from the previous day.  This idea is
supported by the fact  that high ozone concentrations were measured throughout the study
area during the day. Specifically, some high values recorded  on this day were 0.119 ppm
at Summit Bridge, DE (Site  1),  at  1600 EST; 0.105 ppm at Van Hiseville, NJ (Site 5),  at
1100 EST;  0.120 ppm at North Philadelphia Airport  (Site 8),  at  1600 EST; and  0.100  at
                                          127

-------
FIGURE 44.   Surface trajectory for various air parcels located within the study area
              on August 4, 1979. Solid line indicates a trajectory beginning at downtown
              Philadelphia at 0600 EST; dashed lines are parcels ending at the indicated
              site and time.  Dots show positions of the parcels at one-hour intervals.
                                    128

-------
                                                             Site 11
                                                               .135
                                                             1300 EST
FIGURE  45.   Suriace trajectory for various air parcels located within tne study area on
              August  5,  1979.   Solid line  indicates a trajectory beginning  at downtown
              Philadelphia  at  0600 E5T;  dashed  lines  are for  parcels ending  at the
              indicated site and time.  Dots show the positions of the parcels  at one hour
              intervals.
                                      129

-------
Franklin Institute (Site 13) at 1200, 1*00, and  1800 EST.  These sites are located in urban
and non-urban  areas which do  not  appear  to  be  downwind of major  source areas.   The
surface trajectory for Summit Bridge, DE (Figure 45), indicates source areas for these air
parcels to be  primarily  rural.   Pibal  data  at Trenton, New Jersey, and Wilmington,
Delaware,  indicate that  the winds aloft were also primarly westerly or northwesterly
through 2000 m most of the day; however, northerly or northeasterly winds were reported
below  800 m at Trenton  prior to 1100 EST. Therefore, ozone transported aloft probably
originated northwest of the study area.  The magnitude of ozone  levels transported aloft
is discussed below.

     On August 6, the ozone patterns  and airflow  patterns were  very similar to those
observed on August 5.  The peak ozone concentration of 0.136 ppm was again recorded at
Ancora, NJ  (Site 11), again at  1300 EST.  The  mesoscale trajectories shown in Figure 46
indicate that (as on August 5) Ancora, NJ (Site 11), was directly downwind of Philadelphia.
Ozone  concentrations  in  excess  of  0.10  ppm  were  again  recorded  at locations  not
downwind of Philadelphia (e.g.,  Sites 1, 3,  and  5).  The  interesting point about  the
remarkable  similarity between  concentration patterns  on August 5 and 6 is that August 5
was a Sunday and August 6 a Monday, and  the central  business  district  was the apparent
upwind source  area on both days.   However,  there was no indication of  high morning
precursors  on August 5, whereas on August 6, Camden, New  Jersey, measured a 6:00 to
9:00 a.m. LDT average NO  concentration of 0.10  ppm.

     Ozone  patterns on the last  day  of  the four-day episode  revealed high  ozone
concentrations  from the east  to southwest  of  Philadelphia.   The  southwestern-most
station on August 7, Summit Bridge, DE  (Site 1), peaked much earlier than Lumberton, NJ
(Site 3), to the east -- 1300 EST, as compared to 1700 EST.  The mesoscale trajectories
shown in Figure 47 help explain  this phenomenon.  A dramatic flow reversal is indicated at
about  1100 EST.  At this time, air  parcels carried to the southwest were advected back
toward the northeast.  An  air parcel originating downtown at 0600  EST was advected to
near Wilmington,  and  then back up to Claymont, DE (Site 15),  by  1500 EST, where  a
0.120 ppm ozone was recorded.

     The area of origin of the parcel carrying the maximum study area ozone concentra-
tion (0.129 ppm) on August 7, as indicated  in Figure  47, is very near to the area of peak
                                         130

-------
FIGURE 46.   Surface trajectory for various air parcels located within the study area on
              August  6,  1979.   Solid line  indicates a trajectory  beginning  at downtown
              Philadelphia at  0600 E5T;  dashed  lines  are  for  parcels ending  at :he
              indicated site and time.  Dots show the positions of the parcels  at one nour
              intervals.
                                       131

-------
                            PHILADELPHI


                          1500 E5T
FIGURE 47.   Surface trajectory for various air parcels located within the study area on
              August 7, 1979.  Solid line indicates a trajectory  beginning  at downtown
              Philadelphia  at 0600  E5T; dashed  lines  are for parcels ending  at the
              indicated site and time. Dots show the positions of the parcels  at one hour
              intervals.
                                     132

-------
ozone — Lumber-ton, NJ (Site 3).   Morning NO  and NMHC precursor concentrations  at
Lumberton, NJ (Site 3), were  0.04 and  0.2 ppm,  respectively  (6:00  to 9:00 a.m. EOT
average), ievels not usually associated with high ozone concentrations.

1.5.5       NO_2 Patterns

     Peak NC>2 concentrations for the four days (August 4 through 7) ranged from 0.070
to 0.090 ppm.  These peak concentrations occurred in Philadelphia, with the exception  of
August 6,  when   the   peak  NO2  concentration  occurred  east   of   Philadelphia  at
Lumberton, NJ (Site 3).  The NG>2 peak of 0.074 ppm at Site  3  occurred at  the  unusual
time of 0200 EST.  Sites to  the west (Philadelphia) and to the northeast (Robbinsviile, NJ,
Site 4) of Site 3 also reported elevated levels at this time.  Ail  the  peak concentrations
occurring in Philadelphia on  August 4, 5, and 7 were reported in  the late  evening or early
morning hours, reflecting the decomposition of  ozone into NCU  and oxygen.  Daytime NG>2
levels were generally much lower, with rural stations reporting  less than 0.01 ppm.

4.5.6       Concentrations Aloft

     Direct measurement of concentrations aloft by the helicopter was made on all four
study days.   Average ozone  concentrations  aloft  were  determined from  helicopter
soundings at upwind sites and are summarized in Table 31. No early morning soundings
were made on August 4 or 5.

     Transport aloft on August 4 appears to be about average  when compared to other
case study days. Average ozone concentrations of 0.05 to 0.06 ppm were  observed in late
morning soundings made  over  West  Chester, PA,  to  the west  of Philadelphia,  over
Philadelphia, and over  Medford, NJ, to the east of Philadelphia. Moreover, concentrations
aloft  recorded during  transects between West Chester, PA, downtown Philadelphia, and
Medford, NJ, on this flight were between  0.03 and  0.06 ppm.  Surface stations reported
similar levels at 1000  EST, the first hour of uninhibited vertical mixing.   Concentrations
aloft obtained during the  mid-afternoon helicopter  flight present some interesting data.
Evidence was presented earlier in discussing ozone patterns, which indicated that the high
ozone levels recorded at the  surface to the northeast, east, and southeast of Philadelphia
in part resulted from ozone transported from northern New Jersey and New York City.
                                           133

-------
Table 31.    AVERAGE MIXED LAYER OZONE CONCENTRATIONS ALOFT ON
            AUGUST *, 5, 6, AND 7, 1979 (morning concentrations reflect
            levels from the top of that afternoon's mixing layer to the
            top of the layer of ozone depletion).
Date
August *






August 5


August 6






August 7








Time
(EST)
0936
1012
1052
1328
1*10
1*52

0758
0833
0907
0555
0631
0702
0919
09*8
1028
1*18
0557
0628
0656
0855
0925
1020
1328
1*07
1*37
Approximate
Location
West Chester, PA
Downtown
Medford, NO
Downtown
Robbinsville, NJ
15 km SE of
Medford, NO
West Chester, PA
Downtown
Medford, NJ
Pipersville, PA
Pottstown, PA
Downtown
Willow Grove, PA
Downtown
Seabrook, NO
Hammonton, NJ
Pipersville, PA
Pottstown, PA
Downtown
Willow Grove, PA
Downtown
Seabrook, NJ
Downtown
Hammonton, NJ
Seabrook, NJ
Position with
Respect to
Philadelphia
Under
Prevailing
Flow
Crosswind
—
Crosswind

