EUCLIPSE: Reducing uncertainty in climate model projections due to cloud-climate feedbacks
Cloud feedbacks in Global Climate Models (GCMs) or Earth System Models (ESMs) remain the largest source of uncertainty in projections of future climate (see Figure 1). They are also a major contributor to uncertainty in other feedbacks, such as in surface albedo and the carbon cycle. Through interactions with the large-scale circulation, cloud processes also contribute to synoptic circulations and regional climate. They are therefore critical to the prediction of future changes in precipitation patterns, climate variability and extreme events.
Figure 1. Break up of the various contributions to the increase of the global mean surface temperature for 12 CMIP3 Climate models for a 2XCO2 scenario. The dark blue indicates temperature increae due to CO2 doubling only, light blue indicates increase due to water vapour feedback, yellow corresponds to feedback due to changes in the in surface albedo (most notably sea ice) and brown corresponds to cloud feedback. Note that the largest intermodal spread is due to clouds. (Source Dufresne and Bony, Journal of Climate 21, 5135-5144 2008)
The central objective of the project EUCLIPSE (EU CLoud Intercomparison, Process Study and Evaluation project) is to reduce the uncertainty in the representation of cloud processes and feedbacks in the new generation of Earth System Models (ESMs), in support of the upcoming IPCC’s fifth assessment report. EUCLIPSE is an international effort, coordinated by A. Pier Siebesma from the Regional Climate (RK) division of KNMI, funded under Framework Program 7 of the European Union. A special focus is on the representation of boundary layer clouds such as stratocumulus and shallow cumulus clouds, as it are these cloud types that are the main cause of spread in climate model uncertainty of cloud climate feedback.
Figure 2. Overview of EUCLIPSE
The concept of EUCLIPSE is summarized in the figure 2.
- Work Package 1 is responsible for the execution of a number of climate runs such as agreed in CMIP5. In WP1 also the necessary model diagnostics are prepared, most notably the implementation of satellite simulators in climate models that are necessary to make a useful comparison of cloud properties between satellites and climate models.
- Work Package 2 deals with the actual evaluation of cloud properties in climate models with observations using novel evaluation techniques.
- Work Package 3 is exploring process models such as large Eddy Simulation (LES) Models and Single Column Model versions of climate models to better understand the parameterized cloud physics in climate models, to detect deficiencies and to improve them.
- Work Package 4 synthesizes all the knowledge gained in the previous packages, sensitivity studies with alternative formulations of cloud representation in climate models will be explored, new metrics will be introduced and improvement of cloud representation in climate models will be assessed.
RK is actively participating in a number of current activities of EUCLIPSE:
- Currently, we are running a number of CMIP5 climate runs with the Earth System Model EC-Earth in support of the IPCC’s fifth assessment report. The results of these runs will be used to assess, understand and reduce the uncertainty in EC-Earth’s climate model sensitivity for future climate scenario’s. (contact: Frank Selten: selten@knmi.nl)
- We are participating in CGILS (CFMIP-GCSS Intercomparison of Large-Eddy and Single Column Models). CGILS uses idealized large-scale dynamical conditions to evaluate cloud feedback processes in Single Column Model (SCM) versions of GCMs and Large Eddy Simulation (LES) models. The approach has several advantages. It isolates the model physics from dynamics, thus dramatically simplifying the problem to a few selected locations. It therefore allows to link the strength of cloud climate feedback to the physical assumptions that are made in the various parameterization packages of climate models. We are contributing with SCM versions of EC-Earth and RACMO (contact: Roel Neggers: neggers@knmi.nl, Sara Dal Gesso: gesso@knmi.nl )
- We are participating in two EUCLIPSE/GCSS/CFMIP intercomparison studies for the transition of stratocumulus to cumulus for LES models and SCM versions of climate and numerical weather prediction models. One intercomparison is based on observations of the ASTEX experiment. The second intercomparison are in fact three composite transition cases data based on satellite data. We are contributing to this case with SCM versions of EC-Earth and RACMO (contact: Sara Dal Gesso: gesso@knmi.nl ).
- In support of EUCLIPSE we have set up an intercomparison project for radiation codes used in climate models, numerical weather prediction models and process models such as LES models to investigate the spread in the calculation of the albedo for stratocumulus clouds (contact: Sara Dal Gesso gesso@knmi.nl , Piers Siebesma siebesma@knmi.nl
)
)
