Journal of Marine Systems 46 (2004) 169 – 179 www.elsevier.com/locate/jmarsys

Definition, properties, and Atlantic Ocean distribution of the new tracer TrOCA Franck Touratier *, Catherine Goyet Universite´ de Perpignan, EA 1947 LBDSI, 52 avenue Paul Alduy, 66860 Perpignan, France Received 10 July 2003; accepted 20 November 2003

Abstract Natural and anthropogenic tracers in the ocean are widely used not only to better understand water masses circulation and mixing but also to understand and quantify the ocean uptake and storage capacity of greenhouse gases. However, since each tracer is different, it is best to use the complementarity of several tracers to unequivocally identify the various water masses. Here we illustrate the conservative properties and the spatial distribution of the new composite tracer TrOCA (Tracer combining Oxygen, inorganic Carbon, and total Alkalinity) using oxygen (O2), dissolved inorganic carbon (TCO2), and total alkalinity (TA), from the Atlantic Ocean. The significant accuracy improvement of TCO2 and TA measurements since the 1970s, combined to a large effort in measuring these parameters during large scale cruises throughout the Atlantic Ocean, makes this tracer TrOCA an additional tool in analyzing water mass distribution. This tracer is shown to be conservative in intermediate, deep, and bottom waters. For instance, we show that the independence of TrOCA from other tracers provides further information on the origin and mixing of the main Atlantic water masses. Furthermore, TrOCA combined with the composite tracer NO, in particular the ratio TrOCA/NO, can be used to unequivocally identify and separate the Antarctic Intermediate Water, the Antarctic Bottom Water, and the North Atlantic Deep Water. D 2004 Elsevier B.V. All rights reserved. Keywords: Water masses; Carbon cycling; Chemical tracers

1. Introduction The long-term objective of this study is to develop a new concept to estimate the anthropogenic CO2 concentrations in the ocean (Touratier and Goyet, 2004, this issue). However, since this concept relies upon the fundamental properties of a new tracer (TrOCA), here we focus only on the definition, distribution, and properties of this tracer. We will show that TrOCA complements other tracers such as

NO and PO (Broecker, 1974), PO4* (Broecker et al., 1986), DC* (Gruber et al., 1996), N* (Gruber and Sarmiento, 1997), CFCs, and tritium—in best identifying different water masses. In the ocean, the average photosynthesis and aerobic respiration are usually described by the Redfield et al.’s (1963) equation: ðCH2 OÞ106 ðNH3 Þ16 ðH3 PO4 Þ Respiration;decomposition

þ 138O2 V 106CO2 Photosynthesis

* Corresponding author. E-mail address: [email protected] (F. Touratier). 0924-7963/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jmarsys.2003.11.016

þ 16HNO3 þ H3 PO4 þ 122H2 O

ð1Þ

170

F. Touratier, C. Goyet / Journal of Marine Systems 46 (2004) 169–179

Obviously all compounds involved in Eq. (1) are nonconservative tracers since variability in their concentration does not result solely from physics (circulation and mixing), but also from local transformations (i.e. chemical and biological processes). Using data of the 1972 Atlantic Geosecs Expedition, the idea of Broecker (1974) is to combine two of these compounds to obtain conservative properties that could be used to track the origin of the Atlantic water masses. Since the consumption of oxygen is balanced by the production of nutrients during respiration and decomposition, two potential tracers were identified (i.e. NO = 9NO 3 + O 2 ; and PO = 135PO 4 + O 2 ). Broecker’s (1974) paper demonstrates the conservative properties of tracer NO, and its ability to distinguish water masses of different origins. Since the NO3/ PO4 ratio remains constant ( f 15; atomic ratio) throughout the world ocean, only one tracer (NO or PO) is usually used. Craig (1969) first introduced the idea of combining TCO2 (i.e. the sum of carbonate species in seawater) with O2 to form a tracer ‘‘CO’’. However, given the poor accuracy of TCO2 measurements at that time, NO was shown to be a more efficient tracer than CO (Broecker, 1974). More recently, Rı´os et al. (1989) proposed the tracer CAO, which is similar to tracer CO but corrected for calcium carbonate precipitation. However, since the Geosecs expeditions, and especially during the World Ocean Circulation Experiment (WOCE), the accuracy of the TCO2 measurements has improved by approximately one order of magnitude (Lamb et al., 2002). Typically, the accuracy in the measurements of NO3 remained unchanged ( f 0.1 Amol kg 1; e.g. Broecker, 1974; Lamb et al., 1995; Johnson et al., 2002), whereas the accuracy for TCO2 increased from 15– 20 to < 2 Amol kg 1 (e.g. DOE, 1994; Johnson et al., 1995). Thus, this improvement in the TCO2 measurement accuracy provides an opportunity to further test the conservative properties of a tracer based on TCO2 and O2. At depth, TCO2 varies according to respiration and dissolution of calcium carbonate (Eq. (2)). The latter process is estimated using total alkalinity (TA; Brewer, 1978; Goyet and Brewer, 1993). Note that in Eq. (2), the stoichiometric coefficients for O2 and CO2 (i.e. 165 and 123, respectively) differ from those of Redfield et al. (1963); i.e. 138 and 106, respectively; see Eq. (1). The former set of coefficients corresponds