Crosswind
Crosswind

Crosswind
—
Downwind
Upwind
Upwind
—
Upwind
—
Downwind
Downwind
Crosswind
Crosswind
—
Upwind
—
Crosswind
—
Crosswind
Crosswind
Ozone
(ppm)
0.060
0.050
0.060
0.060
0.070
0.110

0.050
0.055
0.060
0.030
0.030
0.025
0.035
0.0*0
0.060
0.090
0.015
0.020
0.015
0.020
0.025
0.020
0.050
0.050
0.060
                                  13*

-------
     Figure 48  shows  ozone   concentrations   recorded  by  the   helicopter  during
mid-afternoon  transects from (a)  downtown Philadelphia  to  Robbinsville, NJ;  (b) from
Robbinsville, NJ, to the site 15 km southeast of  Medford, NJ (Site G in Figure 6); and (c)
from Site G  to  Wilmington, DE, where  the  helicopter landed.  Moderate ozone levels,
between 0.05 and  0.08 ppm were  recorded between Philadelphia  and Robbinsville  and
similar levels were  recorded during a sounding made over Robbinsville.  As the helicopter
transected to the southeast portion of the study area, from Robbinsville to Site G, at an
approximate altitude of 350 m MSL, higher ozone levels, between 0.08  and 0.12 ppm, were
seen; the sounding  made over Site  G revealed  ozone levels above 0.13 ppm at  bout 700 m
MSL.  As the helicopter  flew from  Site G  to Wilmington, DE,  ozone levels  were more
variable than during the two  previous transects, fluctuating between 0.08 and 0.15 ppm.
As  noted earlier,  a zone  of  convergence  characterized  by  towering  cumulus existed
earlier, southeast of Philadelphia at this  time, which helps explain the variation.

     Indications of  transport  aloft on August 5 are ambiguous.  A  mid-morning  vertical
sounding over West Chester, PA, is shown in Figure 49. It shows an average concentration
in the  mixed layer  of  0.05 ppm,  which  is not particularly high.  An inversion based  just
below  500 m  is also apparent with  generally light  northerly winds  at all levels.   The
ground-  level concentration at nearby Downingtown, PA (Site 2), during the first hour of
uninhibited  vertical  mixing  (as determined  from   the  Site 1  acoustic  sounder)   was
0.058 ppm,  corresponding  fairly well  to  the ozone levels  aloft  shown  in  Figure 49.
However,  ground-level  ozone  concentrations   at   Summit  Bridge, DE  (Site  1),   and
Van Hiseville, NJ (Site 5), during the first hour of uninhibited vertical mixing (as deter-
mined  from  the Site 1 acoustic  sounder) were 0.084  and  0.104 ppm,  respectively.  Both
these levels were among the highest  1000 EST concentrations observed at these sites  and
imply high levels of ozone  transported aloft.  While no helicopter data are available for
near these sites in the morning on this day, the helicopter did record occasional values in
excess  of  0.150 ppm  while  transecting from  Philadelphia to  Medford, NJ, at  about
1000 EST, as shown in Figure 50.  Thus, morning ozone levels aloft were not homogeneous
across   the  study  area  and  pockets of  elevated ozone aloft were  apparent.   This
inhomogeneity may have been the  result of the previous day's  convergent wind flow as
discussed earlier.

     Measurements aloft in the afternoon indicate ozone levels of about 0.08 ppm over
central New Jersey, increasing sharply to the southeast and south of Philadelphia.  Very
                                         135

-------
a.
a.
                                          (a)
Ozone Concentration
o o o o
• • « •
0 0 O >—
O .p- OO NJ
Altitude —690 m MSL
11,1
1348 1350 1352 1354
Philadelphia, PA Time

1356
(EST)
	 '
\ \
1358 1400
Robbinsville, NJ
Ozone Concentration (ppm
0 0 O 0
O 0 0 I—
O 4^ 00 K)
(b)
^ 	 -^^

Altitude —350 m MSL
i i i i i iii
1422 1424 1426 1428 1430 1432 1434 1436
Robbinsville, NJ Time (EST") Medford, NJ
                                                (c)
 CL
 Q.

70.12
 o


 20.08
^0.04
o
U

-------
                     O-
                                             NO
-4
in
1_


-------
  0.24
e o.2o
g 0.16

*j
re)


c 0.12
0)
u


O 0.08



o


O


  0.00
      0950        0952

      Philadelphia, PA
0954
  0956

Time (EST)
0958
1000          1002

     Medford, N3
           FIGURE 50. Ozone concentrations reported during helicopter transect

                       from downtown Philadelphia to Medford, N3 on August 5,

                       1979 at an altitude of approximately 550 m MSL.

-------
high concentrations (in excess of 0.200 ppm) were measured aloft during a transect from a
location 15 km southeast of Medford, NJ, to Wilmington, DE, at about 1500 EST.

     On August 6, indications of transport aloft from surface  measurements are  quite
different from helicopter soundings.  Many sites recorded elevated ozone levels during the
first full hour of uninhibited vertical mixing, as determined by the acoustic sounder. At
1000 EST, the ozone levels were 0.080 ppm at Summit Bridge, DE (Site 1), 0.065 ppm  at
Lumberton, NJ  (Site 3),   0.076 ppm at   Robbinsville, NJ   (Site 4),  and  0.074 ppm  at
Van Hiseville, NJ (Site 5).  These were ail  well-above average 1000 EST ozone  concentra-
tions and  indicate significant ozone levels aloft.  Considering that  the flow direction on
August 5 was from the northwest (see Figure 45), it is posible that a reservoir of elevated
ozone was left  aloft to the east and southeast of Philadelphia from  the previous  day.
However,   early  morning  helicopter  soundings  indicate  much  lower  average  ozone
concentrations (about  0.03 ppm) aloft in the northwestern portion  of the study area (see
Table 31).    Transects  made  during the morning  flights also  indicate  low  ozone
concentrations aloft — less than 0.05 ppm. Similar levels were measured at sites to the
northwest  of Philadelphia at  1000 EST;  0.035 ppm   at  Ailentown, PA; 0.038 ppm  at
Lancaster, PA; 0.026 ppm at Reading, PA;  0.046 ppm at Morristown, PA; and  0.021 ppm  at
Bethlehem, PA (these  sites record  ozone  concentrations but did not participate in the
study).     An   afternoon   transect made  at   approximately   430 m   MSL  between
Hammonton, NJ, and Seabrook, NJ, southeast of Philadelphia  (see  Figure  6), at  about
1430 EST  indicated  ozone  concentrations  in  excess  of 0.10 ppm and  a sounding   over
Hammonton, NJ, showed  an  average  ozone  level in  the  mixed  layer  of   0.09 ppm.
Ancora, NJ (Site 11), reported a surface concentration of 0.119 at this time.

     Transport of ozone aloft on August 7 was apparently minimal.  As on August 6, the
early morning helicopter soundings  show  very low  ozone levels aloft  (Figure 51).   This
figure shows vertical profiles of ozone, NO , and  temperature over Pottstown, PA,  along
                                         A
with the vertical wind  profile  obtained  at Robbinsville, NJ  (Site 4). The wind profile is
interesting, considering the dramatic flow reversal  from northeast to  southwest experi-
enced at  the  surface later in  the day.   Average ozone levels observed aloft during ail
morning  soundings were  less  than 0.025 ppm.    Furthermore,  ozone  concentrations
measured  by  the  helicopter while  transecting  between sites  were less than  0.06 ppm
during  the morning and mid-morning flights, and  less than 0.08 ppm  during afternoon
                                          139

-------
                    o.
                                                 NO
Temperature
2500
2000
(7J1500
s
10
-
15(
^

10
>
>
?
>- 5(
3
i i i i i
0    0.04   0.08   0.12
     Concentration (ppm)
0   0.02  0.04  0.06  0.08
     Concentration (ppm)
                                                              i 0
                                                                                                I
                                                                                                  \
                                                                           10
   15    20    25    30
                                                                           Temperature ( C)
FIGURE 51.   Vertical profiles of O,, NO  , and Temperature obtained by the instrumented helicopter
              over Pottstown, PA (bite  J^ at 0628 EST on August 7, 1979. Also shown is the corresponding
              wind profile.  (Key: f = North at 5 m/s)

-------
flights after  uninhibited mixing had occurred.  Thus,  little ozone was available aloft  to
supplement local production.   Ozone levels  observed  at  the surface on this day confirm
the helicopter observations.  The first hour of uninhibited vertical mixing was 1000 EST.
Surface ozone levels observed at this time were 0.047 ppm at Summit Bridge, DE (Site 1),
0.031 ppm  at Downingtown, PA  (Site 2), and  0.042 ppm at Lumberton, NJ  (Site 3),  all
much lower than the levels observed on the previous two days.

