United States
                  Environmental Protection
                  Agency
Atmospheric Sciences
Research Laboratory
Research Triangle Park NC 27711
                                                                                i\
                  Research and Development
EPA/600/S3-87/045  Feb. 1988
SEPA         Project Summary
                   Nonmethane Organic Carbon
                   Concentrations in Air Masses
                   Advected Into  Urban Areas  in
                   the  United States
                  Hal Westberg and Lome MacGregor
                    The purpose of this study was to
                  measure  non-methane  organic
                  carbon  (NMOC) levels aloft in the
                  upwind vicinity of several U.S. cities.
                  An  instrumented aircraft  was
                  employed to collect samples during
                  the  months of July and August in
                  1985 and 1986. NMOC, ozone and
                  oxides  of nitrogen concentrations
                  were measured during the  0400 to
                  0900 period in  background air
                  masses advected  into Dallas-Ft.
                  Worth,  Tulsa,   Atlanta,  and
                  Birmingham in 1985 and Philadelphia
                  and New  York  City in 1986.
                  Measurements were concentrated in
                  the  layer above the morning surface
                  inversion  but below the  normal
                  afternoon mixing height. This is the
                  air mass that would eventually mix
                  with the surface  layer within the
                  urban area after the break-up of the
                  morning inversion. Special emphasis
                  was  placed  on NMOC speciation.
                  Individual   hydrocarbons were
                  identified by gas  chromatography
                  (GC) and  derivatized carbonyl
                  compounds were measured by high
                  performance liquid chromatography
                  (HPLC). Details  concerning the
                  experimental design  and  quality
                  assurance aspects of  the  program
                  are provided  within the report
                    A  tabulation of airborne  data
                  collected during the two year  study
                  period is provided. A data  summary
                  is presented for  each day which
                  includes a flight map, table of ozone
                  and NOX concentrations, a  listing of
 NMOC information, and data recorded
 during aircraft  ascents and/or
 descents. A brief comparison of the
 NMOC results  obtained for the
 various cities investigated is  also
 provided.
   This  Project Summary   was
 developed by EPA's Atmospheric
 Sciences  Research  Laboratory,
 Research  Triangle Park,  NC, to
 announce key findings of the research
 project that is fully documented in a
 separate report of the same title (see
 Project Report ordering information at
 back).

 Introduction
   Characterization of  NMOC levels in
 transported air masses is currently of
 interest in terms  of oxidant production
 and the formation of acidic species in the
 atmosphere. Organic species in the layer
 above a morning surface inversion and
 below the afternoon mixing level are of
 interest because oxidant  precursors in
 this  layer mix  with urban plumes
 following  breakup of  the surface
 inversion. Most photochemical models
 that are designed for regulatory purposes
 incorporate oxidant precursors from aloft.
 Recent modeling studies have indicated
 that  control  requirements  are  quite
 sensitive to the  concentration  and
 composition of organics entrained  from
 aloft. Therefore,  it is important that
 NMOC concentrations  in  this layer be
 defined.
   NMOC concentrations aloft have  been
 determined in several oxidant field study

-------
programs,  however, the emphasis in
these previous  studies was centered
around urban plumes.  Prior to the field
work described  herein,  there  had not
been a study with the primary objective
being to define NMOC levels transported
into urban  areas from aloft.   Organic
species  were measured in  air masses
advected into Dallas, Tulsa, Atlanta, and
Birmingham during the summer of 1985;
Philadelphia and New York City during
the summer of  1986.  In this report we
will summarize  the  types  and
concentrations of organic  species
measured aloft.  Relationships  between
the organics and  other  pollutant  and
meteorological parameters  will  be
described  as well.    The  areas
investigated in 1985 were isolated urban
areas whereas those investigated in 1985
were part of a complex urban area where
major upwind sources may exist under
certain meteorological conditions.

