[Ncep.list.wwatch3.discussion-group] Meeting notes from today's meeting

Wed Sep 9 17:56:45 UTC 2015

Here is what I have from today's discussion. From here on now we will have
to get more involved as we decide on what goes into the next public
distribution. So if you have the time please take a read and help us make
this document as thorough as possible

WWIII Notes -- Sep 9th 2015

This was the first WAVEWATCH III telecon meeting after the Summer break.
The topic of discussion was the coming public release of v 5 of WAVEWATCH
III. v 4.18 was released in March of 2014, so a decision was made a while
ago that we should have another release in March of 2016. Broadly speaking
we would like to stick to this deadline though the deadline can be moved
around a little. However for the moment March 2016 is what we are sticking
with. Keeping that in mind I would like most of the major packages be ready
for integration in the Trunk  by November so we can start doing detailed
regression testing. The packages that are ready earlier should be delivered
early. I am going to go over the different major milestones that we have
identified for v 5 and what we discussed about them. With each milestone I
am going to add target implementation dates (after discussion with the
developers) so we have an idea of status and delivery. Here are the major
packages (ordered as I believe they will be ready for reintegration into
the trunk)

   -

   Sea state dependant stress calculations (switch FLD) -- Development team
   URI : According to Brandon they are very close to having this work
   complete. This package seems to be the closest to being ready and we should
   have this ready for integration into trunk and be tested in the next month
   or so
   -

   Optimize partition tracking algorithm in ww3_systrk -- Development team
   NCEP: This work has essentially been completed by Andre in an offline
   code. It needs to be ported into the subversion server. Since this is an
   independent code it does not preclude any other development and hence its
   implementation is not time constrained. Nevertheless Andre has scheduled a
   completion date by the end of October.
   -

   Implicit scheme for WAVEWATCH III -- Development team Aaron Ronald and
   Mathieu Dutor: I had an offline conversation with Aaron about this. They
   have taken care of most issues associated with this package. This primarily
   affects the unstructured grid option in WAVEWATCH III and they have taken
   some restructuring of arrays that have improved performance and
   significantly enhanced scalability. They expect to have this (with new
   regression tests) to be ready by November.
   -

   ESMF Interface for wave coupling -- Development team Tim Campbell, NRL: Tim
   has added esmf modules to the WAVEWATCH III model. At the moment it works
   for a single grid (grid type regular lat - lon) model. The import feature
   (variables from other models being imported into WW-III) is complete. The
   export feature (variables from WW-III going to other models) is still being
   developed. As soon as that is ready this will be ready to be incorporated
   into the trunk as a first step ESMF capability. This is expected in the
   next month or so.
   -

   Ice Scattering -- Development team, NRL, IFREMER, ENVIRONMENT
CANADA: Waiting
   to hear from Fabrice on when they want the merge back to the trunk to
   happen.
   -

   Converting SMC option to a propagation type from a switch -- Development
   team, UKMO: Jian - Guo has some updates on this. I will create a branch
   so he can provide the updates to the server. They are trying to determine
   how to make the SMC grid option work in multi-grid mode with the other
   grids. For that we may have a series of offline emails / chats.
   -

   Banner et al source term package (switch ST5) -- Development team UNSW: Had
   a chat with Russell before the meeting and they are close to finishing
   their work. Have a few ends to tie up but Russell seems very comfortable
   with the idea of delivering the package by November.
   -

   Space Time Extremes -- Development team NCEP / ISMAR: Waiting to hear
   from Henrique about the status of the project. Validation exercises are
   planned over the fall. Need a rough estimate on when it will be ready for
   integration to the trunk
   -

   TSA -- Development team BIO: Waiting to hear from either Will or Bash on
   the status of this.
   -

   Two way nesting for all grid types -- Development team NRL : This is
   going to be moved to the next implementation as it is seen as a nice
   feature to have but is not at the moment seen as a significant limitation
   of the model.
   -

   Namelists for ww3_multi -- Development team IFREMER / NRL / NCEP: This
   is a major upgrade. Suggestion was made whether it should be left out of
   this development cycle and effort should be made to get all the other
   features in. However there was strong support for the namelist option so
   for now we are going to keep it in the development cycle. Since the idea is
   that this is a design feature to pass information from the ww3_multi.inp
   file to the code, the chance of finding hidden bugs like in some other
   algorithms is limited so the delivery of this feature can be delayed. We
   will make our best effort to get some for of the name list formats in
   ww3_multi. Mickael has some ideas and he will start on this effort and we
   will all try and chime in later to help speed up this process as much as
   possible.
   -

   Memory Footprint and hyperscaling -- Development team NRL/NCEP: This is
   still at an exploratory level. The purpose is to try to reduce the memory
   usage during the initial grid mapping stage. Depending on the level of
   coding needed this may have to be moved to next upgrade. We will take an
   initial look at this and determine what is feasible at this stage.

These are all the points that I have from my notes. If I have missed
anything please let me know.

---------------------------------------------------------------
Arun Chawla
Waves Team Lead
Marine Modelling & Analysis Branch
Room 2101
National Center for Weather & Climate Prediction
5830 University Research Court
College Park, MD 20740
Ph: 301-683-3740
Fx: 301-683-3703
------------------------------------------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://www.lstsrv.ncep.noaa.gov/pipermail/ncep.list.wwatch3.discussion-group/attachments/20150909/280a06d8/attachment.html 