Policy Committee Meeting Summary: June 17-18, 1993

Washington, D.C.

Participants

Members	Representatives	UPC Staff
Robert Fox (Chair)	William Bonner (UCAR)	Sally Bates
Otis Brown	David Fulker (UPC)	Ben Domenico
Russell DeSouza	Clifford Jacobs (NSF)	Linda Henderson
Steven Mullen	Mohan Ramamurthy (Users Committee)	Linda Miller
John Nielsen-Gammon	Douglas Sargeant (NOAA)	Sandra Nilsson
Carlyle Wash	John Snow (UCAR Board of Trustees)	Others
Gabor Vali	Timothy Spangler(COMET)	Richard Greenfield (NSF)

Administrative Matters

The summary of the March 1993 meeting was accepted as written.
The next meetings of the Policy Committee will be
7-8 October, Thursday and Friday, in Boulder, Colorado.
10-11 February, Thursday and Friday, in Boulder, Colorado

Action 1:
Review location of February meeting in October.

The Policy Committee acknowledged the contributions of Gabor Vali and Steven Mullen, whose terms expired with this meeting.

Status Reports

Director's Report
A copy of Dave's report was distributed at the meeting; copies of his on-line slides will be distributed after the meeting.

Discussion

Despite the fact that funding from NSF is significantly lower than requested, staff morale and commitment continues to be high.
The role of Ynot was again discussed. Some members believe that the further development of other packages will overtake Ynot before Ynot can be made into a package of interest to synoptic meteorologists. They argued that there are no signs of interest in the package among synopticians, that its learning curve discourages use, and that the resources devoted to it would be better spent elsewhere. Other members believe that the potential of Ynot to non-traditional users, particularly outside of the synoptic community (e.g., satellite meteorologists), will result in Ynot filling an important niche in the future; dropping support for it now to "save resources" would be mortgaging the future, particularly since the amount of resources currently being devoted to the package is minimal. The broader community that would be interested in Ynot, however, will need applications; writing these will take resources. Some members thought that uses of Ynot might yet be developed and demonstrated at the 1994 Users Workshop. In any case, the Policy Committee has already agreed to a one-year test of Ynot and this is now in progress. The issue now facing the committee is deciding what criteria should be used in the evaluation of Ynot: number of users? research areas of users?

Action 2:
A decision whether or not to proceed with Ynot development will be on the February agenda.

Users Committee Report
Copies of Mohan Ramamurthy's transparencies were distributed at the meeting.

Discussion

The proposal to NSF for the 94 Users workshop had been received by NSF and Jacobs indicated that there were positive preliminary indications for funding.
Users Committee is already drafting user survey as the Policy Committee requested. They hope to have results by early spring.
There was strong interest in this summer's regional workshop. Over 20 people expressed interest in attending when the workshop was first proposed. The draft agenda and application form were mailed to the Unidata community the day before this Policy Committee meeting.
The WetNet products include Antarctica data, temperature and upper-level height fields, vegetation data, and total precipitable water.
A presentation on Unidata to the UCAR Academic Affiliates Program might make sense. In considering appointments to the Users Committee, Fox might want to balance synoptic vs. other kinds of meteorology. Current committee is heavily synoptic.

External Programs Status
Copies of Linda Miller's transparencies were distributed at the meeting.

Discussion

The ICAO (International Civil Aeronautics Organization) is now considering the fate of NOAA's demonstration network. The new profiler frequency has been set for 449 MHz; the demo network now in place uses 404.37 MHz. Retrofitting the old profilers would be very expensive, and no one knows when new profilers, designed to the new frequency, will be built. There is strong support for continuing the demonstration network. A decision will be made by September 30, and the outcome is uncertain.
UPC should determine what is needed from the NIDS vendors and elucidate this in a RFP to all the vendors.
Alden has offered to place FOS data on the Internet using the LDM. Fulker mentioned that he was considering accepting the offer only for the NPS in order to test the system. The committee argued that accepting such an offer would send the wrong signal to the other vendors; it would imply a special relationship between Alden and Unidata and this special relationship does not exist. Instead, the UPC should continue exploring the concept of the UPC becoming a commercial data provider. The point is to retain as much flexibility for the program as possible.
The lightning data currently being distributed on the Unidata/Wisconsin channel is being made available on an experimental basis and the distribution will end in October. Paramax is interested in testing distribution, too. As with the Alden offer, the committee is concerned about avoiding the appearance of a special relationship between the UPC and a data vendor.

