comparison of methods to determine the anthropogenic co2 invasion into the atlantic ocean

FEELY3, 1NOAA Atlantic Oceanographic and Meteorological L aboratory, Miami, FL 33149, USA; 2National Center for Atmospheric Research, Boulder, CO 80307, USA; 3NOAA Pacific Marine and Env

ISSN 0280–6495

Comparison of methods to determine the anthropogenic

CO

By R WANNINKHOF1,*, S C DONEY2, T.-H PENG1, J L BULLISTER3, K LEE1,4 and

R A FEELY3, 1NOAA Atlantic Oceanographic and Meteorological L aboratory, Miami, FL 33149, USA; 2National Center for Atmospheric Research, Boulder, CO 80307, USA; 3NOAA Pacific Marine and Environmental L aboratory, Seattle, WA 98115, USA;4Cooperative Institute for Marine and Atmospheric

Sciences, University of Miami, Miami, FL 33149, USA

(Manuscript received 27 November 1997; in final form 26 October 1998)

ABSTRACT

A comparison of di fferent methods for estimating the anthropogenic CO2burden in the Atlantic

Ocean is performed using referenced, high quality total dissolved inorganic carbon (DIC) data.

The dataset is from two cruises through the center of the basin between 62 °N and 43°S in 1991

and 1993 The specific anthropogenic input is determined utilizing empirical procedures as

described in Gruber et al (1996) and Chen and Millero (1979) to correct for remineralization

and to estimate preanthropogenic endmembers These estimates are compared with output of

the Princeton ocean biogeochemical model and the NCAR ocean model The results show that

the specific inventories of anthropogenic carbon agree to within 20% but with di fferent storage

and uptake patterns The empirical estimates di ffer because of assumptions about mixing and

winter outcrop endmembers The same remineralization quotients (Redfield ratios) were used

for each method Varying these constants within the range of literature values causes changes

in specific inventories of similar magnitude as the di fferences observed with different

methodolo-gies Comparison of anthropogenic CO

2uptake and chlorofluorocarbon ages suggests that the anthropogenic CO 2penetration in the North Atlantic is too shallow following the procedure

according to Gruber et al (1996), and too deep using those of Chen and Millero (1979) The

results support these previous observations in that the uptake of CO

2in the North Atlantic is disproportionate to its surface area This is caused by a combination of deep water formation

and deep winter mixed layers.

increases The ocean has taken up one third to Atmospheric CO2 levels have increased from one half of the anthropogenic carbon released to approximately 280 ppm in pre-industrial times the atmosphere to date This quantification is not (1750) to 359 ppm in 1993 primarily due to release exact, and improved knowledge of oceanic

invent-of anthropogenic CO2 The rate invent-of release has ories will lead to better predictive capacity invent-of increased steadily with half the increase occurring future atmospheric CO2 levels under different

in the past 30 years Oceanic uptake of CO

2 has fossil fuel release scenarios The geographic loca-tions of anthropogenic CO2 storage are important for understanding the pathways and mechanisms

* Corresponding author NOAA Atlantic

of anthropogenic uptake and transport, and the

Oceanographic and Meteorological Laboratory,

possible changes in uptake in response to climate

4301 Rickenbacker Causeway, Miami, FL 33149, USA.

E-mail: Wanninkhof@aoml.noaa.gov and global change

The Atlantic Ocean has long been thought to et al., 1985), the uncertainty in the pCO2a estimate

was approximately 10 to 15matm Subsequent

be a major sink of CO

2(Takahashi et al., 1995).

Because of the large scale meridional overturning work by Chen (1993) extended the analysis from

the Atlantic thermocline to whole ocean inventor-cell (‘‘conveyer belt’’), much of the CO2 taken up

enters the deep ocean and is sequestered on centen- ies of DIC

anthro Recently, improvements in this type of inventory estimate have been proposed by nial time scales Several basin wide estimates of

anthropogenic CO2 (DICanthro) uptake in the Gruber et al (1996) (GSS-96) who employed

better methods to estimate mixing of water masses North Atlantic have been performed to date

(Chen, 1982; Gruber et al., 1996), but they have with different preformed concentrations They also

utilized the transient tracer pair3H–3He to estim-mainly relied on data obtained from the

