Unidata Users Committee Meeting Summary

14-15 March 2002

Boulder, Colorado

 

Members: UPC Staff:
Rich Clark, Millersville Univ, Chair Steve Chiswell
Bill Fingerhut, Lyndon State College Sherri Corpuz
Mark Laufersweiler, Univ of Oklahoma Ethan Davis
Michael Morgan, Univ of Wisc-Madison Ben Domenico
Rajul Pandya, West Chester Univ Steve Emmerson
Larry Riddle, Univ of Calif-San Diego Joanne Graham
James Steenburgh, Univ of Utah Jo Hansen
  Robb Kambic
  Scott Kehoe
Ex-Officio Members: Jeff McWhirter
  Linda Miller
Tom Whittaker, Univ of Wis-Madison, ATAC Rep Don Murray
  Russ Rew
  Jeff Weber
  Stu Wier
  Anne Wilson
  Tom Yoksas
  Yuan Ho

 

 

Meeting Agenda

Administrative Matters

Absent:

ACTION 1: An agenda item should be included for the May 2002 Policy Committee meeting to discuss a replacement for Arlene Laing for the Users Committee

ACTION 2: Research into providing the WRF beta model output on the CONDUIT data distribution should be conducted - Miller will coordinate.

Status Reports

Status reports were provided on-line. The discussion was based on questions or issues from the status reports.

General Discussion:

It was suggested that an important aspect of the next LDM should be the recognition of when data/grids have dropped, there can be automatic retrieval of the data.

ACTION 3: Domenico will follow up with the Users Committee members to identify current problems and future needs of the LDM.

The volume of data available to the community via the IDD has grown rapidly, but there are limits on how much data can be practically delivered with push technologies. A combination of push and pull will be necessary in the near future. DODS, ADDE, and MetApps are all demonstrating the advantages of pulling desired subsets of data from remote servers.

CONDUIT offers another data conveyance, but is really being used for high-resolution model data (not currently available on NOAAPORT) from NCEP.

MetApps Task Force needs more active participation. There were several suggestions for increasing feedback, including having
Unidata staff assist more directly in installing the applications.

MADIS - Patty Miller, Forecast Systems Laboratory

To respond to the Action Item 1 from the last Users Committee agenda, Patty Miller provided an overview of the Meteorological Assimilation Data Ingest System (MADIS) project..

The MADIS site provides support for real time meteorological research for research in data and assimilation with observations and APIs, user documentation and utilities. Initial data sets include 4,225 stations from 15 mesonets (including MesoWest), boundary layer profilers, GPS met network (with Integrated Precipitable Water). Distribution of 5-minute ASOS data is planned. Cooperative data availability with 15 minute temporal frequency, citizen weather observers. Several state departments of transportation (Iowa, Kansas,
Minnesota) are providing road data. The data are available via FTP or using LDM. MADIS is supporting the IHOP project. COAMPS, in conjunction with Mississippi State is being made available with QC checks. The raw data are provided, but QC flags are included if the user wants QC'd data. The archive began last year. Future steps include additional mesonets, COAMPS and WRF data assimilation. Most data are freely available. All that is necessary is to complete a short form to access the data.

Discussion:

The Users Committee thought this was a very valuable resource. Perhaps FSL could consider a DODS server to access the MADIS system. The API is created in Fortran. Much of the WRF is done in Fortran. Pre-built binaries can be accessed. FSL would like other contributors to provide their data to the system.

ACTION 4: The MADIS project should be announced to the Unidata Community. Miller and Miller will work together to create an announcement.

NOMADS - Glenn Rutledge, NCDC

There is no long-term archive for climate and weather models currently available at NCDC. This led to the concept of NOMADS at NCDC. The idea is to provide access to models, product development, and to foster research within geoscience (ocean, water, climate). This will increase interaction between the climate and weather communities. Many groups, such as CDC, NCAR, PMEL, GFDL, COLA GRADS/DODS and Unidata are collaborating on NOMADS. The THREDDS infrastructure is very important to the project. NOMADS will be a THREDDS data provider NOMADS would like to provide NGI access to experimental data sets. NOMADS will be an FY04 NOAA level initiative.

A Unidata university representative is being sought to sit on the science team panel to define which models to archive at NCDC. The question of two members was discussed. The purpose of two would be one representative from weather and another from climate. Discussion with Cliff Mass, Chair of the CONDUIT Working Group, has previously taken place regarding a CONDUIT representative.

ACTION 5: Steenburgh and Clark will discuss possible candidates to participate on the NOMADS Science Panel, via email, and make the recommendation to the Unidata Users Committee for final determination.

Survey Analysis - Discussion led by Clark

ACTION 6: Survey information should be used to inform the Unidata five-year proposal process.

Discussion:

A question posed by a site representative was presented concerning how to get funding for systems administration (if state support is not available) at a university site. The Users Committee discussion included several approaches:

ACTION 7: Check with UCAR to solicit information from the last UCAR community survey on the item of funding for System Administration

Over the past year, approximately 12 academic institutions have independently approached FSL about gaining access to either or both
FX-Linux and FX-Net. This prompted the UPC into evaluating the work that would be required to embrace FX-Linux as a supported Unidata offering, and the User Committee to ask the user community:

"How important is it to your organization to have computer systems that function (nearly) identically to those used in the National Weather Service (e.g., FX-Linux, FX-Net)."

Several survey respondents indicated that it was "worth investigation" systems like FX-Linux and FX-Net.

Tom Yoksas provided an overview of FX-Linux from a Unidata perspective and led an in-depth discussion of FX-Linux in relation to the survey question.

