[Ncep.nhc.nco_contacts] TSB Morning Rounds - Thu Jun 01, 2017

TSB Admin nhc.tsbadmin at noaa.gov
Thu Jun 1 14:55:47 UTC 2017


Here is today's summary of NHC computer operations:
--- Craig Mattocks

----------

TAFB:
1. N-AWIPS issues: Yesterday's 15Z danger graphic for TD Two-E did not have  
the wind speed probabilities (WSPs) overlaid on it. This was an unintended  
consequence of turning off the legacy script that generates the WSP web  
graphics, which also creates the Gempak WSP files. Dave Zelinsky turned the  
legacy crontab entry back on and edited out the send-to-web calls to fix  
this problem.

2. GFE/NCP issues: None of the three forecasters on-shift could not log  
into the Collaboration server. Trouble Ticket #136877 was opened with NCF  
yesterday. Lead Forecaster Jeff Lewitsky began troubleshooting with the  
NCF. The NCF told TAFB to select "chat-ncf" as the server when logging in,  
instead of "Primary chat2-ncf". Forecasters could not enter the "marine"  
chat room because that bookmark also pointed to "chat2-ncf". All users  
should remove that bookmark by right-clicking on it, then selecting the  
upside-down reveal triangle "Public Room Search" and right-click  
on "marine" to add a new/correct bookmark.

3. Nearshore Wave Prediction System (NWPS): The swing shift yesterday  
reported that the 18Z run for the Atlantic/NH2 domain did not complete or  
return to GFE. Investigation revealed that the 18Z run came into AWIPS/GFE  
at 0700Z today (about 14 hours after it was triggered). The 00Z run,  
triggered on the midshift, had no trouble. The 12Z run, triggered this  
morning at 11Z, has not completed or come into GFE. These problems are  
probably caused by Storm Surge Unit activity on the sardine server - very  
few CPU cores available for the Swan wave model. TSB is researching other  
computational platforms on which to deploy the NWPS.

4. An accelerated RFC was implemented by NCO yesterday at 1604Z to upgrade  
the Ballyhoo N-AWIPS workstation to the same graphics driver and system  
configuration as the Cumulus N-AWIPS workstation.

5. Miscellaneous issues: The Storm Surge Unit is using the debout standing  
workstation for innundation graphic creation during the ongoing national  
TCV and storm surge watch/warning collaboration testing.
HSU:
1. ATCF issues: The wind radii guidance was slow to appear for the 15Z  
package yesterday, but was on time for the 21Z package.

2. ATCF issues: The EGRI (UK Met Office model, 6-hr interpolated) track is  
not in the ATCF. Appears that this occurred much of yesterday.

Senior Hurricane Specialist Mike Brennan investigated and determined that  
the 12Z EGRR text tracker (see the GTS bulletin below) dropped TD Two-E  
before 12 hours, so the EGRI/2 trackers did not show up yesterday  
afternoon/evening:

** WTNT82 EGRR 311618 ***
MET OFFICE TROPICAL CYCLONE GUIDANCE FOR NORTH-EAST PACIFIC AND ATLANTIC
GLOBAL MODEL DATA TIME 1200UTC 31.05.2017
TROPICAL DEPRESSION 02E ANALYSED POSITION : 13.8N 100.5W
ATCF IDENTIFIER : EP022017
LEAD CENTRAL MAXIMUM WIND
VERIFYING TIME TIME POSITION PRESSURE (MB) SPEED (KNOTS)
-------------- ---- -------- ------------- -------------
1200UTC 31.05.2017 0 13.8N 100.5W 1006 27
0000UTC 01.06.2017 12 CEASED TRACKING
NEW TROPICAL CYCLONE FORECAST TO DEVELOP AFTER 30 HOURS
FORECAST POSITION AT T+ 30 : 14.7N 100.0W

The tracker output should be available in the ATCF now since the 00Z run  
carried the storm out farther in time.

3. ATCF issues: The 18Z HWRF track only showed the initial location of the  
system with no track thereafter, despite the model showing a coherent  
vortex.

Mike Brennan determined that the 18Z HWRF tracker dropped TD Two-E after  
the initial time, so the HWFI (HWRF-interpolated) tracker was missing at  
00Z last night:

EP, 02, 2017053118, 03, HWRF, 000, 135N, 980W, 33, 1004, XX, 34, NEQ, 0000,  
0000, 0000, 0000, -99, -99, 139, 0

​It has been carrying the storm a little longer in subsequent cycles, but  
the 06Z only goes out to 42 hours. See the TC Vitals file:

http://www.emc.ncep.noaa.gov/gc_wmb/vxt/HWRFForecast/RT2017_EPAC/TWO02E/TWO02E.2017060106/two02e.2017060106.trak.hwrf.atcfunix

This is probably due to a detection rule violation in Tim Marchok's tracker  
software - it is not able to follow a shallow/weak storm or one with a  
large horizontal displacement across multiple vertical levels (severe tilt,  
lack of vertical coherence).

4. Miscellaneous issues: The TC Graphics at 03Z were run immediately after  
the TCM was issued. It plotted the new forecast but showed the old time and  
previous watch area. Reran at 0240z.

TSB Developer Dave Zelinsky reminded the Hurricane Specialists that the TCM  
(forecast advisory) and TCP (public advisory) are required inputs to the TC  
graphics, so both must be transmitted before the TC graphics can be run.  
This is not new in 2017.
Other items:
1. NESDIS STAR IT completed its network maintenance and is in the process  
of resuming normal data flows. Himawari-8 and METEOSAT IODC distribution  
expected to begin around 13:00 EDT; back-filled data to take a little while  
longer.
NHC RFCs:
1. Adjust route filtering on the NCEP core network to allow the routes for  
the Hurricane Hotline to reach the Alaska Region. To be implemented on  
Thursday, June 8 at 1900Z (3 pm EDT).

2. On WCOSS, add new project codes HMON-T2O and HMON-OPS to allow jobs for  
the new HMON (Hurricanes in a Multi-scale Ocean-coupled Non-hydrostatic)  
model to be run with their own accounting codes. To be implemented on  
Tuesday, June 6 at 1600Z (noon EDT) on the WCOSS production supercomputing  
cluster.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://www.lstsrv.ncep.noaa.gov/pipermail/ncep.nhc.nco_contacts/attachments/20170601/26a53e28/attachment.html 


More information about the Ncep.nhc.nco_contacts mailing list