GEOSECS and TTO cruises that took place 15 ate the DIC

anthro on shallow isopycnals that are completely ventilated in the past 40 years, and

to 30 years ago The uncertainty in the total

dissolved inorganic carbon (DIC) data for these they accounted for the pCO2 disequilibrium

between ocean and atmosphere at the outcrops cruises is about five times larger than current

measurements, which are accurate to 1 to An independent approach to the empirical

inventory calculations is to utilize ocean

circula-2mmol kg−1 based on calibration with certified

reference materials (CRM) tion models Box models calibrated using

observa-tions of bomb 14C penetration gave the initial Several different methods of determining the

oceanic anthropogenic inventories have been estimates of global DICanthro uptake by the ocean

(Oescher et al., 1975) More sophisticated general developed in the past The anthropogenic CO

2 signal is relatively small, compared with the nat- circulation models have subsequently been used

to estimate uptake for individual ocean basins or ural DIC spatial variations, and the key to any

such method is how to separate out the natural ocean sections In our comparison we utilize the

NCAR and the Princeton versions of the ocean biological and physical patterns in DIC Brewer

(1978) and Chen and Millero (1979), henceforth biogeochemistry model (OBM), both of which are

derived originally from the Cox and Bryan model referred to as B-78 and CM-79, respectively,

pro-posed methods based on subtracting the soft tissue (Bryan and Lewis, 1979)

The objective of our work is to perform a remineralization and the calcium carbonate

dis-solution components of the DIC by utilizing critical comparison among inventory estimates

along a meridional section in the Atlantic Ocean changes in oxygen and alkalinity (T Alk) with

depth, or along isopycnals, together with the from 62°N to 43°S The section under investigation

was occupied as part of the Ocean-Atmosphere remineralization quotient (Redfield ratio) between

carbon and oxygen The results of this exercise Carbon Exchange Study (OACES) of the National

Oceanic and Atmospheric Administration are critically dependent on the remineralization

quotient used Assumptions about surface alkalin- (NOAA) The southern segment from 5°N to 43°S

along nominally 25/32°W (Fig 1) was occupied ity and DIC values in wintertime outcrop areas

have a significant influence on the inventory estim- in the austral winter (July) of 1991 The northern

section from 5°S to 62°N along 20/25°W was ates as well Mixing of water masses is accounted

for in a simplistic manner by normalizing the performed in July/August of 1993 To estimate

DIC and T Alk outcrop values at high latitude,

T Alk and DIC to constant salinity In these

analyses it is assumed that the oldest waters in we used data from the Transient Tracers in the

Ocean study, TTO (TTO, 1986) and the South the interior do not contain any DICanthro at the

time of measurement The initial work by B-79 Atlantic Ventilation experiment (SAVE/HYDROS)

(Takahashi, pers com.; SAVE (1992)) The carbon and CM-79 had as major objective to estimate

preanthropogenic atmospheric partial pressure of inventory obtained along the cruise track is not

extrapolated basinwide since there are likely CO

2( pCO2a) assuming that the surface ocean was

at equilibrium with the atmosphere appreciable East–West gradients with the Western

Basin containing more DICanthro due to rapid Although, what now appear to be reasonable

estimates of preanthropogenic pCO

2a were transport within the deep western boundary

cur-rent (Chen, 1982; Gruber et al., 1996; Ko¨rtzinger obtained compared to historical atmospheric CO2

records from CO

2in bubbles of ice cores (Neftel et al 1998) The basin wide inventory can be

Fig 1 Data for the analysis were obtained from the observations at the stations depicted The South Atlantic section (circles) was occupied in July 1991 (S Atl-91), and the northern leg (triangles) was occupied in July 1993 (N Atl-93)

as part of the NOAA/OACES program.

performed with high accuracy after completion of accurate measurements of DIC, the fugacity of the CO

2survey sponsored by the US Department CO2 ( f CO2) or total alkalinity (T Alk), oxygen

