NCEP Modeling Synergy Meeting Highlights: June 2, 2008


This meeting was led by Mike Brennan and attended by Jeff Ator, Keith Brill, Mark Iredell, Geoff DiMego, Yuejian Zhu, Vera Gerald, Mary Hart, Jim Hoke, and Joe Sienkiewicz. Stephen Jascourt of UCAR/COMET and David Novak of NWS Eastern Region attended by remote teleconference.


1. NCO


Jeff Ator listed the following scheduled implementations:


·        Alaska gridded MOS : June 10

·        Multi-model wave ensemble: June 24

·        Hi-Res window : August 5


Jeff also listed upcoming implementations for which RFCs have been submitted but implementation dates have not yet been scheduled:


·        Great Lakes wave model using NDFD winds

·        RTMA for Hawaii and Puerto Rico


A detailed list of all upcoming upgrades and implementations can be found here.




2a. Global Climate and Weather Modeling Branch


Mark Iredell discussed the GFS implementation that was recently canceled based on performance metrics and feedback from the operational centers.  The branch will be investigating the following issues:


·        The relationship between the GFS and H-WRF. Specifically, why did the parallel GFS sometimes perform better for TCs while the H-WRF initialized off of the parallel GFS perform worse that the H-WRF initialized off the operational GFS for the same systems?

·        Why were the QPF and hemispheric statistics worse from the parallel than the operational GFS during spring 2008?

·        To attempt to isolate the impacts of the bundled changes, the performance of the new GSI running with the old forecast model and the performance of the old GSI running with the new forecast model will be examined.


Mark is looking for volunteers to help participate in the evaluation of these diagnostic tests. Interested parties should contact Mark at


 2b. Mesoscale Modeling Branch


Geoff DiMego discussed the Hi-Res Window upgrades that will be implemented in August. These changes were largely based on requests from SPC and include:


·        Hourly output

·        The immediate generation of GEMPAK output files as each output time becomes available instead of waiting until the end of the run to begin the GEMPAK processing

·        The computation of updraft helicity


A 75-day change notice was required due to the need to modify data labels in AWIPS.  Parallel output will be made available to the TNCF prior to August 5.


Development has been completed on the RTMA code for Hawaii and Puerto Rico. These are the final NAM-based RTMAs to be implemented. Implementation of the RTMA for Guam will require the development of SmartInit grids from the GFS, which will eventually result in the availability of GFS SmartInit grids to the field.  Work is also underway to increase the resolution of the CONUS RTMA to 2.5 km and the Alaska RTMA to ~ 3.0 km. This RTMA development effort is being performed by NCEP without most of the required funding.


A NAM bundle under development will be frozen in July or August for a 1st quarter FY09 implementation. This bundle includes:


·        NDAS changes to allow heavier use of the global background fields. This is expected to result in improvement in the forecast beyond 36 hours

·        The addition of several new data types, including TAMDAR, Canadian AMDAR, and additional mesonet data

·        Minor land surface changes


2c. Global Ensemble Prediction System


Yuejian Zhu reported on an upgrade for the GEFS scheduled for the 4th quarter of FY08. The following changes are being tested:


·        An increase in resolution for the first 7.5 days of the run from T126 to T190

·        An increase in resolution from T126 to T190 for the entire length of the GEFS run

·        Running the 00Z GEFS cycle out to 31 or 35 days. CPC has been evaluating test runs of this length and found that the ensemble is skillful out to 24.5 days with the first order signal of the MJO

·        Horizontal diffusion changes

·        Stochastic perturbations to increase spread


The following changes to the NAEFS were also discussed:


·        The earlier arrival of CMC ensemble data by up to 80 minutes should result in NAEFS output being available up to an hour earlier

·        Blending of the ECMWF ensembles into the NAEFS is being tested

·        A two-year project to incorporate FNMOC ensemble data into the NAEFS began in April. FNMOC data will be evaluated for the first year and work to implement the new data into the NAFES will begin in the second year


2d. Short Range Ensemble Forecast System


Geoff DiMego reported that the next SREF upgrade will likely occur early in FYY 09. This upgrade will include:


·        An increase in the resolution of all members to 32 km

·        Adjusting the components of the SREF to include more WRF members

·        The new SREF will be composed of:


o       6 Eta members (3 BMJ, 3 KF)

o       5 RSM members

o       10 WRF members (5 NMM, 5 ARW)


2e. Marine Modeling and Analysis Branch


Vera Gerald reported that the branch is awaiting feedback on the wave model ensemble scheduled for implementation on June 24. 




Eastern Region


David Novak asked for further clarification from Mark Iredell on the GFS upgrade cancellation. Mark indicated that the reasons for the lack of improvement shown by the GFS parallel are under active investigation and until understood, there cannot be any timetable for the next implementation.  Dave also asked Geoff DiMego if MMB had noticed continued good performance in the NAM since the last upgrade. Geoff stated that the branch was focused on the next implementation and had not had time to evaluate the current performance, though it was anticipated that the largest benefit of the last upgrade would result from the gravity wave drag parameterization which would have largest effect during the winter season. Geoff re-iterated that with the upcoming NAM upgrade the largest improvement should come after 36 hours, and that improved run-to-run consistency is also expected.


Stephen Jascourt asked for background information on the EMC website crash last week. Geoff DiMego noted that only 3–4 % of the EMC site’s content was lost in the incident. Stephen also asked about problems receiving attachments on the NCEP Model Evaluation listserv. Jeff Ator suggested he contact the NCEP Helpdesk for this issue.


4. The next Synergy Meeting will be held Monday, June 30, 2008, at 12:00 pm EST in Room 209 at EMC, with remote conference capability.