Policy Committee Meeting Summary: September 25-26, 1995

Washington, D.C.

Participants

Members	Representatives	UPC Staff
Robert Fox (Chair)	Harry Edmon (U. Wash./ATAC)	Sally Bates
Otis Brown	Alan Gaines (NSF/EAR)	Ben Domenico
Steve Mullen	Dick Greenfield (NSF/ATM)	Jo Hansen
Carlyle Wash	David Fulker (UPC)	Linda Miller
Julie Winkler	Fritz Hasler (NASA)	Sandra Nilsson
	Clifford Jacobs (NSF/ATM)
	Mohan Ramamurthy (U. Ill./Users)
	Doug Sargeant (NOAA)
	Tim Spangler (COMET)

Administrative Matters

The summary of the May 1995 meeting accepted as written.
The next meetings of the Policy Committee will be:
22-23 February, Thursday and Friday, in Boulder, CO
30-31 May, Thursday and Friday, in Boulder, CO
Fox's term as Policy Committee Chair expires with this meeting; the three-year chairmanship will be assumed by Otis Brown at the February meeting; Fox will fill the remaining two years of Brown's current term. Other membership changes: Russ DeSouza has resigned and Carlyle Wash's term expires; new members will be Colleen Leary (Texas Tech) and John Merrill (Rhode Island). The UCAR Board of Trustees has appointed Donald Johnson (U. Wisc.-Madison) as Trustee Liaison.

Status Reports

Director's Report
Copies of Fulker's report and of his transparencies were distributed at the meeting.

Discussion

Fulker's categorization of Unidata's budget as "green" was questioned. Funding is sufficient for the moment, but the future is problematical. Jacobs indicated ATM funds would be limited; there seemed to be general agreement that Unidata needs to reach out to other geosciences.
Vendors consider Unidata to be in direct competition with them and complain to congressional representatives about it. The Web is adding to the confusion--vendors are slow to realize that they need to add value to the data and universities need to reiterate on their Web pages that their data are NOT operational. Perhaps the UPC needs to write a general article for BAMS explaining Unidata's role in data access; urging universities to include disclaimers on their pages is also a good idea.
COMET is about to begin distributing case studies, displayable by GEMPAK.
There was some discussion about what other data Unidata might distribute. Mentioned were seismic, volcanic, ash tracking, and SAR data.
There was some discussion about what Unidata's role might be within a broader community: Coordination functions can now be distributed because of current communications technologies. Unidata might attempt coordinating work of applications developers--what are people working on? what is being omitted? Unidata's recent decision to work on decoders might be considered an example of a coordinating effort. Also need to identify what's being developed elsewhere that should be brought into the program. (e.g., search engine; data-server).

Action 1:
Bates will include a short reminder concerning Unidata data-usage policy in every issue of the newsletter.

Action 2:
The UPC will encourage Unidata universities to include a disclaimer asserting that all weather data on their Web pages are for non-operational use only.

Action 3:
The UPC will pursue the idea of an article in BAMS re Unidata's role in providing universities with data access.

Budget Report
Copies of Nilsson's transparencies were distributed at the meeting.

Discussion

The potential (impending) budget shortfall raises a strategic issue: should the UPC attempt to keep its activities constant and cut everything else (travel, equipment purchases, etc.) or try to find money for specific activities? The current spending scenario represents a holding strategy that cannot be applied for more than one year.
There was considerable concern expressed about the lack of funds for equipment purchases in FY96. The need for equipment is predicated on the development of new technologies that the UPC will need. These cannot be predicted, but they WILL be needed.
The reduction in subcontract costs is not necessary a good thing. Historically, subcontracts are the mechanism by which new capabilities are brought into the program. Some subcontracts are the result of requests from the Users Committee. The loss of subcontract funds implies the loss of UPC flexibility.
The proposal to NSF for the continuation of the Unidata program is due at NSF in May 1997.

Action 4:
Once a year, the UPC will develop a rolling five-year equipment budget.

Users Committee Report
Mohan Ramamurthy reported that the Users Committee had not met since the May Policy Committee meeting. Despite this, however, the committee's three new members (Jennie Moody, Greg Cox, and Doug Yarger) have already become active. The issue of OS/2 on campuses [Action 1 from the Policy Committee's May meeting] is on the agenda for the Users next meeting in October.

NASA Report
Fritz Hasler showed a video of 1 minute, GOES-9 satellite data of Hurricane Luis. He also presented stereographic images from GOES-8 and 9. He reported that:

NASA's HPCC program is slated for reduction and the Public Use of Remote Sensing Data program might disappear.
NASA's GLOBE program will be funded for another year.

