The Isotopic Composition of Atmospheric
Carbon Monoxide
FINAL REPORT TO:
Coordinating Research Council
Thirty Rockefeller Plaza
New York, New York 10020
and
Air Pollution Control Office, EPA
Principal Investigator:
-------
The Isotopic Composition of Atmospheric
Carbon Monoxide*
C. M. Stevens, L. Krout , D. Walling and A. Venters
Chemistry Division
and
A. Engelkemeir and L. E. Ross
Chemical Engineering Division
Argonne National Laboratory
Argonne, Illinois 60439
*This work supported by Coordinating Research Council--U.S.
Environmental Protection Agency and under auspices of the U.S.
Atomic Energy Commission.
^Present address: 9721 S. 50 Court, Oak Lawn, Illinois 60453.
Present address: Standard Oil Co., P. 0. Box 400, Naperville,
Illinois 60540.
Present address: 506 S. Wheeler Avenue, Joliet, Illinois 60436
-------
-2-
ABSTRACT
The concentration and the carbon and oxygen iso-
topic composition of atmospheric CO of the northern
hemisphere show regular seasonal variations. The iso-
topic pattern is different from that for CO from auto-
mobile engine combustion whether compared to combustion
in the same region as the air sample or to an estimated
world average. These results indicate that there are
several natural sources, one of which is much greater
than anthropogenic emissions. The most likely nature
of these sources are discussed and the production rates
estimated.
-------
-3-
1. Introduction
In recent years, several investigations have revealed that
carbon monoxide may have a more active and complicated role in
nature than had generally been known. Swinnerton et al. [1] and
Seller and Junge [2] have discovered that the oceans may be a
significant source of CO. Seller and Junge [3] have shown that
the tropopause is a sink for CO and Inman et al. [4] found that
soils may also be a significant sink. Junge et. al. [5] have dis-
cussed the global atmospheric budget of CO, assuming only anthro-
pogenic and ocean water sources.
The abundance of the stable isotopes of CO might be a way of
distinguishing CO from different origins and thus identifying the
origins of atmospheric CO. Both carbon and oxygen have quite
variable isotopic compositions in nature [6-8]. At almost every
stage of any natural process involving these elements, there are
isotopic fractionation effects. The combination of isotopic vari-
ations for two elements having different cycles might be of special
diagnostic value.
A program of measuring the isotopic composition of CO in
the atmosphere and from natural sources has been going on at Argonne
since 1969. The atmospheric studies include two separate problems:
(1) the determination of the global-average isotopic composition
of CO emitted by internal combustion engines, and (2) the determi-
nation of the isotopic pattern of CO in the least polluted atmo-
sphere (mostly in Illinois) at different seasons and, where possible,
different latitudes and times of day.
-------
-4-
2. Experimental Procedure
The atmospheric samples were collected in 87-liter aluminum
cylinders which had been evacuated by a mechanical pump connected
through a liquid N- trap. A cylinder was filled with air to a gauge
pressure of 2-3 atm with one or two small rubber-diaphragm compres-
sors, operated from a 12V automobile battery. These compressors
produce less than 0.01 ppm CO.
The atmospheric samples were processed to convert CO to C02
and separate this C02 from all other constituents of air for pre-
cision isotopic analysis. The chemical and mass spectrometric
analytical methods have been previously described in detail by
Stevens and Krout [9]. To summarize this procedure, each sample
is passed through a processing train that first removes moisture,
atmospheric CO2, NO, NO and N02, next oxidizes the CO to CO with
Schutze reagent (I-O ), then purifies the CO,, of CD-origin by freez-
ing with liquid N? and pumping, and finally manometrically measures
the CO,). In the oxidation of CO by I^O^ at room temperature, the
original oxygen atom of CO is retained in the resulting CO,.,, and
the other oxygen atom in CO,, (which is acquired from the 10 in
^ O
the oxidation process) has the same isotopic composition for each
sample. The method is quantitative in the yield of CO oxidized to
C02/ does not produce CO2 from any other known atmospheric compounds
containing carbon, and produces CO2 of CO-origin free from impuri-
ties that could cause large errors in either the concentration or
isotopic analysis.
The samples are isotopically analyzed for both carbon and oxygen
by the usual high-precision technique using a modified Consolidated-
i
Nier isotope-ratio mass spectrometer with a semiautomatic data-
-------
-5-
collecting system. The oxygen isotopic abundances are given rela-
tive to the accepted oxygen isotopic reference - Standard Mean
Ocean Water (SMOW) . To determine the absolute 18O/160 ratios of
the original atmospheric CO, CO was first quantitatively prepared
from SMOW by reaction with graphite at 1000°C; then this CO was
oxidized to CO with I2°[- bY tne same method as was used for the
atmospheric samples. The C02 from both the SMOW and atmospheric CO
samples have one atom of oxygen designated O , contributed by the
I2°5 with a constant isotopic abundance*; hence, the value of the
TO T £
O/ 0 ratio in the original CO can be expressed as the per mil dif-
ference 5R relative to SMOW oxygen as
6R = 2i
ro TR - FT
; ( coo }
L SMOW
X 1000 ' (1)
where (R___t) is the mass-46/mass-44 ratio of the C0_. from oxida-
COO x ^
tion with I~0r of the sample, and (R__.^t) is the 46/44 ratio of
2 5 COO SMOw
CO2 prepared from oxidation (with IpO,.) of CO prepared from SMOW H2
The carbon results are given as per mil differences SR of the
IT 12
C/ C ratio R, relative to PDB** carbon R , where
6R = || 1] X 1000 - A(180/160) X 0.033 . (2)
1° J
The second term [A( O/ 0) X 0.033] is the correction for the dif-
ference in the O/ 0 ratio between the sample C0? of CO-origin
and C0? prepared from PDB carbonate by the usual method using 100%
*0 = 5.8%o relative to SMOW by comparison between (R^__
and (R ) C°°
COO
**The Peedee belemnite used as an isotopic standard for carbon.
-------
-6-
H3PO4 where 6(18O/160) is the difference in per mil of the O/ O
ratio of the CO sample relative to CO,, prepared from PDB carbonate
^ ^
IP ~\ f\
However, since the O/ O ratio is measured relative to SMOW oxy-
gen as described above, the O correction term must take into ac-
count the difference in the 17O/16O ratio between CC>2 from PDB
carbonate and (COC)t)SMOW- In the (CO°1) molecule' the
ratio of one oxygen atom from SMOW H2O is -41% 0 and the other (from
the I2°5 oxidation) is -35.2%0 relative to C02 from PDB. Hence,
the above expression becomes
6R = f| -- l) X 1000 + 1.25 - 6(180/160) X 0.0165 , (3)
\R0 j
where 6 ( O/ 0) is the O/ O ratio difference of the CO relative
to SMOW as determined in equation (1). The external error is es-
timated to be ±0.01 ppm for the CO concentration, ±0.3%0 for the
13C/12C ratio, and ±0.5%0 for the 18O/160 ratio.
Before July 1970 when the oxygen isotopic analyses were started,
about 60 samples were collected and analyzed for CO concentration
13 12
and the C/ C ratio. During the early part of this period, analy-
sis with the Argonne 100-inch mass spectrometer showed that the C00
from the CO contained up to 1% NO. The results on these early
samples were corrected on the basis of measurements of the N O/CO
<£ £
ratio. In addition, some of the samples were purified by treatment
with hot CuO. Meaningful values of the concentration and of
13 12
6 ( C/ C) were obtained for about half of these samples and are
plotted as solid or open squares in Fig. 4 in the section on
atmospheric CO; the other half was unused because of known contami-
nation with engine CO or because the sample was too small or was
lost before the N20 impurity was measured. In July 1970 the
-------
-7-
molecular-sieve trap in the processing train was improved so that
the amountof NO in the final C0~ from atmospheric CO was less than
3 X 10 times the atmospheric NO concentration.
3. Urban Atmospheric Carbon Monoxide
Figure 1 shows the 6(13C/12C) and 6(18O/160) ratios for CO in
urban air collected for a number of major cities in the United States
and Europe. The variable 13C/12C ratios reflect the variable 13C/ C
ratio in petroleum. This is illustrated in the upper portion of
IT I?
