Policy Committee Meeting Summary:
January 27-28, 2000
Boulder, Colorado
Participants
Members |
Representatives |
John Merrill (Chair)
Michael Biggerstaff
Robert Fox
David Knight
Charles Murphy
Mohan Ramamurthy
Julie Winkler
UPC Staff
Sally Bates
John Caron
Steve Chiswell
Ethan Davis
Ben Domenico
Steve Emmerson
Joanne Graham
Jo Hansen
Robb Kambic
Linda Miller
Don Murray
Russ Rew
Jeff Weber
Stuart Wier
Anne Wilson
Tom Yoksas
|
David Carlson (UCAR/ATD)
Harry Edmon (ATAC)
Jack Fellows (UCAR/UOP)
David Fulker (UPC)
Harry Glahn (NOAA/NWS)
Clifford Jacobs (NSF/ATM)
Al Kellie (UCAR/SCD)
Tim Spangler (UCAR/COMET)
Others
Susan Jesuroga (UCAR/COMET)
Joaquim Kuettner (UCAR/JOSS)
Mike Jackson (UCAR/UNAVCO)
Teresa Van Hove (UCAR/GST)
Randolf Ware (UCAR/GST)
|
Administrative Matters
- The Committee welcomed Michael Biggerstaff as a new member and was introduced
to Unidata's new staff members, Anne Wilson and Stuart Wier.
- Correction to last meeting's summary, The seventh bullet
under GEMPAK and AWIPS discussion should read:
"Without shared applications and data, there is
no Unidata community. "
- The next meetings of the Policy Committee will be:
- May 22-23, 2000, possibly at SSEC in Madison, Wisconsin
- October 5-6, 2000, in Arlington, VA
- The May meeting will be devoted in large part to a discussion
of Unidata's contributions to the GEO2000 initiatives.
- Julie Winkler complimented Unidata on the Special Session
at the AMS.
Status Reports
Director's Report
Copies of Fulker's transparencies were distributed at the meeting.
Copies of Graham's transparencies are in the notebook. Graham's
presentation highlighted that the current funding outlook (flat
budget for FY00 with 3% increases through 2003) will create
significant funding problems starting at year-end, FY01.
Discussion
- There was considerable discussion about what the IDD feed-rate
log charts are measuring. The concern expressed at that last
Policy Committee meeting focused on an apparent increase in
latencies, which now seems ephemeral; however, the increase
in data flow rates is still of some concern. It appears that
latency is independent of data volume, although how long this
will hold is unknown; at some future point, the increase in
volume may overwhelm capacity, then latency will be affected,
but no one knows where this point is.
- There is difficulty with the CONDUIT feed from NWS/OSO,
which NWS and NASA/Goddard are working on. It is not
an LDM problem.
- There was concern over the small size of the UPC equipment
budget. Graham and Fulker assured the committee that measures
taken within the program to upgrade when possible rather than
purchasing new hardware make the budget adequate for now.
- Seeking new outside funding for the sake of increasing the budget can add
to workload stresses and eat away at the core mission of the program if the
right synergy between current and new activities is not sought.
Users Committee Report
The summary of the last Users Committee meeting is in the notebook.
Murphy presented the summary for Moody, who was unable to attend.
MetApps Report
Charles Murphy distributed a MetApps status report.
Discussion
- DIFAX will disappear from the NOAA data stream. Michael
Morgan has created scripts for generating DIFAX-like maps
from GEMPAK and has offered to make these available to the
community and/or to add them to the IDD.
- MetApps prototypes will be demonstrated at the summer workshop,
but the Users Committee is still discussing how and where
this will occur.
- The long-term goal of MetApps development, as stated in
the proposal to NSF, was to develop a suite of tools by 2003;
since the resources did not materialize, however, the schedule
cannot be met, but a new schedule has not been established.
NOAA Report
Mary Glackin has moved to NCEP. Harry Glahn will serve as NOAA
representative until further notice. Glahn reported to this
meeting via telephone. He noted that Gen. Kelly has established
a work team to reduce the amount of redundancy in decoder development.
Essentially, selected NWS office will be responsible for the
development of specific decoders/encoders, which will then be
released (made available to any offices needing them) as they
are developed. NCEP is involved in this activity. A copy of
Kelly's policy on this was faxed to Unidata and distributed
at the meeting.
Glahn also reported that there are no new NIDS developments.
The NWS still intends to end the agreements with NIDS vendors
by September 30, 2000. However, that depends on a number of
things being done by the NWS by then. If an extension is necessary,
it will be done in 90-day increments. In the meantime NWS will
continue adding NIDS products to the NOAAport stream in encrypted
form until the agreement ends. (NWS intends, however, to remove
the encryption for short test periods to allow users to test
their ingest and processing of radar products.) The data will
be products from individual radars. They intend to make available
as much data as possible, but face competition for the bandwidth
out of NWS (currently a T1) and there have been no funds identified
to increase the bandwidth in FY 00 or FY 01. High- and low-resolution
mosaics are now being created, including the 10-km mosaic in
the AWIPS requirements, but weather offices don't yet have the
capability to receive these. For the time being, these products
are available on the NWS Web server. The NWS is now working
on producing a 4-km mosaic.