2), temperature (T ), and salinity (S) are

neces-We will first describe the dataset as it pertains sary The f CO

2is the pCO2corrected for a small

to our analysis with emphasis on accuracy, preci- non-ideality of CO2 in air ( f CO2#0.996 pCO2) sion, and internal consistency of the relevant data Furthermore, nitrate or phosphate measurements between the two cruises The methodology for are needed for estimating preformed endmember calculation of DICanthro and the results from the values The analytical methods for the datasets

different estimates is given in Section 3 The are reported in Castle et al (1998), Forde et al agreement and differences between the methods is (1994), and Lee et al (1997) Changes in alkalinity detailed by comparing remineralization quotients are used to correct for increases in DIC due to and correspondence with chloro-fluoro carbon the calcium carbonate dissolution during trans-ages, (tCFC) in the last part of the paper port of water into the interior Here it is calculated

from f CO

2and DIC rather than using the T Alk observations from the cruises The quality of the

2 Description of field data

T Alk data during the 1991 cruise in the South Atlantic, henceforth called S Atl-91, is not as good Since the DICanthro is at most 3% of the DIC

in the surface and rapidly decreases with depth, as subsequent measurements during the North

Atlantic cruise in 1993 (N Atl-93) Best agreement than slightly supersaturated as typical in the

sur-face ocean A likely cause for the offset is a bias between measured alkalinity and calculated

alka-linity from f CO

2 and DIC is obtained using the in the thiosulfate standard reagent To account for

this discrepancy a value of 7.5mmol kg−1 was constants of Mehrbach et al (1973) as determined

with the program of Lewis and Wallace (1998) added to all N Atl-93 O

2values.

For the S Atl-91 dataset the difference between

measured and calculated T Alk from DIC and

f CO

2 is 1.4±10.3 mEq kg−1 (n=223) while 3 Determination of anthropogenic CO

2

for the N Atl-93 dataset the offset is −3.8

±4.8 mEq kg−1 (n=1557) The method of determining the DICanthro by the

empirical method is described in detail elsewhere For both cruises DIC measurements were

per-formed using coulometers with a SOMMA (single (Chen and Millero, 1979; Chen, 1982; Chen et al.,

1990; Gruber et al., 1996) Here we will only operator multi-parameter metabolic analyzer)

inlet system (Johnson et al., 1993) Certified refer- reiterate the important points and emphasize the

differences between the method of CM-79 and the ence materials (CRM Batch #16) were provided

by Dr Dickson of SIO The analyses for each recent method developed by GSS-96 Both

methods have the same basic underlying assump-instrument were corrected to the cruise average

CRM values for the instrument by applying a tion that the DIC

anthro signal can be determined

by differences in preformed carbon along iso-constant offset For each cruise the corrections

were less than 2mmol kg−1 while the standard pycnals, or with depth, by subtracting the

remin-eralization and carbonate dissolution component deviations of all CRMs analyzed during the cruises

were less than 1.5mmol kg−1 DIC measurements of the observed DIC The fundamental equation

for both methods is the same in that:

from the N Atl-93 and S Atl-91 cruises were

compared in the region of overlap between 5°S

and 5°N using deep water where calibration offsets

would be most apparent No offset in DIC was

DIC anthro=DDIC−0.5[DT Alk+RN:O(DAOU)] +RC:O(DAOU),

DX=Xx−X

0,

detected between the N Atl-93 and S Atl-91 data

f CO

headspace with a water sample at constant

temper-ature (20.00°C) and detection with a non-dispers- where DICanthro is the anthropogenic DIC

com-ponent, AOU is the apparent oxygen utilization, ive infrared detector as detailed in Chen et al

(1995) Precision of the f CO

2(20) measurements and RN:Oand RC:Oare the remineralization

quo-tients between NO3 and O2, and DIC and O2, based on 31 replicate samples taken at different

depths was 0.2% for the N Atl-93 cruise, while respectively DX is the difference in the in situ

measurements, X

x and the preformed preanthro-the precision of preanthro-the S Atl-91 cruise was estimated

at 0.3% Again no detectable bias was observed pogenic value at the outcrop, X0, where X is DIC,

T Alk, or AOU The R

N:O(DAOU) is the ‘‘nitrate

in the deep waters in the region of overlap for the

two cruises Oxygen values had a precision of contribution’’ to alkalinity due to soft tissue

remin-eralization (Brewer et al., 1975) A summary of

1mmol kg−1 for both cruises The deep water O2

data from the S Atl-91 cruise showed good corres- symbols can also be found in Section 7

