Facilitating Development of Physical Parameterizations for NOAA's Unified Forecast System

The National Centers for Environmental Prediction (NCEP) has decided to adopt a modularized system architecture for its new Unified Forecast System (UFS), which will be transitioned gradually to operations within the next few years. Applications of the UFS will range from analysis and nowcasting to weather and subseasonal prediction, encompassing both global and limited-area configurations, in deterministic and ensemble forecast mode. While the dynamical core for the UFS has been selected to be the Finite-Volume Cubed Sphere (FV3) developed by NOAA’s Geophysical Fluid Dynamics Laboratory (GFDL), the physics suites for the various applications are currently under review.

A Global Model Test Bed (GMTB) was established by NOAA’s Next Generation Global Prediction System (NGGPS) program to facilitate the community involvement in the process of developing and testing advanced physics. Much of the GMTB work to date has been devoted to the development of the Common Community Physics Package (CCPP), a library of physical parameterizations which is envisioned to contain both the currently operational parameterizations, as well as those under development for future implementation. The CCPP will be distributed through Github and will be available for users (those that want to conduct NWP experiments), developers (those that want to contribute innovations), and operational centers (as a pool of candidates for future operational implementations). The CCPP framework, and its insulation from the dynamical core and overall model configuration through an interoperable physics driver, facilitates addition of new parameterizations and its use with a variety of dynamical cores.

The CCPP is not just a software infrastructure. Instead, it fits within a larger goal of offering a structure to conduct testing of model innovations. To that end, GMTB has put in place a hierarchical test harness, capable of evaluating physics in controlled settings (such as using a single-column model to run cases derived from observational field campaigns) as well as in realistic semi-operational scenarios using a workflow to run the global FV3-based Global Forecast System (FV3GFS) in forecast-only or cycled data assimilation mode using NCEP’s Global Data Assimilation System (GDAS). A myriad of diagnostics and verification results can be produced using tools developed by DTC (using the Model Evaluation Tools; MET), GMTB, and collaborators.

In this presentation we will describe the first public release of the CCPP (targeted for the Spring of 2018) and the process for community users and developers to engage with GMTB for obtaining the code, conducting experiments, and engaging with NOAA in advancing operational physics.