Janet Derrico1,2, Christina Holt1,2, Emily Carpenter1,2, Samuel Trahan1,2, Ligia Bernardet2, Isidora Jankov2, Jaymes Kenyon1,2, Naureen Bharwani1,2
1Cooperative Institute for Research in Environmental Sciences (CIRES)
2National Oceanic and Atmospheric Administration, Global Systems Laboratory (GSL)
The Rapid Refresh Forecast System (RRFS) development team leads at NOAA Environmental Monitoring Center (EMC), NOAA Global System Lab (GSL), and NOAA National Severe Storms Lab (NSSL) recently published a recommendation that the next version of the RRFS transition to using the MPAS dynamic core1. They showed that the FV3 dynamic core is likely the source of large biases in storm intensity and rainfall and that transitioning to the Model for Prediction Across Scales (MPAS) will address the biases.
The NOAA GSL research team has proven eager to get started working with MPAS as an alternative to the FV3-based ufs-weather-model. While efforts are underway to integrate the MPAS dynamical core into the ufs-weather-model, research to better understand the behavior with the desired physics has already begun. The MPAS Application developed at NOAA GSL is the first uwtools-based application built from the ground up that houses a build system for all the necessary research components (WPS, MPAS-Model, MPASSIT, UPP, etc.), a collection of workflow definitions broadly used among scientists, and a testing framework for the end-to-end system.
The MPAS App was used in its prototype form to drive real-time forecasts for the annual Hurricane Forecast Improvement Project evaluation period during the 2024 hurricane season. These cold-start forecasts consisted of a static domain covering a portion of the Atlantic basin, including the Caribbean and Gulf of Mexico and much of the US Coastline often impacted by landfalling tropical systems. The system was driven by the Rocoto workflow manager and included postprocessing via NOAA National Severe Storms Laboratory’s MPASSIT grid interpolation software, the Unified Post Processor (UPP), and the NOAA GFDL Vortex Tracker, along with graphics produced by GSL’s pygraf software.
In this presentation, we’ll discuss the uwtools framework for configuring and running the workflow and its components, as well as some of the features of the MPAS App currently available for general use.
1Carley, Jacob, et al. "Mitigation Efforts to Address Rapid Refresh Forecast System (RRFS) v1 Dynamical Core Performance Issues and Recommendations for RRFS v2", 2023, https://doi.org/10.25923/ccgj-7140
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner