NCEP Modeling Synergy Meeting Highlights: August 25, 2008

 

This meeting was led by Mike Brennan and attended by Ken Campena, Geoff DiMego, Shawna Cokley, Dan Starosta, Bill Bua, Joe Sienkiewicz, Keith Brill, and Yuejian Zhu. Stephen Jascourt of UCAR/COMET, Steve Silberberg of AWC, David Bright of SPC, Rusty Billingsley of Southern Region, and Jeff Waldstreicher of Eastern Region attended by remote teleconference.

 

1. NCO

 

Dan Starosta of TOC was introduced. He will serve as the acting head of the Production Management Branch for the next few months.

 

Shawna Cokley of NCO reported on several upcoming implementations:

 

·        A minor GEFS implementation will occur after the Critical Weather Day (CWD) is lifted

·        Parallel evaluation of the RTMA for Hawaii and Puerto Rico is scheduled to end this week and implemented is planned before the end of FY08

·        A RUC implementation is scheduled for early October

·        A SREF implementation has been pushed back to late October with a parallel beginning in September

·        Minor changes to the GSI are scheduled to be implemented in the next couple of weeks

 

Shawna noted that the usual December “holiday” change moratorium will last through April 2009 due to the implementation of the new supercomputer.

 

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

 

2. NOTES FROM EMC

 

2a. Global Climate and Weather Modeling Branch

 

A new upgrade package for the GFS is tentatively scheduled for implementation in April 2009. This package contains changes that were included in the previous GFS upgrade that was not implemented as well as coupling with an ocean model (GFDL Modular Ocean Model version 4, or MOM4). This package also includes the new GSI that is awaiting implementation.


EMC is looking for early feedback from HPC and OPC on these changes in October, particularly for the retrospective covering the spring 2008 period when the package tested previously performed poorly.

 

2b. Mesoscale Modeling Branch

 

Geoff DiMego said that the evaluation of the RTMA for Hawaii and Puerto Rico has been hindered by an inability to get the data into AWIPS, but hopes that this will not delay the implementation.

 

MMB will be delivering code changes for the NAM to NCO within the next couple of weeks, with an implementation targeted for late November.  This bundle currently includes:

 

  • A change to the NDAS to allow the use of partial cycling; using global background atmospheric fields at T-12 instead of fully cycling with NDAS fields. This change has resulted in improvement in the large scale forecast, especially at longer forecast projections
  • The addition of several new data types, including TAMDAR, Canadian AMDAR, and MetOp satellite data
  • Minor land surface changes
  • Modification of radiation absorption coefficients for cloud water and ice (in response to problems with a cold air damming event last winter)

 

In FY09, no further upgrades will be made to the NAM until the NEMS framework is adopted. This will result in the removal of the NMM from the WRF framework, and moving the model into a common NCEP modeling infrastructure that will be easier to work with.

 

2c. Global Ensemble Prediction System

 

Yuejian Zhu reported on several upcoming changes to the Global Ensemble Forecast System (GEFS) and North American Ensemble Forecast System (NAEFS):

 

  • On July 29 Canadian ensemble data became available earlier, allowing the delivery time of the final NAEFS products to improve by 2.5 hours. Now, for example, the final NAEFS products from the 00 UTC cycle are available by 0530 UTC instead of 0800 UTC
  • A fix in the GEFS for a problem with the GFS model’s hybrid vertical coordinate is awaiting the end of the CWD for implementation

 

A GEFS upgrade is scheduled to be delivered to NCO in early September and includes the following changes:

 

  • An increase in horizontal resolution from T126 to T190 for the first 7.5 days of the forecast. Beyond 7.5 days, the horizontal resolution will remain T126
  • Extension of the 00 UTC cycle GEFS run out to 35 days to assist CPC in MJO forecasting
  • Adopting the use of 8th-order horizontal diffusion to improve short range forecasts
  • A new ESMF version (3.1.0r) that will allow the use of stochastic perturbations

 

Also, there is a plan to increase the number of variables exchanged between NCEP and CMC for the NAEFS to 78 from the current 52. This will require a 75-day notice of change because it will affect the pgrb_a files.

 

Work is underway to finalize the downscaling of NAEFS output for Alaska for max/min temperature, wind speed and wind direction for evaluation at the HPC Alaska Medium Range forecast desk.

 

A 1-year evaluation period of the FNMOC ensemble system is underway for possible inclusion in the NAEFS. This evaluation is being hindered by transfer problems between FNMOC and NCEP. If the evaluation is positive, an additional year will be taken to implement the FNMOC ensembles into the NAEFS.

 

An analogous 2-year effort to evaluate and implement the ECMWF ensembles into the NAEFS is planned. However, there are data processing problems with the 00 UTC ensemble cycle that must be resolved first.

 

2d. Short Range Ensemble Forecast System

 

Geoff DiMego reported that last minute changes to the SREF implementation resulted in code and script changes that are currently being worked on, but will delay the parallel evaluation and implementation. These changes include hourly output through forecast hour 39 at the request of AWC and SPC.

 

2e. Marine Modeling and Analysis Branch

 

No report.

 

3. FEEDBACK FROM OPERATIONAL CENTERS/REGIONS

 

HPC

 

Mike Brennan voiced concerns about the potential large number of back-to-back or potentially overlapping parallel evaluations coming up, including the SREF, NAM and GFS. Mike stated that parallel data need to be available within a couple of hours of the operational data for forecasters to undertake a full evaluation of the parallel data at the service centers.

 

AWC

 

Steven Silberberg re-iterated the above concerns, noting that AWC would also participate in the evaluation of the RUC as well. He also mentioned data volume issues on their end, due to the large amount of hourly RUC data they receive in various formats (GEMPAK, grib, and grib2) for in house use and for product generation for the FAA.

 

4. The next Synergy Meeting will be held Monday, September 29, 2008, at 12:00 pm EDT in Room 209 at EMC, with remote conference capability.