4.5.7      Hydrocarbon Species Data

      Although August 4  was  a Saturday, the early  morning  hydrocarbon levels at the
Franklin Institute site were  higher (but only slightly) than  average and also higher than the
industrial site samples.  Again, some departures from  average speciation were seen.  Both
samples  at Franklin Institute  were  quite high in  aromatics, especially  Cg-C,Q.  In the
0700 to  0800 EST sample, this was  balanced by high levels of  C.-Cc paraffins, probably
from a  local source of  gasoline vapor.  In  the early  Franklin  Institute  sample, the
iso-butane exceeded n-butane, which is  unusual.  In most samples, n-butane  exceeds
iso-butane, but n-butane is smaller than  iso-pentane.   But gasoline brands do differ in the
distribution  among  those four species.   The two industrial  site  samples  were nearly
average  with only  a slight increase in  C^ olefins, especially in the 0700  to  0800 EST
sample.

      Also  on August 4, a sample taken at Flat Rock Park at the outskirts  of Philadelphia
at noon  showed 222 ppbC of  NMHC.   This sample  was quite near  the  overall  sample
average.  The propene/acetylene ratio was 0.33; very near the automobile exhaust  value
and  that for unreacted  air.  Most  samples  in  this program were appreciably  higher  in
propene  than this.

      Helicopter samples were also  taken in the late  morning  of August 4 over the river
near  downtown.  The two samples taken were  similar in composition and quite normal,
except  for  the  450-m  altitude  sample which showed  17.5 ppbC  of iso-propylbenzene
(cumene).  As mentioned earlier,  this is an important industrial hydrocarbon, and sporadic
high concentrations may  reflect  chance sampling of a point source.   The  propene/acety-
lene ratios were again near those of  gasoline  engine exhaust.
                                         141

-------
     A sample taken at 1255 EST in northwest Philadelphia on August 4 had  an NMHC
concentration of only 119 ppbC.  Only 15  hydrocarbons were measurable,  but these were
compounds normally present in the largest  concentrations.

     Although peak ozone values  of 0.135 ppm  occurred  on August 5, both Franklin
Institute and  industrial  site samples  had quite low NMHC  in  all four early morning
samples. The type  distributions were very close to the average. The species distribution
was also very close to average. Since it was a Sunday, perhaps the composition of the air
mass was not significantly affected by the  emissions associated with low traffic density.

     Two helicopter samples were  taken on this Sunday over the Delaware River near
downtown at about 0845 EST. The sample  at 457 m had only 10 measurable hydrocarbons,
totaling  59 ppbC.  The type distribution appears abnormal because only one aromatic and
no olefins (except ethylene, which is  not credible) were measurable.  The very low NMHC
suggests that this sample was taken above the inversion.  The mixing height was just over
147 m  at this time.  At 24 m  altitude, the total NMHC was 447 ppbC,  with a  normal
distribution.   The  iso-propylbenzene was only 1 ppbC  compared  to  17.5 ppbC  on the
previous day (August 4).   A sample  taken in Downingtown,  PA at 0600 EST on  August 5
yielded 149 ppbC with no abnormalities. This total is substantially lower than at the other
sites probably because of its distance from heavy traffic areas.

     The airflow on August 6 when the ozone reached 0.136  ppm was from the northwest.
The mixing  heights in late morning  were  about 600 meters.   At  Franklin Institute and
industrial sites, NMHC was near  or below average and the type distributions were within
the normal range.  All but the 0600 to 0700 EST Franklin Institute samples showed slightly
elevated C,-C. olefins.

     Helicopter samples taken at 1000 EST near downtown at 488 m and at 18 m MSL
were quite similar.  In this case, the upper  sample was taken  just below  the mixing  height,
which was 591 m.  The species and type distributions showed no recognizable abnormali-
ties. Iso-propylbenzene was not measurable (or at least not reported).

     At  0600  EST samples were taken above Pipersville, PA (a rural area 52 km north of
Philadelphia) at 457 m  MSL.  Only seven hydrocarbons were measured totaling 23 ppbC.
                                         142

-------
Near the surface about  127 ppbC were found.  The unusual feature of this analysis is that
methyl styrene was the only olefin reported except for 34 ppbC of ethyiene.  This is surely
an artifact, since the higher elevation sample  showed the same concentration.  Acetylene
was also not reported, otherwise this could be light automobile exhaust.

     A sample taken at an unspecified time in northwest Philadelphia had only 102 ppbC
of NMHC  in a normal species distribution.

     Winds on August 7 were light but the mixing heights were fairly high. The NMHC at
0500 to 0600 EST at the industrial site was, therefore, only 272 ppbC  and only 369 at 0700
to 0800 EST.  The early sample at Franklin Institute is missing but the 0700 to 0800 EST
sample was fairly high (949 ppbC).  A somewhat elevated level of higher aromatics (C > S)
may indicate  an intrusion of a  solvent.  Otherwise, these three samples are  quite normal.
An early afternoon sample taken at Flat Rock Park in  northwest  Philadelphia was quite
low  (NMHC = 108 ppbC).   The paraffin  percentage  is  high  probably  because  many
aromatics  were  too small to  measure, but the  butanes  and  pentanes  were readily
measurable.   Four  helicopter  samples  were  taken on this day.   The two  taken over
Pipersville, PA far north of Philadelphia showed only  37 ppbC (at  0606  EST and 457 m
MSL, and 26 ppbC  at 0608 EST and 46 m MSL).  The compounds detected are just those
usually present in largest concentration. The other two helicopter samples were taken
near downtown  during mid-morning  and had  concentrations comparable  to the Franklin
Institute and industrial  site samples (305 ppbC at 457 m MSL at 0936 EST, and 230 ppbC
at 46 m MSL at 0939 EST).  The distributions were normal, but  it may be noteworthy that
the higher altitude gave the higher concentrations.
                                           143

-------
4.6  AUGUST 10, 1979 (FRIDAY)

     The interesting features  of  the  August 10 case were (1) the  implication of source
areas to the southwest, (2) the advection of the central business district's emissions out of
the study area under the influence of strong winds, and (3) the prefrontal nature of the
high ozone levels.

     The peak ozone concentration recorded on August 10 was 0.170 ppm.  Seven of the
fifteen operating stations recorded ozone concentrations in excess of 0.12 ppm  and an
additional five reported levels  equal to or greater than 0.10 ppm. The peak NG>2 concen-
tration was 0.080 ppm with five stations reporting NCu concentrations > 0,05 ppm.

4.6.1       Synoptic Meteorology

     The synoptic situation is  shown  in Figure  52.  The study area was positioned in the
warm sector  of a frontal  system  after a warm frontal passage on the previous day.   The
Bermuda high-pressure cell, centered  just off the Georgia coast, exerted some influence
and, together with a low pressure center moving north of the Great  Lakes, brought south-
westerly flow to the study area.  Afternoon and  evening thundershowers occurred over the
area, and the cold front  associated with this  cyclone passed through Philadelphia the
following morning (August 11) cleansing out the  study area. Thus, the high ozone levels of
August 10 were associated with prefrontal conditions.

