Application of Hierarchical System Development Approach to Inform Physics Development of the Global Forecast System

Numerical Weather Prediction models continue to suffer from systematic biases, many of which are related to missing or poorly represented physical processes and deficiencies in the underlying model physics. To provide a holistic understanding of the deficiencies of the Global Forecast System (GFS) physics and contribute to NOAAs next generation fully coupled Earth system model, we conducted comprehensive and process-based evaluations of the operational GFS version 16 (GFSv16) physics suite. The evaluation efforts, following the Hierarchical System Development (HSD) paradigm by using both top-down and bottom-up approaches, covered cloud-radiation processes, planetary boundary layer (PBL) evolution, and land-atmosphere interactions. For example, the analysis of diagnostic variables over the entire globe suggests that biased partitioning between liquid and frozen hydrometeors in the microphysics scheme may help explain the biased radiative fluxes at the surface. And the perpetually underestimated warm-season marine stratus is likely associated with model deficiencies in representing selected cloud microphysics and radiation processes. Common Community Physics Package Single-column model simulations for a case at the Atmospheric Radiation Measurements Southern Great Plains site indicate a close relationship among weaker-than-observed nocturnal radiation inversion, overestimated entrainment or capping inversion, abnormally warm and windy conditions after sunset, and excessive clouds, which suggests room for improvement of the PBL and surface schemes, land surface model, and their interactions in the GFSv16 physics. Acknowledgment: This work is supported by the Unified Forecast System (UFS) Research-to-Operation (R2O) project.