Fig. 2, where the 6( C/ C) values of the urban atmospheric CO are
compared with the values for various petroleum sources. In these
samples, taken so that the CO represented averages for many auto-
mobiles, the C/ C ratios in CO samples from San Francisco and
Los Angeles had high values corresponding to the high values for
California oil and low values for Chicago CO corresponding to low
values for Texas sources [7].
The specific objective of the first part of this study is to
determine the carbon isotopic composition of world-average engine CO
in order to compare it with average atmospheric CO. Appendix I gives
13 12
a detailed account of our method of estimating the C/ C ratio
for both U.S.-average and world-average engine CO. The world-average
value is found to be <5 ( C/ C) = -27.4%0. This is admittedly a
crude estimate, but the error is probably not greater than ±0.3%0.
The average measured isotopic composition of oxygen in urban CO is
6( O/ 0) = 24.6%0, which is 1.1%0 higher than the value for at-
mospheric oxygen [14].
4. Data on Non-Urban Atmospheric CO
Figure 3 shows the isotopic patterns of surface atmospheric
-------
CO for minimum pollution conditions at locations upwind from Chicago
(solid points) and at several distant locations in the southern and
northern hemispheres (crosses) in 1970 and 1971. The majority of the
Illinois samples were collected about two miles west of Plainfield,
Illinois, on days when the wind direction ranged from south to north-
northwest. The nearest towns upwind from this location are 5-10
miles away and have populations of less than 1500. Other Illinois
locations (and the corresponding wind directions) were rural sites
near Buffalo Grove (NW to N) and Wilmington (SE to S). The Adler
planetarium promontory on the Chicago lakefront, although not a
rural location, was used on days when the winds ranged from north-
east to southeast.
Figure 4 shows the seasonal variation of the concentration,
and of the same oxygen and carbon isotopic abundance data as shown
in Fig. 3 from August 1970 to the present time. In addition, samples
collected before July 1970 (which were analyzed for only concentra-
tion and carbon isotopic abundance) are indicated as solid squares
(night) or open squares (day) for Illinois rural locations.
The isotopic compositions of a few of the Illinois samples,
c (13C/12C) = =27.8 to -29.0%0, 6(180/160) = +17 to +23%0, show that
30-80% of their CO is from Illinois engine emissions, while the
remainder is from CO whose oxygen isotopic composition is quite
different from engine CO. In three of these cases, the concentra-
tion was proportionately higher and contamination with engine CO
was suspected because of known proximity of automobiles. Except for
these few cases, the isotopic variations indicate the contribution
of local engine emissions is not significant.
-------
-9-
In Figs. 5-11 the data are plotted for seven seasonal periods
as follows: (a) the oxygen isotopic composition versus carbon
isotopic composition, (b) the oxygen isotopic composition versus
CO concentration, and (c) the carbon isotopic composition versus
CO concentration. The periods are approximately as follows: Fig. 5 -
early winter, Fig. 6 - late winter, Fig. 7 - spring, Fig. 8 - early
summer, Fig. 9 - late summer, Fig. 10 - early autumn, and Fig. 11 -
late autumn. The division of these seasonal periods was based on
similarity of isotopic patterns.
5. Isotopic Fractionation Effects
The interpretation of these results must take into account the
possible effects of isotopic fractionation and meteorological factors.
The wide variations in both carbon and oxygen isotopic values
imply a complicated model for atmospheric CO. In addition to mix-
tures of CO from multiple sources which may depend on season and
latitude, there is the possibility of isotopic fractionation in the
scavenging processes. Three scavenging processes have been proposed;
bacterial consumption in soil [4] and photochemical reactions in-
volving OH radical either in the troposphere [15] or stratosphere [3].
We have made several measurements of the isotopic fractionation ef-
fects in CO consumption by soil bacteria and find that this process
favors the lighter isotopes of both carbon and oxygen. Reactions
involving photochemically-produced radicals are probably diffusion
limited, being very fast reactions with low concentrations of oxidant
such as OH radical; thus, kinetic effects would be expected to be
small. Further, isotopic fractionation for both diffusion and kinetic
effects in most simple reactions favors the lighter isotope. A
-------
-10-
decrease in CO concentration corresponding to a greater degree of
scavenging should then be accompanied by an increase in the O/ O
13 12
and C/ C ratios. The majority of samples showed the opposite
effect for oxygen, with decreasing O/ O ratios with decreasing
concentration for both long- and short-term variations. In most
13 1 ?
seasons, the C/ C ratio increased with a decrease in concentra-
tion, but the long-term seasonal trend was the same as for oxygen.
In the summer and autumn, there were some samples with low CO concen-
trations and abnormally high values of both the O/ 0 and C/ C
ratios indicating possible fractionation. Perhaps these were cases
where scavenging by soil bacteria had been predominant.
i Q If.
Since fractionation of the O/ O ratio would be expected to
13 12
be twice that of the C/ C ratio, the isotopic values for oxygen
versus carbon would lie along a line with a slope of two. Examina-
tion of the data in Figs. 5-11 indicates that any such effect was
small compared to variable mixtures of different isotopic species.
On the basis of these considerations, it would seem that iso-
topic fractionation in the scavenging process is of small importance
compared to the variations from the mixing of isotopically different
species.
6. Meteorological Effects
Meteorological factors such as wind speed and atmospheric sta-
bility are known to have an important influence on the concentration
of pollutants in urban atmospheres in different seasons. Our samples
were collected at ground level in country locations selected to be
far upwind of cities and major highways. The sampling times were
mostly 3-5 p.m. or 6-9 a.m. and were selected to be several hours
-------
-11-
later than when the day and night time minimums of auto traffic
would occur in the small country towns upwind of the sampling loca-
tion.
The effects of meteorological factors are difficult to general-
ize. Most of the available data for seasonal variations of the
atmospheric concentration of pollutants are for urban locations.
The large midwestern cities of Chicago and St. Louis show higher
24-hour average concentrations of CO in the summer compared to the
winter [28] . Colucci and Bergeman [29] have shown that in New York
City and Los Angeles, the concentration is highest during seasons
of lower than average wind speeds. For northern Illinois, the
average surface wind speeds in the summer are about one-half of those
in the winter [30] and, therefore, on the basis of the above data,
the concentration of any locally emitted technological CO would be
expected to be higher in summer than winter. The opposite was ob-
served in our studies of atmospheric CO concentration in country
locations between December and July. Data on seasonal variations of
concentrations of pollutants in country are scarce. This is the
first such study for CO. The concentration of SO- measured in
country locations during 1937-1939 in England [31] showed seasonal
variations which were proportional to the seasonal variations in
power demand and residential heating [32]. Finally, the carbon iso-
topic values of most of our country air samples with higher than
average CO concentrations during the period from December through
July became asymototic to an isotopic composition which was very
different from regional engine CO emission.
-------
-12-
The conclusion is that the variation of the long-term minimum
concentration of CO observed in our Illinois samples between winter
and summer was not due to any meteorological influences involving
technological emissions. As will be shown in the observed general
18 16
trend in both the concentration and O/ O ratio was due to seasonal
differences in production and scavenging rates of different varieties
of CO.