     Synoptic  trajectories terminating  at  Philadelphia  on  August 10  are  shown in
Figure 53.  A long,  over-land  trajectory is  evident.   Due to high wind speeds,  the air
parcel in the 100- to 500-m layer passed over  the  Washington, D.C.-Baltimore  area less
than 12 hours prior to its arrival in Philadelphia.

4.6.2       Mesoscale Meteorology

     Winds at all levels  exhibited the highest  speeds of  any  case  study day.  Resultant
wind speeds  were generally > 2.5 m/s at the  surface  all day, while winds  aloft  were
generally about 8.0 m/s,  as compared to 2 to 4 m/s  on other case study days.   Surface
flows were southwesterly in the morning hours with significant horizontal shear in the

-------
                                                I A  1016
                                               1012
                                                 • MONTRE1L
FIGURE 52.   Synoptic situation, 0700 EST, August 10, 1979.

-------
     45 °j
FIGURE 53.   Synoptic trajectory of air parcels arriving at Philadelphia at 1300 EST, August 10,  1979.  Solid line indicates
              parcel in the 500 to 1000-meter layer; dashed line is for a parcel in the 1500 to 2000-meter layer.  Intervals

-------
afternoon.  Winds were southwesterly in the southwestern portion of  the study area with
stations elsewhere reporting northwesterly, all with high speeds.  Winds aloft were more
consistent from the west across the study area.

     Late morning urban mixing heights were lower compared with other case study days,
although afternoon mixing heights were relatively high.  The 0900 EST mixing height (as
determined  by radiosonde  observations) was 339 m, as compared to 1516 m at 1500 EST.
Based  upon  surface temperatures  between 0900  and 1500 EST  and  the 0900 sounding,
mixing heights did not rise above 503 m MSL until 1300 EST and reached a maximum value
of 1516m MSL at  1400 EST. Rural mixing height, as derived from  acoustic sounder data,
did not break  the 1000-m level until 1100 EST, the latest of any case study day.  Vertical
mixing was fair when compared to other case study days,  while horizontal mixing was very
vigorous due to high wind speeds.

4.6.3      Precursor Patterns

     The 6:00 to 9:00  a.m. LOT  averages of  NO   and  NMHC,  along  with the  ratio of
NMHC to NO  , are given in Table 32 for rural and urban  sites.  Morning downtown NMHC
            X
levels  were substantial, as reflected  in the 1.0 ppm 6:00 to 9:00 a.m. LOT  average at
South  Broad Street  (Site 12).  The  corresponding  NOx  average  was 0.050 ppm.  Other
urban sites were above average  when compared to other  case study days, despite the fact
that ventilation was excellent due to the high wind speeds.

4.6.4      Ozone Patterns

     Isopieths of ozone concentrations at 1300, 1400, 1500, and 1600 EST are shown in
Figure 54.  High concentrations (greater than 0.10 ppm) were  recorded at sites east to the
northeast, southeast, and southwest of Philadelphia during the afternoon.  At 1500 EST,
Claymont, DE (Site 15),  reported the highest ozone concentration in  the study area that
day.   It is  interesting  to  note that, with  strong flow from  the southwest,  the  peak
concentrations occurred southwest  of Philadelphia,  suggesting  that emissions  from
Philadelphia were not contributing to those peaks.

     Further  evidence  implicating a  source  area other  than  Philadelphia is  seen  in
Figure 55.   Air parcels tracked backward from Claymont, DE (Site 15), at  1500 EST (time
                                           147

-------
Table 32.    6:00 to 9:00 a.m. LOT (0500-0800 EST) AVERAGES OF NO
           and NMHC (ppm) AND THE RATIO OF NMHC TO NO FOR
           RURAL AND URBAN SITES ON AUGUST 10.
Site
No.
1
2
3
*
5
Rural Sites
Summit Bridge, DE
Downingtown, PA
Lumberton, NJ
Robbinsville, NJ
Van Hiseville, NJ
6:00 to 9:00 a.m. LOT Averages
N0x
0.00
0.01
0.01
0.02
0.00
NMHC
0.2
0.0
0.0
—
—
NMHC/NO
Ratio X
0.0:1
—
—
—
—
Urban Sites
6
10
12
13
1*
Chester, PA
Camden, NJ
So. Broad & Spruce Streets,
Philadelphia, PA
Franklin Institute,
Philadelphia, PA
AM5 Lab, Philadelphia, PA
—
0.04
0.05
0.07
0.08
0.3
—
0.1
—
0.6
__
—
19.4:1
—
7.5:1

-------
              1300 EST
 1^00 EST
             1300 EST
1600 EST
FIGURE 54.   Isopieths of ozone at 1300, 1400, 1500, and 1600 EST, August  10, 1979.
                                    1U9

-------
100C
   FIGURE 55-   Surface  trajectory for various air parcels located within the study area on
                 August 10,  1979.  Solid line indicates  a  trajectory beginning at downtown
                 Philadelphia at  0600  E5T; dashed lines  are  for  parcels  ending at the
                 indicated site and time. Dots show the positions of the parcels at  one hour
                 intervals.
                                           150

-------
of peak  ozone) and forward from the  central business  district  at  0600 EST indicate a
source region far to the southwest and that emissions from the central  business district
are  advected  out  of  the  study  by  1400 EST.   The urban  centers  of Baltimore  and
Washington, D.C., are upwind of the study  area on  this day.   The  high wind  speeds
provided the  necessary  mechanism  for  the  rapid transport of the emissions from these
urban centers into the  study area in less than ten hours.

4.6.5      NQ2 Patterns

      The peak NC>2 concentration  (0.080 ppm)  was observed at the A MS Lab  (Site  14),
during the late morning  hours. South  Broad Street,  PA (Site 12),  another urban site, also
peaked during late  morning.  Such  late morning peaks  are  probably the result  of  the
reaction of high levels of early morning  NO with ozone and occur before  NO_ destruction
by sunlight and reactions with hydrocarbons.  NO2 levels were quite low  at outlying sites
with the exception of Robbinsville, NJ  (Site  4),  which reported elevated concentrations
(0.067 ppm) in the evening (2000 EST) reflecting the scavenging of the day's ozone.

4.6.6      Concentrations Aloft

      Early  morning  helicopter  soundings  show  evidence of  ozone transport  aloft  of
0.06 ppm in the 500- to  2000-m layer. The sounding made near the  area's southwestern-
most site,  Summit Bridge, DE (Site  1),  is shown in  Figure 56.  Peak  ozone levels were
found  at approximately 900 m  MSL  where  the  wind  was from  the west-southwest  at
9.1 m/s. Virtually no NO  was detected aloft, while an inversion based  near the surface is
evident. Ozone concentrations  measured by  the  helicopter while  transecting from West
Chester, PA,  to Philadelphia at about 0630 EST varied between 0.06 and 0.09 ppm. The
mid-morning  helicopter  flight showed higher  ozone levels.  Concentrations between 0.09
and 0.12 ppm  were measured  by the helicopter during a transect between Brandywine, PA
(about 8 km southeast of West Chester, PA), and Philadelphia at about 0916 EST, as shown
in Figure 57.