Discussion

The future of NASA's funding is still unpredictable since there's great disparity between the senate and house bills.
There is talk of recompeting the DAACs; they may become less centralized.

NOAA Report
Doug Sargeant reported that:

Congress is considering dismantling the Department of Commerce. While the situation is chaotic, it appears that NOAA is surviving, but no one knows where the agency (or pieces thereof) will be located.
The AWIPS Pathfinder program has been fielded; it is composed of a PRC broadcast of data via satellite from Silver Spring to receivers at a couple of weather offices. The software hasn't been integrated yet and there are no policies proscribing data use.
The issue of whether to move NOAAport from Ku to C-band should be settled in December.

Discussion

Sargeant did not know whether there are any universities receiving Pathfinder data.
Moving GOES-9 to the west should occur sometime toward the end of the year; the date will be determined by the experiments currently being run using the satellite over the central US.

NSF Report
Cliff Jacobs reiterated the implications of various congressional deficit-reduction strategies. So far, the NSF has not been hit too hard: the allocation for Research and Related Activities is below that requested, but all the others are at the requested amount. In other areas:

Concerning the Unidata equipment grants, NSF funded 14 proposals out of 18 requests. Implementing IDD was a high priority for universities (about 10 of the proposals were specifically for IDD-related equipment.). The amount awarded by NSF was $173,962; when combined with cost sharing by universities, this represents a considerable investment in hardware, mostly for IDD.
Jacobs reported on ATM/UCAR planning. Funding for NCAR in FY96 appears adequate, but FY97 is uncertain. ATM and UCAR are examining strategies for coping. Long-range planning priorities include preserving the base science program for NCAR; there is concern about maintaining NCAR facility infrastructure over the longer term (maintaining all three aircraft, for example, may not be possible), but they still hope to find funds to replace the Electra.
The Division of Atmospheric Sciences within NSF has been reorganized: UCAR and Lower Atmospheric Facilities Oversight Section is now headed by Jacobs, the Lower Atmosphere Research Section is headed by Jarvis Moyers, and the Upper Atmosphere Research Section is headed by Richard Behnke.
There is a possibility of a government-wide shutdown as Congress continues to wrestle with the budgets.
The final report of the task force on the future of the NSF supercomputing centers is scheduled to be submitted to the National Science Board in October. It will probably recommend that NSF maintain but restructure the centers; that the centers be "leading-edge" sites partnered with other sites such as experimental research centers at universities, NSF research centers, and other centers; that the centers find some funding from their partners; that competition for designation as a center be held every five years, that center research be peer-reviewed; that the NSF directorate be more involved; and that NSF should take the lead in higher-performance computing. Jacobs noted that everyone expects there to be a reduction in the number of centers and that the task force obtain input from a wide variety of people.

Discussion

Jacobs noted that the task force may also make some recommendations on simplifying the process of allocating computing resources, but that the process can't be simplified much since each request is weighed by different criteria. He also noted that while networked workstations may work for some disciplines, others would always need supercomputers. Because university research centers are facing financial difficulties, idea that the centers might receive funding from university research centers was met with skepticism.
Concerning future funding for Unidata, Jacobs noted that facilities funding tended to be emphasized over science and since the UPC is a facility and provides a good model for resource use, its funding would probably be stable. Funding for the UPC from within ATM, however, would probably not increase, so Jacobs concurred that Unidata may have problems in FY 97.

IDD Status
A copy of Ben Domenico's and Linda Miller's IDD status reports are in the notebook; they are also available on the Unidata Web server. In addition, Domenico gave a demonstration of new uses of the Web and noted the following:

Vendors and people within the university community both are developing systems that enable users to administer and use UNIX systems, the Unidata LDM, and applications (WXP at the moment) via the World Wide Web on the platform of their choice.
Security concerns may pose a problem. Internet technologies don't work well in firewalled environments, but so far the UPC has been able to overcome problems caused by firewalls.

Discussion

The advent of Java will probably alter Unidata's approach to applications support. U. of Illinois, for example, now has a Web interface to WXP which will include WXP Java applets in the future. Netscape already provides Java support on UNIX with support for Java on Mac due in mid-October. Java should help relieve the pressure on servers since it transfers some processing to the clients. In effect, the Web is capable of providing a unified GUI to Unidata software--a goal long desired by Unidata users.

ATAC Status
Copies of Ben Domenico's presentation are in the notebook; the ATAC summary is available in the notebook and on the Unidata Web server.