NSF Report
Cliff Jacobs described the National Information Infrastructure initiative now before Congress and which seems to be gaining momentum. He noted that FY 93 has not been a good funding year for the agency. He said that Unidata has level funding unless he or someone else can identify other sources of funding within NSF. There was about $80 K for equipment grants and the review panel recommended six proposals. Copies of Jacobs's transparencies will be mailed to the members.

Discussion

NII is an authorization bill only; it has no money attached. Therefore Unidata should monitor its progress but not change any of its program emphases yet.
Ocean Sciences within NSF is very interested in the Unidata Program as a model for community focus and resource.
Commercializing NSFnet means that costs for use of the network will have to be incorporated into grants like any other utility. Effect on universities may be minimal since they already pay regionals. The federal investment in networks (i.e., the government's annual expenditures on networks) is only about 10% of their operating costs. Jacobs believes that the Unidata Program will be relatively unaffected by the change.
The Unidata Program is a model for distributed education. Supporting 100 universities now, encouraging them to support other institutions in their areas, allows society to leverage its resource investment enormously. The Unidata model should be publicized.

NOAA Report
Doug Sargeant reported that NOAA is now interested in providing data via the Internet and is watching the NMC test using the LDM. Email connectivity is now being promoted within the organization and should be established within the year. Meanwhile, NESDIS is creating data archive centers (DACs) and expects an active archive for satellite data to be in operation by the end of the year. This DAC will contain a range of data, including AVHRR; the data will be on a server accessible via the Internet. The DACs, however, will focus on data ordering, not on distribution, and the question of charges for the data have not yet been resolved.

Sargeant reported that NOAAport has been designed to be open ended toward real-time data, so that new data streams can be added when they become available. He said that the New NOAA administration was working on a strategic plan for the year 2003 that focused on the agencies efforts in areas such as Environmental Assessment and Prediction and Stewardship of Environmental Resources. This report is now available in draft form for comment; its entitled 1995-2005 Strategic Plan.

Internet Data Distribution

Copies of Ben Domenico's transparencies were distributed at the meeting.

Discussion

The topic of including a vendor such as Alden in the test of the IDD began again. The UPC is interested in including a vendor such as Alden because no other test site could control errors. This would be useful because it would answer some basic questions about the reliability of the system. The committee felt that a distribution system without error control would mean a more realistic test. If the reliability test is considered essential, however, then the UPC should consider an open invitation to all vendors. From the vendors perspective, whomever participates in the test, gets an Internet connection, network design, and contacts with regional networks.
IP multicasting (sending one packet to many users) is a technology that is evolving quickly. Ultimately, networks may manage distribution systems like IDD.
The IDD system would work best if NSF could fund the hardware at relay sites. The UPC estimate this would cost about $20,000 per site.
The process of hiring two term appointments to help with developing IDD is well underway.
During test phase, why not use NCAR as a data injection point. Jacobs pointed out that from NSF's point of view, NCAR and Unidata should be cooperating more as they seem to share many concerns and expertise and their cooperation would foster the goals of the atmospheric community.

Advanced Technology Advisory Committee

Ben Domenico presented a plan for establishing the ATAC using a core group and four working subgroups. (Copies of his transparencies were distributed at the meeting.) The committee commended Domenico on his plan and there was no further discussion.