Experimental Procedures
   The aircraft  flew arcs upwind  of  the
urban areas during the early morning
hours.  On most days, there were three
arcs flown at two different altitudes above
the morning surface inversion but below
the normal afternoon mixing  height.
Figure 1  provides a schematic illustration
of the flight pattern.   Sample collection
involved  transferring air from a ram-air
manifold inside the aircraft  to the
collection medium.   For hydrocarbons,
this entailed filling  a metal canister to a
positive  pressure with  a metal bellows
pump.   Carbonyl compounds  were
trapped by passing air through adsorbent
cartridges impregnated with dinitrophenyl
hydrazine.  The hydrazones that formed
were analyzed  by HPLC.   An  attempt
was made to measure PAN levels aloft
by collecting air samples in Tedlar bags
and measuring  the PAN content by EC-
GC immediately after the aircraft landed.
Analytical  difficulties  along with  the
absence of a  reliable field calibration
standard compromised the quality of the
PAN data  and  consequently  it  is  not
included in this  report.
   Ozone, NOX, temperature and relative
humidity were  measured continually in
the aircraft.   These  parameters were
employed to define the morning mixing
level  as well  as to provide additional
information  for modeling  the urban
photochemistry.
   A complete  description  of the air
quality  instrumentation used, calibration
methods data processing and validation
procedures are  included in Appendix G
of the report.
  Winds reported at  3000  feet were
obtained  from  the local flight service
station.   This information was  used to
plan  the  location of  upwind  flight paths
each morning.
Results and Discussion

  Table  1  provides a summary of the
average  non-methane hydrocarbon
(NMHC)  and  aldehyde  concentrations
measured aloft over the six cities.  Mean
hydrocarbon concentrations  varied
between  14  and 39  ppbC  with  the
highest average concentrations recorded
upwind of Tulsa.  The  low mean value
determined in Birmingham  may not be
truly representative of  that particular
urban area.   The morning flights  were
plagued with bad flying weather while in
Birmingham and consequently  fewer
samples were  collected.  The  mean
NMHC  concentration  for  all  cities
combined was 22 ppbC with a coefficient
of variance  of approximately  40%.
Paraffinic species were by far the major
class of  hydrocarbons  measured aloft.
These  compounds  comprised
approximately  78%  of the identified
hydrocarbons while  the mean  aromatic
contribution was about 17%  and the
olefins averaged 3%.
   Formaldehyde  and  acetaldehyde
levels in Atlanta and  Birmingham were
generally below the detection limit of the
analytical procedure.  In Dallas,  Tulsa,
Philadelphia and New  York City the two
carbonyls accounted for about 10% of
NMOC that was identified.   A positive
correlation  between  NMHC    and
aldehyde concentrations was  generally
observed.  For example, NMHC levels
upwind of Dallas on the morning of July
6,  1985  averaged  30 ppbC  with  a
corresponding  mean  aldehyde
concentration of about 2 ppb   On  July
14,  Dallas  NMHC levels were down to
about 4 ppbC  and  the  aldehyde
concentrations  ranged  from  below the
detection limit (-0.1 ppb) up to 0.7 ppb
(see Table 2).
   The data base makes it possible to
examine the  variability  in  NMHC
concentrations aloft during the morning
hours since sampling arcs were flown at
different  distances and altitudes upwind
of the urban area.   Generally, one or
more samples were collected along each
of  three  upwind  arcs.  Figure  2
graphically illustrates the daily variations
in NMHC levels observed in each of the
six areas.  The horizontal lines in Figure
2 represent measured NMHC levels in
individual samples.  The vertical  lines
connect  maximum and  minimum  levels
each day. On most days, the high NMHC
measurement exceeds the low value  b
at  least a factor of  two  and on severe
occasions by as much as a factor of ter
As can be seen  in  Figure 2, on  day
when the NMHC concentration range wa
large,  one sample  exhibited a muc
higher  hydrocarbon  level than others
The reason for  this  anomalous behavic
is  not  immediately  obvious.  The  hig
readings  showed  no consisten
correlation with  altitude or distance  fror
the city.
   NMHC compositions  were generall
consistent from  city to   city; howevei
influences from  selected source type
were periodically observed. For exampl
in  Tulsa occasional elevated levels of th
£2 io £4 paraffinic  hydrocarbons  wer
observed, possibly  from the refmerie
located upwind of the  city.  In Atlant
relatively large concentrations of toluen
were occasionally observed. The upwin
source for  this compound was  nc
identified.
   Also of interest is the correlation  (
NMOC aloft  with   other  pollutant!
Generally speaking, there appeared  to b
a  direct relationship between NMO'
concentrations  with  ozone  and  NC
levels.    For  example,  Table
summarizes NMOC,  03 and  NOX  dal
collected upwind  of  Dallas  on  th
mornings of July 6  and   14,  1985.  It
evident that elevated NMOC levels  o
July  6  are associated  with  ozone  an
NOX   concentrations  which   ar
considerably higher  than those recorde
on July 14.
   Philadelphia  and  New York City  wei
chosen for a preliminary examination i
possible correlations between  weath<
patterns and  oxidant precursor levels
the air advected from  aloft.   Becaus
they  are geographically  close,  it
expected that  they would  receive  a
masses  from  the same source  area
under  westerly and  northwesterly flo
conditions
   Two sets of days with similar mornir
wind speed and direction are compare
in Table 3.  In  the  first group,  whic
includes two days  with westerly  wine
exceeding 20  kts,  oxidant  precursi
levels are nearly identical. The maximu
ozone levels recorded at surface statior
in each of the urban areas on July 14 ar
August 8 were very comparable, as we
A similar pattern  exists  for the two da]
listed  with westerly  winds  less than  1
kts.    Precursor  levels  are  vei
comparable as  are  the surface  ozor
concentrations  which in this case  a
elevated  when  compared  to the  hi<;
wind conditions of the  first group.  Tl
fact that higher surface  ozone levels a

