期刊名称：Journal of Advances in Modeling Earth Systems
电子版ISSN：1942-2466
出版年度：2021
卷号：13
期号：6
页码：e2020MS002225
DOI：10.1029/2020MS002225
出版社：John Wiley & Sons, Ltd.
摘要：We demonstrate a new approach for climate model tuning in a realistic situation. Our approach, the mathematical foundations and technical details of which are given in Part I, systematically uses a single-column configuration of a global atmospheric model on test cases for which reference large-eddy-simulations are available. The space of free parameters is sampled running the single-column model from which metrics are estimated in the full parameter space using emulators. The parameter space is then reduced by retaining only the values for which the emulated metrics match large eddy simulations within a given tolerance to error. The approach is applied to the 6A version of the LMDZ model which results from a long investment in the development of physics parameterizations and by-hand tuning. The boundary layer is revisited by increasing the vertical resolution and varying parameters that were kept fixed so far, which improves the representation of clouds at process scale. The approach allows us to automatically reach a tuning of this modified configuration as good as that of the 6A version. We show how this approach helps accelerate the introduction of new parameterizations. It allows us to maintain the physical foundations of the model and to ensure that the improvement of global metrics is obtained for a reasonable behavior at process level, reducing the risk of error compensations that may arise from over-fitting some climate metrics. That is, we get things right for the right reasons. Plain Language Abstract In view of the importance of global numerical models for the anticipation of future climate changes, their improvement is often considered too slow. We present a new approach that we believe could boost model improvement significantly. This approach promotes the use of machine learning techniques developed by the “uncertainty quantification” community for the adjustment of model free parameters, or tuning. These techniques are applied to physics improvement at process scale, represented through parameterizations. In this approach, the tuning of the global atmospheric model is preconditioned by calibration of the model free parameters on a series of well documented cloud scenes for which explicit very high resolution simulations are available. We demonstrate on a real example how the reduction of the parameter space with this approach allows us to save a large amount of computer resources and detract from the long and tedious by-hand phase of model tuning. By automating part of the tuning process, the approach enables climate modeler expertize to focus on understanding and improving the model physics through parameterization.
