Motivation
The simulation of sea ice in CMIP5 models shows a large spread between models, making projections of the timing of a summer ice free Arctic highly uncertain. In order to improve sea ice simulations in climate models, the sea ice modeling community would benefit from a dedicated workshop to come together and discuss ways forward, to learn from CMIP5, and to plan for CMIP6.
Proposed Actions
Year 1: Workshop in year 1 to bring together the sea ice modeling community (developers and model users, with a focus on sea ice modeling in climate models) to discuss model development needs, coordinate model analysis of the source of model biases in CMIP5 models, determine pressing observational need, and compile a list of sea ice variables to be saved for CMIP6.
Large-Scale Sea Ice Simulations Workshop
September 26th, University of Reading, UK
Read Workshop Report
Year 2: Joint workshop with other ongoing sea ice modeling activities (e.g., SIPN) to bring together sea ice observers and modelers to develop a concrete list of observational data that would be most useful for the evaluation and improvement of sea ice simulations by climate models.
Year 2-3: Coordination (through email list and website, as well as meetings of opportunity) of CMIP5 sea ice bias analysis and writing of community paper, as well as regular exchange of new results on sea ice model developments and new observational data products
Goals
The CliC sea ice and climate modeling forum has the goal to improve the development and performance of sea ice models by addressing the following priorities:
- Model evaluation: Compile a list of useful variables to be included in coordinated model experiments going forward (like CMIP6), to facilitate model-model comparisons as well as model-data comparisons. Develop a standardized sea ice simulation protocol and identify data needs
- Causes of model biases: Analyze the cause of the biases in sea ice model simulations, through the analysis of existing simulations (CMIP5, CORE-II) and targeted additional experiments when needed. One goal is to determine the role of sea ice model deficiencies versus biases originating from the sea ice model versus the coupled system
- Model development: Share experience with sea ice model developments and their impact in climate simulations (e.g., prognostic salinity, sophisticated melt pond schemes, rheology choices, etc) and identify the best way forward as well as observational data needs for model development
Deliverables
- A community-reviewed list of sea ice variables desired from CMIP6
- Recommendations for sea ice model evaluation protocols for the future, beyond the sea ice extent, and identification of useful datasets
- Initiate and coordinate a community effort to identify the source of sea ice biases in CMIP5 models, through additional analysis of the CMIP5 archive, resulting in a peer-reviewed article on sea ice model biases in CMIP5, highlighting ways forward
For more information, please contact Alexandra Jahn and Dirk Notz.