-------
                        Wind Direction
                                                                                        Afternoon
                                                                                          Mixing
                                                                                          Height
                     Figure 1.    Schematic diagram of aircraft flight pattern.
                  Table 1. Mean NMHC Concentrations Measured Aloft
                         City
Sampling Dates
Identified NMHC  Formaldehyde   Acetaldehyde
    (ppbC)          (ppb)           (ppb)
Dallas
Tulsa
Atlanta
Birmingham
Philadelphia
New York City
July 5-14, 1985
July 16-24, 1985
Aug. 13-23, 1985
Aug. 26-29, 1985
July8-Aug. 14, 1986
JulyS-Aug. 14, 1986
23
34
25
14
16
19
1.7
3.5
<0.7
<0.7
1.8
1.9
0.2
0.9
<0.3
<0.3
0.7
0.7
recorded  under  more  stagnant
meteorological  conditions agrees with
expectations.
   Figure  3  shows the average NMOC
concentrations measured aloft during this
study  and  compares them  to  the
concentrations  measured  at  a rural,
surface  site  near  State  College,
Pennsylvania and on the ground in
Philadelphia. As  is  readily apparent,
NMOC concentrations at the  rural  site
are comparable to those found aloft and
much  less  than  those  measured in
Philadelphia.  The NMHC levels recorded
in  the  urban area  are on the order of
twenty times  those  observed in  the
background air while the aldehyde levels
are three to four times that found aloft or
at  the rural site.  Since the rural surface
NMOC levels are comparable to those
collected aloft,  it  appears that  surface
data could  be utilized  as  boundary
condition  values  for  modeling  urban
   oxidant production when aircraft data is
   not available.
   Summary and Conclusions

     The NMOC results reported for the six
   cities studied indicate that the air above
   urban  areas for the most part represents
   an aged air mass that is depleted in the
   more  reactive  hydrocarbons.  The
   average NMHC  composition  contains
   78% paraftinic,  17%  aromatic and 3%
   olefmic compounds. The  organic
   carbonyl levels in 4 of the 6 urban areas
   contributed on the average of 10% to the
   total NMOC levels.  The carbonyl levels
   at the other  two sites  were  below
   detectable  limits  of the  analytical
   procedure.  On some flights the NMOC
   levels and  compositions  were  quite
   consistent on each of the three sample
   arcs flown. At  other  times, however,
   NMOC levels and composition changed
   significantly.  A  positive  correlation
                         between  NMHC  and  organic  carbonyl
                         levels was generally observed. The mean
                         NMHC concentration for all cities was 22
                         ppbC with a coefficient of variance of
                         40%. Synoptic meteorological conditions
                         are  likely  to  have an  impact on  the
                         NMOC concentration  level  observed in
                         the background air. From the limited data
                         available, it  appears that the NMOC
                         levels measured aloft agree reasonably
                         well with those measured at remote
                         ground  level locations.  More  data  are
                         required to confirm  this relationship.