to the revisited Redfield ratios estimated by Ko¨rtzinger et al. (2001), using North Atlantic data. The O2 coefficient of 165 is in close agreement with estimates of Takahashi et al. (1985) and Anderson and Sarmiento (1994), whereas the CO2 coefficient of 123 is corrected for the biaising effect of the anthropogenic CO2 (Ko¨rtzinger et al., 2001). The influence of these coefficients on the TrOCA value is further discussed by Touratier and Goyet (2004, this issue).   123 1 17 TCO2 ¼ O2 þ TA þ O2 ð2Þ 165 2 165 Rearranging Eq. (2), and applying the same rules of construction as Broecker (1974) did for NO or PO, leads to the definition of a new tracer called TrOCA (Tracer combining Oxygen, inorganic Carbon, and total Alkalinity):   165 1 TCO2  TA ð3Þ TrOCA ¼ O2 þ 131:5 2 TrOCA ¼ O2 þ 1:2TCO2  0:6TA

ð4Þ

Given the accuracy in O2 ( f 1.1 Amol kg 1; Johnson et al., 2002) and TA measurements ( f 4 Amol kg 1; Millero et al., 1998), the accuracy for tracer TrOCA is of f 3.57 Amol kg 1 (see Touratier and Goyet, 2004, this issue). Considering that the mean value for TrOCA throughout the Atlantic Ocean is approximately 1425 Amol kg 1, this corresponds to a percentage of error of approximately 0.25%. This percentage is four times lower than that estimated by Broecker (1974) for tracer NO. Below, we use data from the CARINA project (CARbon dioxide IN the Atlantic Ocean; http:// www.ifm.uni-kiel.de/fb/fb2/ch/research/carina/). This database includes data from 80 cruises/time series for the 1972 – 2001 period, over the whole Atlantic Ocean (Fig. 1). From this data base, salinity (S), potential temperature (h), O2, NO3, TCO2, and TA data were available at f 95%, 92%, 63%, 60%, 40%, and 26% of the 165,904 sampling points, respectively. Note that in the following work, the pre-WOCE data from the CARINA database are also used to compute tracer TrOCA. Consequently, the TrOCA accuracy is expected to be slightly under 0.25%. However, for the purpose of the present study, the results and overall conclusion remain unchanged since

F. Touratier, C. Goyet / Journal of Marine Systems 46 (2004) 169–179

171

Fig. 1. Map showing the station position for cruises available from the CARINA project.

they do not depend critically upon the absolute value of TrOCA.

2. Is TrOCA a conservative water-mass tracer? When two water masses are mixed, a property is shown to be conservative when its value remains on a

straight mixing line. Any deviation from this line is indicative of local transformation (e.g. chemical or biological processes) and/or exchange with other compartments (e.g. atmosphere, sediments). To provide evidences of the conservative nature of tracers NO and TrOCA, properties from the deep southern Atlantic (from the equator to 40jS; and h V 14 jC) are plotted against salinity (Fig. 2). This

172

F. Touratier, C. Goyet / Journal of Marine Systems 46 (2004) 169–179

Fig. 2. Property-salinity diagrams for deep waters (h V 14 jC) located between the equator and 40jS. The non-conservative properties are (a) NO3, (b) O2, (d) TCO2, and (e) TA; the conservative properties are (c) NO, (f) TrOCA, (g) h, and (h) ratio TrOCA/NO. Note that points located north and south of 25jS are in black and gray, respectively. The main Atlantic water masses are indicated in panel (g): SACW (South Atlantic Central Water), AAIW (Antarctic Intermediate Water), NADW (North Atlantic Deep Water), and AABW (Antarctic Bottom Water).