There are several major sources of uncertainty pondence with the SAVE 5/HYDROS 4 cruises

along similar cruise tracks in the calculation of DICanthro from (1) including

uncertainty in remineralization quotients and pre-However, the N Atl-93 O

2data had a significant

offset compared to the Oceanus-202 cruise along formed concentrations in the outcrop region The

remineralization quotients, R, are not well known the same track (Tsuchiya et al., 1992) This offset

is apparent in the whole water column but can be and possibly vary with depth (Minster and

Boulahdid, 1987) Anderson and Sarmiento (1994) best quantified in the deep water where a difference

of 7.5±1 mmol kg−1 was observed with the assign the following ratios and uncertainties to

the remineralization quotients: P5N5C5–O=

N Atl-93 data being lower We assume that the

N Atl-93 data are biased because the surface water 1516±15117±145170±10 based on a thorough

isopycnal analysis of the world’s oceans In their values are at or slightly below saturation rather

work they excluded the North Atlantic where that NT Alk0 can be estimated for each sample The

preanthropogenic (AOU

0) and current apparent mixing of different water masses confounded their

oxygen utilization are assumed to be 0 at the out-analysis If we assume these are typical

uncertain-crop Oxygen concentrations at the surface could ties for remineralization quotients and propagate

differ by several percent either because of supersat-the error in C5–O, supersat-the uncertainty in R

C:Ois 12%. uration due to bubble entrainment, or because of This has a significant influence on the calculated

undersaturation due to outcropping of undersatur-DICanthro Fig 2 shows the fractional error caused ated water and insufficient time for equilibration

by this uncertainty relative to the observed

with the atmosphere The largest uncertainty in DIC

anthro level for the two empirical methods CM-79 arises because of assumptions in determin-described below The rapidly increasing fractional

ing theDDIC term DIC0 is not known such that error is caused by a combination of increasing

as a first-step DIC

ot, the present-day preformed DAOU and decreasing DICanthro for older waters DIC in the wintertime outcrop, is estimated In The preanthropogenic preformed values in the

CM-79, salinity normalized DIC

ot is regressed wintertime outcrop have to be estimated and

inter-against temperature at the surface such that DICot polated against temperature, or other

quasi-conser-can be determined for all (subsurface) water samples vative parameter It is assumed that the alkalinity

with knowledge of T has not changed due to anthropogenic perturbation

The shortcomings of the CM-79 method have

In CM-79, salinity normalized alkalinity is regressed

been clearly described in Broecker et al (1985) against temperature in the surface mixed layer such

and have been addressed, in part, in GSS-96 While CM-79 accounts for mixing processes by normalizing to salinity, this does not fully account for unique temperature dependences of NT Alk

0 and NDIC

otfor different outcrop regions This is particularly an important issue in the Atlantic where the waters from northern origin mix with waters from the south that have very different preformed quantities GSS-96 estimates preformed alkalinity for the entire Atlantic by creating a multiple linear regression of alkalinity against salinity and the quantity ‘‘PO’’ (Broecker, 1974) This relationship is thought to be valid for all outcrops and shows good agreement with sparse wintertime data as well

The second significant improvement of GSS-96 over CM-79 is the attempt to account for DICanthro

in the upper thermocline where isopycnals have been ventilated during the period of anthropogenic change This is done by determining the water mass age using the helium–tritium (3H–3He) dating technique However,3H–3He ages are not true watermass ages because of mixing (Jenkins, 1988) and tend to underestimate age beyond about

15 years (Doney et al., 1997) This method only yields ages since the large bomb tritium inputs, 30

to 40 years ago at which point half the atmospheric Canthro had already entered the ocean

Fig 2 Fractional error in DIC

anthro(%) plotted against 3.1 T he empirical method according to Chen and DIC

anthroassuming a 12% uncertainty in the C5O ratio. Millero (1979)

(a) DIC anthrocalculated following the methods of CM-79; The first application of the method of CM-79 ( b) DIC anthro calculated following the methods of

GSS-96. was to estimate preanthropogenic pCO

2a values.

