Inversion of Atmospheric CO 2 Concentrations
295
particularly as carbon cycle studies become increasingly involved with the type of
process inversion described in Section 11.2.4.
Carbon isotopes: The use of various carbon isotopes to help elucidate the behavior of the carbon cycle provides an excellent example of the way in which different
types of data can reveal different aspects of the system. Radiocarbon ( 14 C) has long
been used to characterize the global-scale carbon cycle responses (Oeschger et al.,
1975; Enting and Pearman, 1987, for example). Atmospheric nuclear testing, particularly over 1961–1962 produced a large spike of 14 C in the atmosphere—tracking
this perturbation through the carbon cycle has provided important information about
the decadal-scale carbon cycle response. The natural 14 C distribution, derived from
“pre-bomb” data, provides information about the behavior of the carbon cycle on
time scales of centuries or longer, refl ecting the 14 C half-life of 5730 years. Spatial
distributions of the 14 C “spike” from nuclear testing have been analyzed to identify
the seasonal modulation caused by biospheric exchange in the immediate posttesting period when there was a large isotopic disequilibrium between atmosphere and
biosphere (Randerson et al., 2002).
The minority (ca. 1%) stable carbon isotope, 13 C, provides, in principle, a way of
separating terrestrial from oceanic carbon exchanges. This refl ects the differences in
carbon fractionation between photosynthesis and air–sea exchange. The quantitative
interpretation of 13 C data requires incorporation of additional effects, particularly
the so-called “isofl uxes” that arise from isotopic disequilibrium, even in the absence
of net carbon exchange, and also the differences in fractionation between the C-3
and C-4 mechanisms of photosynthesis. A discussion of isotopic budgeting is given
in Enting (2002, Section 10.5).
Some 13 C data were used in initial synthesis inversions, both as a trend (giving a
constraint on the oceanic–terrestrial partitioning of net CO 2 uptake) and as sparse
spatially-distributed data. Inversions making use of larger 13 C data sets followed
Ciais et al. (1995a, b).
Methane: Fung et al. (1991) undertook a study of methane (CH 4 ), apparently coining the term “synthesis” for this type of study. For CH 4 , the main difference from CO 2
inversions is the presence of a sink process (oxidation by hydroxyl radicals) in the free
atmosphere. If the sink rate were known, and strictly proportional to methane concentrations, then it would be possible to defi ne Green’s functions that described the surface
sources and the resulting sink, and combine these in a synthesis (Enting, 2002, Section
15.2). However, most, if not all, studies have treated the sink as an unknown that must
be estimated and performed iterative calculations to obtain a consistent budget.
Mass-balance inversions of CH 4 have been obtained by Law and Vohralik (2001),
with a prescribed atmospheric sink, and by Butler et al. (2004), using a chemical
transport model. The use of a full chemical model can take account of the effect of
methane concentration changes affecting the strength of the methane sink, through
reduction of hydroxyl concentrations.
Halogenated compounds: Various halogenated compounds, originally the chlorofl uorocarbons and more recently their “ozone-friendly” replacements, have been
studied within the AGAGE program (Prinn et al., 2000; see also Enting, 2002, chapter 16). The main aim of the initial inversions was to estimate a destruction rate (or
equivalently, its inverse, termed an “atmospheric lifetime”). The calculations were
© 2010 by Taylor and Francis Group, LLC
295
particularly as carbon cycle studies become increasingly involved with the type of
process inversion described in Section 11.2.4.
Carbon isotopes: The use of various carbon isotopes to help elucidate the behavior of the carbon cycle provides an excellent example of the way in which different
types of data can reveal different aspects of the system. Radiocarbon ( 14 C) has long
been used to characterize the global-scale carbon cycle responses (Oeschger et al.,
1975; Enting and Pearman, 1987, for example). Atmospheric nuclear testing, particularly over 1961–1962 produced a large spike of 14 C in the atmosphere—tracking
this perturbation through the carbon cycle has provided important information about
the decadal-scale carbon cycle response. The natural 14 C distribution, derived from
“pre-bomb” data, provides information about the behavior of the carbon cycle on
time scales of centuries or longer, refl ecting the 14 C half-life of 5730 years. Spatial
distributions of the 14 C “spike” from nuclear testing have been analyzed to identify
the seasonal modulation caused by biospheric exchange in the immediate posttesting period when there was a large isotopic disequilibrium between atmosphere and
biosphere (Randerson et al., 2002).
The minority (ca. 1%) stable carbon isotope, 13 C, provides, in principle, a way of
separating terrestrial from oceanic carbon exchanges. This refl ects the differences in
carbon fractionation between photosynthesis and air–sea exchange. The quantitative
interpretation of 13 C data requires incorporation of additional effects, particularly
the so-called “isofl uxes” that arise from isotopic disequilibrium, even in the absence
of net carbon exchange, and also the differences in fractionation between the C-3
and C-4 mechanisms of photosynthesis. A discussion of isotopic budgeting is given
in Enting (2002, Section 10.5).
Some 13 C data were used in initial synthesis inversions, both as a trend (giving a
constraint on the oceanic–terrestrial partitioning of net CO 2 uptake) and as sparse
spatially-distributed data. Inversions making use of larger 13 C data sets followed
Ciais et al. (1995a, b).
Methane: Fung et al. (1991) undertook a study of methane (CH 4 ), apparently coining the term “synthesis” for this type of study. For CH 4 , the main difference from CO 2
inversions is the presence of a sink process (oxidation by hydroxyl radicals) in the free
atmosphere. If the sink rate were known, and strictly proportional to methane concentrations, then it would be possible to defi ne Green’s functions that described the surface
sources and the resulting sink, and combine these in a synthesis (Enting, 2002, Section
15.2). However, most, if not all, studies have treated the sink as an unknown that must
be estimated and performed iterative calculations to obtain a consistent budget.
Mass-balance inversions of CH 4 have been obtained by Law and Vohralik (2001),
with a prescribed atmospheric sink, and by Butler et al. (2004), using a chemical
transport model. The use of a full chemical model can take account of the effect of
methane concentration changes affecting the strength of the methane sink, through
reduction of hydroxyl concentrations.
Halogenated compounds: Various halogenated compounds, originally the chlorofl uorocarbons and more recently their “ozone-friendly” replacements, have been
studied within the AGAGE program (Prinn et al., 2000; see also Enting, 2002, chapter 16). The main aim of the initial inversions was to estimate a destruction rate (or
equivalently, its inverse, termed an “atmospheric lifetime”). The calculations were
© 2010 by Taylor and Francis Group, LLC