EMC/HPC/MPC/NCO/CPC Synergy Meeting Highlights 12/2/02
This meeting was led by Bill Bua and was attended by Geoff DiMego, John Ward, Hua-Lu Pan, Jun Du, Pete Manousos,, Larry Burroughs and Keith Brill.
1. IBM SP
John Ward reported that in November, there was nothing more consequential than a few bug fixes for operational code. The acceptance test for the new parallel machine ended on December 2, 2002 at noon. No problems of consequence occurred during the test and the computer up-time was 99%, which compares favorably to the performance of the current system.
He noted that some “tweaks” would be performed this week:
• Upgrade to the AIX operating system
• Climate and some hourly operating jobs still need to be transferred to the new system.
• General users should have access to the new machine by 15 December 2002. The names of the new system would likely be “frost (the operational machine)” and “snow (the development machine)”.
• Data was tentatively set to be moved on Monday or Tuesday, 9-10 December. Any data on asp or bsp after that date will need to be migrated over to the new machines by the users.
• The asp machine will continue to be available until the new machine is declared operational. Bsp is devoted to the regional reanalysis.
2. Notes from EMC
a. Global Modeling Group: Hua-Lu Pan reported that a new parallel GFS is running with NOAA-17 satellite data being assimilated. This is being done because of the potential decommissioning of NOAA-16 as a result of data problems. We may have to break the moratorium if NOAA-16 goes out, to implement the NOAA-17 inclusion in GDAS. The final decision on the implementation would come from Dr. Uccellini.
b. Mesoscale Modeling Group: Nothing to report other than a few bug fixes.
c. Global Ensembles: Nothing was reported.
d. Short Range Ensembles: Additional changes have been made to the SREF web page. Accumulated precipitation for 3-, 6-, and 12-hour intervals have been added to the web page. Some bad news, however, is that the Eta-KF runs have been failing over the last few days. While the cause for the failure is being researched, the Eta-KF will not be running in parallel.
A question was asked about probability of precipitation type. Apparently, snow and sleet are combined into one probability, which forecasters participating in the winter weather experiment would like to see separately on the web page. Jun Du said this would be done.
e. Marine Modeling and Analysis Branch (MMAB): The main news is the preparation of a new fog forecast system to be made operational mid-to-late fog season next year (The fog season runs from April through September.). It’ll be run out to 7 days at 3-hourly forecast intervals, with 4 cycles per day (00, 06, 12, 18 UTC), all year round. The new fog forecast system will be based on cloud microphysics and other parameters and their relationship to visibility.
3. Input to EMC and other information from Operational Centers
HPC: Pete Manousos stated that in contrast to last year, this winter season seems to be producing plenty of opportunities to test the SREF and the recent model upgrades in the Winter Weather Experiment (II). SREF data via the web seems to be coming in on a timely basis so that it can be assessed on how well it does (and how useful the ensemble is) in winter situations.
4. Next Meeting Proposed Monday January 6th, 2003 at noon in room 209.