Some of the short-term fluctuations of the CO concentration
in our data were probably due to meteorological factors causing an
increase in the concentration of CO from ground-level sources. The
isotopic composition of these CO varieties rarely showed any contri-
bution from regional engine emissions during the period between
December and August.
Obvious meteorological influences were evident in our autumn
data. The concentration became abnormally high during periods of
atmospheric inversions. Also samples collected between 5-9 a.m.
l ft 1 f\
in 1970 showed on the average higher CO concentrations and O/ 0
values than those collected 3-5 p.m. This indicated there were
surface emissions of CO which resulted in a higher ground level
CO concentration during the hours of greater atmospheric stability.
The isotope composition was variable and showed some contribution
from regional engine CO, but there was a more dominant heavy-oxygen
variety with an isotopic composition different from engine CO.
This variety was most likely emitted from trees and plants and will
be discussed in more detail in the section on autumn CO.
7. Varieties of Atmospheric CO and Model of Seasonal Patterns
An analysis of our data shown that atmospheric CO consists of
several different isotopic varieties, the relative amounts of which
-------
-13-
depend on the season and latitude. The determination of the iso-
topic characteristics of the pure species is complicated by the
fact that in all seasons atmospheric CO consists of mixtures of
two or more varieties.
The data shown in Figs. 5-11 indicate there were five varieties
of atmospheric CO, two with light oxygen and three with heavy oxy-
gen. These are summarized in Table 1 as characterized by either
the season occurrence or isotopic composition. The isotopic compo-
sitions of the different varieties are plotted in Fig. 12 in the
summary. These varieties do not necessarily correspond to separate
and distinct species in all cases - each of the heavier oxygen
varieties III, IV and V are possibly mixtures of more than one
heavy-oxygen species. Also, the same source may contribute to more
than one of the listed varieties.
Carbon monoxide with a light oxygen isotopic composition oc-
curred with variable carbon isotopic composition indicating the
possible existence of two species of CO with light oxygen (I and II
in Table 1). The heavier carbon variety predominated in Illinois
during the winter,- going over to more variable and lighter carbon
in the summer. There is nothing obvious in the data to indicate
whether or not this variation represented different sources or some
carbon isotopic fractionation factor in the production process.
There were three isotopically different heavy oxygen varieties:
variety III occurred in small and variable amounts during the summer
of the northern midlatitudes; variety IV occurred abundantly during
the autumn in the northern midlatitudes; and variety V is the
heavy-oxygen variety which, relative to the concentration of the
light-oxygen varieties, occurred abundantly during the winter and
-------
-14-
decreased in amount as the season progressed to summer. In general,
the concentration of each of these heavy-oxygen varieties was quite
variable.
The general seasonal characteristics of atmospheric CO for the
northern hemisphere fit a pattern of a mixture of a relatively con-
stant amount of 0.10-0.15 ppm of the light oxygen varieties mixing
with variable amounts of heavy-oxygen varieties. The heavy-oxygen
CO varied between 0.15-0.5 ppm in the winter and 0.02-0.3 ppm in
the summer and was not the same isotopically in the different sea-
sons. Figure 12 shows the general seasonal cycle of the isotopic
composition of atmospheric CO in northern Illinois.
It has been suggested by McConnell et al. [19] that oxidation
of CH by photochemically produced OH radical might be an important
source of atmospheric CO. The results of Weinstock [16] strongly
support this suggestion and further show that the oxidation of CO
by OH in the troposphere is highly likely as the scavenging mechan-
ism for CO.
Our data (see section 7.2 - Source Strength of the Light Oxygen
Varieties) suggest that the origin of the light oxygen varieties is
in the atmosphere itself possibly from CH. or other organic con-
stituents. If this CO is both produced by photochemically produced
OH and scavenged by the same oxidation mechanism, then the concen-
tration of these varieties would be independent of the amount of
solar radiation and therefore shows no seasonal effects. Within a
factor of two, this was the case. On the other hand, the atmospheric
concentration of CO from sources whose global production rates are
more or less constant (such as engine emissions and possible marine
emissions) would be enhanced by the reduced solar radiation in the
-------
-15-
northern-hemisphere winter - especially in the midlatitudes which
are estimated to receive 25% of the amount during the summer [21].
The average atmospheric concentrations of the heavy-oxygen varieties
in Illinois during the winter was 5-10 times greater than during
the summer. The highly variable concentration of the heavy-oxygen
varieties in all seasons indicates they are surface emissions.
Surface emissions such as from technological and possibly marine
and terrestrial biosphere sources would tend to produce more vari-
able CO concentrations than a source in the atmosphere because of
the non-uniform distribution of sources and the incomplete mixing
of air masses within times shorter than several weeks.
7.1 Light-Oxygen Varieties (I and II) of Atmospheric CO
The major variety of atmospheric CO in the northern midlati-
tudes during the summer had a light-oxygen composition. In all
other seasons, atmospheric CO showed a component of a light-oxygen
variety which occurred with relatively constant concentration com-
pared to the amount of heavy-oxygen varieties. The limited geo-
graphic data, shown in Fig. 3, indicate that the light-oxygen vari-
eties were the dominant varieties all over the world. They were
always the varieties which were predominant when the concentration
was the lowest.
The carbon isotope abundance of the light-oxygen variety varied
13 12
with the seasons in Illinois, having a high C/ C value in the
13 12
winter and spring, and lower and more variable C/ C in the summer
and autumn. There is nothing to indicate whether the variability
of the carbon isotopic composition is the result of two different
species with light-oxygen, isotopic variations in the source of the
-------
-16-
carbon, or variations in isotopic fractionation in the production
process.
Since there is no clear basis for distinguishing the light-
oxygen varieties, we shall treat them as a composite variety even
though the carbon isotopic composition is variable. There is a
similarity in origin perhaps as the concentration was fairly con-
stant during each season and was nearly the same in summer and
winter, although somewhat higher in the spring (see discussion in
section 7.7 - Atmospheric CO in the Spring).
7.2 Source Strength of the Light-Oxygen Variety of CO
The isotopic dilution method would be useful in estimating
the source strength of the light-oxygen CO varieties in the northern
hemisphere, if the average relative concentration of engine CO on
a global basis can be determined and then treated as an internal
isotopic standard. For accurate results, the isotopic dilution
method requires a uniform mixture of the reference and the unknown
species of CO, and it is evident from the geographic and time vari-
1 p "I £
ations of the O/ 0 ratio and CO concentration that there is in-
complete mixing of isotopically different species of atmospheric
CO. Referring to Fig. 9, one-half of the days sampled in Illinois
during July and August had low CO concentrations (0.12-0.20 ppm) and
18 Ifi
the lightest oxygen isotopic composition [average 6( O/ 0 = 8%o]
of all Illinois samples. These air masses must be the most repre-
sentative of the average midlatitude atmosphere. It is reasonable
to assume that the isotopic variations within this group of samples
are due to small variable amounts of a heavy-oxygen variety mixing
with a light-oxygen variety. While the source of the heavy-oxygen
-------
-17-
variety is not know, an upper limit to the fraction of engine CO
in these samples can be estimated if these variations are con-
sidered as due to variable amounts of engine CO mixing with a light-
oxygen species of CO having a value of 6(180/160) = 5%0 which was
the oxygen isotopic composition of the lightest oxygen variety of
CO observed in the southern hemisphere and is assumed to be the
pure variety. Then the source strength of the unknown species of
CO Q.J. is given by the isotopic dilution equation
Q en, CO
1 " 1816- 6(180/160)I
X [f] X Qen CQton/month ,
I Q I /-
where 5( O/ 0) = 8%0. The factor f is introduced to take
account of any non-uniformity of the engine-CO concentration in
the hemisphere. Since engine-CO emissions are mostly located in
the midlatitudes , their concentration could have an average merid-
ional gradient which is dependent on the ratio of the residence
time of CO to the hemispheric mixing time. For a residence time
long compared with the hemispheric mixing time, there would be
uniform mixing of engine CO throughout the hemisphere. Then the
source strength Q of the predominant variety of CO in the northern
hemisphere summer must be at least five times that of engine CO*
o
or 1.5 X 10 tons/month. This estimate is very likely too low for
two reasons. First, if the residence time is comparable to the
hemispheric mixing time, then the concentration of CO from a source
located in the midlatitudes such as engine CO would be higher than
*2.8 X 10 tons/month for July and August 1971, extrapolated from
gasoline-consumption data for 1965-1968 [10,18] and a CO/gasoline
ratio of 0.35 kg/1 [18].
