Evaluating and constraining storm-resolving climate models

Global storm-resolving models (GSRMs) enable simulating the atmosphere at an unprecedented level of detail. They can directly compute key atmospheric processes which conventional climate models must parameterise (Hohenegger et al, 2023). This significantly improves the accuracy of the simulated cloud and precipitation fields (e.g., Hohenegger et al., 2020), which is a key step towards reducing the

uncertainty in projections of global warming (Randall et al., 2003).

Nonetheless, significant uncertainties in GSRM simulations remain due to substantial differences in model design choices, such as parameterisations of remaining unresolved subgrid-scale processes (Zadra et al., 2018). Adequate evaluation metrics for GSRMs are needed to enable researchers to further improve models and thereby facilitate predicting future climates with confidence (Palmer, 2016).

In this project, we aim to develop robust evaluation metrics for storm-resolving simulations to guide future developments and GSRM improvements. We aim to use the metrics to gain a deeper understanding of the physical processes which control cloud formation utilising GSRMs and satellite observations.