Subsequently, the utility was expanded to estimate ±4 mEq kg−1 for T >20°C It is fit to:

DIC

anthro in the ocean For the initial work, the

anthropogenic perturbation was set, by definition, NT Alk0=2291–2.69(T−20)−0.046(T−20)2

for T<20°C in the North Atlantic, and (4)

at 0 at the surface and decreased with depth

Subsequently, modifications to the approach

were made, using the term defined below as NT Alk0=2291–2.52(T−20)+0.056(T−20)2

for T<20°C in the South Atlantic (5) (DNDIC)min, such that the deep water attain a

DICanthro of 0 and surface waters the showed full The standard deviation is ±5 mEq kg−1 anthropogenic burden Thus, what we describe as

between the fits and the observations These fits the CM-79 method actually includes the

refine-along with N Atl-93 and S Atl-91 surface water ments as discussed in a later publication (Chen,

data (0–40 m) data are shown in Fig 3a The 1993)

algorithms yield a slightly lower alkalinity than The pertinent calculations for the CM-79

observed in our datasets The average surface method are as follows:

water NT Alk for T>20°C for our data is 2295±5 mEq kg−1 versus 2291±4 mEq kg−1 for DICcmanthro=DNDICcm−0.5

the more comprehensive dataset used by Millero

×[DNT Alk+RN:O(AOU)] et al (1998) The differences are not statistically

significant and have little influence on the

DIC anthrocalculation since most of the DICanthro inventory resides in water with T<20°C DNDICcm=NDIC−NDICot−(DNDIC)min For NDICot we chose a relationship for water

(3) originating from the southern hemisphere of NDICot=2227−11.5T and for water from the where DICcmanthro is the anthropogenic component

northern hemisphere of NDIC

ot=2143−7.58T

of the measured DIC according to CM-79

These relationships were obtained by fitting the DNT Alk is difference in salinity normalized T Alk data from the TTO cruises (TTO, 1986) and (=Talk*35/S) between the outcrop and point of METEOR cruises (Chipman et al., 1991) as they measurement.DNDICcm is difference between the include an abundance of samples at T<15°C No measured salinity normalized DIC and the current correction was made to account for changes in day outcrop value, NDICot, corrected for the NDICotcaused by the different times at which the minimum value obtained at depth, (DNDIC)min cruises occurred since the correction would be The (DNDIC)min is the lowest value of significantly less than the uncertainty in the rela-(NDICc−NDICot) observed at depth, where tionships The correspondence of the relationships NDICc=NDIC−0.5[DNT Alk+RN:O(AOU)]+ with our data is reasonable over the common R

C:OAOU This causes a bias if the anthropogenic temperature range (Fig 3b) For the S Atl-91 data, signal has penetrated to the bottom, as occurs in which only extends to mid-southern latitude, our section, or if waters are mixtures of endmem- the

DNDICot (observed–calculated) equals bers with different characteristic NDICot and 20±20 mmol kg−1, while for the North Atlantic

NT Alk0 Typical values of (DNDIC)min are −50 the difference is 6±9 mmol kg−1

to−60 mmol kg−1 The empirical relationships to determine For our current exercise, we used the

remin-NT Alk0 and particularly NDICot are critical and eralization quotients recommended in Anderson can vary significantly depending on the formula-and Sarmiento (1994) Our surface water data do tion used (Fig 3c) NT Alk

0establishes the correc-not cover temperatures less than 10°C, so the tion for CaCO3 dissolution when water ages and surface water NT Alk0 algorithms with temper- moves into the interior The NDICot are the ature were obtained from a compilation of data current preformed values on which the entire from the N Atl-93, SAVE, and TTO cruises for anthropogenic inventory is based The differences the N Atlantic, and the S Atl-91 and GEOSECS in the relationships of NDIC

otversus temperature cruises for the S Atlantic (Millero et al., 1998) can be illustrated from the range algorithms for

NDIC

ot developed for different parts of the The NT Alk

0 is nearly constant at 2291

Fig 3 Relationships between sea surface temperature and salinity normalized (S =35) alkalinity (NT Alk), and total dissolved inorganic carbon (NDIC) used in the analysis according to CM-79 The relationships show large di fferences

in northern and southern endmembers for NT Alk (a) and NDIC ( b) The open circles are the surface observations during N Atl-93 while the solid circles are those from S Atl-91 (c) Shows the relationships for di fferent cruises and temperature ranges and locations (Table 1).

