This page holds information specifically of relevance to the groups conducting the coupled AOGCM-ISM experiments of ISMIP6 and ISMIP7. As for the standalone ISM aspect of ISMIP6, we are now revisiting lessons learned from ISMIP6 and defining the ISMIP7 protocol. To do this, we have started regular zoom meetings on the coupled aspect of ice sheets in climate models. If you’d like to be involved or receive group communications about the coupled ESM-ISM models for ISMIP7, please email: robin.smith@ncas.ac.uk
ISMIP6 ‘Experiments’
The published ISMIP6 description (Nowicki et al, GMD 2016) lays out four experiments for coupled AOGCM-ISM models to complete, which directly parallel other CMIP6 experiments with standalone AOGCMs and ISMIP6 runs with standalone ISMs..
| Tier | New ISMIP6 CMIP6 exp. (Coupled AOGCM-ISM) |
| 1 | Pre-Industrial Control (piControl-withism) |
| 1 | 1% yr-1 CO to quadrupling CO (1pctCO2to4x-withism) |
| 2 | ScenarioMIP SSP5-8.5 up to year 2300 (ssp585-withism) |
| 2 | CMIP6 Historical Simulation (historical-withism) |
However, at the ISMIP6 meeting in Sassenheim it was decided that the standalone ISMs would use a baseline of 1970-2000 as their control period, rather than the pre-industrial. Depending on their initialisation methods, AOGCM-ISM modellers may thus also wish to run their Tier 1 experiments relative to a late 20th century climatology, and omit the historical-withism simulation entirely.
Notes from a discussion at the ISMIP6 Sassenheim meeting between representatives of UKESM, CESM, MPI-ESM, MIROC_ESM, EC-Earth and other interested parties is presented below.
Participants: Robin Smith, Bill Lipscomb, Miren Vizcaino, Sebastien le’Clech, Melchior van Wessem, Leo van Kampenhout, Marie Kapsch, Laura Muntjewerf, Ayako Abe-Ouchi
Since the ISMs will now use a baseline of 1970-2000, does it still make sense to do a coupled preindustrial and historical? MPI-ESM and CESM are planning/doing long coupled spinups under PI conditions.
Everyone would like to see coupled experiments run out beyond 2100 to see greater changes in the ice. A longer 1% run was seen as the best option for that, ensuring compatibility with past and future MIPs rather than extending any CMIP6 ScenarioMIP forcing.
Both UKESM and CESM have seen significant impacts on the North Atlantic when interactive icesheet runoff/calving is supplied due to the impact on the freshwater balance, and snow/ice building up in the north of Greeland
UKESM, CESM and MPI-ESM all use some form of sub-atmospheric-grid downscaling for icesheet surfaces. MIROC-ESM would like to use tiles.
Can we decide on a standard grid/resolution/frequency for supplying tiled output to ISMIP6 for surface forcing. The proposed vertical and time resolutions (5km grids, 100m in the vertical) had been suggested, the datasets will become huge.
Continued communication between the groups would be useful