Risks and Uncertainties under Climate Change

The RISCCI project began in September 2018, and its core development has been the creation of an ensemble platform to explore parametric uncertainty in the MeteoFrance climate model, CNRM CM6.  The simulations informing the scientific deliverables have now been completed, one paper is in final revisions and two are in process.  In parallel with these efforts, the project produced a range of publications on the interpretation of the wider multi-model ensemble and a number of studies using simpler models.  The project remains strongly coupled with the EU projects ESM2025 and PROVIDE, which started in 2021..  

The CNRM-CM6 ensemble project 

The CNRM-CM6 PPE project explores parametric uncertainty in the atmospheric component of the model (ARPEGE), taking a set of 35 uncertain parameters in the model and exploring the climatic sensitivity of the model to univariate and multivariate perturbations. Initial parameter choices, default values and plausible maximum and minimum values were obtained through extended consultation with MeteoFrance developers, and by consideration of output from internal calibration simulations for the beta versions leading up to the release model of CNRM-CM6. 

 Perturbation experiments were then conducted for each of the parameter combinations to produce estimates of net climate feedback and climate sensitivity for each member of the parameter distribution. The resulting ensemble resulted in a significant spread in climate sensitivity, such that model configurations spanned the range of climate sensitivity present in the multi-model archive of climate models contributing to the CMIP project.  This ensemble has been documented in a submitted paper (Peatier et al, in review) at GRL.  A second study, looking at joint behavior over a range of parallel experiments in the UKMO and NCAR modeling centers, is now in process.  This work will continue in the context of the ESM2025 project, where Saloua Peatier (the PhD candidate supervised in the context of RISCCI) will remain at CERFACS as a postdoc.

Emergent constraints in CMIP

The PI has written an extensive review in 2021 on the use of ‘emergent constraints’ in the reduction of climate modeling uncertainties.  This work reviewed a wide range of literature to document how such constraints are used and their potential limitations.  The study was published in Earth System Dynamics in 2021.

Carbon cycle response uncertainty 

The PI is collaborating with CERFACS colleagues and an international team exploring a novel approach for carbon cycle feedbacks which addresses a key issue in the simulation of perturbed parameter sensitivity in land surface models. This research has developed a reduced-form “sparse” configuration of the Community Land Model in order to explore parametric sensitivity of carbon cycle feedbacks to parameter perturbations. This work is ongoing, with initial publications expected in 2022.

Solar Radiation Management

The H2020 PROVIDE project began in 2021, with CERFACS responsible for a task on quantifying uncertainties in the implementation of Solar Radiation Management as a climate mitigation approach.  Though the project will move to CICERO in 2022, a student supported internally by CERFACS funding (Susanne Baur) will continue to work on this theme at CERFACS for the duration of her PhD.