      Another  indication  of  transport  aloft is that Summit Bridge, DE  (Site  1)  and
Downingtown, PA (Site 2)  reported  0.079  and 0.087 ppm,  respectively,  during  the first
hour of uninhibited vertical mixing (1100 EST).   Diurnal profiles of ozone concentrations
                                            151

-------
                                                  NO.
                                             Temperature
  LO
  0)
  5
     2000
     1500
     1000
      500
                                2500
2000
1500
1000
 500
                              2500
                                                               2000
1000
 500
                                          i      i
         0     0.04   0.08   0.12       0   0.02  0.04  0.06  0.08
             Concentration (ppm)
        Concentration (ppm)
               10    15    20   25    30
               Temperature ( C)
FIGURE 56.   Vertical profiles of O., NO , and Temperature obtained by the instrumented helicopter
              over Summit Bridge, DE (stfeA) at 0535 EST on August 10, 1979. Also shown is the corresponding
              wind profile. (Key:  )• - North at 5 m/s)

-------
a.
^x


o
*j
n)
u

£


§
o
O

ID
o
   0.20
   0.16
   0.08
   0.04
   0.00
            09U

       Ikandywine, PA
                                     0916
                                                 Time (EST)
0918
 0920

Philadelphia, PA
 FIGURE 57.  Ozone concentrations reported during helicopter transect from Brandy wine, PA to Philadelphia

              on August 10, 1979 at an altitude of approximately 850 rn MSL.

-------
at these sites are given in Figure 58. Ozone peaks eariy in the day are apparent at both
sites.  Both sites exhibit similar ozone levels  at the time  of  initial mixing  from aloft.
However,  Summit Bridge concentrations continue to increase in response to  transport
from the southwest, but Downingtown levels do not, suggesting that little of this transport
affects the western portion of the study area.

4.6.7       Hydrocarbon Species

     At   Franklin  Institute,  the  hydrocarbon  species  type  distribution  showed high
paraffins, mostly propane and butanes.  The  C., to C^  olefins  were slightly elevated as
well.  It is not clear  what these should be ascribed to.

     In contrast, the 0500  to  0600 EST sample at the industrial site showed a large
percentage of aromatics; most of  this was benzene (185 ppbC).  This must be attributed to
a local source but it was transitory since the  0700 to 0800 EST  sample was nearly normal
in distribution (20.3  ppbC benzene).  Surprisingly, both samples showed six carbon olefins
(1-hexene  and  C-2-hexene)  although  in  this  case  the later  sample  was  higher  in
concentration.  It is  difficult to explain this since we lack signatures for possible sources.
To maintain perspective it must be kept in  mind that these "intruders" were in no case a
major  part of the NMHC.

     West Chester is a small city about 25  miles west of Philadelphia.  A sample taken at
an unspecified time on August 10 showed  a  high total NMHC of 907 ppbC.   This was
heavily weighted with light  paraffins.  The butanes and pentanes made up about 4396 of
this  total.  This is  probably a  case of  gasoline vapor  since minimal concentrations  of
higher  aromatics were measured and is probably due to a local source.

     An afternoon sample was  taken at 1300 to 1400 EST at  the industrial site.   The
distribution was  normal  with a total NMHC  of  390 ppbC.  It  is surprising  that the two
aromatic  olefins —  styrene  and  methyl styrene -- were the  only  olefins,  except  for
propene, reported in this sample.

     At  0621  EST on  this date, a sample  was taken at 488 m  MSL near West Chester,
Pennsylvania.  Only  99 ppbC was  reported in  a  normal pattern  although neither ethylene
                                           154

-------
0.18
                                              Summit Bridge, DE
                                              /     (Site  1)
            0300
0700
1100         L500
   Time (E5T)
1900
2300
     FIGURE 58.  Diurnal profiles of ozone  at Summit Bridge,  DE (Site  1) and
                  Downingtown, PA (Site 2) on August 10,  1979.
                                     155

-------
nor acetylene was listed.  At about 0940 EST, samples were taken over downtown at two
altitudes.   These were similar in composition, the upper  sample taken just  below the
mixing height of 339 m, showing 875 ppbC NMHC at 488 m MSL, and the lower sample
showing 702 ppbC at 18 m MSL.  Both showed high percentages of paraffin due to high
propane concentrations.  Propene levels were also slightly elevated.
                                         156

-------
4.7  AUGUST 22, 1979 (WEDNESDAY)

     Some  of the  interesting features of  this day  were:   (1)  high  ozone leveis in
Philadelphia proper,  as well as downwind, (2) the evidence of significant wind shear aloft,
and (3) a low-level  synoptic  trajectory with a  long residence time  over  the  Northeast
Corridor.

     The maximum  ozone and NO2 concentrations reported on August 22 were 0.140  and
0.100 ppm, respectively.  Four of the fourteen operating  stations  reported ozone concen-
trations  equal to or greater  than  0.12 ppm  and one other station reported 0.100 ppm.
Seven sites reported NCX concentrations > 0.05 ppm.

4.7.1       Synoptic  Meteorology

     The synoptic  situation  on  this  day is  illustrated  in  Figure 59.   Although, like
August 10, the wind flow was predominantly from the southwest, the synoptic situation
was  quite different.  A  low-pressure center had passed through the study area from  the
northwest on  August 21  with  precipitation in the study  area  on  August 20  and 21.  On
August 22, a  high-pressure ceil  was rebuilding  over the  area.  The  center of this high
pressure was over southern Quebec at 0700 EST on August 22, moving offshore toward  the
southeast.

     The synoptic trajectory for  an air  parcel in the  100-  to  500-m  layer, shown in
Figure 60, is somewhat puzzling.  While surface flow over the study area was consistently
from the southwest and, as will be shown later, surface study area trajectories confirmed
this  southwesterly  flow, the  trajectory of an  air parcel  in  the  100- to 500-m  layer
approached from the northeast during the  12 hours preceding 1300 EST. Winds aloft data
show wind shear but only at  very  high  levels (i.e., 1500 m MSL).  Prior to the erratic
behavior and  long residence  time  in  the vicinity of major  Northeast Corridor  urban
centers,  the  synoptic trajectory  shown in Figure 46 exhibits a  long, over-land, anti-
cyclonic  approach to Philadelphia.  A trajectory  for a parcel in the 1500- to 2000-m layer
approached more directly from north-northwest.
                                           157

-------
00
                                                                                                                        40°
                                                                                         75'
                                       FIGURE 59-   Synoptic situation, 0700 EST, August 22, 1979,

-------
                                                                                                                45'
FIGURE 60.   Synoptic  trajectory of air parcels arriving at Philadelphia at 1300 EST, August  22, 1979.  Solid line indicates
               parcel in the 500 to  1000-meter layer; dashed line is for a parcel in the 1500 to 2000-meter layer. Intervals
               are six hours.

-------
4.7.2       Mesoscale Meteorology

     The prevailing surface wind flow was generally southwesterly all day, although some
variability was noted during the  late morning hours, particularly at Sites 7, 8, and 21.
These  sites  showed  some  tendency toward northerly winds, although  the  prevailing
directions for the  day were south to southwest.  Winds aloft were  northwesterly  in the
early morning backing to southwesterly by 1500 EST with northeast winds at 1500 m  MSL.

     Urban mixing heights  were generally  higher than  on  other  case  study days.  The
maximum afternoon mixing  height was 1707 m MSL at 1500 EST.  Rural acoustic sounder
data indicate mixing through the 1000 m MSL layer by  0915 EST,  with  excellent vertical
mixing throughout the day.

4.7.3-      Precursor  Patterns
     The 6:00 to 9:00 a.m. LOT averages of NMHC and  NO. along with the  ratio of
                                                           J\
NMHC to NO  , are given in Table 33 for rural and urban sites.  Morning precursor levels
            X
were  above  average  throughout the  study  area.   The 6:00  to  9:00 a.m. LDT  NO
concentrations at ail sites  near  the central business district were above 0.05 ppm; even
Robbinsville, NJ  (Site 4),  experienced  NO   concentrations  three  times  the  normal
average.  NMHC levels were also somewhat  elevated with South Broad Street  (Site 12),
and the  AMS  Lab (Site 14) reporting  6:00 to 9:00 a.m. LDT levels  of 1.1  and 0.7 ppm,
respectively, resulting in 6:00 to 9:00 a.m. LDT NMHC/NOx ratios  of 12.2:1 and 5.4:1,
respectively.  Summit Bridge, DE (Site 1), Downingtown, PA (Site 2), and Lumberton, N3
(Site 3),  reported  higher-than-normal  NMHC  concentrations — 6:00  to  9:00 a.m. LDT
averages of 0.4, 0.3, and 0.3 ppm, respectively.

