Romain Fiévet
| Department | Service Climate Physics |
| Group | SCLab Computational Infrastructure and Model Development Climate Surface Interaction |
| Position | Research Scientist |
| phone | |
| romain.fievet@mpimet.mpg.de | |
| Room | B 415 |
I am a research scientist with a background in computational fluid dynamics and aerospace engineering. Here at MPI, I specialize in storm-resolving climate modeling, cloud microphysics and hm-scale regional simulations. I work on the full research pipeline: from developping and running the ICON model, to processing and analyzing large complex datasets. By running at ever-refined spatio-temporal resolutions, I seek to understand how fine-scale atmospheric processes - such as cloud microphysics and turbulence - can, upon improved representation, reverse-cascade and ultimately feedback onto larger scales.
Google Scholar Profile: View complete publication list
Professional Network: LinkedIn
Work samples
Limited-area modelling for multiscale cloud physics
During summer 2024, the ORCESTRA (Organized Convection and EarthCARE Studies over the Tropical Atlantic) field campaign was launched to advance our understanding of the physical mechanisms governing mesoscale tropical convection organization. The campaign investigated how convective organization interacts with tropical waves and air-sea processes, and examined its broader implications for climate dynamics and Earth's radiation budget.
In parallel with the observational campaign, we conducted high-resolution numerical simulations using the storm-resolving ICON Limited Area Model in its Sapphire v1.0 configuration. These simulations employed explicit convection at kilometer-scale resolution, with high-frequency model outputs specifically designed to match the temporal sampling of ORCESTRA observations from dropsondes, radiosondes, and satellite-based radar measurements.
This coordinated approach allows for detailed model validation by comparing similarly-sampled observational and numerical datasets. By revealing the biases present in the model, this work contributes to improving it. Conversely, this work shows how storm-resolving models can be used on their own to investigate the physics of small-scale processes driving ITCZ multi-scale dynamics.
Global-nested simulations
Rather then run ICON over a limited-area domain, having to prescribe and construct imperfect boundary conditions, we can run the model with several domains at once, nested within a global mesh. This approach allows us to perform numerical experiments zoomin-in on any part of the global, while observing the effect of each mesh-refinement level.
Large-eddy simulations of cold pools
Convective cold pools (CPs) are regions of cool air forming beneath precipitating clouds. As rain droplets evaporate during their fall, they cool down the surrounding air, thereby increasing its weight and causing it to sink down. As the cold sinking airflow hits the ground, it spreads outward in all directions, creating strong wind gusts which are important mediators of the weather system. Indeed, such gust fronts can suppress cloud formation in some locations, by cooling the near-surface air, while simultaneously triggering cloud formation along their edges by wedging underneath and lifting the ambient air. Numerical weather models should ideally account for these effects. Unfortunately, the models' grid resolution is often too coarse to represent the thin CPs, and it is not yet known what resolution should be used. In this study, high-resolution simulations of idealized CPs are carried out at different mesh resolutions, from 800 to 25m, allowing a systematic exploration of CPs' sensitivity to grid resolution. Our simulation results may help reveal configuration requirements for high-resolution simulations and guide climate model development.
Peer-reviewed publications