EMC/HPC/MPC/NCO Synergy Meeting Highlights - 03/27/00
For one hour each month representatives of NCEP's Environmental Modeling Center (EMC), Hydrometeorological Prediction Center (HPC), Marine Prediction Center (MPC), and NCEP Central Operations (NCO) meet informally over lunch in the World Weather Building (WWB) in Camp Springs, Maryland, in what has become known as the Synergy Meeting. Primary purposes of the meeting are to keep the HPC and MPC forecasters apprised of EMC's model development plans and NCO's computer and model implementation plans, as well as for EMC and NCO staff to learn about model performance and forecaster needs.
THE FOLLOWING ARE HIGHLIGHTS FROM THE MOST RECENT MEETING. INFORMATION IN THESE HIGHLIGHTS SHOULD BE REGARDED AS TENTATIVE AND FOR PLANNING PURPOSES ONLY. THE INFORMATION IS SUBJECT TO CHANGE AND SHOULD NOT BE RELIED UPON OR REFERENCED AS FIRM DEADLINES OR ESTABLISHED POLICY.
This meeting was led by Peter Manousos (Science and Operations Officer - HPC) and followed a generic agenda denoted by the subtitles below. Attendees included Carl Staton (Director - NCO), David Reynolds (Forecast Operations Branch Chief - HPC), John Ward (Production Management Branch Chief - NCO), Jim Hoke (Director - HPC & MPC), Steven Jascourt (COMET Post Doc from the Office of Meteorology residing at WWB), Hua-Lu Pan (Global Modeling Branch Chief - EMC), William Bua (COMET Post Doc from the Office of Meteorology residing at WWB), Edwin Danaher (Development Training Branch Chief - HPC), Keith Brill (Research Meteorologist - HPC), and Steven Lord (Director - EMC).
III. Meeting Highlights
1. Status of Code Transfer onto IBM SP
John Ward stated this task is complete. The J-90 machines have been purposely deactivated to ascertain if any code was unintentionally not transferred to the SP. Official announcement of code transfer completion will be sent out by NCO this week. A very significant benefit of code conversion onto the IBM SP is that there is minimal variability in product delivery time from cycle to cycle.
2. Planning of Phase II IBM SP
The Phase II computer system under the IBM SP procurement is slated to be delivered in September 2000 and includes an approximate doubling of disk space, an increase from 2 processors per node on 384 nodes to 16 processors per node on 128 nodes, and an increase of processor speed from 300 MHz to 375 MHz. A team will be established within 2 weeks of this synergy meeting to help identify modifications of current code to accommodate these scheduled improvements. The improvement associated with Phase II sets the stage for the use of higher resolution operational models including a 10-km Eta and a T254 AVN/MRF as well as a coupled ocean/atmosphere GFDL hurricane model run out to 5 days 4 cycles per day.
3. Status of Ongoing Parallel/Testing Efforts from the Global Modeling Branch
Hua-Lu Pan provided an outline to the meeting attendees of the T170 parallel test slated for implementation May-June 2000. This included the following:
a. Model Changes
- Reducing the resolution toward the poles in forecast mode (but not analysis mode).
Currently, the resolution at the poles is 1.2 km but will be reduced to 7 km
- A small change to the convective scheme will improve the threat score and reduce
the bias in the QPF. This should also reduce the false alarm rates of
model-generated tropical systems.
- Use of a 30-second topographic data base as opposed to the currently utilized
10-minute data base.
b. Data Preparation
- A new scheme will be implemented to relocate tropical storms. Instead of "bogusing"
in the storm, the scheme will relocate a model-generated system to the location
of the observed storm.
- Ozone analysis modification. The Climate Prediction Center (CPC) found that the
AVN/MRF depleted the levels of ozone over the Antarctic too quickly, something the
new scheme will alleviate.
- Inclusion of GOES-10 sounder data.
- Accounting for horizontal translation during ascent when specifying the location of
4. Forecaster Feedback to the EMC
HPC - The apparent wet bias of the AVN was noted, as well as the re-occurrence of convective feedback problems. Ideally, the changes slated for implementation stated above in item 2 will mitigate these problems.
5. Next Meeting Proposed April 24, 2000 at noon in WWB EMC room 209.