4.7.4      Ozone Patterns

     Peak study area  ozone levels were observed at stations in Philadelphia and to the
east and northeast.  The AMS Lab  in  Philadelphia (Site 14)  recorded  the highest ozone
concentration  of the  day with 0.140 ppm at 1400 EST.  Figure 61 shows ozone  patterns
over the  study area at 1400 and 1500 EST. Unfortunately, an area-wide investigation of
afternoon ozone concentrations  in the  study area is plagued  by missing afternoon data at
                                          160

-------
Table 33.    6:00 to 9:00 a.m. LOT (0500-0800 EST) AVERAGES OF NO
            and NMHC (ppm) AND THE RATIO OF NMHC TO NO  FO&
            RURAL AND URBAN SITES ON AUGUST 22.       X
Site
No.
1
2
3
*
5
Rural Sites
Summit Bridge, DE
Downingtown, PA
Lumberton, NJ
Robbinsville, Nd
Van Hiseville, Nd
6:00 to 9:00 a.m. LDT Averages
NO
X
0.02
0.02
0.0*
0.09
0.01
NMHC
0.*
0.3
0.3
—
—
NMHC/NO
Ratio X
20:0.1
13.5:1
7.5:1
—
—
Urban Sites
8
10
12
13
1*
Northeast Airport,
Philadelphia, PA
Camden, NJ
So. Broad & Spruce Streets,
Philadelphia, PA
Franklin Institute,
Philadelphia, PA
AMS Lab, Philadelphia, PA
0.07
0.15
0.09
0.06
0.13
—
—
1.1
—
0.7
—
—
12.2:1
—
5.^:1
                                 161

-------
            EST
1500 E5T
FIGURE 61.   Isopieths of ozone at 1*00 and 1.500 EST, August 22, 1979.
                            162

-------
most  northeasterly stations.  Concentrations upwind (to the southwest) are low.  None of
the stations to the southwest, west, or southeast of Philadelphia reported ozone levels in
excess of 0.10 ppm at any time during the day.

      Mesoscale  surface  trajectories  for  two air  parcels  of  interest  are shown in
Figure 62.  An air parcel arriving at the AMS Lab, Philadelphia (Site 14), at 1400 EST with
0.140 ppm  of ozone  passed through Philadelphia  the previous  hour and over precursor
source areas along the Delaware  River earlier in  the morning.  The 0600 EST emissions
from  Philadelphia's central  business district were advected  slowly to the northeast and
were  in the vicinity of Bristol, PA (Site 7), by about  1400 EST.  Unfortunately, no ozone
data  for that  site is available  past  1200  EST when a concentration of 0.099 ppm  was
recorded.  Nearby, North Philadelphia Airport (Site 8) reported 0.130 ppm  of  ozone at
1400 EST.    The  mesoscale  trajectories  show that surface advection  was  similar in
direction to  August 10 but  wind speeds were much lower.  Still, source areas other than
Philadelphia, but  within the study  area, are again suggested as  the explanation  of high
ozone levels.

4.7.5       NO2 Patterns

      The highest  NO^ concentrations of the day occurred in Philadelphia and Camden, NJ
(Site  10), in  the  early  afternoon (e.g.,  0.10 ppm at South  Broad  Street  (Site 12) at
1300 EST)  due  to  the availability of both NO and ozone for  NO-, production.  The peak
concentration of 0.095 ppm  at  Camden, NJ (Site 10), occurred at 0700 EST and  corres-
ponded  to  a  drop in  NO concentration from 0.179 ppm (the  highest  level of the  day) to
0.062 ppm  and  preceded a large increase in ozone levels.  Other concentrations in excess
of 0.05 ppm occurred in the morning at outlying sites.

4.7.6       Concentrations Aloft

      Although vertical  pollutant  profiles  aloft  are lacking,  some  indication of  early
morning  ozone and precursor  transport aloft can once  again  be  found by  a review of
surface  concentrations during  the  first  complete hour of  uninhibited  vertical  mixing
which, according  to the acoustic sounder at Summit Bridge,  DE (Site 1), was 1000 EST.
The diurnal profile of ozone at  Summit Bridge, DE, and Downingtown, PA, is shown in
                                          163

-------
FIGURE 6Z.   Surface trajectory for various air parcels located within the study area on
              August 22, i979.  Soiid line  indicates a trajectory beginning at downtown
              Philadelphia  at 0600  E5T;  dashed  lines are  for  parcels ending  at  the
              indicated site and time.  Dots show the positions of the parcels at  one hour
              intervals.

-------
Figure 63.  The rapid increase in ozone levels characteristic of  ozone mixing down from
aloft  can be  seen, especially at Summit Bridge.  Concentrations (ppm) at these sites at
1000 EST were:
Location
Summit Bridge, DE (Site 1)
Downingtown, PA (Site 2)
Ozone
0.077
0.062
NO
0.000
0.000
NMHC
0.50
0.10
      These data indicate substantial levels of ozone aloft.  The 0.077 ppm ozone level
observed  at Summit  Bridge, DE (Site 1), at  1000 EST was  the maximum for the day.
Although  substantial NMHC is present (0.50 ppm, an unusually  high level  for  1000 EST),
indicating at least some of the raw material of further  ozone production is available, NO
and NO  are lacking.  The apparently substantial levels  of  transported ozone  and  low
maximum  concentrations  upwind of Philadelphia are puzzling.  Unfortunately, no actual
measurements  of ozone aloft are available.
                                            165

-------
   0.1*


   0.12
~ 0.10
 o
 •M

   0.08,
 u
   0.06
 c
 o
O 0.04
   0.02
   0.00
                           Downingtown, PA
                          j     (Site 2)
                                         Summit Bridge, DE
                                              (Site 1)          \
                                                                           \
               0300
0700
  1100
Time (EST)
1900
2300
        FIGURE 63.  Diurnal profiles of ozone at Downingtown, PA (Site 2) and
                    Summit Bridge, DE (Site 1) on August 22, 1979.
                                     166

-------
                                  5. CONCLUSIONS

     The analysis of the ten case study days discussed in the previous chapter has led to
the following findings:

           Ozone levels aloft  in the early morning  on  days of high surface ozone vary
           from 0.015 to 0.097 ppm.

           Significant nighttime ozone  transport aloft can be a major cause of high late-
           afternoon surface ozone concentrations in  urban and rural areas.

           Peak ozone levels do not always occur in the Philadelphia urban plume.

           The  ozone transport phenomenon in the  Northeast Corridor involves  local,
           mesoscale, and synoptic scale transport.

5.1  MORNING OZONE LEVELS ALOFT

     During the ten case study days, early morning  ozone levels observed aloft were as
low as 0.015 ppm and as high as 0.097 ppm.   Table 34 summarizes  the  ozone concentra-
tions  aloft  observed generally upwind of Philadelphia for  the  case study  days.   Two
methods were  used  to estimate these concentrations aloft:  (1)    the  actual  average
concentration measured  aloft by helicopter above the  layer of surface ozone depletion and
below the top of the  afternoon mixed layer; and (2) the average  concentration observed at
the surface during the first full hour  of complete vertical mixing through  the 1000-m
layer, as determined by the acoustic sounder  located at Summit, DE (Site 1).  An assumed
complete vertical mixing time of  1000 EST, based upon the average  time  of complete
mixing on other case study days, was used for 3uly 12, 13, and 16, due to the unavailability
of acoustic  sounder  data.  Average concentrations detected aloft  ranged from 0.015 to
0.020 ppm on August 7 to 0.060 ppm on August 10. Ozone concentrations measured  at the
surface at  the  time of complete  vertical  mixing ranged from 0.031 to 0.047 ppm on
August 7 to  concentrations in excess of 0.080  ppm on July 13, 19, and August 10.
                                          167

-------
Table 34. SUMMARY OF MORNING OZONE CONCENTRATIONS (in ppm) OBSERVED GENERALLY UPWIND
         ON CASE STUDY DAYS.
Date
July 12

July 13

July 16

July 19

August 4

August 5

August 6


August 7


August 10

August 22

Concentrations Aloft
Site
Location
_.

