NCEP Synergy Meeting Highlights: January 31, 2011


This meeting was led by David Novak (HPC) and attended by Geoff DiMego (EMC); John Ward (EMC); Yuejian Zhu (EMC); Chad Carry (EMC); Jamese Sims; Mary Hart (EMC); Stephen Jascourt (COMET/UCAR); Bill Bua (COMET/UCAR); Kathy Gilbert (MDL); David Bright (AWC); Jason Levit (AWC); Steve Silberberg (AWC); Steve Weiss (SPC); Andy Dean (SPC); Mike Brennan (NHC); Keith Brill (HPC); Ed Danaher (HPC); Joe Sienkiewicz (OPC); Bill Ward (PR); Bob Ballard (PR); Eugene Petrescu (AR); Ken Pomery (WR); Warren Blier (WFO MTR); Pete Browning (CR); Bernard Meisner (SR); David Radell (ER), and Brian Miretzky (ER); Ken Smith (AFWA); John Kelly (NOS); Aiyun Zang (NOS); Frank Aikman (NOS)


1. NCO (Becky Cosgrove)


BUFR Update

Bufrlib upgrade was implemented on 1/25. The new BUFR format files can be ingested into AWIPS.

More details available in the TIN. (


CFS V2.0

CFSv2.0 has been postponed.  New start of the 30-day clock was 1/27.  30-day evaluation period ends 2/26, with scheduled implementation on 3/8. See:


High Res Window Runs

A 30-day parallel evaluation is expected to start Feb. 3. Implementation is scheduled for 3/15. The implementation updates the models, and includes physics changes. The implementation also adds a Guam run and associated products, changes the Puerto Rico grid domain, adds additional elements, and changes the directory location of the data. More details available in the TIN.


Additionally, the post-processed product “HREF” will also be added. See below.




NAEFS upgrade is scheduled for March 8. Changes to NAEFS will be technical in nature, including:

1) Switch to CMC's GRIB2 raw ensemble data instead of current GRIB1.

2). Receiving additional (29 more) CMC's raw forecast variables.

3). bias corrected additional 13 variables (12 in stratosphere), and produce NAEFS probabilistic forecast (see attached new products - pink). Existing NAEFS products should be nearly identical.

The FNMOC Global Ensemble will not be included in the derived products as mixed performance was observed upon inclusion. Access to the FNMOC ensemble will be provided via NOMADS.


For the latest schedule updates see:





2a. Global Climate and Weather Modeling Branch (GCWMB) (John Ward)


Climate Forecast System v2.0

A major upgrade to the Climate Forecast System (CFS v2.0) is expected in March, 2011. The upgrade includes resolution changes, assimilation changes, atmospheric and ocean model changes, addition of sea-ice and land modeling, coupling, and other significant changes. Version 1.0 may need to continue running for some time.



-Inconsistencies in the display of MSLP and winds in AWIPS were found to be due to smoothing (by EMC). This smoothing has been improved.


The major upgrade to the GFS in August 2010 introduced three biases:

-2 m temp warm bias

-cold bias in the stratosphere

-low wind bias

Corrective changes are expected to be implemented in April 2011.


A resolution increase, use of a semi-lagrangian model formulation, and possible physics improvements are expected in FY12.



Updates to the datatypes are expected to be made in the April 2011 bundle.



Resolution improvements and physics changes are expected in the summer 2011 timeframe. Resolution increases from T190L28 to T254L42. The physics will be updated to be consistent with the current operational GFS.


Hurricane WRF

Changes are expected in May. More information should be forthcoming.


2b. Mesoscale Modeling Branch (MMB) (Geoff DiMego)


High Resolution Window Runs

A major upgrade to the High Resolution Window Runs is planned for March 2011.  Plans include upgrading the WRF code version (from v2.2 to v3.2) and adding a twice-per-day (00z & 12z) run to cover Guam (which would use GFS initial and boundary conditions). The newer WRF versions for both NMM and ARW use a more conservative advection scheme for the passive variables (e.g. water vapor, hydrometeor fields, turbulent kinetic energy) which has the effect of reducing QPF bias. Pacific region has authorized the new run over Guam to replace the current RSM run over Hawaii.  It will also provide a much improved first guess for the Guam RTMA.  Additionally, the Puerto Rico domain will be expanded to cover Hispaniola to provide QPF guidance for the NOAA/IA supported Hydrologic Research Center (San Diego, CA)’s flash flood support program for Haiti.



