Earth System Model Evaluation to Improve Process Understanding
Earth System Models (ESMs) are highly-ambitious — they attempt to help understand and project how the physical climate and biogeochemical cycles of the globe changes under a wide-range of anthropogenic and natural forcings. Given the challenge of this task, it is perhaps unsurprising that different models project rather different degrees of climate change over the next 100 years. However, it is vital that this range is reduced to improve the quality of the information that policymakers, businesses, and the public, have to act upon. This workshop aims to contribute to reducing the spread of climate projections by enabling more complete evaluation of model outputs against observations, and by identifying Emergent Constraints – observable aspects of the contemporary Earth System that are most closely related to future projections.
Motivation: Earth System Models (ESMs) are highly-ambitious — they attempt to help understand and project how the physical climate and biogeochemical cycles of the globe changes under a wide-range of anthropogenic and natural forcings. Given the challenge of this task, it is perhaps unsurprising that different models project rather different degrees of climate change over the next 100 years. However, it is vital that this range is reduced to improve the quality of the information that policymakers, businesses, and the public, have to act upon. This proposal aims to contribute to reducing the spread of climate projections by enabling more complete evaluation of model outputs against observations, and by identifying Emergent Constraints – observable aspects of the contemporary Earth System that are most closely related to future projections.
Description: The experimental design of CMIP6 is now finalized. One purpose of the Diagnostic, Evaluation and Characterization of Klima (DECK) experiments and the CMIP6 historical simulations is to provide a basis for documenting model simulation characteristics. Towards that end infrastructure is being developed to allow analysis packages to be routinely executed whenever new model simulations are contributed to the CMIP archive at the Earth System Grid Federation (ESGF), utilizing observations from obs4MIPs and related efforts. Examples of available tools that target routine ESM evaluation in CMIP6 include the Earth System Model Evaluation Tool (ESMValTool) and the PCMDI Metrics Package (PMP). These tools can be used to comprehensively characterize agreement with observations (or “performance”) for the wide variety of models that will contribute to CMIP6, with first results being produced in time to be discussed at the workshop. The workshop will also focus on scientific improvements that are required to further advance ESM evaluation. Topics will include a better consideration of observational uncertainty, internal variability and model tuning. It will also emphasize advanced process- or regime-oriented diagnostics that can be used to understand the sources of errors and uncertainties in models, thereby highlighting specific areas requiring model improvements. A related open scientific question is the relation between present-day model performance and future projections. We will review and discuss Emergent Constraint studies that use observations to constrain climate sensitivity. Innovative methods that can be used to more objectively weight multi-model climate projections will also be discussed, taking into account both model performance and model inter-dependence. Advancing on these important topics requires the modelling community to get together with the observational community, scientists involved in CMIP6-Endorsed MIPs, statisticians and users of climate model output.
Relevance: The climate projections considered in IPCC AR5 were largely based on ESM experiments defined and internationally coordinated as part of CMIP5. CMIP is providing understanding of past, present and future climate variability and change, but intelligent use of CMIP results requires an awareness of their limitations. It is essential, therefore, to subject models to a systematic evaluation against observations and to further improve evaluation methods and techniques. The benefits of a more efficient and systematic approach to model evaluation are clear. The recording of a set of informative diagnostics and metrics would enable anyone interested in CMIP research to obtain a broad overview of model performance and simulation behavior. Emergent Constraint studies can guide model development priorities towards processes crucial to the magnitude and spread of future climate change, inform future observational priorities, and ultimately reduce uncertainties in climate sensitivity. These efforts would also substantially support the IPCC AR6 by facilitating the assessment of the models in IPCC WG I and by supporting WGs II and III through a better quantification of related uncertainties.
Expand to see available videos and presentations
10:25 am Part I: Historical View of Model Performance, and the Role of Models in Understanding the Climate System
1:55 pm Part II: Opportunities from New Observations and Observational Uncertainty in the Context of Model Evaluation
3:55 pm Part III: Application of New CMIP Evaluation Tools to CMIP5/6 Models, Workflow
8:55 am Part IV: Large-Scale Persistent Systematic Biases
10:55 am Part V: Progress in Process Understanding (Sectorial)
2:30 pm Breakout Groups 1-3
In November 1995, after three days of deliberations in Madrid’s Palacio de Congresas, the Intergovernmental Panel on Climate Change (IPCC) reached the historic finding that “the balance of evidence suggests a discernible human influence on global climate”. This sentence changed the world. While other individuals and national scientific organizations had reached similar conclusions before Madrid, the “discernible human influence” statement marked the first time that the international climate science community had spoken so clearly and forcefully.
The reaction was swift. The “discernible human influence” conclusion led to Congressional investigations, charges of “scientific cleansing” allegations of corruption of the peer-review process and professional misconduct, and claims of political tampering. Santer spent several years addressing such criticism. This lecture is a reflection on the top ten scientific and personal lessons Santer learned after publication of the IPCC’s 1995 Report. Many of these lessons still have relevance in today’s world.
8:55 am Part VI: Emergent Constraints on Climate and Earth System Sensitivities
11:10 am Breakout Groups 4-6
8:55 am Part VII: Weighting Based on Performance and Interdependence
10:55 am Part VIII: User and Policy Oriented Metrics and Applications
2:30 pm Breakout Groups 7-9
4:00 pm Breakout Groups
8:55 am Part IX: Key Points and Synthesis
Peter Michael Cox
The attendee list and participant profiles are regularly updated. For information on participant affiliation at the time of workshop, please refer to the historical roster. If you are aware of updates needed to participant or workshop records, please notify AGCI’s workshops team.