-------
-18-
in the southern latitudes. Robinson and Robbins [18] observed an
overall meridional concentration gradient in November and December
1967. The main source of the CO variety which produced this gradient
was probably terrestrial plants and trees in the midlatitude regions
and not engine CO (see section 7.5 - Heavy-Oxygen Variety of Atmo-
spheric CO in the Autumn). Nevertheless, the results show a
meridional concentration gradient from a non-uniform distribution
of sources and a residence time less than or comparable to the hemi-
sphere mixing time. Secondly, engine CO was assumed to be the only
species of CO causing the variations in the O/ O ratio during
the summer. There is evidence that this was not the case. The vari-
18 16
ations in the 0/0 ratio versus concentration for the summertime
Illinois samples shown in Fig. 9 suggest there is a source which
is different from engine CO and produces a heavy-oxygen species of
CO (see section 7.4 - Heavy-Oxygen Variety (III) of Atmospheric CO
in the Summer).
If the isotopic-dilution calculation is adjusted using a factor
of two for the non-uniformity of the meridional concentration of
engine CO, the production rate of the light-oxygen species is at
Q
least 3 X 10 tons/month during July and August. We believe that
this is a very conservative estimate. The rate could be much greater
depending on the production rates of other natural sources during
the summer with a heavy-oxygen isotopic composition.
7.3 Nature of the Source of the Light-Oxygen CO Variety
This predominant variety must be of natural origin since engine
CO constitutes 88% of known anthropogenic emissions. If this CO
were produced by surface emissions such as the oceans, lakes, trees,
-------
-19-
O
etc., the rate would have to be greater than 1 std cc/hour/m .
This is 1-2 orders of magnitude too much to be accounted for by any
known surface sources in the biosphere. Junge et. al. [5] esti-
mated that the supersaturation of CO in ocean waters produced 0.05
2
std cc/hour/m . We have found CO in lakes and rivers in amounts
equal to or less than in ocean water, depending on the degree of
pollution. Some species of trees and plants emitted CO, but the
amounts were much too low' to account for such a rate. Also, the
oxygen isotopic compositions of all these surface-emitting species
were 10-15%0 higher than those of the light-oxygen variety. Thus,
this predominant source is most likely produced directly in the
atmosphere.
Recently McConnell et al. [19] have suggested oxidation of CH.
by OH as an important source of atmospheric CO. Little is known
about the sources and geographical and seasonal variations of atmo-
spheric CH.. Estimates of total global emission rates of CH. from
biological activity in paddy fields, swamp lands, and tropical
o
regions are 3.1 - 14 X 10 tons/year [18,20]. Since oxidation of
CH. would be dependent on photochemical production of OH, there
would be a seasonal variation in the rate of formation of CO. Taking
into account the seasonal and meridional variation of the daily solar
o
radiation intensity [21], our estimate of 3 X 10 tons/month in the
9
summer would integrate to 2.7 X 10 tons/year for the northern
hemisphere assuming a constant concentration of CH.. This is com-
parable to the highest estimated global CH. production rate.
Very little information about the carbon isotopic composition
13 12
of atmospheric CH. is available. Craig [6] obtained S( C/ C) =
-40.3 and -10.7%0 for two samples, but had reservations about the
-------
-20-
purity and absence of fractionation in the samples. Bainbridge
et al. [22] obtained -39%o. Our values for the CO species with the
lightest oxygen varied between -31 and -22%0 with no correlation
between the carbon and oxygen isotopic abundances. This could
indicate that there are two sources of CO with a light oxygen abun-
13 12
dance, one of which may be atmospheric CH4 with 6( C/ C) as low
as -39%0. The other possibility is that the carbon isotopic abun-
dance of atmospheric CH. is variable, but this seems unlikely be-
cause the residence time for CH4 [18] is significantly longer than
hemispheric mixing times. In view of the limited data on atmospheric
CH., one cannot be sure whether or not atmospheric CH. is the prin-
cipal source of atmospheric CO.
7.4 Heavy-Oxygen Variety (III) of Atmospheric CO in the Summer
In previous sections, it has been pointed out that the heavy-
oxygen variety of CO occurring during the summer in Illinois appeared
to be emitted from the surface and had an oxygen isotopic abundance
which is significantly different from engine CO. Carbon monoxide
from marine sources [1,2] is undoubtedly a constituent. Wilks [23]
has observed CO release from bagged tree branches as well as from
cut green plant structures. Swinnerton et. al. [24] have reported
that rain water is supersaturated with CO. We have confirmed these
sources of CO and find that the O/ 0 ratio for these species of
-| Q if:
CO is greater than 6(0/0) = 16%o. Thus, this variety of
heavy-oxygen CO may consist of a mixture of several species of CO
including engine CO. All of these are minor constituents of the
summer atmospheric CO. It is planned to discuss these species of CO
in more detail in a later report.
-------
-21-
7-5 Heavy-Oxygen Variety (IV) of Atmospheric CO in the Autumn
The increase in the 180/160 ratio in September and October for
Illinois samples and the simultaneous increase in concentration is
most likely related to the end of terrestrial plant life in the
autumn.
Increased atmospheric stability and lower wind speeds in this
season would result in higher concentrations of any CO from ground
emissions. The contribution of the light-oxygen species was still
very noticeable until the end of November. The heavy-oxygen variety
was considerably different from world-average engine CO and espe-
cially Illinois engine CO indicating a different and large CO source.
It has been suggested by Loewus and Delwiche [25] and Katz [26] that
CO emissions may possibly accompany the degradation of chlorophyll
in plants and trees. There is a definite diurnal effect which can
be attributed to the effect of atmospheric stability on the concen-
tration of surface emissions. Samples collected in the early
18 1 fi
morning (5-7 a.m.) showed higher CO concentrations and higher O/ 0
values than those collected in the late afternoon (3-5 p.m.).
November 1970 showed a decrease in concentration corresponding to
either dispersion of this CO throughout the northern hemisphere or
the effects of the scavenging process. In late October 1970, there
occurred an eight-day peak in the CO concentration; the maximum of
nearly 1 ppm coincided with an atmospheric subsidence. The fact
that the carbon isotopic abundance at that time was 3-5%0 higher
than regional engine CO indicates that these were due to surface
emissions which consisted mostly of non-engine CO. Samples col-
lected in regions of large forests (Sioux Narrows, Ontario and
Alleghany State Park, New York) in the autumn showed both high CO
-------
-22-
13 12
concentrations and high 6( C/ C) values, the same as the general
trend of atmospheric CO in the autumn in Illinois.
The amount of this seasonal CO production of variety IV in the
Q
northern hemisphere is estimated at 2-5 X 10 tons emitted over
1-1/2 months.
7.6 Heavy-Oxygen Variety (V) of Atmospheric CO in the Winter
Except for the October maximum, the average monthly CO concen-
tration had a maximum of 0.35 ppm in December and then decreased
gradually during January and February to 0.24 ppm in March. The
make-up of this CO is not certain. Undoubtedly it consists partially
of a light oxygen variety similar to the dominant variety in the
summer.