Discussion

The UPC doesn't know how much data-recovery activity Unidata generates. UPC might want to request that people log data requests and responses with support@unidata.ucar.edu.
Unidata excels at developing the community "glue" and at once-only development (such as the proposed decoders).
ATAC Resolution 9501:
The ATAC recommends that the Unidata Policy Committee request that the National Weather Service provide users with the following information on the Family of Services data stream:
on-line station information (in ASCII format) which includes
the same range of information that is provided by the U.S. Air Force
additional information on county and zone codes
information on controlling WFO numbers
on-line sample database of data formats
WMO headers with descriptions
It was agreed that Linda Miller will work with the UPC staff and COMET (Peggy Bruehl) to provide a comprehensive list which, after review by the ATAC, will be submitted to the Unidata Policy Committee.
Policy Committee response: The topic of ATAC resolution 9501 should be considered a UPC action item. Other discussion: The NWS is becoming more proactive in supplying information on-line, but they will not be interested in identifying older headers since they are extremely busy due to their modernization efforts and their efforts to cope with new instrumentation.

Action 5:
Bob Fox will write a letter to the National Weather Service asking that more information be made available on the NWS Web server.

Action 6:
Bob Fox will look into making the SSEC's copy of the master station list available to the Unidata community.

ATAC Resolution 9502:
The ATAC recommends that the UPC undertake the creation of decoders that write netCDF files, as resources permit, and that responsibility for applications reading decoded netCDF files is a lower priority.

Policy Committee discussion:

Reading netCDF files is the developer's problem; the resolution focuses on getting netCDF decoders for surface and upper-air data (they exist already for gridded data). The ATAC did not consider satellite data.
The problem with the UPC modifying applications to READ netCDF files is that it would be on-going: each new version of the application would require remodification.
While COMET is committed to working on adding netCDF support to GEMPAK, Spangler did not know how committed NMC is to the same task.
SSEC is building an ADDE-server that employs data calls that contain data format information; they are including support for netCDF.
N-AWIPS is considered an interim software, to be succeeded by AWIPS, which is being developed jointly by PRC and the government. It is not clear at this point what GEMPAK's role will be within the NWS in the future.
For the near term, the UPC is planning to commit one-half to one-third FTE to decoder development; the UPC does not know what resources will be required to maintain the decoders over the long term.
The FOS stream will continue "as long are there are users," but the costs must be borne by the users. This is a self-destructing path that will affect the UPC: how will Unidata acquire the data universities want? At what point does it become economical for the UPC to install a NOAAport receiver?
How can Unidata achieve end-to-end netCDF compliance?

Resolution 1:
The Policy Committee reaffirms its commitment to a fully netCDF-compliant suite of applications. In light of ATAC resolution 9502, the Policy Committee advises the UPC to develop decoders for conventional data that produce netCDF files.

Resolution 2:
The Policy Committee, after considering the current state of netCDF development, recommends that netCDF operator implementation be terminated and that program effort be applied to enhancing the netCDF data model, library, language interfaces, and application interfaces.

ATAC Resolution 9504:
Unidata's future development strategy for needed new applications--to be built via partnerships between UPC and community experts--should emphasize UPC responsibility for building blocks that employ contemporary methodologies, such as class or template libraries, and that operate on multiple platforms.

ATAC Action 9508:
UPC should actively investigate RMP; several sites will help in this effort; UPC will also check with SCD to ascertain their interest in this. ATAC members interested in being a test site should send e-mail to Brian.

Policy Committee discussion:

The Policy Committee reviewed ATAC resolution 4 and Action 8 and there was no disagreement with the sentiments expressed in both.
In investigating RMP, the UPC might want to consider seeking vBNS money to support the effort.

Platform Support

A copy of the proposed platform support policy is in the notebook.

Discussion

Is the UPC really supporting what the universities want? This is doubtful, since most universities have many Windows and Mac and some NT machines. Almost all of the operating systems in the supported platforms table are simply variants of UNIX. Problem is that Unidata is constrained by what the applications developers support.

Resolution 4:
The Policy Committee adopts the policy on platform support drafted by the UPC on 15 September 1995.

Unidata's Name

There are some other institutions with the name Unidata, including one in Denver that has trademarked the name. The potential for confusion is increasing. Should Unidata change its name?

Discussion

Simply adding UCAR to the name is a possibility, but universities feel ownership of Unidata and might be offended by the change.

Action 9:
The question of Unidata's name should be on the agenda for the next Policy Committee meeting.

