Policy Committee Meeting Summary: March 3-4, 1994
Boulder, Colorado
Participants
| Members | Representatives | UPC Staff |
| Robert Fox (Chair) | William Bonner (UCAR) | Sally Bates |
| Otis Brown | Bill Buzbee (NCAR/SCD) | Ben Domenico |
| Russell DeSouza | David Fulker (UPC) | Linda Henderson |
| John Nielsen-Gammon | Robert Gall (NCAR/MMM) | Linda Miller |
| Perry Samson | Clifford Jacobs (NSF) | Sandra Nilsson |
| Carlyle Wash | Mohan Ramamurthy (Users Committee) |
| | John Snow (UCAR Board of Trustees) |
| | Timothy Spangler (COMET) |
Administrative Matters:
- Minutes amended to correct spelling errors.
- Next meetings of the Policy Committee will be:
28-29 June in Boulder
6-7 October in Washington, D.C.
- Agendas for future meetings should contain only topics that comprise
essential informationm particularly areas requiring policy decisions.
Status Reports
Director's Report
Copies of Dave Fulker's Director's Report and slides were distributed at the
meeting.
Discussion
- The Policy Committee questioned whether the loss of a systems
administrator (Karl) left the UPC short-handed; Fulker and Domenico stated
it was not yet a problem.
- There was considerable discussion on the role of OS/2 and PCs in the
program. The issue is one of cost, not of platforms, but in terms of
systems administration. Unidata needs to continue support for small schools
with limited resources, but the implication of OS/2 support is that the UPC
must ensure that OS/2 sites be able to receive IDD data.
- There was considerable discussion of the IDD and Unidata/Wisconsin data
streams. U/W stream is mostly GOES images with a small percentage being
conventional data. SSEC now has UNIX software for building an on-site
database of OS/2 FOS data.
- It is not yet clear how information servers will affect the program. The
needs of some smaller sites may be served by Web servers; even at powerful
research universities, many students prefer obtaining information via the
Web to generate their own displays. UPC must be in the position to help
universities use information technologies. The evolution of these
technologies may, at some point, remove the need for Unidata to support
OS/2.
- The increased availability of and interest in products (as opposed to
raw data) is indicative of the growth in a new service industry. However,
Unidata has always emphasized data rather than products.
Action 1:
The Policy Committee recommends that the UPC:
- Investigate at what point will there be replication of data in the IDD
and Unidata/Wisconsin data streams;
- Define what the alternatives are for supporting the OS/2 stream as
separate from the IDD
- Estimate the resources required to build an on-site OS/2 database of the
Unidata/Wisconsin stream.
Action 2:
The Policy Committee recommends that a User Committee subcommittee look into
the problem of proliferating servers and the need for mirror sites for the
most popular servers.
Budget Report
Copies of Sandra Nilsson's transparencies are in the notebookwere distributed.
Discussion
There was no discussion.
Users Committee
Mohan Ramamurthy summarized the February Users Committee meeting (a copy of
the summary appears in the notebook), including Users Committee reactions to
the results of last fall's community survey. He brought the committee up do
date on planning for the June workshop, noting that there were 33
applications to attend, not including committee members and UPC staff.
Discussion
- The committee members perused the results of the survey and found no
surprises.
- There were questions about the COMET case studies; Ramamurthy reported
that the subcommittee found them to have too many format problems and too
limited data to be of use as currently publishes.
- There was considerable discussion on the workshop. The lack of
attendance is attributed to the many conflicting opportunities in mesoscale
meteorology. Both COMET and NSF indicated they were still planning to
support the workshop.
- There was concern about the lack of K-12 involvment. It was suggested
that a regional workshop might be a better arena for K-12 which could
emphasize network publishing technologies, for example.
- There was discussion about the discontinuation of the Wind Profiler
Demonstration Network.
Resolution 1:
The Policy Committee recommends that the UPC plan a regional workshop on
K-12 outreach by the atmospheric science community.
Action 3:
Bob Fox and Bill Bonner will develop a letter of concern addressed to the
appropriate individuals uring the continuation of the Wind Profiler
Demonstration Network.
External Programs Status
Linda Miller announced that she has accepted a part-time position with
Project Globe, remaining part-time in Unidata as well. Her Unidata time will
be devoted to the Users Committee, data access issues, and associate
participation.
Discussion
- There was discussion of the status of international associates in the
current international climate of nationalization and commercialization of
weather data. Since associate members are not entitled to data access,
international associates are possible. There was concern, however, that
demand by international organizations could swamp Unidata resources.
Action 4:
John Snow will alert the UCAR Board of Trustees to the Policy Committee's
concerns about data exchange and data access in the current international
environment.