The isotopic abundance of the dominant heavy-oxygen variety
indicates a large contribution from engine emissions. We have dis-
cussed briefly in an earlier section the possibility of the reduced
solar radiation in the winter decreasing the scavenging rate for
CO and consequently increasing the concentration of a source with
a constant emission rate. The concentration of CO from engine
emissions, which are principally localized in the northern midlati-
tudes, could be enhanced by a factor of as much as eight times the
average hemispheric concentration because of the combined effect
of the reduced solar radiation and a non-uniform meridional distri-
bution. Carbon monoxide from the combustion of fuel for residential
heating which Robinson and Robbins [18] have estimated to be 4% of
total annual anthropogenic sources would be enhanced in the winter
in the northern midlatitudes both by a large seasonal factor of 4-5
and by the same effects of non-uniform distribution and reduced
-------
-23-
scavenging rate as engine emissions. We estimate that atmospheric
CO in the winter in the northern midlatitudes could consist of as
much as 10% from residential heating and 50% from total anthropo-
genic sources, even though the average contribution of the latter
to total technological sources throughout the northern hemisphere
is less than 10%. The measured average oxygen isotopic composition
of CO at this time was in agreement with such a mixture. However,
13 12
the C/ C ratio was 1-2%o higher than would be expected for such
a distribution. Other possible constituents would include marine
CO and residual amounts of the autumnal emissions not yet fully
scavenged, both of which may account for the higher C/ C ratio.
7.7 Atmospheric CO in the Spring
The CO occurring in the spring consists of a mixture of a
light-oxygen heavy-carbon variety similar to variety II occurring
in the winter and a mixture of heavy-oxygen varieties intermediate
between the winter and summer heavy-oxygen varieties (V and III,
respectively).
The total CO concentration varied between 0.2-0.4 ppm, the
same range as during the winter. However, the concentration of the
light-oxygen variety was about 0.15 ppm or 1-1/2-2 times as much
as in the winter with the concentration of the heavy-oxygen vari-
eties correspondingly lower. This decrease in the concentration of
the heavy-oxygen varieties is consistent with an increase in the
scavenging rate from increasing solar radiation in this season. The
cause of the increase in the concentration of the light-oxygen variety
to a value higher than the value either in the winter or summer is
not understood. We have suggested that this variety is related to
variety I, the light-carbon light-oxygen variety, and that the source
-------
-24-
or sources of these varieties are in the atmosphere itself and
consist of methane and possibly other organic compounds. At this
state in our understanding of these sources, it is only possible
to speculate on what factors could cause the higher concentration
of variety II. Some of these speculations are: 1) the change
of air masses in midcontinental U.S. in this season, with lower
frequency occurrence of polar-continental air and higher frequency
of tropical-marine air compared with winter, may result in a higher
concentration of any marine organic compound which might be the
precursor of this CO variety, 2) the start of the growing season
with the possibility of an increase in plant expirations of organic
substances, or 3) the increase in the amount of atmospheric water
with higher temperatures which may alter the production rate of CO
from methane in the mechanism suggested by McConnell et al. [19].
8. Atmospheric CO in the Southern Hemisphere
The atmospheric CO concentrations for Tonga Islands and
American Samoa in August were higher than those for Australia in
May. This is in agreement with the corresponding seasonal dif-
ferences observed for Illinois. These very limited data, if compared
with those for the northern hemisphere during the same season, con-
firm the findings of Robinson and Robbins [18] and Seller and
Junge [3] that the concentration is lower in the southern hemisphere
than in the northern hemisphere. The very high concentration (0.75
ppm) obtained near San Paulo in May is analogous to the high con-
1 O 1 £
centrations in Illinois in the autumn. The O/ 0 ratio is lower
than the values observed in Illinois. It is also considerably lower
18 16
than the O/ 0 ratio in urban atmospheric CO indicating that
-------
-25-
pollution sources were not the major source of the high CO concen-
tration. The difference in the 0/0 ratio between the samples
from American Samoa and Tonga Islands of about 6%0 may be due to
the presence of some heavy-oxygen variety, possibly marine CO,
emitted in the region of American Samoa.
9. Residence Time
The conclusion (based on evidence from isotopic compositions)
that natural emissions of CO are many times the amount of engine
CO emissions implies that the residence time is shorter than had
been indicated previously by Junge e_t al. [5] and Dimitriades and
Whisman [27]. The residence time of CO in the summer is less than
Q
1.0 month if we use our estimate of at least 3 X 10 tons/month
for the production rate of the predominant species of CO in the
northern-hemisphere summer and an average concentration of 0.12 ppm
for this species. Weinstock [15] has calculated a value of 0.1 year
14
based on the concentration of CO in the atmospheric CO collected
at Buffalo, New York, between January and March. This result sup-
ports our estimate of an upper limit of the residence time. Further-
more, if the scavenging process is inversely proportional to the
amount of solar radiation, as indicated by our data, then the sum-
mer residence time based on Weinstock's winter value is about 10
days. The production rate of the predominant light-oxygen species
p
in the summer would then have to be 8 X 10 tons/month. We have
indicated that the production rate based on isotopic dilution by
engine CO could be as much as this.
10. Summary
The seasonal variation of the isotopic composition of
-------
-26-
atmospheric CO in northern Illinois is summarized in Fig. 12. The
following conclusions are drawn from the results of this study:
18 16
(1) Varieties of atmospheric CO with a low O/ O ratio
dominate during the northern hemisphere summer and are similar to
atmospheric CO seen in the limited number of samples collected in
the southern hemisphere where technological sources are many times
smaller than in the northern hemisphere. It is deduced that this
form is produced year round and originates from the atmosphere itself.
The variability of the carbon isotopic composition indicates the
possibility of two types of sources. Oxidation of atmospheric
methane is a suggested source of the light carbon species of this
variety and oxidation of the other organic constituents of the atmo-
sphere from either marine or terrestrial origin may account for the
heavy carbon species. More investigation is needed to resolve the
origins of these species of CO. In the northern hemisphere, the
9
production rate is estimated to be > 3 X 10 tons/year as compared
Q
to 2.7 X 10 tons/year from man-made sources.
(2) A large burst in concentration of a heavy oxygen variety
of CO is seen in the autumn, quite probably derived from the degrada-
tion of plant life chlorophyll at that time of the year. The total
amount emitted in the northern hemisphere over a 1-1/2-month period
O
is estimated at about 2-5 X 10 tons.
(3) Another major variety with a heavy oxygen composition
occurs during the winter and spring in the northern midlatitudes and
may be a mixture of CO from several different sources. Anthropogenic
emissions from engines and residential heating could be significant
constituents of this variety. The atmospheric concentrations of
these sources would be enhanced during the winter of the northern
-------
-27-
midlatitudes by the combined effects of a seasonal slowing-down
of any scavenging process dependent on solar radiation, the locali-
zation of emissions in the northern midlatitudes, and the increased
seasonal emissions of heating fuels. The magnitude of the pro-
duction rate of this collective species including the anthropogenic
emissions is estimated to be in the range of 3-6 X 107 tons/month
during the winter.
(4) There appears to be a small variable amount of a heavy-
oxygen variety of CO occurring during the summer in the northern
midlatitudes. This would include any anthropogenic emissions, but
the oxygen isotopic values indicate they are not a major constituent
of this variety.
The isotopic composition of the atmospheric CO in all seasons
is more variable than can be explained by assuming various mixtures
of only two isotopic species. Assuming there is no problem in the
analytical procedure, there has to be some other factor causing the
wide variations of both the carbon and oxygen isotopic compositions.
