Jump to content
  • Member Statistics

    17,606
    Total Members
    7,904
    Most Online
    ArlyDude
    Newest Member
    ArlyDude
    Joined

Wx Computer Model upgrades for this winter


Recommended Posts

NCEP only has a NAM and an HRRR upgrade scheduled. I don't know about GGEM UKMET and EURO.

 

http://www.nco.ncep.noaa.gov/pmb/changes/

 

The GFS is going to 13km this year (probably sometime in fall).  Some of the package has changed, but some of then plans can be found here:

http://www.emc.ncep.noaa.gov/GFS/impl.php

 

It is a T1534 L64 package, move from Eulerian to Semi-Lagrangian dynamics, and a suite of physics changes.  There are also some DA changes notably increasing the resolution of the ensemble used to prescribe the background error covariance in the hybrid DA.  A change/bug fix in a cloud detection algorithm has led to pretty significant improvements in terms of use of satellite data along ice sheets (this has yielded big improvements in SH scores in particular).  I"m sure details will be made available soon as a lot of the retrospective testing is already completed.

 

I recall seeing that ECMWF has a spatial resolution increase (T15xx or T17xx or something?) in the works but I don't know when it is going in (probably sometime this year IIRC).  

Link to comment
Share on other sites

The GFS is going to 13km this year (probably sometime in fall).  Some of the package has changed, but some of then plans can be found here:

http://www.emc.ncep.noaa.gov/GFS/impl.php

 

It is a T1534 L64 package, move from Eulerian to Semi-Lagrangian dynamics, and a suite of physics changes.  There are also some DA changes notably increasing the resolution of the ensemble used to prescribe the background error covariance in the hybrid DA.  A change/bug fix in a cloud detection algorithm has led to pretty significant improvements in terms of use of satellite data along ice sheets (this has yielded big improvements in SH scores in particular).  I"m sure details will be made available soon as a lot of the retrospective testing is already completed.

 

I recall seeing that ECMWF has a spatial resolution increase (T15xx or T17xx or something?) in the works but I don't know when it is going in (probably sometime this year IIRC).

Have they worked out the government hardware issues i.e. IBM to Lenovo or are they keeping the existing Commodore 64's it's been running on since inception (sarcasm).
Link to comment
Share on other sites

Have they worked out the government hardware issues i.e. IBM to Lenovo or are they keeping the existing Commodore 64's it's been running on since inception (sarcasm).

There is nothing wrong with the current hardware.  The WCOSS machine (IBM idataplex servers) has been operational for a while now (over a year).  The 13km GFS will fit on the current hardware.

 

The issues come if there are delays to the WCOSS expansion (i.e. phase 2 and phase 3 upgrades), which are at risk partly because of the sale of their server division.  I don't think a final decision has been made.  If delayed, the upgrades planned to take place on the bigger machine will not happen in time (vertical resolution increase to the GFS, extension of GDAS hybrid DA to 4D, etc.).

Link to comment
Share on other sites

The GFS is going to 13km this year (probably sometime in fall).  Some of the package has changed, but some of then plans can be found here:

http://www.emc.ncep.noaa.gov/GFS/impl.php

 

It is a T1534 L64 package, move from Eulerian to Semi-Lagrangian dynamics, and a suite of physics changes.  There are also some DA changes notably increasing the resolution of the ensemble used to prescribe the background error covariance in the hybrid DA.  A change/bug fix in a cloud detection algorithm has led to pretty significant improvements in terms of use of satellite data along ice sheets (this has yielded big improvements in SH scores in particular).  I"m sure details will be made available soon as a lot of the retrospective testing is already completed.

 

I recall seeing that ECMWF has a spatial resolution increase (T15xx or T17xx or something?) in the works but I don't know when it is going in (probably sometime this year IIRC).  

Thanks, didn't know about that page. I assume these will be run in parallel for a while until the switchover.

Link to comment
Share on other sites

Thanks, didn't know about that page. I assume these will be run in parallel for a while until the switchover.