Atlantic, and with different datasets Because the NDICotrelationships can be attained by creating

relationships for smaller geographic regions calculated NDICot is strongly dependent on T

(Table 1), the choice of relationship can alter the The choice of using northern or southern

NT Alk0 and NDICot endmembers for a particular total inventories In particular, since most of the

ocean has temperatures less than 10°C, differences water parcel will have a significant influence on

the results as well For the thermocline and

sub-in the relationship at low T , where we have little

surface water data, will have a large influence on thermocline waters with temperatures between

20°C and 3°C, we chose 15°N as a division based the DIC

anthro calculation Improvements in the

Table 1 Algorithms of NDIC

otversus temperature

All surface waters S Atl-91 and N Atl-93: NDIC

ot(±9.2)=2142–7.226T (data from OACES S Atl-91 and N Atl-93)

S Atl-91: 15 °S–42°S: NDIC ot(±7.8)=2143–7.596T

(data from OACES S Atl-91 and N Atl-93) T =10–24°C

S Atl-91 and N Atl-93: 15 °S–15°N: NDIC

ot(±9.0)=2390−16.4T (data from OACES S Atl-91 and N Atl-93) T =24–30°C

N Atl-93: 15 °N–60°N: NDIC ot(±8.2)=2146−7.48T

(data from OACES S Atl-91 and N Atl-93) T =8–25°C

South Atlantic METEOR (Chipman et al., 1991): NDIC

ot=2227−11.5T

T =0–22°C TTO N Atlantic (TTO, 1986): NDIC

ot=2143−7.58T

T =0–22°C TTO N Atlantic and N Atl-93: NDIC ot=2119.7−6.3T

Atlantic and Southern Ocean (Chen, 1982): NDIC

ot=2219−11.8T , T <19.5 NDIC

ot=2115−6.4T , T >19.5 Greenland and Norwegian Sea (Chen et al., 1990): NDIC

ot=2171−5.7T

on the sharp change of salinity along isopycnals cropping and intermediate/deep water formation

areas Higher penetration is apparent in the north-Large tritium and CFC gradients also suggest

slow exchange across this boundary (Broecker ern latitudes compared to the south The large

DICcmanthro signal of greater than 50mmol kg−1 in and Ostlund, 1979; Doney and Bullister, 1992)

For deep waters (T<3°C), the equator was chosen the subpolar North Atlantic with a subsurface

maximum corresponds to the recently ventilated

as a division based on the predominance of

Antarctic waters south west of the Mid Atlantic Labrador Sea Water The low values in the top

1000 m subtropical Atlantic cannot readily be Ridge and water of northern origin north (east) of

the Ridge as indicated by large silica gradients In explained and could be an artifact of our choice

of NT Alk0 and NDICot The separation of waters reality, there will be a large region with an

admix-ture of waters from northern and southern origin ventilated to the bottom in the north and more

isolated southern waters is apparent between 15° Once the relationships of NDIC

otand NT Alk0 have been established, the difference between and 20°N A uniform penetration of DICcmanthro is

observed between 10°N and 30°S with a very NDICc and NDICot is determined The value will

be 0 at the surface and reach a minimum (negative) rapid drop off and barely detectable DICcmanthro

levels below 800 m Further southward in the value at depth, (DNDIC)min Subsequently, this

minimum value is subtracted from each subtropics penetration increases to over 1500 m

A rigorous analysis of uncertainties is difficult NDIC−NDICot determination to estimate

DICcmanthro The deep water values are averaged because of somewhat subjective choices of

pre-formed values and remineralization quotients over 20° latitude bands south of 20°N and this

(DNDIC)min is subtracted from all the values Previously, the uncertainty has been estimated

ranging from 15mmol kg−1 (Chen, 1982) to above From 20°N northward, (DNDIC)min

between 0° and 20°N is used The data is sub- 10mmol kg−1 (Chen, 1993) with the improvement

largely because of smaller uncertainties in the sequently gridded corresponding to the Princeton