Discussion
- There was discussion of the implications for Unidata of
the volume of NIDS data. Unidata does intend to distribute
as much of the data as possible, but priorities may have to
be established. The NWS has asked Unidata to be a test facility
for NOAAport data beginning in May, which should bring any
problems to the fore.
- The timing of the cessation of the NIDS contract with vendors
is awkward for the universities. They will need to decide
before October whether to continue their own contracts for
another year.
- McIDAS and GEMPAK have been adapted to read current NIDS
products.
NASA Report
George Serafino was unable to attend the meeting.
NSF Report
Cliff Jacobs updated the committee on new leadership in the
Geoscience Division (Margaret Leinen is now in place as the
director; Jarvis Moyers has been appointed as director of ATM
for two years, and Hermann Zimmerman is the director of Earth
Sciences). On NSF's budget, Jacobs reported that GEO doesn't
yet have its operating budget for this year (but most of the
7% increase to NSF went to ITR, biocomplexity, and education,
so GEO is expecting a flat budget this year); the budget for
2001 should be released Feb. 7, and unconfirmed reports indicate
that it includes a proposed 17% increase within five major themes
(ITR, biocomplexity, 21st century workforce, nanoscale science
and engineering, and unmet needs).
Jacobs urged the committee to consider Unidata activities in
the context of NSF's strategic activities. He believes Unidata
is mature enough to move beyond mesoscale meteorology to a broader
context that is consistent with the Geoscience research agenda
(See Geo2000).
Jacobs then summarized the Findings
and Recommendations of the National Science Board's Task
Force on the Environment Interim Report.
For Bernard Grant (who was unable to attend), Jacobs reported
that the equipment grants announcement is scheduled for release
in February. It will have a May deadline for proposals, which
NSF intends to review in June and award in July.
Discussion
- Jacobs indicated that Unidata adopting strategic goals congruent
with NSF's would provide Jacobs with ammunition for prying
funding out of other (non-ATM) NSF groups. He also believes
that participating in these activities would enable Unidata
to broaden its community.
- Unidata is clearly already aligned with some endeavors:
the NSB Task Force's call for virtual centers was cited as
one example (Recommendation 8). The call for more cross-disciplinary
efforts would require examining Unidata's disciplinary focus
(atmospheric science). There were questions about whether
other disciplines had created their own "Unidatas."
CORE and IRIS were cited.
- The Users Committee has long discussed the need to provide
access a range of retrospective data. Methodologies for dealing
with these would be applicable to many disciplines. There
was a suggestion that a joint Unidata/UCAR (SCD) discussion
on this with NSF might be productive. There was considerable
discussion about whether Unidata's emphasis of real-time data
needs to be revisited.
Action 1: The Unidata/SCD collaboration on access
to data archives will be on the agenda for the next meeting,
depending on Kellie's schedule.
Action 2: Following consultation with Jacobs,
Fulker will draft one or more letters to NSF articulating a
role for Unidata in NSF's major research initiatives for the
coming decade. The letters will go out over Policy and Users
Committee signatures
SuomiNet
Randolph Ware gave an overview of the SuomiNet project. A copy of the forthcoming
BAMS article on SuomiNet is in the notebook. In the course of his overview he
noted that 170 sites have registered, 27 of whom are from outside the U.S. They
have received 33 letters of commitment covering 50 sites. (The letters formally
acknowledge the site's commitment to provide infrastructure support for the
receiving stations for five years.) Their proposal calls for purchasing equipment
for between 70 and 100 sites; they are now in the throes of testing equipment
to determine which will be purchased. Mike Jackson, who defined the specifications
for the needed system, passed around one possible GPS receiver. GST expects
to purchase equipment in May and to send it to participants in midsummer.
Discussion
- There was some discussion of how the data would be distributed.
GST is considering a star network, whereby all raw data could
be passed through selected relay sites to UNAVCO, which would
provide quality assurance. Community members could elect to
receive raw data or qualified data.
- GST personnel have identified some potential products but
also have some information that they don't yet know how to
use.
- There were questions about how to use the data. Much of
the information can be turned into 2-D vertical profiles that
can be assimilated into current numerical models. Committee
members noted that many would be interested in using the data
in regional MM5 models. Ware noted that effort would be needed
to enable these new data to be assimilated into a range of
models; he views this as part of the research agenda for the
SuomiNet project.
- Ware was questioned about the meteorological package to be used. He noted
that two have been evaluated and one is being tested. SuomiNet specifications
require the ability to gather both meteorological and GPs data.
Archive Data at SSEC
Bob Fox provided a handout on the archive satellite data available
from SSEC to universities. He noted that the system for provided
these data is ready to go and that SSEC would like feedback
on how well it works.
Discussion
- Fox was questioned about what amount of data would be considered
"reasonable." He responded that this was yet to
be determined.
- Committee members noted that an online facility with browse
capability of the "most popular data" would be wonderful.