Strategic Planning

Over the next year, the Policy Committee needs to develop a long-range (five-year) plan as a framework for Unidata's next proposal to NSF (due to Jacobs's office in May 1997). The last hour of the meeting was devoted to "brainstorming" how this might be accomplished, how strategic planning might be facilitated, what the plan needs to address. Some of the ideas expressed included:

There is a need for community input, where community may be more broadly defined than currently. It was again suggested that the UPC hold forums at both the AMS and AGU.
The issues that need addressing are data diversity, program limitations, and evolving technologies.
The Policy Committee may want to engage in a focused strategic-planning retreat.
Policy Committee meetings need to be more focused, with less time spent on status reports so that more time can be devoted to policy issues.
The Policy Committee needs to address the breadth of the Unidata Program; the Users Committee needs to identify the unsatisfied needs within the user community, and the ATAC needs to address the potential future of technological developments.
The availability of new observations will transform the program: is the Unidata system scaleable by two orders of magnitude?
Unidata needs to keep track of events occurring within government agencies.
The UPC has historically appeared concerned about supporting field studies. The issue seems to be how large-scale processes would be supported. And a data infrastructure would be required--the community would want access to results.
The scope of the Unidata program is defined by its sponsors.
We're seeing the beginning of an explosion of environmental data; Unidata will not be able to cope with them all, but Unidata is in a unique position, having already developed some of the requisite infrastructure. The program now needs to choose between being the locomotive or simply another car on the train: the program can sit down with other disciplines and work toward providing data to a wider range of disciplines or focus on becoming more efficient at providing selected data.
One problem with widening the Unidata community is the question of software support: other disciplines use other software; what should Unidata support? Are data formats a problem? Unidata might want to set up partnerships with other communities and focus on whatever technologies are shared.

Action 10:
Henceforth, status reports will be limited to the morning of the first day as follows: all UPC reports, 1-1/2 hours; committee reports, 15 minutes; agency reports (NOAA, NASA), 15 minutes each; NSF report, 1/2 hour. Policy Committee members will read the status reports in the notebooks and be prepared to discuss particular items or raise specific questions as needed. Topics needing policy decisions will become separate agenda topics.

Action 11:
The ATAC should begin considering what technology will require of the Unidata program in the year 2003 .

Action 12:
The Users Committee should seek ideas from the community concerning where the program should be in the year 2003. How this is to be accomplished will be on the agenda for the next Users meeting (October).

Action 13:
Bob Fox and Otis Brown will explore the question of where Unidata's boundaries should be.

Action 14:
Bob Fox and Otis Brown will work with the UPC on creating a forum for the AGU.

Action 15:
The UPC will contact the new Policy Committee members about the dates for the next Policy Committee meetings.

List of Resolutions and Action Items

Resolution 3:
The Policy Committee adopts in principle the policy on sponsored participation drafted by the UPC on 15 September 1995.

Resolution 4:
The Policy Committee adopts the policy on platform support drafted by the UPC on 15 September 1995.

Action 1:
Bates will include a short reminder concerning Unidata data-usage policy in every issue of the newsletter.

Will begin with next issue.

Action 2:
The UPC will encourage Unidata universities to include a disclaimer asserting that all weather data on their Web pages are for non-operational use only.

Done.

Action 3:
The UPC will pursue the idea of an article in BAMS re Unidata's role in providing universities with data access.

In progress.

Action 4:
The UPC will develop a rolling five-year equipment budget every month.

In progress.

Action 5:
Bob Fox will write a letter to the National Weather Service asking that more information be made available on the NWS Web server.

Became unnecessary as the NWS, without prompting, began providing material via the Web.

Action 6:
Bob Fox will look into making the SSEC's copy of the master station list available to the Unidata community.

Done.

Action 7:
The UPC will redraft the policy to incorporate words on international data exchange issues and will circulate the revised draft to the Policy Committee.

Action 8:
The UPC can begin acting on the sponsored participation policy.

Done.

Action 9:
The question of Unidata's name should be on the agenda for the next Policy Committee meeting.

No time at this meeting; will carry over to next meeting.

Done.

Action 11:
The ATAC should begin considering what technology will require of the Unidata program in the year 2003.

In progress.

Done.

Action 13:
Bob Fox and Otis Brown will explore the question of where Unidata's boundaries should be.

In progress.

Action 14:
Bob Fox and Otis Brown will work with the UPC on creating a forum for the AGU.

Done.

Action 15:
The UPC will contact the new Policy Committee members about the dates for the next Policy Committee meetings.

Done.

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 .

Policy Committee Meeting Summary: September 25-26, 1995

Participants

Administrative Matters

Status Reports

Sponsored Participation

Platform Support

Unidata's Name

Strategic Planning

List of Resolutions and Action Items