Parametrisation of Cloud Processes in Climate Models

To fine-tune the global kilometre-resolution coupled ICON climate model, a crucial step involves calibrating cloud microphysical parameters. To achieve this, we’re exploring an innovative approach using machine learning for optimal calibration. We use an perturbed-parameter ensemble (PPE) of limited-area atmosphere-only ICON simulations focused on the North Atlantic ocean.

In the initial stage, we focus on calibrating the autoconversion scaling parameter, and we achieve this by analyzing satellite-retrieved top-of-atmosphere radiation fluxes. For this purpose, we performed limited area simulations of the North Atlantic using the ICON model. During these simulations, we experiment by altering different cloud microphysical parameters to evaluate their potential impacts on the radiation fluxes‘ output.

More to it:

TU
Universität
Max
Leibnitz-Institut
Helmholtz
Hemholtz
Institut
Fraunhofer-Institut
Fraunhofer-Institut
Max-Planck-Institut
Institute
Max-Plank-Institut