EOMF-09. Assessment of Simulation Uncertainties of the Unified Forecast System Medium-Range Weather Application v1.0 on Multiscale Atmospheric Processes

Abstract
The Unified Forecast System (UFS) is a community-based comprehensive Earth modeling framework that is designed to contain the source code of NOAA’s operational numerical weather prediction model. The UFS Medium-Range Weather (MRW) Application focuses on the prediction of weather phenomena out to approximately two weeks. We assess the model uncertainties of the MRW App in multiscale forecasting situations that have been identified as challenging for the operational configuration of the MRW App, namely the Global Forecast System (GFS). In the synoptic scale, there is right-of-track bias for the 2019 Hurricane Lorenzo over the North Atlantic Ocean. The misrepresentation of the position and strength of the North Atlantic subtropical ridge is a key hypothesis to explain Lorenzo's early recurvature. In the mesoscale, the underestimated convective available potential energy (CAPE) at one 2017 summertime storm event is partially attributed to the dry soil moisture. In the local scale, we found that the overestimated surface friction velocity leads to overmixing near surface in one radiation inversion case. This leads to a warmer near surface temperature and thus to a weakened simulated inversion strength. To facilitate physics innovations and model development, we provide a platform to share the results and resources of these multiscale case studies with the broad community.