Specific Priorities
Link research to operations
RAL will provide a continuous link between the research and operational communities to rapidly transfer new NWP technology into operations. This type of technology flow-through is the essence of the WRF concept and must be done efficiently in order for the WRF program to become successful.
Improve accessibility
RAL will provide easy and complete access to all components of the WRF community model system used in the U.S. for both research and operations. Accessibility by all principal investigators interested in working with the WRF model is essential to its success. This includes access to a variety of data sets including output from model and system prediction tests, field program data sets for model and model component evaluation, operational model system output for comparison with experimental systems as well as routine meteorological data to initialize and evaluate model and weather prediction system tests.
General testbed facility
RAL will provide a facility for testing and evaluating a wide range of weather forecasting systems including hybrid systems that post-process NWP output and other data for presentation to decision makers. It will also develop and test new and existing verification schemes for validating and evaluating output from a wide variety of prediction systems including NWP models, hybrid systems and expert systems that combine information from several sources.
The WRF Developmental Testbed Center (DTC)
The WRF Developmental Testbed Center (DTC) is a major component of the JNT and fully established, though at a modest level. The DTC is a distributed facility with components at NCAR (called the NCAR DTC), a component at the NOAA Global Systems Division called the GSD DTC, and a component at the Naval Research Laboratory called the NRL DTC. The figure below depicts this structure and emphasizes both the distributed nature of the DTC and its central location between the research and operational communities. It also takes advantage of extensive computing resources at various super-computer centers such as NCAR, NOAA, GSD and the naval computer center at Bay Saint Louis, Mississippi.
The Mission of the WRF DTC is to establish a national collaborative test bed where NWP research and operational communities interact to test and evaluate numerical weather prediction models in order to accelerate transfer of these technologies to operational users, and to test research models and techniques used as tools to provide a better scientific understanding of atmospheric physical processes.
Maintenance of the WRF model system: A major role of the DTC is to maintain the primary code system for the WRF model, known as the Reference Code and its documentation. This includes all components of the model system necessary to construct a working model such as initialization procedures, the dynamic model (or core) various physics options and a post-processing and verification system. Currently the Reference code includes two cores called the ARW and NMM cores, and a variety of physics options. In the near future it will also include the COAMPS core from the Naval Research Lab (NRL). The DTC in Boulder is currently maintaining the NMM core, MMM maintains the ARW and the NRL DTC maintains COAMPS.
Testing and evaluating new NWP technology: The DTC is a facility where new NWP technology is tested and evaluated for possible advancement into operations. This will include testing new codes that are under consideration in the WRF Reference Code repository. In addition, The DTC will evaluate new NWP technology to provide guidance to the operational centers for improving the operational systems, and to the research community for use of the models in research applications.
Simulated reflectivity from the DTC Winter forecast experiment where the model was run at 5-km resolution.. Comparison is made with the composite radar reflectivity map.
Transferring NWP technology into operations: The DTC will work with the operational centers at NCEP, AFWA and FNMOC to perform tests prior to decisions for changes in the operational system. This will include verification that the changes will provide measurable improvements in the operational model through a series of tests known as pre-implementation testing.
Conducting tutorials on the use of WRF modeling system: The WRF system is a community model that will allow the research and operational communities to work with a common model. The WRF Program encourages wide usage of the model. It provides a tutorial program on the model intended both for new users, to instruct them on getting started with the WRF system, and for existing users, to familiarize them with the latest new features in WRF.
Connecting to the research community through the visitor program: The DTC is intended to be the place where the research community and operational community can interact to advance the science of numerical weather prediction. To facilitate this interaction, the DTC will annually release announcements of opportunity for members of the broad NWP community (scientists from universities, mission agencies and the private sector) to work with the DTC on priority projects. This is a visitor program where the visitor can work with members of the DTC at NCAR, the other components of the DTC located at FSL and NRL, any of the operational centers (NCEP, AFWA, FNMOC, NRL) and be located at any of these centers during their visit.
Developing and testing verification systems: Since testing and evaluation of weather prediction systems including NWP models is central to the DTC, it will also work with developers of new verification systems to evaluate those systems and determine whether the new techniques should become part of a standard verification package. The DTC will maintain verification software for use by the community in evaluating the output of NWP models or other weather prediction systems. The DTC will also maintain a suite of data sets for use in the verification system, as well as examples of forecasts from existing operational models and experimental forecasts to use as comparison for testing new weather prediction technology.
Bias and Equitable Threat Scores from the DTC Winter Forecast Experiment that tested both the NMM and ARW models at 5-km resolution during the winter. Confidence intervals were added to the traditional scores to indicate whether the differences in the scores between the models were statistically significant. Note that the bias scores are statistically different for high thresholds, but the ETS scores are not for any threshold.