OBM depth bins for comparison with the model analytical precision We estimate our uncertainty

at a similar level as the recent work by Chen results Fig 4a shows the expected features of deep

penetration of DICcmanthro in the thermocline out- Fig 5a shows the specific inventory of DICcmanthro

Fig 4 Cross sections of DIC

anthrofor the Atlantic basin following the cruise track shown in Fig 1 using the methodo-logy of CM-79 (a); the approach of GSS-96 (b); the output from the Princeton OBM (c); and NCAR OBM (d) The data shown in (a) and ( b) were gridded in the same bin sizes as the OBM (5 ° horizontal and 12 depths increasing

in size going downward) before contouring The 0.1 pmol kg −1 CFC-11 isopleth (corresponding roughly to a tCFC

of 30–35 years) is shown by the plus symbols.

From our data in the top 100 m, we determined

a relationship for T Alk

0:

T Alk 0=(278.4+57.01S+0.0074NO)

where NO=O2+R

N:ONO3 (Broecker, 1974). This result is similar to the calculated T Alk0 obtained by GSS-96 from older data in the

N Atlantic The standard deviation between the observed T Alk0 and predicted T Alk0 from (8) is 5.2mEq kg−1 DICeqfor our work were calculated from the derived T Alk

0(8) and pCO2a=280matm

Fig 5 Specific water column inventories using the meth- using the dissociation constants of Mehrbach et al

odology of GSS-96 (crosses); CM-79 (solid circles); (1973) These resulting DIC

eq are subsequently

output from the Princeton OBM (open squares); and linearized as in GSS-96 using a least squares linear

NCAR OBM (open triangles).

fit:

binned in 5° latitude bands The specific inventory

varies by almost a factor of ten with very high

DIC eq=2082.787+0.782249(T Alk0−2320)

−5.10296(S−35)−8.99668(T −9) r2=1.00, n=335

specific inventories in the north, intermediate

levels in the south and lower inventories at low

(9) latitudes

with a standard deviation between the calculated 3.2 T he empirical method according to Gruber, DICeq and the linearized DICeq of 1.4 mmol kg−1 Sarmiento, and Stocker (1996) The temperature, T is in°C

The DICGanthro is subsequently calculated, The analysis according to GSS-96 employs the

accounting for the air–sea disequilibrium, same general principles as the earlier methods but

expressed in terms of DIC (DCdis), by two different with some significant improvements The method

methods depending if the particular isopycnal in employs a quasi-conservative quantity C*:

the interior of the Atlantic contains measurable C*=DIC−0.5(T Alk+RN:OO2)−RC:OO2 (6) CFCs or not For isopycnals that are CFC-free in

the interior (CFC−11<0.005 pmol kg−1), the

A similar equation can be set up for C*0 in which

disequilibrium, DCdis is the DC* of the interior DIC, T Alk, and O

2 are replaced by their prean- endmember of the particular isopycnal in the thropogenic preformed values The difference is

interior and DICGanthro=DC*−DCdis For the defined asDC*:

Atlantic the density surfaces referenced to 2000 dB, DC*=C*−C*0 =DDICG−0.5DT Alk s2 of greater than 36.85 are CFC-free in the

interior For isopycnals that have measurable +(RC:O−0.5R

N:O)AOU, (7) CFC-11 on the interior endmember, the

water and the corresponding atmospheric CO2 DDICG=DIC−DICeq(S, T , T Alk

0)pCO2a=280. level, pCO

2(tCFC), for that age:

DDICG is the difference in measured DIC and the

DCdist=DDICGt −0.5DNT Alk DIC in equilibrium with a pCO

2aof 280matm at

in situ S, T , and preformed alkalinity value, T Alk0 +(RC:O−0.5R

DC* is similar to the definition of DICanthro out- where

lined above in (1) More precisely DC* is the

DICanthro if the surface ocean was in equilibrium DDICGt=DIC−DICeq(S, T, T Alk0)pCO2(tCFC)

DICGanthro=DC*−DC

dist.

with the atmosphere with a pCO

2a of 280matm, discounting any influence of non-linear mixing

effects along isopycnals ThetCFC is calculated by converting the CFC-11