NOAA Report
Fox spoke with Doug Sargeant and filed the following report:
- On international data access: the rush to commercialize weather data
appears to be slowing. This topic is on the European Community agenda in
June. The U.S. positions remains strongly in favor of the free exchange of
data.
- NOAAport Schedule:
Jan. 95: First test of the broadcast
July 95: First field site equipped
March 96: First 8-site deployment
- NEXRAD: deployed at Norman, OK on 1 March, first commission has 46
radars being delivered to the National Weather Service.
Discussion
- COMET will install a NOAAport receiver about six months after the first
deployment.
- The UPC want on site on the Internet ingesting NOAAport data for the IDD
system. UPC also intends to provide ingest software that alllows users who
purchase NOAAport receivers to use the data in their Unidata systems.
- Need to find out what NOAAport includes: FOS? only regional broadcasts?
Action 5:
Schedule plenty of time for a NOAAport discussion at the next Washington
meeting of the commitee (October).
National Science Foundation Report
Cliff Jacobs presented a dramatic summary of the budget situation at NSF,
including previews so-called Branscomb Report (due to be released this
spring) that paints a vision of the Foundation for the coming years.
High-Performance Computing is essential to this vision. Jacobs also
summarized how National Information Infrastructure (NII) developments are
affecting the High Performance Computing and Communications programs,
specifically the component called Information Infrastructure Technology and
Applications (IITA). He noted that Unidata's mission falls under this
rubric. Copies of Jacob's transparencies were distributed at the meeting.
Discussion
- The NSF vision appears to be a call for more interdisciplinary research
and education. There were concerns expressed about what policy issues this
poses. There also may be policy issues in seeking funding outside of NSF
ATM, such as IITA, although in Jacobs' view IITA and ATM interests were not
in conflict.
Action 6:
Schedule a discussion of Unidata's role in the context of NSF's overall
istitutional vision and goals as opposed to Unidata's role as an NSF/ATM
program.
IDD Status
Hard copies of Ben's slides were distributed at the meeting.
Discussion
- There were several basic questions about the IDD that remain to be
answered: scalability (UPC doesn't know how many sites have sufficient
bandwidth); deployment (the UPC is having difficulty finding relay sites;
has yet to identify a final topology); reliability (what's really
acceptable?), data (what data should the UPC be responsible for)?
- Community expectations may be unrealistic. There was concern that the
search for relay sites would bring in a flood of applicants by sites
simply wanting IDD access immediately. Committee felt that an
implementation plan should be created and publicized within the community
indicating that full deployment was still in the future.
- There is interest in IDD beyond the atmospheric sciences
community. Unidata is building a prototype IDD; the prototype will
prove/disprove the concept on a limited scale. If the IDD wants to expland
beyond atmospheric sciences, the IDD can then be passed off to others to
expand.
- May want to bring network providers (e.g., MCI) into the IDD deployment
since they can use the IDD to sell products.
- NSF/CISE will be interested in end-to-end network performance figures;
they don't gather statistics this way. For example, satellite broadcast and
IDD statistics aren't comparable since error-free seconds and packets lost
are different. UPC may need to look at products instead of packets. Toward
this end, IDD developers have discussed adding unique headers to products;
may accomplish this in LDM5.
Action 7:
By the next Policy Committee Meeting, the UPC will address the following:
- Define the scope of the IDD undertaking:
- data sets involved
- community served
- Define the philosophy of operation. For each layers:
- how will the IDD function in normal mode?
- how will it continue to operate under various degraded
situations?
- Define the requirements for each level of node:
- people
- power
- management
- network
- Develop an Implementation Plan
- Develop a Node Selection Plan (i.e., define the deployment criteria
- Gather statistics on server congestion problems in the underlying network.
Project Globe
The Policy Committee heard a briefing by Sushel Unninayar on the Globe Program,
a Vice Presidential initiative that is a multi-agency planning effort involving
NOAA, NASA, EPA, NSF, OEP, OSTP, DoEd, et al. The program will be announced
on Earth Day, 22 April, by V.P. Gore. The purpose of the project is to engage
the world community in environmental observations and analyses. More specifically,
Globe aims to involve some of the 2 billion students world wide in making scientifically
valid observations and using the observations in their curricula to learn about
their environment. FY 94-96 will be focused on a coordinated planning and rapid
prototyping effort. They hope to have 1,000 schools involved world-wide by the
end of FY 96. UCAR will take part in the initial planning during 1994 and create
a plan for FY95-96, which will be integrated into the Washington, D.C. planning
effort.
Discussion
- Software,networking, and observing are only part of the problem. Education
considerations (i.e., meeting national/state curriculum guideslines), providing
technical support to education administrators, and training teachers are
harder to do. These are not issues of money, but of knowledge and attitude.
