EMC/HPC/MPC/NCO/CPC Synergy Meeting Highlights 4/27/03
This meeting was led by Bill Bua and was attended by Hua Lu Pan, Pete Caplan, Eric Rogers, John Ward, Kevin McCarthy, James Hoke, and Larry Burroughs.
1. IBM SP and CCS
New system will not be installed until August 2004, with implementation not until Dec. or 2005.
2. Notes from EMC
a. Global Modeling Group: Development is starting now on Grid-scale Statistical Interpolation (GSI) Data Assimilation System. This new GDAS will take quite awhile to evaluate, thus we are not sure about any implementation date. Testing is currently being done at low resolution. The reasoning for going from Spectral (SSI) to GSI is that it gives more flexibility regarding specification of error covariance. SSI error covariances get spread across the globe. Also, we are working toward global and regional DAS using the same code. Other development includes sigma-p hybrid coordinate and a resolution increase, in that order, after SSI.
b. Mesoscale Modeling Group: DGEX is now in a production slot, 2 times per day, plus an AK run. The new 0-84 hour Eta forecast is now running okay. New AWIPS build will get DGEX information out to the field. Until then, the data will have to be obtained from NCEP. The field has been looking at the data qualitatively. Eric Rogers has been looking at statistical verification and it seems to be reasonably close. These new requirements are being driven by NDFD requirements at WFOs.
c. Global Ensemble Prediction System: Nothing new to report.
d. Short Range Ensemble Prediction System: No report.
e. Marine Modeling and Analysis Branch (MMAB): HYCOM charter has been adjusted to account for resources in NCO and the lateness of the new computer system.
3. Hydrological Prediction Center (HPC)
Pete Manousos just filed a report re: HPC coordinated evaluation of the DGEX. In general, DGEX and GFS forecasts are similar. Forecasters overall preferred the GFS to the DGEX, however the DGEX was favored when the GFS and DGEX showed similar solutions mostly because the detail offered in the DGEX QPF. Also it was noted the DGEX does not suffer grid scale feedback as dramatically as GFS.ike the DGEX
A question was brought up about making files available from the offtime MREF for HPC (nmap and ntrans). This will be followed up with NCO.
4. Next Meeting Monday June 7, 2004 at noon in room 209.