The Ross Sea (Fig. 1) is a deep bay of the Southern Ocean that borders the Amundsen Sea near the Cape Colbeck and the marginal seas of East Antarctica at Cape Adare. The southern boundary of the Ross Sea abuts the Ross Ice Shelf (RIS), which is the largest ice shelf on the Earth with an average thickness of 370 m (Smith et al., 2012). The Ross Sea (and RIS) overlies a deep continental shelf with an average depth of ~ 530 m (Smith et al., 2012), which is cut by several troughs including the Drygalski Trough, the Joides Trough, and the Glomar Challenger Trough running approximately in the northwest-southeast direction. These troughs facilitate the on-shelf intrusion of the circumpolar deep water (CDW; Klinck and Dinniman, 2010; St-Laurent et al., 2013), a relatively warm (θ > 1.2°C) and salty water mass (S > 34.4) (Orsi and Wiederwohl, 2009) that circulates along the Antarctic Circumpolar Current (ACC) and enters the cyclonic circulation of the Ross gyre at its eastern limb. Warm CDW upwells from deep layers to 200–300 m near the continental slope and intrudes onto the shelf. It then mixes with the shelf water and the Antarctic surface water (AASW), forming the modified Circumpolar Deep Water (mCDW) (Gordon et al., 2000; Jacobs and Giulivi, 1999; Orsi and Wiederwohl, 2009), which has temperature of \(－\)1.5°C to 1.0°C (Orsi and Wiederwohl, 2009) and is an important heat source to the continental shelf (Budillon et al., 2000; Dinniman et al., 2003).
Observations over the past few decades have shown a significant reduction in the salinity of shelf water (Budillon et al., 2011; Fusco et al., 2009) and surface salinity within the Ross gyre (Jacobs et al., 2002). Fresher surface water would increase the strength of the pycnocline (vertical density gradient), retaining heat in the subsurface layer, resulting in ocean warming at depths of ~ 300 meters north of the continental shelf (Jacobs et al., 2002), which corresponds to the CDW layer. Poleward transport of warm CDW across the continental shelf is thought to supply most of the heat involved in the basal melt of several ice shelves along the coast in the Amundsen Sea (Hellmer et al., 1985; Jacobs et al., 1996; Jenkins et al., 1997; St-Laurent et al., 2015; Wåhlin et al., 2010; Walker et al., 2007), the Bellingshausen Sea (Jenkins and Jacobs, 2008; Potter and Paren, 1985; Talbot, 1988) and around East Antarctica (Saari et al., 1987). Though there are multiple pathways for the CDW intrusion in the Ross Sea (Fig. 1a–b), the narrow eastern shelf makes it possible for mCDW to spread all the way to the RIS area resulting in a higher temperature of the eastern shelf than that of the western shelf (Fig. 1c, Orsi and Wiederwohl, 2009), which serves as a potential heat source to the ice shelf basal melting (Jacobs et al., 1985; Jacobs et al., 1979; Smethie and Jacobs, 2005). If more heat is retained in the subsurface layer north of the Ross Sea shelf due to enhanced stratification resulting from surface freshening, then the CDW intrusion would supply more heat to the shelf and cause greater basal melting of RIS. However, the RIS seems to be in near equilibrium with much less basal melting (Shepherd et al., 2010). There must be other factors offsetting the heat supply from CDW intrusion, which are not well understood up to now.
Based on an ocean–sea ice–ice shelf model driven by projected atmospheric forcing from phase 3 of the Coupled Model Inter-comparison Project (CMIP3), Dinniman et al. (2018) found that the increase in wind in the future will increase the on-shelf transport of CDW, but meanwhile will also enhance the vertical mixing on the shelf, inducing heat loss from the subsurface layers to the surface layer. The later process can offset the heat supplied by the CDW intrusion and result in little basal melting of RIS. This work revealed one factor that can work against the heating caused by the CDW intrusion and that helps to maintain a not too warm shelf. On the other hand, the Ross Sea is also an important source region of a cold water mass — the high salinity shelf water (HSSW) (Mathiot et al., 2012), which is the precursor of the Antarctic bottom water (AABW) (Budillon et al., 2011; Orsi et al., 1999) that supplies the lower limb of the global overturning circulation and ventilates the abyssal ocean. HSSW is mainly formed in the coastal polynyas resulting from brine rejection during new ice production, which is most prominent in the Terra Nova Bay (TNB) polynya (Fusco et al., 2009; Jendersie et al., 2018; Rusciano et al., 2013) and the western portion (between 170°E and 178°E) of the RIS polynya (Orsi and Wiederwohl, 2009). This water mass may also serve as a sink of heat on the shelf, and play a significant role in modulating the subsurface layer heat budget of the eastern Ross Sea shelf (ERSS). The contribution of HSSW production and transport to the heat budget of the Ross Sea shelf is rarely discussed by previous studies and thus remain poorly understood.
This work aims at revealing the dominant processes for the heat budget variations in the subsurface layer of ERSS, which would be important for us to understand the potential mechanisms modulating the stability of the RIS. In this study, six-year simulations from the Southern Ocean State Estimate (SOSE) product are employed to analyze the interannual variations of the heat budget and the controlling dynamical processes. It will be shown that the interannual variation of the ERSS heat content is mainly affected by variations of heat advection associated both with the CDW intrusion and the HSSW transport, while occasionally vertical mixing also plays a role. The manuscript is organized as follows. In section 2, the SOSE product is introduced and validated against observational data. The definition of the study area is provided, and the methods for computing the heat budget and relevant physical properties are described. In section 3, interannual variations of the heat budget terms and heat content in the subsurface layer of ERSS are presented, and the dynamical processes controlling these variations are analyzed. In section 4, discussions are provided on the combined effects of CDW and HSSW transports on the annual heat content change of ERSS, the mechanisms for the HSSW transport and higher-frequency variability of the heat budget terms. In Section 5, we summarize the main findings and their significance in the context of climate change in the high latitudes of the Southern Hemisphere.