20040526: LDM/IDD issues at GoMOOS

>From:  Philip Bogden <address@hidden>
>Organization:  GoMOOS
>Keywords:  200405261932.i4QJWUH1002801 LDM IDD CONDUIT

Hi Philip,

Jo Hansen passed your comments along to me for comment.

>I'll share with you, so that you can pass it along to the right person, the
>following email excerpt.  Timely since we were discussing LDM just
>yesterday.  Can you get someone to respond for me, please?  I'd like to
>understand the issue...
>
>
>> There were two major issues with LDM. The first
>> was dependability and predictability. Our
>> environmental program folks observed large
>> variations in time of day that particular
>> products were received, with some products
>> being missed entirely some days. The environmental
>> program group moved to downloads from NCEP
>> ftp site and said it was vast improvement over
>> LDM [since we were looking for something better
>> than ftp downloads, it did not seem like a
>> winner for us:-)].

I would guess that the folks writing this had tried using an old
version of the LDM, LDM-5.x, that we replaced with our current
offering, LDM-6.0.14.  I would further guess that they were trying to
get the high resolution model output that we are delivering in our IDD
CONDUIT datastream using the old LDM.  If both of these assumptions are
correct, then the upgrade of the LDM to LDM-6 would correct at least
one of the problems that they were seeing.  The number of users getting
the CONDUIT datastream through the IDD has grown rapidly since our
update of LDM to LDM-6.  Perhaps it is time for your or your colleagues
to revisit the use of the LDM.  I would be interested in helping
you/them run a test to demonstrate the validity of my assertions.

>> The other issue was that the particular
>> model output files that we were using were not
>> available via LDM.

That is a different matter altogether.  If we don't have a particular
set of data flowing through the IDD, it won't matter how fast the LDM-6
delivery is, you simply won't be able to get it :-)  However, if there
are model data that one wants to see added to the CONDUIT stream, 
folks can lobby for this to happen.  We have an email list for users
interested in CONDUIT and CRAFT data issues that you may want to
participate in:

address@hidden

You (and/or others) can subscribe online at:

http://my.unidata.ucar.edu/content/support/mailinglist/mailing-list-form.html

I would like to encourage you and your colleagues to signup and lobby
for the addition of the data that you are now only able to get using
FTP.  By the way, only subscribed members of lists we maintain are
allowed to post to those lists.  We do this to keep email spams our of
our lists.

>> Both issues were show stoppers for us using LDM

The first issue should no longer be a show stopper.  The second issue,
if it is still true, can be overcome through lobbying.

>Thanks!

No worries.

Cheers,

Tom
--
+-----------------------------------------------------------------------------+
* Tom Yoksas                                             UCAR Unidata Program *
* (303) 497-8642 (last resort)                                  P.O. Box 3000 *
* address@hidden                                   Boulder, CO 80307 *
* Unidata WWW Service                             http://www.unidata.ucar.edu/*
+-----------------------------------------------------------------------------+