Possible factors are: (1) isotopic fractionation of the production
and scavenging processes, (2) the existence of three or more iso-
topcially different species, and/or (3) variations in the isotopic
composition of the carbon and oxygen in the CO from the source. A
more detailed treatment of these data can be made. In addition,
the results of an unfinished study of the isotopic composition of
CO from fresh and marine waters, trees and plants, and rain may
help to decipher the atmospheric data.
The limited geographic and seasonal sampling of the atmosphere
hampers the interpretation of these results. The complete description
of atmospheric CO becomes very complicated because of geographic and
-------
-28-
seasonal variations of the production rates of the several sources
as well as the scavenging processes. Regular sampling of the
atmosphere in several locations, including regions of the trade
winds and of the westerlies in each hemisphere, would provide the
data for a better understanding of this complex subject.
-------
-29-
Acknowledgements
The work would not have been possible without the cooperation,
help and advice of very many people. We gratefully acknowledge
the encouragement of Bernard Weinstock, Donald Stewart and Lewis
Friedman; the very helpful discussions with William Chupka, Harmon
Craig, Louis Kaplan and Kenneth Wilzbach; the gas-chemistry advice
of Ben Holt; the air sample collections by Evan Appelman, Bernard
Abraham, Peter Jeffries, R. F. Weiss and David Hess; the isotopic
reference samples from Robert Clayton and Toshiko Mayeda; the
meteorological advice by Harry Moses, Donald Gatz and James Carson;
the editing and comments of the manuscript by Francis Throw; the
frequent assistance of John Sevec; the patient cooperation in
typing the manuscript of Nancy Bertnik, Brenda Grazis and the Argonne
Chemistry Division Typing Pool, Ruth Bernard and the Argonne Graphic
Arts Typing Pool, Beverly Webner and Sandra Tasharski; and the
preparation of the figures by David Kurth.
-------
-30-
Appendix I Estimation of Carbon Isotopic Abundance of World-
Average Engine CO
Two methods have been used to determine U.S.-average engine
CO, which is 50% of the world's emissions. One method is to com-
bine the regional 1968 automobile registrations [10] with the
corresponding regional isotopic abundances of engine CO. The U.S.
was divided into three sections (western, midwestern and eastern)
and the measured isotopic compositions of CO for Los Angeles,
Chicago and New York City were taken as representative of the
1312
respective regions. The resulting average is 6( C/ C) = -27.4%o.
The second method combines the quantity of petroleum produced in
13 12
each state [11] with the C/ C ratios for the corresponding oil
fields [7]. The data in Fig. 2 indicate that isotopic fractiona-
tion in going from crude oil to engine CO is very small. The U.S.-
13 12
average composition of engine CO then was found to be 6( C/ C) =
-27%0. The agreement between the two methods is surprising in view
of the broad assumptions made in the computations, especially
13 12
since the spread of C/ C ratios is ten times the difference be-
tween the results of the two methods.
13 12
Only meager data on the C/ C ratio are available for the
other principal regions of the world. The world-average is esti-
mated from the 1968 automobile registrations of the U.S., Europe
and the U.S.S.R. [12], which represent 90% of the world automobiles
The average value for London and Paris urban CO is probably quite
representative for the C/ 2C ratio of European engine CO. There
To 12
is a published value 6( C/ C) = -29%0 for average petroleum from
-------
-31-
the Volgograd region of the U.S.S.R. [13]. The world-average for
engine CO is found to be 6(13C/12C) = -27.4%0. This is admittedly
a crude estimate, but the error is probably not greater than ±0.3%0
-------
-32-
References
[1] J. W. Swinnerton, V. J. Linnenbom and R. A. Lamontague,
The ocean, a natural source of carbon monoxide, Science
167, 984-986 (1970).
[2] W. Seller and C. Junge, Carbon monoxide in the atmosphere,
J. Geophy. Res. T5_, 2217-2226 (1970).
[3] W. Seller and C. Junge, Decrease of carbon monoxide mixing
ratio above the polar tropopause, Tellus 21, 447-449 (1969).
[4] R. E. Inman, R. B. Ingersoll and E. V. Levy, Soil: a natural
sink for carbon monoxide. Science 172, 1229-1231 (1971).
12
[5] C. Junge, W. Seller and P. Warneck, The atmospheric CO and
14CO budget, J. Geophy. Res. 76_, 2866-2879 (1971).
[6] H. Craig, The geochemistry of the stable carbon isotopes,
Geochim. et Cosmochim. Acta 3_, 53-92 (1953) .
[7] S. R. Silverman and S. Epstein, Carbon isotopic compositions
of petroleums and other sedimentary organic materials, Bull.
of the Amer. Assoc. of Pet. Geologist 42, 998-1012 (1958).
[8] M. Dole, The natural history of oxygen, J. Gen. Physiol.
£9 (1) , Pt. 2, 5-27 (1965) .
[9] C. M. Stevens and L. Krout, Method for the determination of
the concentration and of the carbon and oxygen isotopic
composition of atmospheric carbon monoxide. Int. J. Mass
Spect. and Ion Phys. 8_, 265-275 (1972).
[10] Statistical Abstract of the United States, 1969, U.S. De-
partment of Commerce, Bureau of the Census, U.S. Government
Printing Office, Washington, D. C. (1969) p. 550.
-------
-33-
[11] Ibid, p. 670.
[12] Ibid, p. 853.
[13] F. A. Alekseev, V. S. Lebedev and R. A. Krylova, Isotopic
compositions of carbon of natural hydrocarbons and some
questions of their genesis, Geokhimiya 5. 510-518 (1967).
[14] P. Kroopnick and H. Craig, Atmospheric O : isotopic com-
position and solubility fractionation, Science 175, 54-55
(1972) .
[15] B. Weinstock, The residence time of carbon monoxide in the
atmosphere, Science 166, 224-225 (1969).
[16] B. Weinstock and H. Niki, Carbon monoxide balance in
nature, Science 176, 290-292 (1972).
[17] J. Pressman and P. Warneck, The stratosphere as a chemical
sink for carbon monoxide, J. Atmos. Sci. 27, 155-163 (1970).
[18] E. Robinson and R. C. Robbins, Sources, abundances and fate
of gaseous atmospheric pollutants. Final Report Stanford
Research Institute, Project PR-6755 (1968).
[19] J. C. McConnell, M. B. McElroy and S. C. Wofsy, Natural
sources of atmospheric CO, Nature 233, 187-188 (1971).
[20] T. Koyama, Gaseous metabolism in lake sediments and paddy
soils and the production of atmospheric methane and hydrogen,
J. Geophys. Res. 6B_, 3971-3973 (1963).
[21] D. M. Gates, Energy exchange in the biosphere, Harper and
Row Biological Monograph, Ed. Allan H. Brown, Harper and
Row, New York (1962) p. 8.
[22] A. E. Bainbridge, H. E. Suess and I. Friedman, Isotopic
composition of atmospheric hydrogen and methane, Nature 192,
648-649 (1961).
-------
-34-
[23] S. S. Wilks, Carbon monoxide in green plants, Science 129,
964-966 (1959).
[24] J. W. Swinnerton, R. A. Lamontagne and V. J. Linnenbom,
Carbon monoxide in rainwater, Science 172, 943-945 (1971).
[25] M. W. Loewus and C. C. Delwiche, Carbon monoxide production
by algae, Plant Physiol. .38(4), 371-374 (1963).
[26] J. J. Katz, personal communication (1970).
[27] B. Dimitriades and M. Whisman, Carbon monoxide in lower
atmosphere reactions, Environ. Sci. Tech. 5_, 219-222 (1971).
[28] Air Quality Data from the National Air Sampling Networks and
Contributing State and Local Networks 1964-1965. U.S. Depart-
ment of Health, Education and Welfare, Public Health Service,
Division of Air Pollution, Cincinnati, Ohio (1966).
