EMC/HPC/MPC synergy meeting highlights 1/29/01
This meeting was led by Peter Manousos and followed a generic agenda denoted by the subtitles below. Attendees included Geoff Dimego, Hua Lu Pan, Jim Hoke, Dave Reynolds, Steve Jascourt, Kevin McCarthy, Bill Bua, and Keith Brill.
1. IBM SP
a. Status of Phase II: Geoff Dimego reported that the second super computer ("development machine") to be installed in accordance with the completion of Phase II will be ready for use by the end of February 2001. After this time the 12km Eta parallel will be run on the production machine with the possibility of another Eta parallel being run in sigma coordinates at this resolution on the development machine. A T254 parallel MRF may also run on the production machine starting sometime over the summer after the changes now being tested in the MRFY are implemented.
2. Notes from EMC
a. Mesoscale Modeling Group: Geoff Dimego stated that the 22km parallel Eta will be implemented on January 30th 2001 (most likely at the 12z cycle) including the Ocean Modeling Branch's new 2D-VAR SSTs, but will not include three changes listed in the latest "Intent to Change" memo released last week. These three changes include the way the 10km winds are post processed, a feature to allow a restart of the 84 hour extension, and boundary conditions needed for the nested runs. These 3 changes along with the extension to 84 hours will be implemented into the operational model by the end of February 2001. Thereafter, by early to mid spring, three more changes will be implemented as a bundle to the operational Eta including a retuned 3DVar, assimilation of precipitation observations, and an upgrade to the land surface package. This will lead into parallel testing of the 12km Eta. Other projects under development include removing from the analysis radiances from channels weighted predominantly in the upper stratosphere, applying the automated parameter estimation technique of Purser and Parrish to 3D-VAR tuning parameters, and modifying the EDAS balance constraint to limit the divergence tendency (currently the constraint is simply thermal wind balance).
b. Global Modeling Group: Hua-Lu stated that the Winter Storm Reconnaissance decoding software has been approved by CAFTI and will be implemented operationally by the end of February 2001. The GMB will also implement analysis changes including incorporation of GOES 10 and NOAA 16 satellite data, the inclusion of freezing levels (similar to the Eta) in the post processed output, and posting sea level pressures through a T80 filter at that time. Also implemented operationally to the global model by the end of February will be analysis changes including incorporation of GOES 10 and NOAA 16 satellite data, the inclusion of freezing levels (similar to the Eta) in the post processed output, and passng sea level pressures through a T80 filter. The MRFy (which features convective momentum transport and prognostic cloud water equations) has a targeted implementation of June 2001. This is however contingent upon 1) Running the MRFy retrospectively on a month worth of data that featured both convective feed back over the CONUS in the operational MRF as well as spurious spin ups in Central and South America, and 2) Having a positive validation by HPC on these retrospective parallel runs prior to implementation in June. Finally, there has been a request by TPC to run the 06 and 18Z runs of the AVN out to 120 hours to support 5-day hurricane track forecasts, though EMC questions whether the skill in 5-day track guidance warrants this use of resources.
3. Status of Ensembles
Geoff Dimego reported that the completion of Phase II will allow the 10 member SREFs to potentially be run operationally at 48km resolution by late spring or early summer of 2001. Most likely the SREFs would be run from 9Z and 21Z, though there will be tests against later SREFs to assess impact of having to run before the 12z and 00z RAOBs come in. The problem is finding a time slot; running after 12z and 0z would require waiting until the operational Eta finishes. There was some discussion of the upcoming meeting at NCEP led by EMC Deputy Dir. Ralph Petersen (on how to get the SREFs operational by May 2001).
HPC International Desk Meteorologist Michael Davison reported that the MRFy is performing well over the S.H. HPC Peter Manousos asked on behalf of MPC James Partain if the SST analysis that will be implemented in the operational Eta will also be utilized by the global models. Hua Lu Pan reported that yes, this is planned, but a 2 week parallel test and evaluation is needed. Finally, Peter Manousos asked Geoff Dimego if since the Eta will be utilizing the new SSTs, will the NGM (since it is run off of the EDAS) as well, and what potential effect that might have on MOS (countering the "frozen" model aspect prescribed to the NGM). However, Geoff stated that the Reynolds SST analysis (not the improved SST analysis) will remain in the NGM and therefore not effect MOS.
5. Next Meeting Proposed February 26, 2001 at noon in room 209.