[Ncep.nhc.nco_contacts] TSB Morning Rounds - Mon Aug 07, 2017
nhc.tsbadmin at noaa.gov
Mon Aug 7 01:26:07 UTC 2017
Here is today's summary of NHC computer operations:
--- Craig Mattocks
1. N-AWIPS issues: Problem (error) executing SWELL_GlobalWave [formatter?]
on the Pacific side on Saturday. TSB AWIPS II Sysadmin Chris Mello will
2. GFE/NCP issues: Issues remain on LX3 for user account hcobb and logging
into GFE. Called NCF, ticket number 146417 opened. Chris Mello will follow
up on this issue.
1. Other Techincal issues: Significant differences exist between GFS-TABM
and ECMWF-TABM forecast tracks; TABM-EC is closer to the ECMWF
deterministic positions than TABM-GFS is to the GFS sfc positions. Why?
Explanation from Mark DeMaria (paraphrased): The ECMWF SHIPS and related
models (TAB, LGEM, RII) still use the 2016 configuration. For the TAB
model, the averaging radius for computing the steering flow is 450 km. The
GFS TAB uses a smaller radius (350 km) in the 2017 configuration. This
results in slightly lower track forecast errors, but it retains too much
curvature from the GFS vortex, which sometimes introduces significant
trochoidal wobbles in the TAB track. If the TAB model starts off a litttle
slow for a westward moving TC, that introduces a northward bias due to GFS
representation of the TC that was not removed properly. The EC TAB uses the
ECMWF fields, while the GFS TAB uses the GFS fields. If the represenation
of the TC or the steering is different in the two models, TAB wll also be
different. Matt Onderlinde and I working on changing the radius of the GFS
TAB back to 450 km to reduce the vortex influence. We hope to have code to
submit to NCO for operational deployment by the end of this week.
2. ATCF issues: No EC-SHIPS product for Atlantic invest storm al992017 at
18Z Sat Aug 05 2017. TSB will consult with Mark DeMaria to diagnose the
3. Early Wind Speed Probability (WSP) failiure. The 00Z 20170805
nhc_windprob_early_00 job for the west Pacific storm wp072017 failed on
WCOSS because the e-deck (track forecast uncertainty) file
/nhc/save/guidance/storm-data/zfst/ewp072017.dat was huge in size
(approximately 23 MB, compared to the normal size of 5-10 kB), due to
either an error in JTWC's ATCF processing or the addition of new models to
the file. NRL's Buck Sampson was contacted.
The WSP can still run without an e-deck. The code reverts back to
basin-wide track error statistics, usually resulting in only a small change
in the answer. NCO System Programmer Analyst (SPA) Simon Hsiao trimmed the
e-deck and the job ran successfully to completion.
1. From the SDM's log: The 12Z nhc_guidance job failure on Monday 31 July
2017 was due to 0-byte input data. NHC's Monica Bozeman investigated and
found the home directory on their server where the data is processed was
full, so they were unable to copy the data over. The disk space on the
server in question is not monitored by NCO. IWSB Sysadmin Doug Gaer is
compiling a list of all volumes that will be added to the NCO Big Brother
1. RFC 249 - 'get_rtgsst.bash' v1.0.2 – NHC does not need to send the daily
Real Time Global (RTG) SST files to WCOSS anymore since the NHC guidance
suite no longer uses them. This send will be turned off. To be implemented
on Monday, August 7 at 1200Z (8 am EDT).
2. HMON implementation delayed. The new Hurricanes in a Multi-scale
Ocean-coupled Non-hydrostatic (HMON) model v1.0.0 will become operational
on Tuesday, August 15, 2017 at 1200Z (8 am EDT):
3. The 30-day evaluation for the Probabilistic and Deterministic
Extra-Tropical Storm Surge models (PETTS/ETSS) was restarted on Friday,
August 4 because tides were not being calculated at some secondary tidal
locations. During the previous evaluation, an infinite loop was discovered
and corrected in the tidal post-processing code and the precision was
increased to 64 bits. Feedback should be sent to Steven Earle prior to
Tuesday, September 5.
The Service Change Notice (SCN) for the ETSS and PETSS may be viewed at:
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Ncep.nhc.nco_contacts