Strategies for Antarctic Ice sheet simulations and paleo data-model comparison
The INStabilities & Thresholds in ANTarctica (INSTANT) is a Scientific Research Programme of SCAR that aims to quantify the Antarctic ice sheet contribution to past and future sea-level change. This type of program has the capacity to organise workshops and other collaborative activities.
There is a need to discuss strategies of ice sheet models simulations within INSTANT and we propose a group to do so. It could be an across theme subcommittee of INSTANT but several points, especially data-model and paleo-climate, could be tagged as "Hub discussion" (platform of interactions between themes) because they really involve other communities than ice sheet modellers. This group should be also connected to the expert group ISMASS (Ice sheets mass balance and Sea level). We would like to know who is interested in getting involved in this group.
The objective is to define best pratices on how to implement ice sheet simulations in order to compare simulations between each others and with data. This is a necessary step to make simulations of the past a reliable way to reduce uncertainty for the future. We recognize that there are several ice sheet models types and several model/data approaches. It is important to ensure that all these approaches feed into each other along the duration of INSTANT instead of running in parallel.
The points that should be discussed within this group (non exhaustive)
- Databases and approaches for Model-data comparison
- Review available data, their uncertainty (time and space), how they should be prepared for modelers. As a first step, this group could help organizing a Hub workshop with both data (mainly paleo proxy) and model people.
- Discussion on methods : data assimilation, ensemble of simulations, parameters calibration, Bayesian inference, emulators. A key question is: how data uncertainty propagates into projections ?
- Strategies for simulations.
- How to manage the coupling/forcing with climate models (links with PMIP)
- Is it possible to define time slices during which all kind of ice sheet models could be compared ? and how to overcome initialization issues and resolution differences
- How to design simulations to study processes
- Should we consider intercomparison experiments (paleoicemip) ? There are on-going ISMIP6 and later ISMIP7, MISOMIP, …. How to be complementary without too much overlap ?
- A first trial could be to focus on LGM reconstruction.
- further in the project: Last deglaciation extending to future.