Fox responded that the data holdings at SSEC are vast and
that determining which are "most popular" would
be difficult. The current retrieval system is straightforward:
a request is submitted online, the requested images are loaded
on a server and the requester notified. The images remain
on the server for a predetermined length of time, then are
removed; it is the users responsibility to download them within
the specified time period.
Note: Fox was roundly applauded by the committee for
this initiative.
Discussions on Collaborations
Fulker distributed a table of current and potential collaborations
and asked the committee to brainstorm on how these might impact
Unidata and to evaluate Jacobs' request to consider how and
whether Unidata might participate in NSF's long-range goals.
He noted that, at the staff retreat in October 1999, staff members
had expressed the desire to keep Unidata at about its current
size and they had questioned whether all the potential collaborations
fit into the Unidata mission.
Discussion
- There was discussion of the impact of adding staff to the UPC to accommodate
collaborations. . Spangler noted that a program's size matters. At about 40,
a program encounters problems in communications but gains flexibility. At
even larger sizes (60-80), Fox added, turf wars are inevitable.
- There was considerable discussion surrounding the question
of whether Unidata's current five-year proposal is expressed
in a manner that provides the program with a long-enough vision,
or enables it to set priorities.
- When considering the expansion of Unidata, several approaches
were suggested: expansion in number of services, or in disciplines
served, or in geography, etc.; by examining thematic areas
such as enabling technologies or information networks; or
to categorize collaborations as enabling Unidata to enhance
the needs of its core users or to acquire additional resources
or to reach out to users in other disciplines.
- Carlson suggested that Unidata's core mission of providing
data and software to users will not change; change, instead,
will be forced by the evolution of technology. Implementing
new technology takes 2-4 years. In this timeframe, will the
funding environment change? He suggested that the strategic
question facing Unidata is: should the program bend to NSF's
changing vision or simply look for funding to meet the ongoing
and evolving needs of its users. Fulker agreed that the constants
in Unidata are the community's needs for data and support,
but noted that technology change can also be viewed as a constant
and that Unidata is driving some of this change.
Participation Policy
Fulker prepared a
draft matrix of participation. The new concepts presented in the matrix
are the lack of any geographical constraints and the requirement of a nonprofit
status for receiving support. He noted that the licensing discussion is based
on defining who the Unidata community contains. Ben Domenico, Joanne Graham,
and Robb Kambic provided an overview of and development plan for the user-tracking
system that Unidata is developing.
Discussion
- For support to be possible, users must have the ability
to discuss technical matters in English. This requirement
was added to the matrix.
- There was considerable discussion about the definition of "nonprofit."
What about "commercial" universities? What should be Unidata's role
vis-à-vis for-profit institutions? NSF's view is to engage the broader
university community for wide dissemination. There were also concerns expressed
about the Freedom of Information Act and how this would affect Unidata.
- Concerning the required qualification of having a research
and education mission, members noted that policing data distribution
beyond the first IDD tier seems unrealistic. Members also
felt a service component needed to be part of the requirement.
This was added to the footnotes.
- Concerning the required qualification of offering courses
for credit, members questioned why data couldn't flow directly
to state agencies and whether the universities (many of whom
direct information to these) would be responsible for monitoring
what happens to the data. The broader question of defining
the scope of IDD was raised: does it end, for support, with
universities?
- The "point and click" license will need to include a set of rules
for data use, including restricting the right to redistribute data. Exceptions
to or violations of these will need to be dealt with on a case-by-case basis.
The specter of monitoring was raised again.
- While the university should be the point of contact, this
hasn't proved practical; therefore matrix allows for individuals.
Individual participation, however, entails support problems
since Unidata cannot provide support to students. Sorting
requests by both individual and institution is one way to
help define who should get support Therefore, individual affiliation
is and the identification of an individual point of contact
for an institution is a requirement.
- Concerning constraints on usage, public service was added to education and
research and redistribution was limited to nonprofit education, research,
or public service purposes.
- Unidata staff would like to expand its user base beyond
synoptic mesoscale meteorology.
Action 3: The Users Committee will discuss the
support issue as it pertains to the "point and click"
environment.
Action 4: Prior to the next meeting of the Policy
Committee, Fulker will revise the Participation Policy to reflect
the consensus achieved on the participation matrix.
Action 5: Unidata will provide an demonstration
of the on-line user-information tracking system at the Spring
or Fall Policy Committee meeting.
LDM Development
In response to action item 4 from the last Policy Committee
meeting, Russ Rew and Anne Wilson updated the committee on LDM
development. In terms of short-term development, they identified
two major concerns:
- The shortage of feedtypes for new data feeds; and
- Performance problems with the LDM product queue
Over the longer term, areas under evaluation for the next-generation
IDD include:
- Routing protocols
- Improving local data management
- Simplifying configuration
- Incorporating "pull" technology
Action 6: Unidata will prepare discussion guides
in advance of the May meeting and distributed them to meeting
participants. Additional guests from NOAA, NASA, UOP, and others
will invited.
Unidata Homepage
This page was Webified by Sally
Bates.
Questions or comments can be sent to <sally@unidata.ucar.edu>.
This page was updated on .