F. Touratier, C. Goyet / Journal of Marine Systems 46 (2004) 169–179

173

Fig. 3. Latitudinal variation of (a) NO3, (b) O2, (c) NO, (d) TCO2 and TA, (e) TrOCA, and (f) h of deep waters characterized by h V 2.5 jC and 34.88 V S V 34.89 (see text for details).

174

F. Touratier, C. Goyet / Journal of Marine Systems 46 (2004) 169–179

area is selected because of the largest longitudinal variability in O2 content (from f 250 Amol kg 1 close to the Brazilian coast to < 20 Amol kg 1 in the Angolan and Namibian deep waters). Given the four most important water masses present in this area (Fig. 2g), it is clear that tracers NO and TrOCA are much more conservative than their intrinsic components (i.e. NO3, O2, TCO2, and TA). Contrary to the h– S and NO – S diagrams, however, the TrOCA values repre-

sentative of SACW ( < 1425 Amol kg 1; Fig. 2f) deviate significantly from an expected straight line. Since TrOCA increases with depth, the noted deviation (i.e. the non-conservative portion of the TrOCA – S relationship) is related to a process occurring approximately in the upper 1500 m of the water column (probably due to the air – sea CO2 flux). NO and TrOCA can also be used to discriminate waters of different origins. For example, Fig. 2c and f shows that

Fig. 4. Latitudinal variation of the TrOCA – h diagram.

F. Touratier, C. Goyet / Journal of Marine Systems 46 (2004) 169–179

profiles located north and south of the latitude 25jS have distinct NO and TrOCA signatures, as it is the case for h (Fig. 2g). Another example is given by the deep water properties characterized by S and h ranges of 34.88 –34.89 and V 2.5 jC, respectively, over the whole Atlantic (Fig. 3). This corresponds to the 34.886 isohaline (70% NADW; 30% AABW) analyzed by Broecker (1974), using the Geosecs data between 35jS and 35jN. Broecker’s conclusion that NO remains conservative at these latitudes is confirmed by the post-Geosecs data; it however deviates significantly from conservation at latitudes >40jN (Fig. 3c). Similar conclusions are reached for tracer TrOCA (Fig. 3e). Note however that the h signature of waters located north of 65jN (h < 0 jC; Fig. 3f) indicates that these waters are not representative of the previous NADW/AABW mixture. These cold and dense waters, partly formed within the Greenland and Norwegian seas, belong to the Arctic Bottom Water (ABW; e.g. Tomczak and Godfrey, 2001).

175

3. Distribution of TrOCA Using the NO – h diagram, Broecker (1974) shows that clear differences exist between waters forming at the northern and southern end of the Atlantic. The latitudinal variations of the TrOCA – h diagram (Fig. 4) do not provide information of this nature. For latitudes ranging from 50jN to 50jS, linear or sigmoid relationships could be used to describe the decrease in TrOCA with increasing h. However, at higher latitudes similar relationships cannot be used because of the large TrOCA variability within the reduced h range. Results illustrated in Fig. 4f between 50jS and 65jS should be analyzed with caution since few data are available for this region (Fig. 1). The highest TrOCA values (>1600 Amol kg 1; Fig. 4e, g) are systematically located in the Greenland and Norwegian seas (close to 75jN), and it is obvious that the formation of deep waters in the surface layer of this region (e.g. Anderson et al., 1998, 2000) could alter the conservation of TrOCA through net atmospheric inputs of CO2.

Fig. 5. North – south distribution of tracer TrOCA (Amol kg 1) in the Atlantic Ocean.

