@article{3057565,
    title = "Improved Representation of Clouds in the Atmospheric Component LMDZ6A of the IPSL-CM6A Earth System Model",
    author = "Madeleine, J.-B. and Hourdin, F. and Grandpeix, J.-Y. and Rio, C. and Dufresne, J.-L. and Vignon, E. and Boucher, O. and Konsta, D. and Cheruy, F. and Musat, I. and Idelkadi, A. and Fairhead, L. and Millour, E. and Lefebvre, M.-P. and Mellul, L. and Rochetin, N. and Lemonnier, F. and Touzé-Peiffer, L. and Bonazzola, M.",
    journal = "Journal of Advances in Modeling Earth Systems",
    year = "2020",
    volume = "12",
    number = "10",
    publisher = "Wiley-Blackwell Publishing Ltd",
    issn = "1942-2466",
    doi = "10.1029/2020MS002046",
    keywords = "Earth (planet);  Parameterization, Atmospheric components;  Cloud parameterizations;  Cloud radiative effects;  Earth system model;  Global distribution;  High-latitude regions;  Low-level cloud covers;  Object oriented approach, Climate models, atmospheric convection;  CALIPSO;  climate modeling;  cloud cover;  cloud radiative forcing;  CloudSat;  observational method;  PARASOL",
    abstract = "The cloud parameterizations of the LMDZ6A climate model (the atmospheric component of the IPSL-CM6 Earth system model) are entirely described, and the global cloud distribution and cloud radiative effects are evaluated against the CALIPSO-CloudSat and CERES observations. The cloud parameterizations in recent versions of LMDZ favor an object-oriented approach for convection, with two distinct parameterizations for shallow and deep convection and a coupling between convection and cloud description through the specification of the subgrid-scale distribution of water. Compared to the previous version of the model (LMDZ5A), LMDZ6A better represents the low-level cloud distribution in the tropical belt, and low-level cloud reflectance and cover are closer to the PARASOL and CALIPSO-GOCCP observations. Mid-level clouds, which were mostly missing in LMDZ5A, are now better represented globally. The distribution of cloud liquid and ice in mixed-phase clouds is also in better agreement with the observations. Among identified deficiencies, low-level cloud covers are too high in mid-latitude to high-latitude regions, and high-level cloud covers are biased low globally. However, the cloud global distribution is significantly improved, and progress has been made in the tuning of the model, resulting in a radiative balance in close agreement with the CERES observations. Improved tuning also revealed structural biases in LMDZ6A, which are currently being addressed through a series of new physical and radiative parameterizations for the next version of LMDZ. ©2020. The Authors."
}