NCEP Modeling Synergy Meeting Highlights: November 3, 2008


This meeting was led by Bill Bua and attended by Hendrik Tolman, Chris McGee, Mary Hart, Yuejian Zhu, Keith Brill, James Hoke, Ed Danaher, Joe Sienkiewicz and Geoff DiMego. Stephen Jascourt of UCAR/COMET, Steve Silberberg of AWC, and Jeff Waldstreicher of Eastern Region attended by remote teleconference.


1. NCO


Chris McGee of NCO reported that the NAM-NMM, SREF, and RUC are all running in parallel presently. The RUC is to be implemented the week of 10 November, while it is hoped the NAM and SREF can be implemented before the holiday moratorium around the 15th of December. Great Lakes wave model will be implemented 18 November, including the use of NDFD winds to force the waves.  GFS will not have anything new implemented until after the holiday/new MPP moratorium which will end in April 2009. A detailed list of all upcoming upgrades and implementations can be found here.


The new operational MPP will be under acceptance testing this month, with the new development MPP build scheduled to begin April 2009.  Everything regarding new MPP hardware is currently on schedule.




2a. Global Climate and Weather Modeling Branch


No report except for global ensemble (see below).


2b. Mesoscale Modeling Branch


Geoff DiMego said that the new RUC implementation will be briefed to Dr. Uccellini on 4 November for implementation the following Tuesday.  See Stan Benjamin’s presentation on the RUC website for further details.


NAM parallel is currently being run by NCO.  Details of changes are included in previous synergy meeting notes.  Upper air scores are improved overall, with the improvement increasing with forecast lead time.


A unified RTMA is running in parallel over all regions:  CONUS, AK, HI, and PR.  Since RTMA is the basis for ensemble downscaling, improvements should be noted in the downscaled ensemble data as well. Surface pressure and terrain height are being added to the RTMA data files.


The NAM will be transitioning in structure from the WRF to the National Environmental Modeling System (NEMS) framework.  Eventually, all modeling at all scales will be done using this framework.


Switch to WRF-MOS and termination of NGM MOS:  MDL is implementing MOS for the NAM.  Eta-based MOS (METMOS) will be shut off on 9 December 2008 at 12 UTC, while the NGM MOS (FWCMOS) will be shut off on 3 March 2009 at 12 UTC.


2c. Global Ensemble Prediction System


Yuejian Zhu reported that the new global ensemble forecast system (GEFS) is ready to be run in parallel, as soon as there is room on the MPP to do so.  The new GEFS will increase in resolution from T126 to T190 (approximately equivalent to increase in resolution from 90- to 70-km).  Also added will be “stochastic physics” (perturbing the physics time tendencies for forecast variables to represent the uncertainty in the physical parameterizations) and 8th order horizontal diffusion. In retrospective tests from August and September 2008, there was an increase in GEFS skill of about 1 day resulting from these changes.  Hurricane track forecasts seemed to be improved from the GEFS as well.


Downscaling of the GEFS and the combined NWS/Canadian North American Ensemble Forecast System (NAEFS) data to 6-km resolution in the Alaska domain is also scheduled to be implemented.


2d. Short Range Ensemble Forecast System


SREF implementation includes making all members about 32 km in resolution and increasing NAM membership from 6 to 10 (2 each for the ARW/UCAR and NCEP/NMM versions) while decreasing Eta membership from 10 to 6.  Precipitation fields have shown more structure and upper air has shown better scores, but precipitation verification by standard methods have shown mixed results.


2e. Marine Modeling and Analysis Branch


Hendrik Tolman reported that the operational HYCOM model tends to generate “fake features” unique to ocean models that need to be eliminated, such as spurious eddies. This will be addressed through an increase in horizontal resolution. Sea-level height altimetry (from JASON) will be getting into HYCOM through its data assimilation system at mid-2009 as well. On the Global Wave Model, new files are being created that combine wave GRIB with GFS GRIB data.  The next major change to the wave model will be an increase in resolution in 2010, which should improve wave model physical parameterizations and thus the wave forecast as a whole.






No feedback at this time.




Steven Silberberg reported that the AWC is very happy with the new RUC being implemented next week; otherwise nothing to report.


4. The next Synergy Meeting will be held Monday, December 1, 2008, at 12:00 pm EST in Room 209 at EMC, with remote conference capability.