176

F. Touratier, C. Goyet / Journal of Marine Systems 46 (2004) 169–179

Fig. 6. Longitudinal distribution of tracer TrOCA (Amol kg 1) in the Atlantic Ocean: (a) for the latitude belt 40 – 50jN; and (b) for the latitude belt 10 – 40jS.

F. Touratier, C. Goyet / Journal of Marine Systems 46 (2004) 169–179

In the Atlantic Ocean , tracer TrOCA varies between 1087 and 1693 Amol kg 1. According to Fig. 2f and g, typical TrOCA values of >1500, f 1440, f 1450, and < 1400 Amol kg 1 can be assigned to AABW, NADW, AAIW, and SACW, respectively. The high salinity waters belonging to the North Atlantic Central Water (NACW) are also clearly characterized by TrOCA values < 1400 Amol kg 1. The distribution of the tracer TrOCA throughout the Atlantic Ocean is shown in Fig. 5 (north– south section) and Fig. 6 (two longitudinal sections). The highest TrOCA values (>1500 Amol kg 1) are located in the deepest part of the southern ocean (spreading of AABW) and at all depths for waters located north of ca. 65jN (Fig. 5). The detailed analysis of the TCO2, TA, and O2 properties in the Greenland and Norwegian seas, strongly confirms the idea that oceanic pumping of atmospheric CO2 is the primary process by which deep invasion of CO2 and thus elevated TrOCA concentrations at depth could be explained. Surface and subsurface waters of temperate and low latitude areas are characterized by TrOCA values

177

< 1400 Amol kg 1 (i.e. SACW and NACW). Since the TrOCA signatures for AAIW and AABW are almost similar ( f 1440 – 1450 Amol kg 1; see Fig. 2f), it is difficult to identify these water masses from the north – south section (Fig. 5). In the northern hemisphere, a longitudinal section corresponding to the latitude belt 40jN–50jN (Fig. 6a) indicates that the western basin is characterized by high TrOCA values within the layer 1500 –3000 m. These high TrOCA values probably result from the influence of the Deep Western Boundary Current (DWBC) flowing southward. Variability of TrOCA in the surface and subsurface layers is also much more important in the west than in the east. In the southern hemisphere (latitude belt of 10 –40jS; see Fig. 6b), however, the longitudinal variability of TrOCA in the upper 1000 m is much lower. On the abyssal seafloor, AABW spreads northward in the western basin (depth < 4000 m), as noted by Durrieu De Madron and Weatherly (1994) and Stramma and England (1999). As well known, AABW is identified as far as 40 – 50jN in the western basin (Tomczak and

Fig. 7. North – south distribution of the ratio TrOCA/NO in the Atlantic Ocean.

178

F. Touratier, C. Goyet / Journal of Marine Systems 46 (2004) 169–179

Godfrey, 2001). This water mass is characterized with slightly diluted TrOCA values ranging from 1490 to 1500 Amol kg 1 (Fig. 6a).

genic CO2 signal, which is a critical issue in the context of global change (Sarmiento and Gruber, 2002).

4. Properties of TrOCA

Acknowledgements

In the Atlantic Ocean, tracer TrOCA has been shown to be conservative except in the upper layer where interactions with atmosphere occur (this is also true for other tracers like S, h, or NO). In contrast with the ratio NO/PO which remains nearly constant, the ratio TrOCA/NO varies by more than a factor 2 between f 2.5 and 7 (Fig. 2h). Fig. 7 illustrates the variability of its distribution in the Atlantic Ocean. In the upper 1000 m, the northern and southern subtropical gyres are well characterized with the highest TrOCA/NO values, and this ratio decreases with increasing depth throughout the Atlantic Ocean. As mentioned above, if tracer TrOCA cannot be used alone to discern easily waters belonging to AABW and NADW, it is very useful to separate AAIW from AABW (Fig. 2f). As a complement, ratio TrOCA/NO is best used to separate NADW and AABW (Fig. 2h). Therefore, given that the ratio TrOCA/NO is highly variable, and that tracers TrOCA and NO are independent, these three entities are each informative and complementary concerning the properties of the main Atlantic water masses. In the latitude range 40 – 50jN, tracer TrOCA seems even more conservative than NO which shows abrupt change in its values despite nearly constant temperature and salinity values (h f 2 jC, S f 34.886; Fig. 3). The above results illustrate that tracer TrOCA is an independent quasi-conservative property complementary to other tracers such as NO, PO, PO4*, DC*, N*, CFCs, or tritium. As seen above, TrOCA deviates from conservation only in the surface layer, and within the entire water column of the Greenland and Norwegian seas (Figs. 3 and 4). These are areas contaminated with anthropogenic carbon (e.g. Chen and Millero, 1979; Gruber et al., 1996; Orr et al., 2001). Consequently, these deviations were expected since by definition the TrOCA tracer is conservative only in anthropogenic CO2 free waters. Tracer TrOCA is thus of importance since it provides a new tool (1) to identify water masses, and (2) to potentially unravel the anthropo-