—

—

—

West Chester, PA

West Chester, PA

Pottstown, PA
Pipersville, PA
Willow Grove NAS, P
Pottstown, PA
Pipersville, PA
Seabrook, NJ
Summit Field, DE
West Chester, PA
—

Time
(EST)
__

—

—

—

0930

0758

0631
0555
A 0919
0628
0577
1020
0535
0612
—

Concen-
tration
_.

—

—

—

0.060

0.050

0.030
0.030
0.035
0.020
0.015
0.020
0.060
0.060
—

Surface Concentrations
Site
Location
Summit Bridge, DE
Downingtown, PA
Lumberton, NJ
Van Hiseville, NJ
Lumberton, NJ
Van Hiseville, NJ
Lumberton, NJ
Van Hiseville, NJ
Downingtown, PA
Robbinsville, NJ
Downingtown, PA
Robbinsville, NJ
Downingtown, PA


Summit Bridge, DE
Downingtown, PA

Summit Bridge, DE
Downingtown, PA
Summit Bridge, DE
Downingtown, PA
Site
Number
1
2
3
5
3
5
3
5
2
4
2
4
2


1
2

1
2
1
2
Time
(EST)
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000


1000
1000

1100
1100
1000
1000
Concen-
tration
0.068
0.075
0.097
0.064
0.049
0.061
0.082
0.083
0.058
0.067
0.058
0.076
0.061


0.047
0.031

0.079
0.087
0.077
0.062

-------
     Concentrations observed at the surface at  the time  of complete mixing tend to be
somewhat higher than  concentrations measured aloft earlier.  Although the bulk of ozone
production occurs during the late morning  and early afternoon, some ozone is probably
produced  between  the time of  the helicopter  sounding  and  the  time of  the  surface
concentrations shown in Table 34.

     The measures of transport  presented in Table 34 correspond to  the geographical
extent of peak ozone  concentrations in the study area.   High transported levels imply
widespread levels later in the day.  July 13, 19,  and August 10 were the only case study
days on which surface  concentrations higher than 0.08 ppm were observed during the first
hour of uninhibited vertical mixing.  In addition, August 10 was the only case study day
during which average  ozone concentrations of  0.06 ppm were reported aloft in the early
morning.

     A  review   of  peak  surface ozone  concentrations  shows  that July 13,  19,  and
August 10 were also the  only case study days in which high ozone  levels were widespread
across the study area (i.e., more than five sites with concentrations > 0.12 ppm and ten or
more sites with  concentrations > 0.10 ppm).   Although the sample size is limited, the
implication  is that significant  nighttime ozone transport  aloft can be a major cause of
high late-afternoon surface ozone  concentrations  in urban and rural areas.

5.2  OZONE LEVELS  IN THE PHILADELPHIA URBAN  PLUME

     During  the  ten case study days, peak  ozone levels did not always  occur within the
urban  plume. Table 35 summarizes the locations of  peak  ozone concentrations observed
with respect to  Philadelphia.   On five of the  ten case study days,  peak  concentrations
were found  downwind  of  Philadelphia under prevailing flow.   On one occasion, the peak
concentration occurred upwind,  on one occasion  in Philadelphia, and on two occasions in
other areas.  It should  be pointed out that, while fourteen to fifteen monitoring sites were
generally operating in the study  area, the peak  concentration due to  the urban plume
could have been missed, particularly when  northerly flow  prevailed.  The frequency with
which peak ozone concentrations occur outside  the urban plume is surprising. Some of the
conditions which contribute to this phenomenon were found to be:
                                           169

-------
Table 35.  LOCATIONS OF PEAK OZONE CONCENTRATIONS WITH
         RESPECT TO PHILADELPHIA.
Date
July 12
July 13
July 16
July 19
August *
August 5
August 7
August 10
August 22
Peak Ozone Concentrations (in ppm)
Downwind
0.12*
0.123
0.151
0.157
0.102
0.135
0.120
0.129
—
Upwind
0.11*
0.1*1
0.067
0.10*
0.123
0.082
—
0.170
0.095
Philadelphia
Proper/
Cam den
0.111
0.161
0.100
0.138
0.130
0.120
0.122
0.137
0.1*0
Other
Areas
0.1*7
0.183
0.131
0.160
0.1**
0.119
0.129
0.12*
0.120
                            170

-------
     o    Surface  advection  of  photochemical products from  other  uroan  centers into
          the study area (August 4 and 10).

     o    Very light or  variable surface winds which cause  high ozone levels in or near
          Philadelphia Duly 13 and August 22).

     o    Surface  transport from  other high  emission  density areas in  the study area
          (July 12).

     Of  the  five  days during  which  peak  ozone  levels  were  found downwind  of
Philadelphia, one day (August  7) experienced complete surface  flow reversal so that  no
true upwind area existed.  For the remaining  four days, the difference between upwind
and downwind ozone peaks  ranged from 0.044 to 0.084 ppm.  For this  very small sample,
this 0.04 to  0.08 ppm range represents the  additive effect of the urban plume, while the
late morning surface ozone levels representing  ozone  aloft (Table 34) were in the 0.05 to
0.08 ppm range.

5.3  MESOSCALE TRANSPORT

     Mesoscale  surface trajectories developed  for the case study days revealed both local
(0 to  50 km) and mesoscale (50  to 100 km) surface  transport.  Table 36 presents the
locations  of surface  source  areas  for  peak  ozone  levels  and  areas  impacted  by
Philadelphia's central business  district emissions.   These locations were determined from
surface trajectories.  The central business district's emissions could be directly related to
study area peak ozone levels on three days:  July 15  and August 5  and August 6.  (It  is
interesting to note  that August 5 was a Sunday and August 6  a Monday.  Downwind ozone
levels are virtually identical for the two days.)  A source region outside the study area
was apparent on August 10 and suggested on August 4. High emission density areas within
the study area  were possible source regions on July 12 and August 22. The  source areas
for the remaining  three case days (July 13, 19, and  August 7) were  rural or near rural
regions.  July 13 and 19 showed evidence of high  ozone levels aloft,  while August 7 was  a
puzzling case with a very pronounced flow reversal.
                                            171

-------
            Table 36.  SOURCE AREAS FOR PEAK OZONE LEVELS AND AREAS IMPACTED BY EMISSIONS FROM
                     PHILADELPHIA'S CENTRAL BUSINESS DISTRICT AS DETERMINED BY MESOSCALE SURFACE
                     TRAJECTORIES.
Date
July 12

July 13

July 16

July 19

August 4

August 5

August 6

August 7

August 10

August 22

Peak
Ozone
(ppm)
0.147

0.183

0.151

0.160

0.144

0.135

0.136

0.129

0.170

0.140

Location
Ancora, NJ

Chester, PA

Downingtown, PA

Claymont, DE

Robbinsville, NJ

Ancora, NJ

Ancora, NJ

Lumberton, NJ

Claymont, DE

AMS Lab, Philadelphia

Source
Area
(0600 EST)
10 km SE of
Wilmington, NJ
25 km ESE of
Philadelphia
Central Business
District
15 km SE of
Lumberton, NJ
55 km NNE of
Robbinsville, NJ
Central Business
District
Central Business
District
5 km SW of
Lumberton, NJ
Baltimore/
Washington, D.C.
Claymont, DE

Area
Impacted by
Central
Business
District
20 km S of
Robbinsville, NJ
20 km NE of
Downingtown, PA
5 km NW of
Downingtown, PA
10 km W of
Downingtown, PA
40 km SSW of
Ancora, NJ
25 km SSE of
Ancora, NJ
20 km SE of
Ancora, NJ
15 km W of
Chester, PA
>100km NEof
study area
5 km NE of
Bristol, PA
Approximate
Ozone
Concen-
tration
(ppm)
0.12

0.12

0.15

0.16

0.12

0.14

0.14

0.12

	

>0.10

N)

-------
     Ozone levels  found  downwind  of the central business district were,  in all cases, in
excess of 0.10 ppm.  (No data were available for August 10 during which  the  central
business  district's emissions were advected out of the study area.)  On four of the case
study days, peak ozone levels downwind of the central business  district were more than
0.03 ppm lower than the peak level observed in the study area.