Procedures on Participation by Institutions "with Potential Synergy"

The UPC has not established procedures for accepting or rejecting requests for participation by institutions that fall outside the Unidata Core (see the minutes of the December 1992 Policy Committee meeting). The Policy Committee agreed to assess on a case by case basis each request for participation from institutions having "potential synergy" with the program. However, no procedures for doing this have been established. The UPC currently has requests from a university in Brazil and from Lincoln Labs and Lawrence Livermore Labs.

Discussion

In allowing foreign institutions to participate, Unidata must be careful about honoring international data agreements. Unidata should have a list of nations that have data agreements with NOAA. Since Unidata currently only supplies data through Zephyr, data access isn't yet a problem. This will be a concern when IDD is implemented. The IDD will introduce further issues, such as whether Unidata wants to provide hemispheric data products. These data would be of interest to, e.g., Brazilian institutions but not to the majority of U.S. universities. At the same time, UCAR is interested in promoting international cooperation.
The case-by-case evaluation could be made by the UPC in email consultation with the Policy Committee; recommendations could then be made to UCAR for final decision.

Action 3:
Doug Sargeant will check with the NWS concerning restrictions on the distribution of FOS data outside the U.S.

Resolution 1:
The Policy Committee recommends that applications for participation in the Unidata Program from non-core institutions be evaluated by the Unidata Program Center and UCAR and that the applications and the UPC's recommendations be circulated to the Policy Committee for comments.

The Future of DIFAX in the Unidata Program

In May, there was considerable discussion on the community mail list about DIFAX. Copies of the email messages are in the notebook. Dave Fulker said that he doesn't believe this technology has any long-range place in Unidata and recommended staying out of the issue.

Discussion

Alden has software that creates bitmaps of DIFAX products. There was a suggestion that Unidata consider making these available to the community. Furthermore, while the current applications can generate most of the products, it requires knowing your software very well, so creating the products may be hard for small institutions. Others suggested that the UPC remain uninvolved since the products will continue to be available through Alden (users subscribing to DIFAX must buy into the entire system, however, so this solution is not cheap).
In the age of AWIPS, digital data will be sent over the net and display products will be created locally.

Resolution 2:
Given other Unidata's high-priority activities, the Policy Committee advises Unidata not to distribute or support DIFAX. DIFAX is readily available commercially, and many DIFAX products are available elsewhere for those that need them.

Action 4:
Dave Fulker will respond to email on DIFAX and explain the Policy Committee's resolution on DIFAX to the community. He will also discuss this with Alden.

COMET Case Studies

A selection of COMET and UPC staff members met to discuss the requirements for making COMET case studies accessible to the Unidata tool set. Dave Fulker's report on the meeting and his recommendations are in Section B on the notebook.

Discussion

COMET plans to expend considerable effort on creating case studies. They may number into the 80s or 90s. A small investment by Unidata now may therefore yield large returns in the future.
There were questions about vector graphics: none of the current applications incorporate graphic primitives and standards haven't been set for these. This may or may not be a problem. Unidata may have to deal with vector graphics even without the case studies; any delay in adopting suitable standards will simply delay Unidata's distribution of case studies.
All the satellite images are missing navigational information because the current case studies are designed for the FSL workstation; however, COMET is evolving away from this platform, so this may not be a limitation in the future. There was also a question whether anyone wanted geo-referencing in a case-study environment.
The Users Committee is now involved in evaluating current case studies. This effort will continue to determine whether any would be of interest to the universities now.

Resolution 3:
The Policy Committee supports the UPC's stated plan to cooperate with COMET in developing data formats that might allow Unidata to use COMET case studies in the future.