During the course of the discussion, the issue of the current IDD headers being altered (each product's CCB is currently being stripped) for distribution in the datastream was considered. Modifying the NOAAPORT products to keep the CCBs would allow site receiving IDD feeds to be able to use FX-Linux if they so desired. This has been an action item that Unidata has been planning to undertake, but has not been able to make a priority issue.

In Yoksas' opinion, one full-time support person would be required to adequately support FX-Linux system use in the Unidata university community. Furthermore, significant work would need to be done so that NOAAPORT GINI imagery (broadcast in NOAAPORT channels 1 & 2) could be delivered to a wider audience via the IDD. Use of the wavlet compression techniques employed by FSL in support of the FX-Net workstations should be considered.

ACTION 8: The Users Committee advises the UPC to modify NOAAPORT products injected into the IDD by _not_ stripping the product CCBs. Associated with this change will be needed modifications to the LDM queue scanner, pqact, to equip it with the (default) capability of stripping off the CCBs.

For now, the Users Committee asked Unidata to continue monitoring the progress with FX-Linux and to provide updates on activities.

ACTION 9: Connect survey results to the Unidata Strategic Plan, adopted February 2002.

ACTION 10: The Users Committee should provide input on key issues of the Users and Organizational surveys to Rich Clark by 1 April 2002. The results should be summarized and provided to the Unidata community in an email from the members of the Unidata Users Committee.

Composite Radar Product - Chiswell, Yoksas

By combining GEMPAK's grids and McIDAS compressed area files, a 1 km composite has been created using compressed area files.

Three radar base reflectivity composites have been added to the FNEXRAD feed. They are:

  1. 6km national NOR
  2. 1km regional NOR floater
  3. 10 km national RCM

The NIMAGE stream 1km national NORM is not generally available, due to the bandwidth requirements. Their size is from 1MB to 14MB.

No VIL or Precip composite has been created.

Testing is being done on "motherlode" and six community servers.

Discussion:

Several issues need to be addressed:

MetApps - Don Murray

Much of the motivation for conducting the next Users Committee meeting for two full days is to allow for the Users Committee members to have hands-on experience with the MetApps.

Strategic Planning and Proposal Discussion - Domenico

Ben summarized the discussions which took place during the Unidata Policy Committee meeting. Focus on the Core Values and Constituentcy (Section 2) in the Unidata Strategic Plan were discussed. They are:

In Section 3.1 Addressing Crucial Scientific and Social Problems, the Grand Challenges and the study of the water cycle are key grand challenges evolving for the Unidata community in coming years. Building on the WWW for scientific data is an important area, as discussed in Section 3.2, Simplified Discovery and Use of Geoscience and Geographic Data. Section 3.3 Collaboration and a Strengthened Unidata Community draws on the document written by Charlie Murphy, in his role as Chair of the MetApps Task Force. Section 3.4 guides us in the direction of running high-resolution local models, using the infrastructure built by the UPC. In Section 4.0 Goals and Objectives, the committee cautioned Unidata to be careful in writing the proposal to identify two levels of support when depicting the goals. Section 5, Focused Strategies, current data sources are important to continue, but high-resolution local models, including local and coupled models, hydro/atmos models and ensembles from these models, are also very important. Other types for consideration are flood, landslide, plume, dispersion, pollutants, volcanoes, pollens, etc. In Section 6. Anticipated Change, Unidata needs to:

ACTION 11: Discuss Support separately from the "engagement of diverse users" as depicted in the Strategic Plan.

Discussion:

GIS software is being discussed and considered in many programs. Some of the issues raised were:

The Salt Lake City Olympics used a dispersion model which fed into the model output through a GIS and provided highway information, etc.

Access to the NCAR Mass Store was discussed. This was a topic that was raised during the CONDUIT meeting in January too. Steve Worley, SCD, is working on the issue, and Bob Gall, USWRP, has talked with Al Kellie, SCD Director, about improving access to a broader community.

ACTION 12: A two-part action:

  1. Unidata will revisit the community data portal access to the Mass Store and discuss possible DODS and/or THREDDS server access.
  2. The Users Committee should identify specific datasets desired from the NCAR Mass Store.

Equipment Grants - Graham

Joanne distributed a "draft" proposal for the purpose of review before submission to the NSF for funding to be provided to reinstate the equipment grants for the Unidata community. The NSF can no longer oversee the equipment grants program, due to the overhead and general resources required at the NSF to conduct the peer reviews, awards, etc. Unidata, with guidance from the Users Committee, has been researching methods to conduct the "call for proposals," with peer reviews and hardware grants for the Unidata community. The funding would be sought from the NSF for (5-years) and the effort would be coordinated by Unidata, with at least one member of the Users Committee participating on the peer review panel. The following salient points were made:

The final draft proposal will be sent to the Users Committee in April for review, prior to submission to the NSF.

Summer Workshop "2003"

A general discussion of a theme for the summer workshop resulted in the proposed title of, "Expanding Horizons Using Environmental Data and Model Output for Education, Prediction, and Decision Making." This theme has the potential to stimulate various interests, including:

Michael Morgan and Mark Laufersweiler volunteered (that's a bit of a stretch) to co-chair the workshop. Congratulations!

ACTION 13: The Users Committee would like the agenda and links, to relevant agenda items, in one file for downloading to their laptops, PDAs, etc. . Robb Kambic offered to check into a different method for the next meeting.


Linda Miller - lmiller@unidata.ucar.edu
External Liaison, Unidata
University Corporation for Atmospheric Research
P.O. Box 3000
Boulder, CO 80307-3000
303 497-8646 fax: 303-497-8690