This work was supported by a grant from the Lawrence Livermore National Laboratory, University of California, U.S.A. (Subcontract No B516279) and the ACTION project (Anthropogenic Carbon: Temporal Increase, Observations and Numerization), funded by the French PROOF program (PROcessus bioge´ochimiques dans l’Oce´an et Flux). We wish to thank all the scientists and personnel who were involved in the measurements as well as those who created the CARINA program and data-base.

References Anderson, L.A., Sarmiento, J.L., 1994. Redfield ratios of remineralization determined by nutrient data analysis. Glob. Biogeochem. Cycles 8, 65 – 80. Anderson, L.G., Chierici, M., Jones, E.P., Fransson, A., Olsson, K., 1998. Anthropogenic carbon dioxide in the Arctic Ocean: inventory and sinks. J. Geophys. Res. 103, 27707 – 27716. Anderson, L.G., Chierici, M., Fogelqvist, E., Johannessen, T., 2000. Flux of anthropogenic and steady state carbon into the deep Greenland Sea. J. Geophys. Res. 105, 14339 – 14345. Brewer, P.G., 1978. Direct observation of the oceanic CO2 increase. Geophys. Res. Lett. 5, 997 – 1000. Broecker, W.S., 1974. ‘‘NO’’, a conservative water-mass tracer. Earth Planet. Sci. Lett. 23, 100 – 107. Broecker, W.S., Ledwell, J.R., Takahashi, T., Weiss, R., Merlivat, L., Memery, L., Peng, T.-H., Ja¨ne, B., Munnich, K.O., 1986. Isotope versus micrometeorological ocean CO2 fluxes: an order of magnitude conflict. J. Geophys. Res. 91, 10517 – 10527. Chen, C.-T.A., Millero, F.J., 1979. Gradual increase of oceanic CO2. Nature 277, 205 – 206. Craig, H., 1969. Abyssal carbon and radiocarbon on the Pacific. J. Geophys. Res. 74, 5491 – 5506. DOE, 1994. Handbook of methods for the analysis of the various parameters of the carbon dioxide system in seawater, version 2. In: Dickson, A.G., Goyet, C. (Eds.), ORNL/CDIAC-74. Durrieu De Madron, X., Weatherly, G., 1994. Circulation, transport and bottom boundary layers of the deep currents in the Brazil Basin. J. Mar. Res. 52, 583 – 638. Goyet, C., Brewer, P.G., 1993. Biochemical properties of the oceanic carbon cycle. In: Willebrand, J., Anderson, D.L.T. (Eds.), Modelling Oceanic Climate Interactions. NATO ASI Ser., I Glob. Environ. Change, vol. 11. Springer-Verlag, Berlin, Heidelberg, pp. 271 – 297.

