Process-Oriented Diagnostics to Inform the Physics Suite of Future GFS Implementations using NOAA's Unified Forecast System

To support NOAA’s Environmental Modeling Center (EMC) in selecting an advanced physics suite for Version 16 of the global weather configuration of NOAA's Unified Forecast System (the Global Forecast System - GFS), the Developmental Testbed Center (DTC)’s Global Model Test Bed (GMTB) project team conducted an array of analyses and diagnostics for evaluating four physics parameterization suites. The diagnostics are process-oriented and cover the following atmospheric processes: precipitation and moist physics, planetary boundary layer (PBL) evolution, and the transition from large scales to meso- and smaller scales. Most of these diagnostics were beyond the scope of both EMC’s verification statistics database (VSDB) and the Model Evaluation Tools (MET), where these new diagnostics provided valuable information to the independent evaluation panel for a comprehensive assessment. The diagnostics offered insight into possible issues with the convection, PBL, and land-surface schemes, the interactions among parameterizations, and with the dynamical core. For example, the metrics helped identify a dry bias in the tropics in three of the physics suites, while a wet bias was present in the fourth physics suite. It suggested that the dry bias may be associated with precipitation triggering too early. The PBL analysis showed that all four suites exhibited different levels of an overmixed lower troposphere. And the differences in the kinetic energy spectra implied that the physics-dynamics coupling varied by physics suite. These and other results will be described more fully during this presentation. In addition, the DTC has published a diagnostic report for this test (https://dtcenter.org/eval/gmtb/phytest2019/GMTB-2019-Adv_Phys_Test_Diagnostic_Final_Report.pdf).