parent nodes: Publications

Paper 2358 - Session title: EarthCARE Exploitation

11:10 Antarctic precipitation in the LMDz and MAR climate models : comparison to CloudSat retrievals and improvement of cold microphysical processes

Madeleine, Jean-Baptiste (1); Gallée, Hubert (2); Vignon, Etienne (2); Genthon, Christophe (2); Krinner, Gerhard (2); Palerme, Cyril (2); Claud, Chantal (1)
1: Laboratoire de Météorologie Dynamique (LMD), CNRS/UPMC, Paris, France;
2: Laboratoire de Glaciologie et Géophysique de l'Environnement (LGGE), CNRS/UGA, Grenoble, France

The APRES3 project (“Antarctic Precipitation : Remote Sensing from Surface and Space”, see the partner presentation by Genthon et al. in this conference) aims at providing constraints on the precipitation rates over Antarctica by using the CloudSat CPR data, and paves the way for a similar use of the EarthCare data. A key application of precipitation observations in Antarctica is the improvement of global climate models (also known as GCMs), which are poorly constrained in this part of the world. GCMs show multiple responses in Antarctica and do not agree on the future precipitation rates over the ice sheet. Indeed, their cold microphysics schemes are often built on concepts and datasets developed in non-polar regions, and the mixed phase and ice phase microphysical processes of polar environments are still poorly known. These microphysical processes play an important role in polar precipitation but also more generally in predicting accurately the cloud radiative properties and resulting energy budget. For example, many GCMs underestimate cloud cover and therefore miss essential climate features in polar regions.

In this study, the measurements of precipitation made by CloudSat in the context of the APRES3 project are compared to the results of the LMDz GCM (Laboratoire de Météorologie Dynamique zoomable GCM) and MAR model (Modèle Atmosphérique Régional). The LMDz GCM is used as the atmospheric component of the IPSL Earth System Model, which is involved in the IPCC CMIP exercises. The LMDz simulations are run using prescribed sea surface temperatures over the CloudSat observation period. In order to distinguish the influence on the results of dynamics from that of the physical parameterizations, the GCM is also run in nudging mode where the winds are adjusted toward the ECMWF reanalyses. The main biases of the LMDz GCM as well as the misrepresented physical processes are identified. Comparisons are also made to simulations performed using the MAR model. MAR is a limited area model developed for polar regions which can computationally afford a more advanced microphysics scheme than the LMDz GCM. The formation of cold clouds and solid precipitation in the LMDz GCM is often based on laws deduced from observations of high tropical clouds, and improving the representation of precipitation in polar regions might be necessary. Developments are underway to improve the cold microphysical processes of the LMDz GCM in polar regions while keeping the short computational time that is required by global climate simulations.