Antarctic Basal Melt

We are currently looking for a PostDoc to support the group. If you are interested, please see here.

Overarching objective of the Antarctic ice–ocean focus group

Ocean-driven melting underneath the Antarctic ice shelves is the main driver of ice loss in Antarctica at present and will play a central role in determining the future contribution of the Antarctic ice sheet to sea level rise. The Antarctic ice-ocean focus group provides guidance on how to choose and calibrate melt parameterizations, and processes the CMIP ocean outputs to provide input that can be used for these parameterizations in the ISMIP7 ice-sheet model projections.

What ice-shelf basal melt parameterization should be used for ISMIP7? 

In contrast with ISMIP6, there is no standard parameterization in ISMIP7, and modellers are free to choose the one they want as long as it is driven by ocean temperature (and salinity) from the CMIP models. A range of parameterizations have been developed in recent years (e.g., Jourdain et al., 2020; Lazeroms et al., 2018; Reese et al., 2018; Pelle et al., 2019; Lambert et al., 2023). These parameterizations have been assessed by Favier et al. (2019), Burgard et al. (2022) and Burgard and Lambert (2025). There is no strong consensus on a single parameterization that would outperform the others, with only an emerging body of evidence that a simple linear parameterisation (Beckmann and Goosse, 2003) is generally not the best choice.

How will the melt parameterizations be calibrated in ISMIP7?

All of the aforementioned parameterizations rely on tunable parameters. A number of approaches to optimise these parameter values exist (Burgard et al., 2022; Jourdain et al., 2022; Reese et al., 2023;Menthon et al., 2025), and ISMIP6 revealed that these choices have a major influence on projected melt rates and sea-level rise (Jourdain et al., 2020; Seroussi et al., 2020; Reese et al., 2021; Edwards et al., 2021). There is no consensus on a single best way to tune these parameters, and the best parameter values may also depend on the ice-sheet model resolution. In contrast with the standard ISMIP6 runs, we therefore do not provide any set of pre-tuned parameters to the ISMIP7 modellers.

We nonetheless provide a protocol and a dataset to calilbrate the melt parameterizations, which should be done using the exact model implementation and resolution as in all ISMIP7 ice-sheet simulations. The calibration protocol aims to minimise the parameterization biases and considers melt rates from remote sensing data, agregated both in the individual IMBIE drainage basins (Mouginot et al., 2017) and in the areas important for buttressing (Reese et al., 2018). Given the increasing availability of ice-shelf basal melting projections from ocean models (Timmermann and Goeller, 2017; Mathiot and Jourdain, 2023; Pelle et al., 2021; Naughten et al., 2018; 2021, 2023; Jourdain et al., 2022), we also use the outputs of these simulations to constrain the sensitivity of melt parameterizations to ocean warming.

How are the CMIP model outputs processed for ISMIP7?

The processed CMIP outputs will be provided to ice sheet modellers. The ismip7-antarctic-ocean-forcing python package has been developed for this, consisting of remapping the ocean temperature and salinity to a standard stereographic grid, extrapolating to locations not resolved in the CMIP models and where ISMIP7 models could produce ice shelf drafts, and correcting the CMIP model biases. The bias correction will be based on the new ocean climatology of Zhou et al. (2025).

Toward a better understanding the ISMIP7 results through MeltMIP

We are also calling for a Melt parameterization Model Intercomparison Project (MeltMIP) which, analogously to initMIP for the model initialisation (Seroussi et al., 2019), tests in idealised experiments for the influence of the choice of melt parameterisation and parameter values. We hope that this preliminar intercomparison will help understand the diversity of the main ice-sheet model projections in ISMIP7. MeltMIP will consist of common idealized temperature perturbations, imposed instantaneously in ice-sheet 1-time-step calculations.

Proposed Approach for ISMIP7 Presented during November 2025 Webinar:

Follow these links to the slides, recording and documentation of the proposed protocol for Greenland ocean presented during the November 2025 webinar.