-------




0
<3
5
0
a:
5


100 -
90 ~
80 ~
70 ~
60 ~

50 ~

40 ~

30 ~
20 ~
10 ~
n










I .
•








I
I .
•
Atlanta









m






: I
                    \               \




0
-Q
CX
EX
Cj
3:
|








100 -1
90 ~
80 ~
70 ~
60 ~


50 ~

40 ~
30 ~
20 -
10 ~


Birmingham






• •


I
I
1 I
1 1 \ \
co ^ QQ oj
CN 
-------




o
-Q
&
Cj
1
700~
90-
80-
70-
60 ^

50-
40-
30~
20~
10~
Tulsa






1 -
* ~





-,-




i I -




I
                    \       \      \      \
                                          Date
  700
   90 ~
   80 ~
   70~
%  60-
|5°J
|  40~\
   30
   20 H
   10 ~\
    0
                                       /Vew York
T i
 !•
CM f
•X
rs r-

i
•»
M O
x \
s 00

\
§
\
oo
Date
\
§
\
00

\
o
N
00

l^
»*,
\
00

\
^
\
oo

\
«»
»-.
X
CO

Figure 2.
            Continued.

-------
             100-

             90-

             80-

             70-
         O
         •Q
                                                Philadelphia
§: 50-
o
| 40-
^ 30-
20-
w-



I
\
§
\





I
\
<3

rs




I
\

(S




1 I
* 1 ;






' I .
1 1 1 ' 1 1 ^
s;?10^^^^?
v v^\\\\ \
^^^K^ri^^
Date
    Figure 2.    Continued.
Table 2.  Summary of Airborne Data Collected Upwind of Dallas on July 6 and 14, 1985


   Date    Arc (mi)     Alt (ft)
                                NMHC
                                ppbC
                    Formaldehyde    Acetaldehyde
                        (ppb)	(ppb;	03 (ppbj   NO* (Ppb)
July
My 14
              15
              20
              25
             25
             20
              15
2000
3000
2000
2000
3000
2000
22.4
36.7
15.3
                                 3.8
                                  4.7
1.5
1 6
1.4
              0.1
              0.3
              0.7
0.5
07
<02
              <0.2
              <0.2
               0.7
                                                                           -50
                                                                           -20
                                                                                        15
                                                                                       ~5

-------
Table 3. Oxidant Precursors and Wind Data

                          Wind
Date
07/14
08/11
07/16
07/22
City
P
NY
P
NY
Speed
(Ms)
25
22
8
7
Direc-
tion (deg)
290
270
290
290
Ident. NMHC
(PpbC)
20 1
21 8
15.5
15.6
Aldehyde
(ppb)
2.8
2.7
50
3.3
NO, (ppb)
3
4
3
3
Peak Surface
03 (ppb)-
62
64
114
107
P = Philadelphia.
NY  = New York.
The O3 data was taken from SAROAD sites 310720003F01 for 7/14, from 330280002F01 for 8/11, from
311760002F01 for 7/16, and from 313260001F01 for 7/22.
500 -
400 •<
300 -

200 -
700 -
n -



ppbC


X/>TvN.\
~i^S/.rs
r~ss, r
'/•'ft'/ ^
Ip
l^
£/
'%//,







                          NMHC
                                   Aloft
                                                ppb
 Formaldehyde

Rural
   Acetaldehyde

Urban
        Figure 3.    Comparison of average NMOC concentrations measured at 11) a rural surface
                    site in Pennsylvania, /2) urban Philadelphia, and (3) aloft in the vicinity of Atlanta.
                    Birmingham, Dallas, Tulsa, New York City and Philadelphia.

-------