CONDUIT-CRAFT (C2) Summary

AMS-Atlanta Meeting

31 January 2006

Meeting Agenda  

Introduction:

Kelvin Droegemeier, C2 Chair, opened the meeting with remarks about CONDUIT and CRAFT and the importance of active participation in the decision making of the projects. Louis Uccellini, NCEP Director, provided an historical perspective of CONDUIT (1997) via the USWRP Science Steering Committee. He discussed the link between NCEP and Unidata, and the importance of the research being conducted through the use of operational models, and the usefulness of research results feedback for improving models. NCEP does not serve the academic community directly so the connection with Unidata is important. He also emphasized the importance (recognized in 1999) of real-time access to hi-resolution model data. The requirements are important toward elevating the project for sustainability and visibility.

Two questions were posed:

  • Why CONDUIT
  • What are the current issues

Priorities need to be set to enhance and improve data sets provided through CONDUIT. There is a need to hear from the 50+ recipients who are receiving the CONDUIT feed. What data are missing? A coordinated response and agreement is necessary from the recipients. This process could be done by making use of the email list that has been set up for discussion. Some community users have made their desires known. We will post those suggestions to the user list for discussion.

The email list is c2@unidata.ucar.edu Interested people who are not members of the list can subscribe at:

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

Steve Chiswell, Unidata, provided an overview of the current status.

Brent Gordon, NCEP, provided the NCEP status. NCEP Operational Model Suite as of January 2006

Mohan Ramamurthy, Unidata, provided a glimpse into the future of data systems and services.

Discussion:

Some data sets will be made available in higher resolution versions. We need to be cognizant of datasets that are placed on NOAAPORT. We do not want to replicate any data sets available on NOAAPORT.

Dr. Uccellini indicated that they are creating a change management board and would like to have an outside voice. The question of what notification is required was discussed. Should there be a 30 day, 60 or 90 day notification? An important aspect of change management notices for users is to provide time for changes needed on their specific systems/computers. Some sites are still not prepared to use GRIB2. We must emphasize the need to transition from GRIB to GRIB2 or they will be left behind.

Fred Branski, NWS/TOC ( Telecommunications Operations Center), discussed the “moratorium” which involves the NCEP to TOC feed, and the fact that full operation should be underway sometime in February. The NCEP to TOC upgrade of infrastructure is taking place with the Andrew File System (AFS) going away.

The LDM server is out of date and cannot handle the additional load required by the CONDUIT feed. There is a need to “operationalize” the system. Unidata will discuss this issue and report on proposed options.

Discussion:

Jim Steinberg, University of Utah, impressed the need for the university perspective. Again, the email list is one important route to being heard by the group and facilitating needs and discussion.

It was suggested that a data meeting be held during the year to discuss the future and to focus on:

  • future data sets
  • security issues
  • spectrum of users/how to increase user base i.e., number of users
  • things to be done now
  • modernize approach
  • seek to remove redundancies

What can be done short-term?

    • The LDM server decision must be made very soon.
    • Change management procedures can be identified and implemented
      • 30 day, 60 day, 90 day

Level II discussion:

  • Question of KMet group – two radars around the Caribbean were supposed to become available

ACTION: Contact Tim Crum for status.

ACTION: should create a BAMS article to increase awareness of CONDUIT. Could include CRAFT, and possibility discuss some future activities, such as Grid computing in the article?

Discussion:

Should the article be submitted for major BAMS article or one of the Nowcast articles in the BAMS.

Linda Miller
Community Services, Unidata

(with thanks to Jo Hansen, Unidata)