U.S. Dept. of Commerce / NOAA / OAR / PMEL / Publications

Warming of global abyssal and deep southern ocean waters between the 1990s and 2000s: Contributions to global heat and sea level rise budgets.

S. G. Purkey1,2 and G. C. Johnson1,2

1Pacific Marine Environmental Laboratory, National Oceanographic and Atmospheric Administration, Seattle, Washington, USA

2School of Oceanography, University of Washington, Seattle, Washington, USA

Journal of Climate, 23(4), 6336–6351, doi: 10.1175/2010JCLI3682.1.
© 2010 American Meteorological Society.

1. Introduction

A warming climate is unequivocal, with the global top of the atmosphere radiative imbalance currently on the order of 1 W m −2, very likely due to anthropogenic greenhouse gases (Solomon et al. 2007). Over the past few decades, roughly 80% of the energy resulting from this imbalance has gone into heating the oceans (Levitus et al. 2005), which have a large heat capacity compared with the land or the atmosphere. This warming is important in sea level rise (SLR) and other climate projections (Bindoff et al. 2007).

The earth's radiative imbalance affects SLR in two ways (e.g., Cazenave et al. 2008; Trenberth and Fasullo 2009).Much of the heat raises ocean temperature, causing thermal expansion, termed thermosteric SLR. A much smaller portion of the heat acts to melt continental ice, adding mass to the ocean. Highly accurate satellite measurements from Ocean Topography Experiment (TOPEX) / Poseidon and Jason altimetry have reported an average rate of SLR of 3.1 mm yr−1 between 1993 and 2003, with roughly half of that being due to thermal expansion and half due to mass changes, mostly from the melting of continental ice (Bindoff et al. 2007). However, there is still debate over the exact breakdown of SLR between thermostatic and mass addition components. Some recent SLR budgets using observations of upper-ocean warming and mass changes either do not close or have high uncertainty (Miller and Douglas 2004; Raper and Braithwaite 2006), especially post-2003 (Willis et al. 2008). Other SLR and global energy budgets rely on poorly constrained deep ocean heat uptake for closure (Domingues et al. 2008; Murphy et al. 2009).

Models contain a delay between greenhouse gas forcing and surface temperature increase because of the long equilibration time of the ocean (Hansen et al. 2005). Therefore, even if greenhouse gas concentrations were kept constant at current levels, ocean temperatures and sea level would continue to rise for centuries. Furthermore, model fluxes of heat from the ocean surface layers to the deep ocean differ dramatically depending on model details. This climate sensitivity affects predictions of the magnitudes and rates of future SLR and global atmospheric warming (Raper et al. 2002; Meehl et al. 2005). Indeed, uncertainty involved in deep-ocean heat uptake may be the largest cause of variation among climate projections (Boé et al. 2009), making it vital to close observed heat and SLR budgets, including the deep ocean, for the purposes of adequately constraining predictions of and preparing for future climate change.

The deep ocean is ventilated by dense water sinking at high latitudes. North Atlantic Deep Water (NADW) is a mixture of water masses formed through deep convection processes in the Nordic and Labrador Seas (LeBel et al. 2008). Antarctic Bottom Water (AABW) is formed by a complex interaction of water masses and physical processes, with varieties produced in at least three source regions: the Weddell Sea, the Ross Sea, and the Adelie Coast (Orsi et al. 1999). While pure AABW is largely confined to the Southern Ocean, here we will refer to deep and bottom waters primarily ventilated near the Antarctic as AABW for simplicity. NADW and AABW feed the deep and abyssal limbs of the global ocean meridional overturning circulation (Lumpkin and Speer 2007).

Recent studies have revealed property changes in AABW near its source regions. In the Weddell Sea, the deep water has warmed at a rate of 0.009°C yr−1 between 1990 and 1998, followed by a period of cooling (Fahrbach et al. 2004). In the Ross Sea, shelf and surface water has freshened, along with warming at middepths (190–440 m) and in the deep layer within the gyre (Jacobs et al. 2002; Ozaki et al. 2009; Jacobs and Giulivi 2010). Deep water off the Adelie Coast along 140°E has also shown cooling and freshening on isopycnals (Aoki et al. 2005; Rintoul 2007; Johnson et al. 2008a; Jacobs and Giulivi 2010).

