2.4.2. Flask Measurements of Carbon Monoxide

The expansion of the number of sites in the CMDL cooperative air sampling network from which CO was determined continued to increase during the past few years.  Only glass flasks fitted with glass pistons and Teflon O-rings were used for CO analysis.  These flasks have been gradually introduced to the network by replacing the previously used glass flasks, which had greased ground glass stopcocks, and by providing all new sites with the greaseless flasks.  By 1996 nearly all network sites used flasks suitable for CO measurements.  The annual mean CO mole fractions for 1996 and 1997 are given in Tables 2.8 and 2.9.  With many more sites measuring CO than in previous years, its global distribution is now better determined.  In particular, more locations in the high southern latitudes are being sampled.  Mole fractions measured at the antarctic sites are typically the lowest found on Earth.  As observed in the high latitudes of the northern hemisphere, the time series from individual sites in the high southern hemisphere also show very similar features (Figure 2.17b).  The seasonal cycles in both the arctic and antarctic are thought to reflect the effects of both transport of polluted air from lower latitudes plus photochemical production and destruction of CO.    

The seasonal CO maximum in the southern hemisphere was quite depressed in 1996.  However, many sites in the tropics and extratropics showed a strong recovery during 1997 (Figure 2.18).   The relatively large annual mean CO values reported for 1997 in the low latitudes of the South China Sea (Table 2.9) reflect very high CO levels during August, September, and October.  The magnitude of the seasonal maximum in the southern hemisphere is believed to be driven largely by seasonal biomass burning [Novelli et al., 1998].  Widespread fires, fostered by dry weather typical of El Niño conditions, occurred in Indonesia during June to November 1997.  The enhanced CO levels found in the south during spring 1997 most likely resulted from these fires.  Samples of air collected from a commercial jet over Indochina during this period show elevated CO in the middle troposphere, presumably resulting from the vertical transport of gases from the polluted boundary layer [Matsueda et al., 1998, H. Matsueda, personal communication, 1998]. 

TABLE 2.8.   Preliminary 1996 and 1997 Annual Mean CO Mole Fractions from Land Sites  

Site

Annul Mean CO (ppb)

Code

Station

1996

1997

ALT

Alert, Canada

     127.5

   124.1

ASC

Ascension Island

       64.4

     71.8

BAL

Baltic Sea

     166.5

   177.8

BME

Bernuda (east coast)

     122.6

   118.0

BMW

Bermuda (west coast)

     118.9

   122.8

BSC

Black Sea, Constanta

     265.8

   235.8

BRW

Point Barrow, Alaska

     131.8

   121.5

CBA

Cold Bay, Alaska

     129.8

   123.8

CGO

Cape Grim, Tasmania

       45.5

     47.6

CHR

Christmas Island

         [ ]

       [ ]

CMO

Cape Meares, Oregon

     133.5

   128.1

CRZ

Crozet Island

      46.2

       [ ]

EIC

Easter Island, Chile

       55.3

     52.9

GMI

Marianas Island, Guam

       84.0

     93.8

GOZ

Gozo, Malta

     155.9

       [ ]

HUN

Hegyhatsal, Hungary

     261.2

   235.3

ICE

Vestmannaeyjar, Iceland

     125.3

   121.4

ITN

Grifton, North Carolina

     171.7

   165.2

IZO

Izana, Tenerife

       97.2

     94.7

KEY

Key West, Florida

     108.2

   112.7

KUM

Cape Kumukahi, Hawaii

       97.7

     94.2

LEF

Park Falls, Wisconsin

     149.8

   139.3

MBC

Mould Bay, Canada

     129.4

      [ ]

MHT

Mace Head, Ireland

     130.8

   116.9

MID

Midway Island

     119.8

     95.8

MLO

Mauna Loa, Hawaii

       83.3

     94.2

NWR

Niwot Ridge, Colorado

     117.4

   117.3

PSA

Palmer Station

       44.9

     44.5

QPC

Qinghai Province, China

     117.2

   124.0

RPB

Ragged Point, Barbados

       91.3

     88.1

SEY

Seychelles

       77.0

     90.1

SMO

American Samoa

       58.8

     56.0

SPO

South Pole, Antarctica

       43.5

     43.6

SYO

Syowa, Antarctica

       44.0

      [ ]

TAP

Tae-ahn Penisula, South Korea

     224.3

   234.5

UUM

Ulaan Uul, Mongolia

     145.8

   138.8

WIS

Negev Desert, Israel

     152.8

   145.7

ZEP

Ny-Alesund, Svalbard

     129.8

   124.1

 TABLE 2.9.   Provisional Annual Mean CO From Shipboard Programs

Latitude

Annual Mean (ppb)

 
 

1996

1997

Pacific Ocean*

N30

111.8

-

N25

114.1

-

N20

95.8

-

N15

89.8

-

N10

81.6

-

N05

66.7

-

000

61.2

-

S05

59.7

-

S10

54.4

-

S15

53.4

-

S20

52.4

-

S25

51.2

-

S30

50.0

-

S35

51.2

South China Sea

N21

247.3

221.9

N18

157.2

155.8

N15

146.1

153.5

N12

130.5

145.3

N09

124.6

154.8

N06

108.5

164.9

N03

129.3

161.6

   *Annual mean values are not reported for 1997 as no samples were collected during December 1997.

It is clear that there is considerable variability in levels of tropospheric CO as is evident, for example, in the 15-year record from Cape Point, South Africa [Scheel et al., 1996].  The shorter record of global measurements also shows periods of CO increase and decrease.  To study global trends in CO, the time series from all marine background sampling locations were combined and smoothed to provide a surface representing tropospheric CO as a function time and latitude.  From this surface, global trends were extracted. This surface also provides averaged time series for the two hemispheres and four semi-hemispheres, from which zonal trends were also calculated.  After a sharp decline in global CO levels during 1992-1993  [Novelli et al., 1994], there was some recovery during 1994-1995 [Novelli et al., 1998].  However, the recovery in the southern hemisphere was short-lived, and another decline in CO occurred during 1996.  The temporally averaged global rate of change was -2.1 ppb CO yr-1 from 1991 through 1996.  This is equivalent to a decrease of approximately 2.4% yr-1 relative to average mole fractions in 1991.  The zonally-averaged decrease in the northern hemisphere (-2.3 ppb yr-1) was greater than that in the southern hemisphere (-1.9 ppb yr-1).  These decreases are believed to be due primarily to changes in emissions from fossil fuel emissions and biomass burning [Novelli et al., 1994, 1998].

  Time series from the Seychelles, Ascension Island, and American Samoa

Fig. 2.18.   Time series from the Seychelles (4ºS), Ascension Island (8ºS), and American Samoa (14ºS).

[BACK]    [CONTENTS]    [NEXT]