The Rapid Refresh Forecast System (RRFS) is the new NOAA convective-allowing model (CAM) and ensemble (CAE) system for North America. DTC support for RRFS development began in 2020, with work directed at benchmarking the performance of NOAA’s Unified Forecast System (UFS)-based regional configuration of the Finite-Volume Cubed-Sphere (FV3) atmospheric dynamical core, with an emphasis on convective-allowing resolutions (3 km). These evaluations have evolved to include partnering with the Spring Forecast Experiment (SFE) component of NOAA’s Hazardous Weather Testbed (HWT) to evaluate state-of-art CAEs, including Model for Prediction Across Scales (MPAS)-based configurations. To quickly assess output from RRFS “prototypes” (a specific “snapshot” of code used to create a set of forecasts), the DTC Verification Workflow is used to facilitate the rapid and straightforward comparison of new RRFS prototypes to current operational capabilities. Powered by the enhanced Model Evaluation Tools (METplus) capabilities, this verification workflow can output a wide array of deterministic, ensemble, and probabilistic verification. Objective verification results are communicated to model developers, decision-makers, and stakeholders, providing them the information necessary to identify areas for model improvement and potential innovation. These timely evaluations have proven to be a crucial component for on-time delivery of the RRFS into operations.
The DTC is also working to help address critical decisions about the design for NOAA’s CAE system. DTC evaluations have addressed time-lagging, stochastic physics, and analysis uncertainty. Originally, these evaluations were conducted in the context of regional FV3 configurations. More recent work is focused on connecting the UFS stochastic physics capabilities to standalone MPAS and providing model evaluation to support NOAA’s efforts to include MPAS within UFS.
Sponsors: NOAA
Collaborators: NOAA OAR (GSL, NSSL), NOAA NWS (OMD, SPC), NSF NCAR
