The Cloud Feedback Model Intercomparison Program has designed a protocol to evaluate clouds in climate and weather prediction models based on satellite observations (https://www.geosci-model-dev.net/10/359/2017/gmd-10-359-2017.pdf)
Why is it important to evaluate the cloudiness simulated by numerical models ?
The representation of clouds and cloud-related processes is critical for climate prediction over a wide range of space and time scales : for Numerical Weather Prediction (NWP), for climate prediction at seasonal to decadal timescales, and for the prediction of the long-term climate response to anthropogenic perturbations at the global and regional scales. This stems from the strong interaction of clouds with the global Earth's radiation balance, the local energy balance, the atmospheric circulation and the hydrological cycle. The representation of clouds in models thus constitutes a key component for the prediction of climate sensitivity, climate variability, extreme events, and the prediction of precipitation patterns in present-day and future climates.
Why using satellite simulators to evaluate modelled clouds ?
The definition of clouds or cloud types is not unique. It differs among observations (e.g. clouds detected by one particular sensor may not be detected by another sensor), and between models and observations (e.g. models may predict clouds at any atmospheric level where condensation occurs, while observations many not detect clouds overlapped by thick upper-level clouds). A comparison between modelled and observed clouds thus requires a consistent definition of clouds and cloud diagnostics, taking into account the effects of viewing geometry, sensors' sensitivity and vertical overlap of cloud layers.
This CFMIP-OBS website gathers observational data and diagnostics that are fully consistent with outputs from the CFMIP Observations Simulator Package (COSP) for the evaluation of clouds and radiation in numerical models : CALIPSO, CLOUDSAT, CERES, ISCCP, MISR, PARASOL.
What is COSP ?
The Cloud Feedback Model Intercomparison Project (CFMIP) has developed COSP (the CFMIP Observations Simulator Package), a community tool developed by several centers in the UK, in France and in the USA (MetOffice Hadley Centre, LMD/IPSL, LLNL, CSU and the University of Washington) to facilitate the comparison of clouds simulated by numerical models with observations from passive or active remote sensing.
COSP diagnoses from model outputs some quantities (e.g. infrared and visible radiances, radar reflectivities, lidar backscattered signals) that would be observed from space if satellites where flying above an atmosphere similar to that predicted by the model. Diagnostics about the presence and the properties of clouds can then be applied consistently to observations and to simulator outputs, ensuring a consistent model-data comparison.
The latest version of COSP (v2.0) can be downloaded here.
The previous version of COSP (v1.4) can also be downloaded here.
For more specific inquiries about COSP, please contact Alejandro Bodas-Salcedo.
Who uses COSP ?
COSP is already being used by several climate and NWP modeling groups (e.g. UKMO, LMD/IPSL, MPI, ECMWF, GFDL, CCCMa, CCSR, CESM, MIROC, ...).
Recognizing the importance of the evaluation of clouds and cloud-related processes in climate models, the WCRP Working Group for Coupled Models (WGCM), with the endorsement of GEWEX and WGNE, has recommended that COSP be used for some of the CMIP5 and CMIP6 simulations (either in-line or off-line) to be assessed by the 6th Assessment Report of the IPCC.
Namelists containing the configuration of COSP for long and short CMIP5 simulations can be found here.
What observations may be compared to COSP outputs ?
Currently, COSP enables the comparison of model outputs with several observations :
Additional information available:
Who is taking care of CFMIP-obs ?
CFMIP-OBS dataset is the result of a community effort.
- S. Bony (IPSL/LMD), H. Chepfer (IPSL/LMD), M. Chiriaco (IPSL/LATMOS), J-L. Dufresne (IPSL/LMD), S. Klein (LLNL), N. Loeb (NASA/LarC), R. Marchand (University of Washington), R. Pincus (University of Colorado), D. Tanré (LOA), M. Webb (UKMO), D. Winker (NASA/LarC), S. Xie (LLNL), Y. Zhang (LLNL).
Thanks are due to NASA and CNES for satellite observations. This work was financially supported by CNES and by ENSEMBLES. The data access and computation are due to ICARE and CLIMSERV.