[29] J. M. Colucci and C. R. Begeman, Carbon monoxide in Detroit,
New York and Los Angeles air, Environ. Sci. Tech. _3, 41-47
(1969)
[30] H. R. Byers, General meteorology, McGraw-Hill Book Co., Inc.,
New York (1944) pp. 352-355.
[31] A. R. Metham, Atmospheric pollution, its origins and pre-
vention, Third Edition, A Pergamon Press Book, The Macmillan
Co., New York (1964) p. 214.
[32] Ibid, pp. 59 and 103.
-------
-35-
Figure Captions
Figure 1 — The isotopic composition of carbon and oxygen in CO
of urban air.
Figure 2 -- Comparison between the isotopic compositions of carbon
in CO of urban air (present work), petroleum from
different sources by Silverman and Epstein [7] and
Alekseev et al. [13], and natural sources by Craig [6].
Figure 3 -- The isotopic composition of carbon and oxygen in CO
in non-urban air.
Figure 4 -- Seasonal variations in atmospheric CO: (a) concentra-
tion, (b) <5(180/160), and (c) 6(13C/12C). The collection
sites and times are indicated.
Figure 5 -- Non-urban atmospheric carbon monoxide in the period
December 4 to January 18. In this plot and in Figs.
T Q "I C 1719
6-11 (a) is 6(0/0) versus 6( C/ C), (b) is
6(180/160) versus CO concentration, and (c) is 6(13C/12C)
versus CO concentration. In all figures, the isotopic
composition of world-average engine CO is shown - in
(a) as an open square, and in (b) and (c) as a dashed
line. All solid lines in all figures represent mixing
curves between different varieties as indicated by the
data. Dotted lines represent theoretical mixing curves
of the light-oxygen variety with variable amounts of
engine CO.
-------
-36-
The data are plotted with open and closed circles as
day and night Illinois collections resp. in 1971, open
and closed triangles as day and night in Illinois in
1970,- open and closed inverted triangles in Illinois
in 1969, winged circles as collections made in Illinois
by airplane at 4000-5000 ft and crosses as collections
in non-Illinois locations as noted below:
1. Northwestern United States
2. Northeast Coast, Oahu, Hawaii
3. Matsue, Japan
4. Brittany, France
5. Sioux Narrows, Ontario
6. Caroline, New York
7. Atlantic Ocean, 74°W - 29°N
8. Caribbean, 78°W - 12°N
9. Pacific Ocean, 140°W - 24°N
10. Pt. Reyes, California
11. Big Sur, California
12. Parguera, Puerto Rico
13. Sugar Loaf Key, Florida
14. Yanchep Beach, Australia
15. Ely, Minnesota
16. Cotia, Brazil
17. Horseshoe Lake, Mississippi River
18. Pago Pago, American Samoa
19. Tonga Island
20. Allegany State Park, New York
21. Eastern Iowa
22. LaCrosse, Wisconsin
The southern hemisphere samples are plotted six months
out of phase with the actual collection dates.
Figure 6 -- Non-urban atmospheric carbon monoxide in the period
January 19 to February 27. The format is the same as
in Fig. 5.
Figure 7 -- Non-urban atmospheric carbon monoxide in the period
March 4 to June 2. The format is the same as in Fig. 5
-------
-37-
Figure 8 — Non-urban atmospheric carbon monoxide in the period
June 5 to July 7. The format is the same as in Fig. 5.
Figure 9 -- Non-urban atmospheric carbon monoxide in the period
July 8 to September 14. The format is the same as in
Fig. 5.
Figure 10 -- Non-urban atmospheric carbon monoxide in the period
September 15 to October 13. The format is the same
as in Fig. 5.
Figure 11 -- Non-urban atmospheric carbon monoxide in the period
October 17 to December 10. The format is the same as
in Fig. 5.
Figure 12 — The seasonal isotopic composition of atmospheric CO
in northern Illinois. The dashed line indicates the
seasonal cycle of the average isotopic composition of
atmospheric CO. The estimated isotopic composition of
the five varieties of atmospheric CO are shown by the
numbered circles.
-------
Table 1 - Varieties of Atmospheric CO
Variety
6(180/160)
6(13C/120
Principal Seasonal
or Meridional
Occurrence
Probable Sources
Estimated Production
Rate in Northern
Hemisphere
-30
II
-24
III
16-18
^-28
IV
26-33
-26 - -22
V
20-25
^-27
Major species oc-
curring during
summer in northern
midlatitudes.
Occurs in varying
amounts with I in
northern midlati-
tudes. Most abun-
dant in winter and
spring. Also oc-
curs in marine air
of low northern
latitudes.
Lesser abundant
heavy-oxygen vari-
ety occurring dur-
ing summer in
northern midlati-
tudes .
Major variety oc-
curring during
autumn in northern
midlatitudes.
Major variety oc-
curring during
winter and spring
in northern mid-
latitudes .
Atmospheric CH
Unknown
>> 3 X 10 tons/year
i
LO
CD
I
Unknown
= 5 X 10 tons/month
during summer
Degradation of
chlorophyll.
Consists partially
of anthropogenic
emissions. Other
species unknown.
0
2-5 X 10 tons dur-
ing autumn
3-6 X 10 tons/month
during winter
-------
+ 40
-39-
+ 35
• BOTH 8(Cl3/Cl2)ond «0I8/016)
MEASURED
«0I8/0I6)NOT MEASURED
3 +30
II
+25
o:
LU
^ +20
(0
NEW YORK CITY
LONDON
+15
SAN FRANCISCO
LOS ANGELES
SOUTH SAN FRANCISCO
PASADENA
PARIS
WORLD AVERAGE ENGINE CO
+ 10
+ 5
-30 -25 -20 -!5
8(CI3/CI2)PER %o (PDB =0)
-10
Figure 1
-------
-40-
H WORLD AVERAGE ENGINE CO
CHICAGO rN.Y.C.
I97hn
\yf\j-]
I969-, I
-LONDON
rPARIS
pCALIF
in I M URBAN ATM. CO
ST. LOUISJ
TEXAS-
r-VENEZUELA
-LOUISIANA
I C.AMO—I l-LUUIOIHFMH
UTAHni i iOH~CALIF; PETROLEUM SOURCE
USSR (KRASNODAR; *"•* "^NEAREAST?
VOLGOGRAD)
NATURAL GASES MARINE.PET.
NON-MARINE PET.^ COAL
MARINE PLANTS
LAND PLANTS
ATM.C02
-50 -40 -30 -20 -10 0 10
8(CI3/CI2)PER MIL (PDB = 0)
Figure 2
-------
-41-
-1-35
-1-30
+ 25
ALLEGANY STATE PARK
- r-ILLINOIS ENGINE CO 1970-71
f
WORLD AVERAGE ENGINE CO
ALLEGANY STATE PARK
+ 20
O
(O
o
00
?L +15 -
+ 10 -
+ 5 -
0
• RURAL LOCATIONS 40-50 MILES
UPWIND OF CHICAGO
X AS INDICATED
HAWAII
MATSUE
JAPAN
BRITTANY,
'AGO PAGO
X- PT REYES, CALIF
KEYS, FLORIDA
X- PUERTO RICO
^YANCHEP BEACH, AUSTRALIA
'TONGA Is.
