Flow routing is a fundamental process of Earth System Models’ (ESMs) river component. Traditional flow routing models rely on Cartesian rectangular meshes, which exhibit limitations, particularly when coupled with unstructured mesh-based ocean components. They also lack the support for regionally refined models (RRMs). While previous studies have highlighted the potential benefits of unstructured meshes for flow routing, their widespread application and comprehensive evaluation within ESMs remain limited. This study extends the river component of the Energy Exascale Earth System Model (E3SM) to unstructured Voronoi meshes. We evaluated the model’s performance in simulating river discharge and water depth across three watersheds spanning the Arctic, temperate, and tropical regions. The results show that while providing several benefits, unstructured mesh-based flow routing can achieve comparable performance to structured mesh-based routing, and their difference is often less than 10%. Although the unstructured mesh-based method could address several existing limitations, this research also shows that additional improvements in the numerical method are needed to fully exploit the advantages of unstructured mesh for hydrologic and ESMs.

The representation of submesoscale processes in ocean models is critical for accurately representing the exchange of mass and tracers between the surface mixed layer and subsurface waters. Comparisons of these processes using multiple models is an important step for the validation of different numerical schemes and identifying possible future opportunities for model coupling. This study investigated the representation of submesoscale processes using two ocean models, the structured regional ocean model ROMS and the unstructured global ocean model MPAS-O in an idealized domain. Results from the two models are highly similar for multiple statistical metrics of submesoscale dynamics including frontal processes, and also show that the impact of ocean model choice is smaller than the choice of spatial grid resolution. Overall, these results show promise for the successful coupling of the two ocean models, and highlight additional differences in the representation of along- and cross-frontal divergence estimates for further exploration.