F. Touratier, C. Goyet / Journal of Marine Systems 46 (2004) 169–179 Gruber, N., Sarmiento, J.L., 1997. Global patterns of marine nitrogen fixation and denitrification. Glob. Biogeochem. Cycles 11, 235 – 266. Gruber, N., Sarmiento, J.L., Stocker, T.F., 1996. An improved method for detecting anthropogenic CO2 in the oceans. Biogeochem. Cycles 10, 809 – 837. Johnson, K.M., Wallace, D.W.R., Wilke, R.J., Goyet, C., 1995. Carbon dioxide, hydrographic, and chemical data obtained during the R/V Meteor cruise 15/3 in the South Atlantic Ocean (WOCE section A9, February – March 1991). ORNL/CDIAC82, NDP-051. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, Oak Ridge, TN. Johnson, K.M., Wills, K., Koertzinger A., Neill, C., Wallace, D.W.R., 2002. Carbon dioxide, hydrographic, and chemical data obtained during the R/V Meteor cruise 28/1 in the South Atlantic Ocean (WOCE section A8, March 29 – May 12, 1994). In: Kozyr, A., Beaty, T.W. (Eds.), ORNL/CDIAC-135, NDP-079. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge, TN. Ko¨rtzinger, A., Hedges, J.I., Quay, P.D., 2001. Redfield ratios revisited: removing the biasing effect of anthropogenic CO2. Limnol. Oceanogr. 46, 964 – 970. Lamb, M.F., Feely, R.A., Moore, L., Atwood, D.K., 1995. Total carbon dioxide, hydrographic, and nitrate measurements in the Southwest Pacific during austral autumn 1990: results from NOAA/PMEL CGC-90 cruise. ORNL/CDIAC-84, NDP-052. Carbon Dioxide Information Analysis Center, Oak Ridge Natiol Laboratory, Oak Ridge, TN. Lamb, M.F., Sabine, C.L., Feely, R.A., Wanninkhof, R., Key, R.M., Johnson, G.C., Millero, F.J., Lee, K., Peng, T.-H., Kozyr, A., Bullister, J.L., Greeley, D., Byrne, D.H., Chipman, D.W., Dickson, A.G., Goyet, C., Guenther, P.R., Ishii, M., Johnson, K.M., Keeling, C.D., Ono, T., Shitashima, K., Tilbrook, B., Takahashi, T., Wallace, D.W.R., Watanabe, Y.W., Winn, C., Wong, C.S., 2002. Consistency and synthesis of Pacific Ocean CO2 survey data. Deep-Sea Res., Part 2, Top. Stud. Oceanogr. 49, 21 – 58.

179

Millero, F.J., Dickson, A.G., Eischeid, G., Goyet, C., Guenther, P., Johnson, K.M., Key, R.M., Lee, L., Purkerson, D., Sabine, R.G., Schottle, R.G., Wallace, D.W.R., Lewis, E., Winn, C.D., 1998. Assessment of the quality of the shipboard measurements of total alkalinity on the WOCE Hydrographic Program Indian Ocean CO2 survey cruises 1994 – 1996. Mar. Chem. 63, 9 – 20. Orr, J.C., Maier-Reimer, E., Mikolajewicz, U., Monfray, P., Sarmiento, J.L., Toggweiler, J.R., Taylor, N.K., Palmer, J., Gruber, N., Sabine, C.L., Le Que´re´, C., Key, R.M., Boutin, J., 2001. Estimates of anthropogenic carbon uptake from four three-dimensional global ocean models. Glob. Biogeochem. Cycles 15, 43 – 63. Redfield, A.C., Ketchum, B.H., Richards, F.A., 1963. The influence of organisms on the composition of seawater. In: Hill, M.N. (Ed.), The Sea. The Composition of Seawater, vol. 2. Wiley, New York, pp. 26 – 77. Rı´os, A.F., Fraga, F., Pe´rez, F.F., 1989. Estimation of coefficients for the calculation of ‘‘NO’’, ‘‘PO’’ and ‘‘CO’’, starting from the elemental composition of natural phytoplankton. Sci. Mar. 53, 779 – 784. Sarmiento, J.L., Gruber, N., 2002. Sinks for anthropogenic carbon. Phys. Today, August, 30 – 36. Stramma, L., England, M., 1999. On the water masses and mean circulation of the South Atlantic Ocean. J. Geophys. Res. 104, 20863 – 20883. Takahashi, T., Broecker, W.S., Langer, S., 1985. Redfield ratios based on chemical data from isopycnal surfaces. J. Geophys. Res. 90, 6907 – 6924. Tomczak, M., Godfrey, J.S., 2001. Regional Oceanography: An Introduction WWW page, http://www.es.flinders.edu.au/~mattom/ regoc/pdfversion.html. Touratier, F., Goyet, C., 2004. Applying the New TrOCA Approach to Assess the Distribution of Anthropogenic CO2 in the Atlantic Ocean. Journal of Marine Syst. (this issue).