5.*  SYNOPTIC TRANSPORT

     While the  case study days were selected to  provide a variety of prevailing flows, it
is interesting  to  note that these flows were spawned  by a  variety of synoptic situations
and that synoptic trajectories  approached the study area in a variety of ways.

     While most of the  trajectories examined showed over-land paths,  July 16 and  19
showed over-water paths.  Six  of the  ten  case study days  demonstrated  significant
anticyclonic curvature in  the trajectory.  Two of the days examined (July 19 and August 4)
demonstrated  practically no relationship between trajectories in the 100- to 500-m  and
1500-  to 2000-m layers.

     The three  days on  which the highest morning  ozone levels  were observed which
suggested synoptic transport were July  13 and 19, and August 10.  The synoptic trajec-
tories, particularly low-level trajectories, showed evidence of tracks which passed upwind
over or near major urban areas.

5.5  DAYS RECOMMENDED  FOR iMODEL VERIFICATION

     To properly evaluate the effects of ozone control strategies on ozone concentrations
downwind of  Philadelphia, cases  in which  ozone levels occur  downwind  of  Philadelphia
must be selected for modeling.   To give confidence  to the  results, the  model must be
validated on  these types  of days.  Of the ten  case study  days analyzed, the study area
peak levels occurred unambiguously  downwind of Philadelphia on three days (July 16, and
August 5 and  6).

      To consider the  effects of  transported  versus locally-produced ozone, July 13,  19,
and  August 10 are candidates for  model verification.   July 19 is especially suitable since
                                          173

-------
very  high  ozone  levels  were recorded  downwind  of  Philadelphia,  while significant
transport aloft resulted in widespread high ozone levels.

      Finally, for the most precise definition of boundary and initial conditions aloft, the
days with helicopter data must be considered:  August 4 through 7, and August 10.
                                            17*

-------
                                6.   BIBLIOGRAPHY


Baxter,  R., G.Y.  Lowe, and T. Nguyen (1978):  Acoustic Radar Chart  Preparation and
     Interpretation.  AeroVironment Inc., Pasadena, California.

Chan, M.W., D.W. Allard, and I. Tombach (1979):  Ozone  and precursor transport into an
     urban area:  evaluation of measurement approaches.  EPA-450/4-79-039.

Code of Federal Regulations, Volume 40, Title 50, Appendix F.

Federal Register (1978): Volume 44, Page 8221-8233, February 8.

Hancock, C.W. (1978):  Philadelphia oxidant study,  1978 helicopter  platform  data report.
     Northrop Services, Inc., Report No. E5G-TR-78-15.

Maynard, 3.B.,  and W.N.  Sanders  (1969):   Determination of  the  detailed  hydrocarbon
     composition and  potential atmospheric  reactivity  of  full range motor  gasolines.
     3APCA 19, 7 (505-510).

Mayrsohn, H., 3.H. Crabtree, M. Kuramoto, R.D. Sothern, and  S.H. Mano (1977): Source
     reconciliation of atmospheric hydrocarbons 1974. Atmos. Environ. _H_ (189-192).

Reynolds, S.D., T.W. Tesche, and L.E. Reid (1978):  An  introduction to the SAI Airshed
     Model and its usage.  Systems Application, Incorporated, Report No. EF78-53R3.

Sanders, W.N. and  3.B. Maynard  (1968):   Capillary Gas chromatographic method for
     determining the C^-C,-, hydrocarbons in full  range  motor gasolines.  Anal. Chem.
     40, 3 (527-535).   J   u

Stephens, E.R. (1973).   Hydrocarbons in polluted air summary report.   Coordinating
     Research Council Project CAP A-5-68.

Westberg, H.  and P.  Sweany  (1980):  Philadelphia oxidant  data enhancement study:
     hydrocarbon  analysis.    Washington   State  University  Chemical  Engineering
     Department, February.
                                          175

-------
                                   TECHNICAL REPORT DATA
                            {Please read Instructions on the re:iene before
  REPORT NO.
  EPA-450/4-81-011
                                                          13. RECIPIENT'S ACCESSION N <_>
<». TITLE AND SUBTITLE
                                                         -  5. REPORT DATE
  Philadelphia  Oxidant Data Enhancement Study •
  Analysis and  Interpretation of Measured Data
                                                       March 1981
                                        6. PERFORMING ORGANIZATION CODE
 . AUTHOR(S)                                                  8. PERFORMING ORGANIZATION REPORT NO.

 Douglas Allard, Michael Chan, Chris Marlia, and Edgar Stephens  ! AV-FR-80/588R
  PERFORMING ORGANIZATION NAME AND ADDRESS
  AeroVironment, Incorporated
  145 Vista Boulevard
  Pasadena, California 91107
                                                           10. PROGRAM ELEMENT NO.
                                        11. CONTRACT/GRANT NO.

                                          68-02-3332
12. SPONSORING AGENCY NAME AND ADDRESS
  Monitoring and Data Analysis  Division
  Office of Air Quality  Planning  and Standards
  Environmental Protection  Agency
  Research Triangle Park, North Carolina 27711
                                         13. TYPE OF REPORT AND PERIOD COVERED
                                           Final
                                         14. SPONSORING AGENCY CODE
 15. SUPPLEMENTARY NOTES
  Project Officer: Norman  C.  Possiel
 16. ABSTRACT
 The Philadelphia
 ozone, precursor,
 The data will be
 1) to verify and
 their use, and 2)
Oxidant Data Enhancement  Study was conducted to obtain  a  data base of
 and meteorological measurements in the vicinity of  Philadelphia, PA.
used by the U.S.  Environmental  Protection Agency for two  purposes:
apply various photochemical  models and to help develop  guidance for
 to better understand  the  photochemical and meteorolonical  processes
 associated with peak ozone  and  N02 concentrations in the  Philadelphia area.  Philadel-
 phia was selected for  study because it is large enough to  be  capable of generating
 high ozone levels, and  it may be  subject to significant ozone and/or precursor trans-
 port.  During the period of July  2, 1979 through September 18,  1979, AeroVironment,  Inc
 (AV) operated five surface  monitoring stations in the vicinity  of Philadelphia to
 collect the required data.   Eleven other surface monitoring stations operated by State
 and local agencies collected data for inclusion in the data base.   An instrumented
 helicopter, operated by  EPA's Environmental  Monitoring and Support Laboratory in
 Las Vegas, collected data aloft during a four-week period  in  July and August.  In
 addition, upper-air meteorological  measurements were obtained by Beukers Labs.  This
 report describes the data collection efforts of AV and includes an analysis and
 interpretation of the  field study data base.  The analyses are  primarily directed
 toward identifying the  concentrations of ozone and precursors transported  in the city
 and ozone formed in the  urban plume downwind.	
 17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS
                                                         COSATl Field/Group
     Ozone Measurements
     Non-methane Hydrocarbon Measurements
     Oxides of Nitroaen Measurements
     Transport of Pollutants
     Urban Plumes
     Trajectory Analyses
 18. DISTRIBUTION STATEMENT
                            19. SECURITY CLASS fThis Report!
                                                                         21  NO O F D A G e 5
                                                                           187
                                              20 SECURITY CLASS tThis paqei
                                                      122. PRICE
 EPA Form 2220-1 (Rev. 4-77)   PREV.OUS EOIT.ON 15 OBSOLETE

-------