EMC is already running it in near real time (in addition to the several retrospectives that have already been completed).  If you look at the regular EMC monitoring page, scores are being plotted ("HW14"):

http://www.emc.ncep.noaa.gov/gmb/STATS/html/new_acz5.html

 

I'm no longer at EMC so I do not know the status of the code handoff to NCO nor the "official 30 day evaluation" parallel.  I would expect that to be coming quite soon....

Link to comment
Share on other sites

EMC is already running it in near real time (in addition to the several retrospectives that have already been completed).  If you look at the regular EMC monitoring page, scores are being plotted ("HW14"):

http://www.emc.ncep.noaa.gov/gmb/STATS/html/new_acz5.html

 

I'm no longer at EMC so I do not know the status of the code handoff to NCO nor the "official 30 day evaluation" parallel.  I would expect that to be coming quite soon....

 

I haven't heard anything about the GFS yet, though I don't hear everything working in ops.  It appears from a glance that the GFS would fit fine on WCOSS in parallel even with the NAM/NDAS/NLDAS and several others running.  Could be shooting for some time in FY15 Q1.

Link to comment
Share on other sites

  • 2 weeks later...

Heard from a buddy in the UK that the UKMET is getting a decent overhaul relatively soon. Not sure if its just for the tropical season or what, but I've always liked the UKMET due to it's likeness to the ECMWF. 

 

The UKMET received a significant upgrade on July 15.  Virtually everything was upgraded, from the dynamical core to model physics to data assimilation.  It appears that TC genesis and intensification have been improved significantly, although it tends to under-weaken slow moving TCs over shallow thermoclines due to a lack of upwelling (still uncoupled).  It will be interesting to see whether or not east coast wintertime cyclogenesis is also improved. 

 

Public release:

http://www.metoffice.gov.uk/research/news/2014/endgame-a-new-dynamical-core

 

Full details:

http://www.metoffice.gov.uk/media/pdf/s/h/ENDGameGOVSci_v2.0.pdf

 

Looks like it is verifying well, just slightly below ECMWF:

post-378-0-32213000-1407173364_thumb.png

Link to comment
Share on other sites

The UKMET received a significant upgrade on July 15.  Virtually everything was upgraded, from the dynamical core to model physics to data assimilation.  It appears that TC genesis and intensification have been improved significantly, although it tends to under-weaken slow moving TCs over shallow thermoclines due to a lack of upwelling (still uncoupled).  It will be interesting to see whether or not east coast wintertime cyclogenesis is also improved. 

 

Public release:

http://www.metoffice.gov.uk/research/news/2014/endgame-a-new-dynamical-core

 

Full details:

http://www.metoffice.gov.uk/media/pdf/s/h/ENDGameGOVSci_v2.0.pdf

 

Looks like it is verifying well, just slightly below ECMWF:

 

If so, then having the UKMET and The King in your back pocket would be a noteworthy boost to forecaster confidence.  

Link to comment
Share on other sites

  • 2 weeks later...

The GFS is going to 13km this year (probably sometime in fall).  Some of the package has changed, but some of then plans can be found here:

http://www.emc.ncep.noaa.gov/GFS/impl.php

Prelim. schedule has the 13km SL GFS going live in November.  

 

NAM and it's relatives NAM4km and NAM 1.33km Fire Nest just got an upgrade. Most notable of changes so far is the improvement in nighttime temperatures and 10m wind due to the new longwave radiation scheme.

 

http://www.nws.noaa.gov/os/notification/tin14-29namcca.htm

The winds are likely improved because of GWD and roughness length changes.  A lot of the "synoptic scale" improvements are probably from the DA-related changes (move to hybrid var/enkf assimilation using perturbations from the global/gdas enkf, use of DFI in the NDAS, improved satellite data assimilation).  

Link to comment
Share on other sites

  • 2 months later...

Levi Cowan has the GFS Parallel available on his site and running @ T1534 (~13km). Dr. Maue will also be providing the GFS-para for WxBell subs as well in full resolution. For those not familiar with the Dec 9th implementation, here is the upgrade notice. 

 

http://www.tropicaltidbits.com/analysis/models/

 

cor_day5_HGT_P500_G2NHX.png

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...