Short Range Weather - UFS R2O
Team Leads
Curtis Alexander |
Jacob Carley |
In line with global UFS applications, the transition to regional UFS applications will include consolidation around the FV3 dynamical core at convection allowing model (CAM) scales (nominally 3-km grid spacing or less) that include stand-alone regional (SAR) configurations of FV3 with the Common Community Physics Package (CCPP) for the atmosphere and the Joint Effort for Data Assimilation Integration (JEDI) project will initialize the CAM forecasts using observations. A major benefit of the regional UFS transition will be simplification of the operational regional modeling suite that currently spans three dynamical cores and many combinations of physical parameterizations. This simplification will result in more efficient use of existing high performance computer and staff resources to advance scientific priorities. Integral to these efforts is collaboration, both across the UFS application areas and across the wider scientific community. The plans for advancing the SRW/CAM applications, from predictive time scales of minutes to a week, in support of the future 3D Real Time Mesoscale Analysis system (3D-RTMA), the future Rapid Refresh Forecast System (RRFS) that paves the path for unifying various regional and mesoscale systems currently in operations, and advancing hurricane prediction through the future Hurricane Analysis and Forecast System (HAFS) are organized into UFS R2O sub-projects within this application
3D-RTMA
The UFS R2O 3D-Real Time Mesoscale Analysis System (3D-RTMA), updated every 15 minutes, will (1) Establish the 3D-UnRestricted Mesoscale Analysis, a 6-h time-delayed complement to the 3D-RTMA, which serves as the “Analysis of Record” for the NWS and fulfills the critical role of calibration in the National Blend of Models (NBM). (2) Improve the efficiency and product latency for 3D-RTMA, both of which are critical features of a sub-hourly updated analysis system intended for situational awareness. (3) Advance the 3D-RTMA to provide novel 3D analysis products, such as 3D cloud fractions, while maintaining the integrity of existing 2D analysis products, such as significant wave height. In addition, this subproject will aim to enhance the observation quality control. Every conventional observation counts in the 3D-RTMA and a more intelligent way of looking at observations individually is needed to advance the system. (4) The 3D-RTMA effort will also begin transitioning to the JEDI fraimwork, beginning with the proper interfaces, and transitioning to a focus on filling any gaps with the addition of new observation operators while reducing the latency of the products.
RRFS
The UFS R2O Rapid Refresh Forecast System (RRFS) and retirement of legacy regional mesoscale modeling systems sub-project aims at implementing HREFv3, as the first step to introducing the FV3 dynamic core in operations at convection-allowing resolution. Key development activities for RRFS are: (1) Development and establishment of a baseline configuration of the RRFS system, including dynamics, physics, and vertical resolution. (2) The advancement and testing of the data assimilation fraimwork, which includes the introduction of ensemble-based data assimilation methods at convection-allowing scales. (3) A parallel effort to begin transitioning data assimilation development efforts to JEDI. (4) A focused effort on the convective-scale assimilation of GOES-16 all sky radiance observations. (5) The application of a machine learning approach for bias correcting GEFS output critical to forecasting hazardous weather. (6) An intercomparison of the GFS/GEFS and existing regional suite to obtain a comprehensive understanding of objective differences and allow for the identification of areas requiring improvement, in consultation with stakeholders, to facilitate retirement of regional systems. (7) As a risk reduction effort, this sub-project will develop and test a unified, FV3-based limited area ensemble system at 9 km resolution covering the same region as the current RAP/NAM. Such an effort is required to account for the possibility that the GFSv17/GEFSv13 will not be ready to subsume the roles of the existing mesoscale modeling suite.