The upcoming Q3FY2011 major implementation will transition the WRF-based NMM to a NEMS-based NMM-B.  NEMS = NOAA Environmental Modeling System which is, in turn, based on the basic principles of ESMF (Earth System Modeling Framework).  NMMB = Nonhydrostatic Multiscale Model on a B-Grid where multiscale comes from the fact the NMMB can be run globally or regionally and, with telescoping nests, it can run with synoptic scale, mesoscale to stormscale grid-spacing. This transition allows the long-awaited establishment of multiple internal nests: 4 km over CONUS, 6 km over Alaska, 3 km over Hawaii and 3 km over Puerto Rico/Hispaniola; while the North American parent domain remains at 12 km. An internal EMC parallel has been running for several months with the four nests.  Reinstatement of the Fire Weather / IMET Support run will be accomplished by running a single FWIS nest either within the CONUS domain at 1.33 km or within the Alaska domain at 1.2 km.  The parent 12 km will run out to 84 hours and will populate all existing NAM guidance products.  The 4 child nests will run to 60 hours and distribution is TBD, but one idea is for their output to replace the existing NAM DNG.  Finally, the FWIS runs will extend to 36 hrs with data expected to be distributed via Fx-Net.  Full complements of output will be available on NOMADS and anonymous ftp servers in Silver Spring, MD. 




A major upgrade to SREF is planned now for late FY11 which eliminates the members based on the legacy Eta and RSM models and adds members based on the NEMS-NMMB.  Membership will remain at 21 by having 7 members generated each by NMMB, WRF-ARW and WRF-NMM. This upgrade would also increase resolution from 32-25km to 22-25 km, add a BUFR output, and may transition to an ensemble perturbation approach.  The WRF Development Testbed Center (DTC) is helping evaluate different initial perturbation possibilities. The Eta and RSM models will be discontinued, pending all dependencies on Eta and RSM models can be removed.



Initial data from the hybrid ensemble approach of Jun (2004) ( applied to the high-res window runs is available to NCEP Service Centers. This involves adding the forecast perturbations from the 21 SREF members to the Hi-res window NMM and ARW forecasts.



3 km Alaska RTMA being established along with an additional domain for the Juneau area at 2 km or finer and an additional small domain covering southwest Canada and northwest CONUS (to cover the RFC in that area).


Post Processing

SPC and EMC collaborating to provide CAPE and CIN calculations using virtual temperature. These will be new elements, in addition to the historical CAPE and CIN calculations using the sensible temperature.


2c. Marine Modeling and Analysis Branch (Hendrik Tolman)


Q3 SST of the RTG update planned to include AMSR-E SST data.


Wave Model

FNMOC ensemble information is being incorporated for the probabilistic wave information. This change is expected at the end of FY11.


Additional major physics upgrade in the wave model is expected at the end of FY11.  This is the first major physics upgrade in a decade and will address some of the known issues of WAVEWATCH III



An implementation of the RTOFS Global (HYCOM based) is expected in FY11Q4. The 1/12th degree data is now available on NOMADS and in netCDF format. 


2d. NOS

NOS has established the Great Lakes Operational Forecast System (GLOFS) which will be coming online shortly. Similar models for the Chesapeake Bay, and Tampa Bay are under development and will become operational soon.



3a. MDL (Kathy Gilbert)

-2.5 km CONUS grids are pending work by NCO.


-HRMOS (deterministic and probabilistic MOS QPF) development is nearing completion.


-Working with HPC and NCO to test prototype Day 8–10 MOS guidance.


3b. NCEP Centers and NWS Regions


NCO-PMEL have signed an agreement to support an operational Tsunami model.



4.High Resolution Ensemble Forecast (HREF) presentation  - Jun Du (EMC)

Initial data from  the hybrid ensemble approach of Jun (2004) ( applied to the high-res window runs is available to NCEP Service Centers and will be an operational product as part of the upcoming SREF implementation. This involves adding the forecast perturbations from the 21 SREF members to the Hi-res window NMM and ARW forecasts. Put differently, the product adds the variance from the SREF to the Hi-res Window runs. Ultimately the end goal is an ensemble of high-resolution members. Since HREF only uses postprocessing and does not involve any additional model runs, it can be done now with limited computer resources. The skill is unknown.


Discussion focused on establishing objective and subjective evaluation of the products. A key question is whether the HREF is improved versus the SREF. EMC plans to start objective verification.


The distribution mechanism is not yet identified. It was asked whether NCO can set LDM up to help serve data. This will need to be explored.



5. The next Synergy Meeting will be held at 2:30 pm EST on Monday, February 28, 2011 in room 209, with remote teleconferencing capability.