Key points are not available for this paper at this time.
Machine learning (ML) has been increasingly applied in climate modelling on system emulation acceleration, data-driven parameter inference, forecasting and knowledge discovery, addressing challenges such as physical consistency, multi-scale coupling, data sparsity, robust generalization and integration with scientific workflows. This paper analyses a series of case studies from applied ML research in climate modelling, with a focus on design choices and workflow structure. Rather than reviewing technical details, we aim to synthesize workflow design patterns across diverse projects in ML-enabled climate modelling: from surrogate modelling, ML parameterization and probabilistic programming, to simulation-based inference and physics-informed transfer learning. We unpack how these workflows are grounded in physical knowledge, informed by simulation data and designed to integrate observations. We demonstrate a framework for ensuring rigour in scientific ML through more transparent model development, critical evaluation, informed adaptation and reproducibility, and aim to contribute to lowering the barrier for interdisciplinary collaboration at the interface of data science and climate modelling. This article is part of the theme issue 'Statistical workflow'.
Zheng et al. (Thu,) studied this question.