Budgeting Unidata under Various NSF Funding Scenarios

Sandra Nilsson presented the current status of Unidata's financial resources. Copies of her transparencies are in the notebook and there were no questions from the committee. Nilsson and Dave Fulker then presented scenarios for various funding levels:

As submitted, the five-year proposal called for:

Support community use
Maintain and upgrade systems
Respond to needs for new capabilities
Foster contributions and involvement
Building consensus, advocate for the community 4 major new initiatives
1. Distribute data via NREN
2. Access new data sources
3. Enhance analysis capability]
4. Integrate analysis/synthesis tools]

To be achieved via 7 classes of UPC activity requiring 23 staff members:

4 senior/administrative
14 programming/technical specialists
2 secretarial/clerical
2 part-time students
1 visitor

The funding requested was:
$ 3.1 for FY 1994

UPC assesment of the programmatic changes necessary to cope with NSF alternate funding scenarios:

Scenario A: $ 2.3 M for FY 1995
Fulker noted that there is little flexibility in the ongoing endeavors. Some resources could be saved by cutting back on endeavors 1, 3, and 4. No significant savings could be identified by reasonable reductions in the other two endeavor areas. The greatest reduction in costs would therefore fall on the new initiatives: allocations to initiatives 1 and 2 would be reduced and initiatives 3 and 4 would be dropped altogether.

There would be no resources expended for:

Adapting Unidata standards (e.g., coping with the vector graphics problem)
Standardizing geo-referencing
Satellite data
Advanced visualization
MDA collaboration.

Scenario B: $ 2.0 M for FY 1995
In addition to the reductions detailed for Scenario A:

Reducing the UPC force
Reducing the number of supported platforms to one.

In-Between Scenario.
Funding Unidata's current program will cost $2.1M. The addition of $200,000 will allow the two term appointments now being hired to continue. Without these term appointments, the IDD schedule (deployment by the third quarter of 1994) cannot be met.

Fulker has been asked to submit a revised budget for Scenarios A and B for all five years, and in the next few weeks he will be submitting a formal statement describing the impact of funding reductions.

Discussion

Equipment for relay sites? This is not included in any current budget projections; in both scenarios, funds for these would have to be added.
Scenarios entail changes to basic program? Yes, scenario B would eliminate support for multiple platforms. The program is already bare-bones: the netCDF operators have already been put on hold, and there is currently little effort expended on software integration.
How were priorities developed? Through interactions with the users: personal, in committees, and during the proposal-review process.
Dropping support for multiple platforms would impose a severe hardship on many McIDAS-OS/2 sites and would not be well-received by the community. In the short term, dropping support does not mean denying access to the Unidata/Wisconsin channel. In the long term, however, SSEC is planning architectural changes and is currently maintaining one-version compatibility with OS/2. Without Unidata support, this effort would be dropped and the changes would not be folded into the OS/2 software at all. Ultimately, this would render the data stream inaccessible to OS/2 users.
How would dropping support for OS/2 affect support efforts? Tom Yoksas is currently easing sites' transition to UNIX and has historically created special menus, documented routines, and fixed bugs faster than SSEC can do. Without support, would these activities continue? Who would answer users questions? Advise given on the mailing list pales in comparison to Yoksas's activities.
The committee discussed potential sources for additional funds. Fulker pointed out that NOAA is already contributing to the program through the FSL monies for IDD and through the NIDS money for data distribution to universities. Other sources discussed: OAR (may be interested in expanding access to data), NASA--Global Change (Jay Fine or Torben Nielsen might help). However, the Policy Committee has consistently resisted chasing funds; sentiment has been that the program should be focused and should not expend energy on proposal-writing.
While the Unidata Program was initiated and fostered by ATM, perhaps Unidata's mission today belongs more properly in another area dealing with NII.
Maybe Unidata should broaden its community of users beyond synoptic meteorology.

List of Resolutions and Status of Action Items

Action 1:
Review location of February meeting in October.

Done.

Action 2:
A decision whether or not to proceed with Ynot development will be on the February agenda.

Not applicable at this meeting

Action 3:
Doug Sargeant will check with the NWS concerning restrictions on the distribution of FOS data outside the U.S.

Report Forthcoming

Action 4:
Dave Fulker will respond to email on DIFAX and explain the Policy Committee's resolution on DIFAX to the community. He will also discuss this with Alden.

In progress

Index

Unidata Homepage

This page was Webified by Jennifer Philion.
Questions or comments can be sent to <support@unidata.ucar.edu>.

This page was updated on .