-30 -25 -20 -15
8 [I3C/I2C] in %o (PDB = 0)
Figure 3
-------
-42-
i.o
SEPT OCT NOV DEC JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV
1970 1971
+ 35
+ 5
SEPT OCT NOV DEC JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV
1970 1971
SEPT OCT NOV DEC JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV
1970 1971
Rural Locations 40-50 Miles
Upwind of Chicago
• Night 1970-71
O day 1970-71
• Night 1969-70
Q Day 1969-70
Downwind and Urban Chicago
+ 1970-71
ffl 1969-70
X Non-Illinois Locations
1. Northwest, U.S. 11.
2. Northeast Coast, Oahu, Hawaii 12.
3. Matsue, Japan 13.
4. Brittany, France 14.
5. Sioux Narrows, Ontario 15.
6. Caroline, N. Y. 16.
7. Atlantic Ocean, 74°H 29°N 17.
8. Caribbean, 78°W 12°N 18.
9. Pacific Ocean, 140°W 24°N 19.
10. Pt. Reyes, Calif. 20.
Big Sur, Calif.
Parguera, Puerto Rico
Sugar Loaf Key, Florida
Yanchep Beach, Australia
Ely, Minn.
Cotia, Brazil
Horseshoe Lake, Miss. River
Pago Pago, Samoa
Tonga Is.
Allegany State Park, N. Y.
Fig. 2. Seasonal Variations in Atmospheric Carbon Monoxide during 1969-1971: (a) concentration, (b) 6(180/160),
(c) 6( C/ C). The collection sites and times are indicated above.
Figure 4
-------
HU
O
u
o ^O
CO
^ 20
c
"o1
\ 10
0
CO
1
CO
1 1 1 1 1 1
_ (a) .
0*
- Q . -
°o
^ M
— ^ *.• ~
1 1 1 1 1 1
-40 -35 -30-25 -20 -15
S[I3C/I2C] in %0 (PDB =0)
EARLY WINTER
«+u
o
II
§30
CO
^ 20
c
ID
\ 10
o
00
0
u 1 R
" ID
00
§ -20
jl -25
. '". -30
^ -35
o
i 1 I
_ (b)
A 0
• ...
III 1 1 1 1 1
) O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
CO CONCENTRATION (PPM)
I 1 1
-------
tu
0
ii
§30
^20
.E
^o1
to
\ 10
O
CO
i
CO
\
1 1
- (a)
D —
* 0 ~~
•A
0 °
o
x
x — 19 ~
1 1 1 1 1
-40 -35 -30 -25 -20 -15
8[I3C/I2C] in %0 (PDB =0)
_ATE WINTER
*fU
o
|30
c/)
** 20
.E
\ 10
0
GO
O
|| — | ^N
CD
£ -20
&l -25
c.
^-30
u
0
^ 4O
1 1 1 1 1 1 1 1 1 1
_ (b)
"~ 0 *°^^ ~~
^ x 19
! 1 1 1 1 I 1 1 1 1
) O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 i.O
CO CONCENTRATION (PPM)
1 1 1 1 I I I I I 1
~~ X o§*%0Oq/»» * . ~~
o *
_ N,
19 — x
1 1 1 1 1 1 1 1 1 1
1
*
£>
«>
Q.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
CO CONCENTRATION (PPM)
Figure 6
-------
40
_ (a)
o
30
c
"o1
o
oo
• i
CO
20
0
0
D
40
_ (b)
O
CO
c
"o1
30
20
10
o
00
-40-35 -30-25 -20 -15
8[I3C/I2C] in %o (PDB = 0) 5
0
I
*>.
Ul
I
0 O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
CO CONCENTRATION (PPM)
II —I
SPRING
CD
Q
Q.
o
J5*
O
•20
•25
•30
•35
CJ
ro
I
I
O O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
CO CONCENTRATION (PPM)
Figure 7
-------
HU
O
II
|30
CO
o
* 20
.E
tO
\ 10
O
00
OQ •
\
_ (a)
D
o
o —
0
° o *
1
-40 -35 -30 -25 -20 -15
8 [I3C/I2C] in %o (PDB = 0)
EARLY SUMMER
\ 10
O
GO
OQ
O
ii -IR
" \D
CD
g -20
5 -25
~ -30
o
^ -35
0
ro 40
1 1 1 1 1 1 1 1 1 1
_ (b)
_ o _
10— x **
0° * _
A 0
O
1 1 1 1 1 1 1 1 1 1
) O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
CO CONCENTRATION (PPM)
I I 1 I 1 I I I 1 1
10— x 'I
9 — x ° 8 o
0 0
T°
° -7
1 1 1 1 1 1 1 1 1 1
(Ti
I
0 -O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
CO CONCENTRATION (PPM)
Figure 8
-------
O
n
O
CO
o
O
(£>
_ (a)
o
oo
20
10
0
n
x-
A
X v-
!x^ x
«°«f*^
o
I30
en
(£>
20
0
O
00
-40-35 -30-25 -20 -15
S[I3C/I2C] in %o (PDB=0) 5
0
- (b)
0 O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
CO CONCENTRATION (PPM)
LATE SUMMER
DO
a
CL
&
c
o
N
\
o
ro
-\-J
-20
-25
-30
-35
4O
I 1 1 I I I I I I 1
~ I0! X19 ,
-^^f^^//^^A -
d^>o* °Tyx) ^ ° itpo •
o \
• 4
0
— —
1 1 1 1 1 1 1 1 1 1
0~ O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
CO CONCENTRATION (PPM)
Figure 9
-------
HU
O
n
§30
CO
^20
c.
I
to
\ 10
O
00
— j
<5O
1 1
_ (a)
4 A
0 A ^ ^
A
OA
°0
*
O
— o —
cP
1 1
-40-35 -30-25 -20 -15
S[I3C/I2C] in %o (PDB=0)
EARLY AUTUMN
^u
o
§30
CO
** 20
'"o1
(£>
\ 10
0
00
6O
O
n 1 R
ii 13
CD
Q
o_ ~20
&l -25
. '". -30
o
^ -35
o
K) A.r\
1 1 1 1 1 1 1 1 1 1
. (b)
A
' '* A
0 A A
A
- - *A
A 0
o o
0 00
c
o<£°d»
— o —
0 o
1 1 1 1 1 1 1 1 1 1
) O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
CO CONCENTRATION (PPM)
' X
1 I 1 1 1 I 1 1 1 1
A
— »A A —
& _ _A * C^,.. _ 4i *
* V fJ n v On* A
o" o °A 0 A A
O
I 1 1 1 1 1 1 1 1 1
I
£>.
CO
I
O O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
CO CONCENTRATION (PPM)
Figure 10
-------
H-U
6
II
^ 3O
O °^
2
CO
>l ™
^ 20
c
^
0)
\ 10
O
00
S or*
o^ 2O
""o"
<£>
\ 10
o
GO
x>
O
ii -IS
11 i «j
00
§ -20
»l -25
C
."".-30
o
^ -35
o
to /\r\
.— . 'fU
(
1 1 1 1 1 1 1 1 1 1
~ (W
A
ort v
" /
*
4 oo
^ 1 ^s ° ~
A *w\ ^ x ir
o \ ° '6-
\
17
X-^^-14
1 1 1 1 1 1 1 1 1 1
) O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
CO CONCENTRATION (PPM)
I I I 1 1 I 1 1 I I
_ (0
4
* *^ *oo°-b ' 16 720 X
" v H*A/£-JL . .• ' / /• . ~
- *^,4 \
17
1 1 1 1 1 1 1 1 1 1
J O.I 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
£>.
^>
I
CO CONCENTRATION (PPM)
Figure 11
-------
-50-
40
35
30
O
it
0 25
20
WORLD -AVERAGE
LATE AUTUMN
10
EARLY AUTUMN
LATE SUMMER
©
WINTER
-LATE WINTER
V^ — SPRING
EARLY SUMMER
-40 -35 -30 -25 -20
s['3C/'2c] in %0 (PDB = 0)
-15
Figure 12
------- |