Object-based Verification of the HRRR, FV3-based RRFS, and MPAS-based Configurations during the 2024 Hazardous Weather Testbed Spring Forecast Experiment

With the operational implementation of the Rapid Refresh Forecast System (RRFS), NOAA will be retiring several regional deterministic and ensemble modeling systems. While initial efforts focused on a Finite Volume on a Cubed Sphere (FV3)-based RRFS, momentum is shifting towards inclusion of the Model for Prediction Across Scales (MPAS) within the Unified Forecast System (UFS) for use in the RRFS. As such, it is essential to understand and quantify the performance of FV3- and MPAS-based convective-allowing model configurations against observations as well as current operational baselines, such as the High-Resolution Rapid Refresh (HRRR) model. NOAA’s Hazardous Weather Testbed Spring Forecast Experiment (HWT-SFE) provides a rich venue to test and evaluate new and/or emerging technologies targeted toward improving the prediction of severe weather; as such, the 2024 HWT-SFE included WRF-, FV3-, and MPAS-based convective-allowing model guidance, providing a wealth of data to evaluate. Traditional verification of the HRRR, FV3-based RRFS, and several MPAS configurations for the 2024 HWT-SFE period will be examined in a presentation by Ketefian et al. (2025). This poster will focus on object-based verification using the Method for Object-based Diagnostic Evaluation (MODE) within the enhanced Model Evaluation Tools (METplus). With well-documented biases related to convective storm structure and intensity in the FV3-based RRFS, we will be focusing our evaluation on comparing and contrasting performances of the HRRR, FV3-based RRFS, and MPAS configurations for composite reflectivity and accumulated precipitation. Object-based evaluation can provide crucial information with respect to model errors related to spatial scale, coverage area, displacement, orientation, and intensity, which can then provide insight back to model developers and stakeholders.