A parameterization of ice shelf–ocean interaction for climate models

https://doi.org/10.1016/S1463-5003(02)00019-7Get rights and content

Abstract

Model results from a regional model (BRIOS) of the Southern Ocean that includes ice shelf cavities and the interaction between ocean and ice shelves are used to derive a simple parameterization for ice shelf melting and the corresponding fresh water flux in large-scale ocean climate models. The parameterization assumes that the heat loss and fresh water gain due to the ice shelves are proportional to the difference in freezing temperature at the ice shelf edge base and the oceanic temperature on the shelf/slope area of the adjacent ocean as well as an effective area of interaction. This area is proportional to the along-shelf width of ice shelf and an effective cross-shelf distance, which turns out to be rather uniform (5–15 km) for a variety of different ice shelves. The proposed parameterization is easy to implement and valid for a wide range of circumstances. An application of the proposed scheme in a global ice ocean model (CLIO) supports our hypothesis that it can be used successfully and improves both the ocean and sea ice component of the model. This parameterization should also be used in models of the climate system that include a coupling between an ice sheet and an oceanic component.

Introduction

Fresh water fluxes play an important role for the oceanic water mass transformation in high latitudes (e.g., Foster and Carmack, 1976; Foldvik et al., 1985; Carmack, 2000). In addition to the net atmospheric PE fluxes, ice shelf melting contributes significantly to the fresh water balance in the shelf areas around Antarctica (Jacobs et al., 1992; Timmermann et al., 2001). Sub-ice shelf melting typically reaches a few tens of centimeters per year up to a few meters per year, and thus represents a major factor in the mass balance of ice sheets (see, e.g., Huybrechts, 2002). The total melt water input at the Antarctic coast has been estimated to about 25 mSv (Jacobs et al., 1996). This is equivalent to a freshwater flux of 0.5 m/a over the circumpolar continental shelf area (the edge being defined by the 500 m isobath), exceeding PE by a factor of at least 2. Also, the injection of this fresh water occurs at the base of the ice shelf edge, i.e., 200–400 m depth, and has therefore a different impact on the stability of the coastal ocean than the surface forcing.

The sensitivity of the coupled sea ice–ocean system in the Southern Ocean to changes in fresh water input has been shown by Stössel et al. (1998), Goosse and Fichefet (1999), Beckmann et al. (1999) and Marsland and Wolff (2001), and it seems desirable to include these interaction processes in today’s ocean climate models, as well as ice sheet models.

Today’s ocean climate models (for a recent overview, see Griffies et al., 2000) do not include the sub-ice shelf cavities around the Antarctic continent and Greenland, because it would require substantial modification of the model code (see, e.g., Beckmann et al., 1999; Holland and Jenkins, 2001) and extension of the model domain beyond 75° S (to about 82° S in the Weddell Sea and 86° S in the Ross Sea). In addition, with horizontal resolution of a few degrees the proper representation of ice shelves seems hardly possible, even for the larger areas, the Filchner–Ronne ice shelf (FRIS) in the Weddell Sea and the Ross ice shelf (RIS), and the northern part of Larsen ice shelf (LIS) and Amery ice shelf (AIS). The majority of the comparatively small shelves at ice coasts (in the eastern and northeastern Weddell Sea, the Amundsen and Bellingshausen Seas and along Adelie Land) however are clearly sub-gridscale.

The interaction between ocean and ice shelves, however, is necessary for long-term sea ice–ocean climate studies as well as ice sheet modelling and for the development of Earth System models (coupling atmosphere, ocean, sea ice, ice sheets, carbon cycle, …), both for an estimation of the melting of ice shelves and the impact on water mass modification. In the past, part of the effects have been implicitly included by nudging to surface salinity (e.g., DeMiranda et al., 1999), or by prescribing additional fresh water fluxes corresponding to estimates of present-day shelf melting (e.g., Goosse and Fichefet, 2001). However, this approach is only valid if the system is assumed to have no evolution through time. It cannot be used to study climate variability, climate change or in paleoclimate studies. Consequently, a more adequate parameterization is necessary.

Section snippets

Sub-ice shelf circulation and melting

This section gives a brief overview of the salient features of ice shelf–ocean interaction.

The interaction between ice sheets and ocean is a complex phenomenon, which has been first studied in two-dimensional (xz) configurations by Hellmer and Olbers (1989) and Hellmer et al. (1998). The pressure dependence of the melting point of sea water leads to melting at the grounding line of the ice shelves, rising of the freshened and cooled water and hence an overturning circulation: the ice pump (see

Parameterization of net melting

In this section we investigate the possibility of parameterizing the net effects of the complex three-dimensional flow field and the water mass modification inside the ice shelf cavities. Such a parameterization will greatly help to improve both the Antarctic ice sheet models and ice–ocean models of the Southern Ocean.

Application of the parameterization in CLIO

In this section we present results from a climate model where the above parameterization has been implemented. We focus on the differences between a reference experiment with 30 cm/a ice shelf related surface fresh water input on the Antarctic continental shelf and the parameterization as outlined in the preceeding section. The resulting changes in sea ice, mixed layer depth and zonally averaged fields are shown.

Conclusions

We have derived a first order parameterization of the interaction between ice shelves and the adjacent ocean. It turns out that ice shelf melting can be reasonably well estimated knowing the oceanic temperature at the ice shelf edge as well as a characteristic length scale in the direction perpendicular to the edge. This length scale ranges typically from 5 to 15 km. It is very similar for all the ice shelves regardless of their size and much smaller than their actual cross-shelf extent.

In a

Acknowledgements

Helpful discussions with members of the BRIOS team and two anonymous reviewers are gratefully acknowledged. HG was supported by the Second Multiannual Scientific Support Plan for a Sustainable Development Policy (Belgian State, Prime Minister’s Services, Federal Office for Scientific, Technical, and Cultural Affairs, Contracts EV/10/7D and EV/10/9A).

References (32)

  • T. Fichefet et al.

    A numerical investigation of the spring Ross Sea polynya

    Geophysical Research Letters

    (1999)
  • A. Foldvik et al.

    Bottom currents near the continental shelf break in the Weddell Sea

  • P.R. Gent et al.

    Isopycnal mixing in ocean circulation models

    Journal of Physical Oceanography

    (1990)
  • R. Gerdes et al.

    Ocean circulation beneath Filchner–Ronne ice shelf from three-dimensional model results

    Journal of Geophysical Research

    (1999)
  • H. Goosse et al.

    Importance of ice–ocean interactions for the global ocean circulation: a model study

    Journal of Geophysical Research

    (1999)
  • H. Goosse et al.

    Open-ocean convection and polynya formation in a large-scale ice–ocean model

    Tellus

    (2001)
  • Cited by (0)

    View full text