EMC/HPC/MPC synergy meeting highlights 11/27/00

This meeting was led by Peter Manousos and followed a generic agenda denoted by the subtitles below. Attendees included Jim Hoke, Pete Caplan, Geoff Dimego, Hua Lu Pan, John Ward, Steve Jascourt, James Partain, Bill Bua, Keith Brill, Ed Danaher, Kevin McCarthy and Naomi Surgi.

1. IBM SP

a. Status of Phase II: John Ward stated that the transition to Phase II will be complete by December 7^th 2000. In the meantime, Phase II product delivery has been extremely comparable to Phase I (if anything, faster). Phase II will be throttled in order to maintain the current product delivery times and difference of on the order of minus 10 minutes is expected. There have also been some difference in solutions between models run on Phase I and Phase II and this is primarily due to the inclusion of MORE observational data into the Phase II runs.

b. New Web Page: NCO has put together a new web page that allows a user to monitor the status of each model as it is run on the SP. See http://www.ncep.noaa.gov/NCO/PMB/nwprod/prodstat/. Archive capability of this web page will be investigated by NCO.

2. Notes from EMC

a. Mesoscale Modeling Group: Geoff Dimego stated that the 22km parallel Eta is not always able to be run during the upgrade to Phase II. The extension to 84 hours might be implemented as soon as January. The assimilation of observed precipitation data is being tested in parallel runs at 48km resolution. A spring (March/April 2001) model upgrade is anticipated, including precipitation data assimilation, revised land surface package, and re-tuning of the 3DVar. Thereafter, running the 12km parallel on the production side in pseudo real time will occur before the planned implementation near November 2001. Further research needs to be conducted on the use of sigma vs Eta coordinates at that resolution as well as how to run this out to 84 hours (given the computational resources required to run at this resolution). Geoff suggests that HPC access the parallel Eta grids to be viewed operationally.

b. Global Modeling Group: Hua-Lu stated that one of the two current tests (satellite data ingest) should be implemented ASAP after the upgrade to PHASE II. The second test (prognostic cloud water and convective scheme changes - momentum mixing and multiple cloud tops) will be rerun on warm season data to assess potential impact. This will ideally be implemented before the 2001 hurricane season. Thereafter, the T254 parallel is on target for 2002 implementation (allowing sufficient time to assess this impact of T254 resolution on both warm and cold season data PRIOR to implementation).

3. Status of the Ensembles

There was good discussion on allocation of resources for SREF given the requirement for an operational Eta at 12km resolution. This led into the discussion of major highlights from the NCEP SOO Conference held 14-17 of November dealing with Ensemble Prediction Systems. It was stated that the NCEP SOOs did determine that the ensemble methodology is viable and can be utilized to support and improve BOTH the current and future NCEP operational forecast product suite. The field recognizes it needs to learn to better use existing ensemble products, looks forward to new products including SREF, and is pleased that EMC had the foresight to develop these before receiving field requirements for ensembles. EMC and NCO stated that in order to support operations, they need to know what the NCEP service center ensemble requirements are in order to best determine how ensemble (both SREF and Global) prediction systems will be tasked on the SP and capabilities of future computers. It was noted that such a requirements list is being drafted by NCEP SOOs and will be presented to the NCEP Director imminently.

4. Next Meeting Proposed Monday January 29, 2000 at noon in room 209 (sometime around our first big snow).