NCEP Synergy Meeting Highlights: January 5, 2009


This meeting was led by Bill Bua and attended by Chris Magee, Keith Brill, John Ward, Geoff DiMego, and John Derber. Stephen Jascourt, Steve Silberberg of AWC, Steve Weiss of SPC, Richard Pasch of HPC, Dave Novak and Jeff Waldstreicher (Eastern Region), Dave Myrick and Andy Edman (Western Region), Bill Ward (Pacific Region), and Carven Scott (Alaska Region) attended by remote teleconference.


Checkers: Please pay particular attention to highlighted items.


Pre-NCEP centers reports:


Attention was brought to an item from the NCEP 2008 Annual Model Review, to change the meeting time to something later than noon, to accommodate the regions in the western CONUS and OCONUS. The main purpose would be to allow all regions to participate, especially with respect to pending implementations. Some thought a decision had been made at the meeting to change the time to 3 pm ET the last Monday of the month, but not all were aware of this. Bill Bua took an action to get agreement on a new day and time that would maximize participation by all.


1. NCO


Chris McGee of NCO reported that the NAM, NAM MOS, RUC and RTOFS changes went in as discussed at the last Synergy meeting on 1 December 2008. There will be a parallel GFS run with changes to the Global GSI; AWC, HPC, and CPC have agreed to participate in the associated evaluation.


We are in a moratorium generally until the new IBM Cirrus 6 is installed; earliest lifting of the moratorium is April 2009.




2a. Global Climate and Weather Modeling Branch


A new physics package for the GFS is ready for testing with the goal of implementing prior to the 2009 hurricane season. The GCWMB also hopes to implement a more significant GFS change late in 2009. The Climate side of the shop is (and will be) running the GFS reanalysis/reforecast on the Power 5.


2b. Mesoscale Modeling Branch




A small bug in the code just implemented related to the LSM and initialization of snow cover, has been found and will be fixed upon the return of Eric Rogers. As noted by NCO, the implementation of NAM with partial cycling on the global data assimilation system was successfully launched on 16 December.


No major changes will be made for the remainder of 2009, as the MMB concentrates on changing to the NEMS modeling framework. A GSI upgrade may be done in Fall 2009 if necessary, which may include a change to the balance constraint if a consistently better formulation can be found.


2c. Global Ensemble Prediction System


No report.


2d. Gridpoint Statistical Interpolation (Data Assimilation) System


No report.


2e. Short Range Ensemble Forecast System


No report.


2f. Marine Modeling and Analysis Branch


No report.




NCEP Centers and NWS Regions


Western Region (Andy Edman) reported that the GFS recently has been struggling/flip-flopping on western US storms. John Ward asked for cases, and Andy Edman will get him some. Andy also mentioned hes already shared some cases (specifically, 18 and 21 December 2008) with Zoltan Toth for testing the GEFS for improvements in forecasting for them.


NHC (Richard Pasch) reported that they are aware of the web pages for the upcoming GFS/GSI implementation performance on hurricanes. NHC will review these and give feedback. He asked and was tentatively granted an extension on the GFS/GSI evaluation until mid-to-late February because the holiday period slowed down review.


General discussion:


Case study reruns: John Ward said such runs would take several months to perform as part of an evaluation of a parallel, because of the data cycling/data assimilation issues and computer availability, so it may not be feasible at present.


30-day evaluation periods: Internal briefing to the Change Control Board (CCB) would be made available to the evaluators to assist them in knowing


  1. What to look for in the parallel versus operational runs, and
  2. Whether the changes in the parallel would be expected to significantly impact forecasts used by the evaluator (e.g. SPC and an analysis change in the GFS might not be needing to evaluate).


It was suggested that a link be provided to a website containing real-time forecasts and any cases that have been rerun.


It was noted that not having the parallel in AWIPS is an inhibiting factor in having WFOs participate in parallel evaluations; ways to remove that impediment should be examined.


4. The next Synergy Meeting will be held at a time to be determined by the expanded Synergy Group, per initial discussion at the 5 January 2009 meeting. E-mail will be sent once the time is determined.