They could hinder the implementation of a project like Globe.
- The committee questioned Unidata's role in the program. The UPC has agreed
only to help in setting up interactive computing systems and to help set
the networking standards. The role is advisory only, although IDD may become
the underlying technology.
- Naval forecasters have adopt-a-school programs; Globe may want to connect
with them.
- High-school science programs are difficult to change; students lack enthusiasm.
Platform Support Policy
Action 8:
The UPC will draft a policy on platform support and circulate it via email to
the Policy Committee.
University Proposals with UPC-Components
Action 9:
The topic of University Proposals with UPC-Components will be on the agenda
for the next Policy Committee meeting.
NIDS and FOS Procurement
Dave Fulker reported that Unidata selected two vendors: Alden for FOS data and
WSI for NIDS data. The UPC is pleased with the RFP results.
Discussion
- Community will be anxious to begin accessing the NIDs data. Need to inform
them what NIDS data are, what the costs will be, that there is no display
software, and when the data will be available.
- Need to claify usage contraints. UPC's role is to outline the constraints;
final decisions, however, are to be negotiated between the university and
the vendor.
Resolution 2:
The Policy Committee recommends that the UPC publish a "special edition" of
the Unidata Newsletter as soon as practical covering those aspects of the
FOS and NIDS contracts that the community needs to know and that serves as
a reference for acquiring and using FOS and NIDS data and products within
the spirit and letter of the contracts. The impending newsletter should advertise
the upcoming special edition if possible.
Action 10:
The Winter Newsletter should include a note that NIDS and FOS vendors have
been chosen.
MOSAIC Demo
Mohan Ramamurthy demonstrated the new World Wide Web server developed by the
University of Illionois.
Lightning Data
Dave Fulker reported that SUNY-Albany is willing to make lightning data available
via the Internet. Site desiring these data need to make individual arrangements
with SUNY-Albany. This is not a "supported" data stream.
Discussion
- Need to differentiation between the IDD and the Internet being used by
someone with an LDM.
- Since bandwidth may be limited, UPC may need to define criteria for deciding
which data stream to "support."
List of Resolution and Action Items
Resolution 1:
The Policy Committee recommends that the UPC plan a regional workshop on K-12
outreach by the atmospheric science community.
Resolution 2:
The Policy Committee recommends that the UPC publish a "special edition" of
the Unidata Newsletter as soon as practical covering those aspects of the
FOS and NIDS contracts that the community needs to know and that serves as
a reference for acquiring and using FOS and NIDS data and products within
the spirit and letter of the contracts. The impending newsletter should advertise
the upcoming special edition if possible.
Action 1:
The Policy Committee recommends that the UPC:
- Investigate at what point will there be replication of data in the IDD
and Unidata/Wisconsin data streams;
- Define what the alternatives are for supporting the OS/2 stream as separate
from the IDD
- Estimate the resources required to build an on-site OS/2 database of the
Unidata/Wisconsin stream.
In progress.
Action 2:
The Policy Committee recommends that a User Committee subcommittee look into
the problem of proliferating servers and the need for mirror sites for the
most popular servers.
Postponed until after the Users Workshop
Action 3:
Bob Fox and Bill Bonner will develop a letter of concern addressed to the
appropriate individuals uring the continuation of the Wind Profiler Demonstration
Network.
Done.
Action 4:
John Snow will alert the UCAR Board of Trustees to the Policy Committee's
concerns about data exchange and data access in the current international
environment.
Done.
Action 5:
Schedule plenty of time for a NOAAport discussion at the next Washington meeting
of the commitee (October).
Not applicable to this meeting
Action 6:
Schedule a discussion of Unidata's role in the context of NSF's overall istitutional
vision and goals as opposed to Unidata's role as an NSF/ATM program.
On current agenda
Action 7:
By the next Policy Committee Meeting, the UPC will address the following:
- Define the scope of the IDD undertaking:
- data sets involved
- community served
- Define the philosophy of operation. For each layers:
- how will the IDD function in normal mode?
- how will it continue to operate under various degraded situations?
- Define the requirements for each level of node:
- people
- power
- management
- network
- Develop an Implementation Plan
- Develop a Node Selection Plan (i.e., define the deployment criteria
- Gather statistics on server congestion problems in the underlying network.
See IDD materials in Status Report and in Section A tabs of
this notebook.
Action 8:
The UPC will draft a policy on platform support and circulate it via email
to the Policy Committee.
Postponed.
Action 9:
The topic of University Proposals with UPC-Components will be on the agenda
for the next Policy Committee meeting.
On current agenda.
Action 10:
The Winter Newsletter should include a note that NIDS and FOS vendors have
been chosen.
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
.