Warming of AABW is not limited to the Southern Ocean. In the deep basins north of the Antarctic Circumpolar Current (ACC), AABW has also shown signs of warming, although at a reduced rate compared to the warming near the source regions. In the Atlantic Ocean, the abyssal water in the Scotia Sea, Argentine Basin, and Brazil Basin—all fed by AABW originating from the Weddell Sea—has warmed over the past two decades (Coles et al. 1996; Johnson and Doney 2006; Zenk and Morozov 2007; Meredith et al. 2008). In the southeastern Indian Ocean, warming has been observed in the Australian–Antarctic Basin but little change has been seen to the north (Johnson et al. 2008a). Finally, the abyssal southwest and central Pacific basins have both significantly and widely warmed over the past two decades (Fukasawa et al. 2004; Kawano et al. 2006, 2010; Johnson et al. 2007). In addition to the warming, there is some evidence of recently reduced abyssal circulation in the North Atlantic (Johnson et al. 2008b) and North Pacific (Kouketsu et al. 2009; Masuda et al. 2010).

Furthermore, the upper 1000 m of the water column throughout much of the Southern Ocean has also warmed over the last few decades, apparently at a faster rate than the upper-ocean global mean (Gille 2002, 2008; Böning et al. 2008; Sokolov and Rintoul 2009). This warming may be partly associated with a poleward migration of the ACC due to an increase in the strength and southward shift of the westerly winds that drive the ACC.

Here we make quantitative global estimates of recent (1990s to 2000s) deep and abyssal ocean warming, mostly within or originating from the Southern Ocean. We use repeat hydrographic section data to quantify temperature trends in two regions of the world's oceans: the global abyssal ocean, defined here as >4000 m in all deep basins (excluding the Arctic Ocean and Nordic seas), and the deep Southern Ocean, defined here as the region between 1000 and 4000 m south of the Subantarctic Front (SAF). AABW, as defined by a water-mass analysis (Johnson 2008), dominates much of the abyssal (Fig. 1a) global ocean and deep (Fig. 1b) Southern Ocean. The abyssal Pacific and Indian oceans are primarily composed of AABW with only a small fraction of NADW present (Fig. 1a; Johnson 2008). However, in the abyssal Atlantic, AABW dominates only in the Weddell–Enderby, Cape, Argentine, and Brazil basins, with a small fraction persisting near the bottom of the other basins, where NADW is endemic. In the deep Southern Ocean, south of the SAF, AABW also dominates (Fig. 1b; Johnson 2008). Thus, our fixed control volume (necessary for evaluating a change in heat content) contains primarily, although not solely, AABW, and it allows for an estimate of warming due to changes in waters mostly of southern origin.

figure 1

Fig. 1. (a) Tracklines of the 28 repeated sections studied (black lines) with WOCE designators noted adjacent. Basin boundaries are outlined (gray lines) over the depth-averaged fraction of AABW below 4000 m (color bar) after Johnson (2008). The Subantarctic Front (SAF; Orsi et al. 1995) position (magenta line) and the 4000-m isobath (thin black lines) are also shown. (b) As in (a) but a polar projection with tracklines of the nine repeated sections that extend south of the SAF plotted over the depth-averaged fraction of AABW from 1000 to 4000 m with the 1000-m isobath and without basin boundaries.

We describe the repeat hydrographic data and screening methodologies used in section 2. In section 3, we describe regional rates of temperature change and estimate their uncertainties from these data. In section 4, we use these rates to calculate the contributions of warming primarily of southern origin to heat and SLR budgets in two ways: first, we calculate local abyssal contributions for individual ocean basins and the deep Southern Ocean (section 4a); and second, we combine abyssal contributions from all 32 basins (assuming unsampled basins do not change) and the deep Southern Ocean for a global assessment (section 4b).We explore the effects of variations in our 4000-m boundary between global abyssal and deep Southern Ocean volumes near the end of section 4. We conclude with a discussion of the results in section 5.

Return to Abstract

Go to Next section

PMEL Outstanding Papers

PMEL Publications Search

PMEL Homepage