Arlington, Virginia
| Members |
UPC Staff |
|---|---|
| John Merrill (Chair) Representatives Richard Clark, (Users Committee) Harry Edmon (ATAC)Wayne Faas (NOAA/NESDIS) David Fulker (UPC) Bernard Grant (NSF/ATM) Clifford Jacobs (NSF/ATM) |
Ben Domenico
|
Users Committee Report
Chair, Rich Clark, who is beginning a three-year term in this position, reported on the 16 June 2000 Users Committee meeting and subsequent workshop. The business of the meeting was to firm up plans for the workshop, "Shaping the Future: Unidata Users as Leaders." The workshop, held the week of 19-23 June, consisted of 17 presentations and seven lab sessions. Lab sessions featured a diversity of topics from case studies to web-based teaching materials. Emphasis was placed on the visualization of atmospheric processes for the purposes of education and research, and included exercises on Met Apps and VisAD. Presentations were equally diverse. Topics included using Unidata applications for air chemistry research and educational applications in mesoscale modeling. Broader views were presented in other plenary sessions by recently-appointed NCAR Director, Tim Killeen (the keynote speaker), and by Cliff Jacobs, NSF. The Users Committee at its next meeting (9-10 November 2000 in Boulder) will focus on preparing the BAMS article reporting on the workshop.
NOAA Report
WSR-88D Level II Archive NOAAPort Data Archive and Retrieval System (NDARS)
This is a collaborative effort with NWS, FSL, and Unidata whose goal is to centralize
NCDC ingest operations and provide real-time climate monitoring capability and
increased data availability to the operational and research communities. NCDC
is receiving all four channels of the NOAAPort data stream with a 10-day retention
in place. LDM is being used for pushing the data to pre-selected users and N-AWIPS
is currently being used for display.
NOAA Operational Model Archive and Distribution System (NOMADS)
Faas described this project as a collaboration with GFDL, NCAR, NCDC, and NCEP.
Its major goal is to enable government and university NWP and GCM researchers
to collaborate using the large data volumes from the primary collaborators,
both in real-time and retrospectively. DODS may be used as the format-independent
data exchange, along with a focus of long-term data exchange needs of the operational
and research communities. Proof-of-concept DODS servers at NCEP and NCDC will
be implemented.
Discussion
NSF Report
Cliff Jacobs's NSF report included three major focus areas: the ITR Initiative, the FY2001 budget, and the Grand Challenges in the Environmental Sciences.
ITR (Information Technology Research)
The ITR initiative has a 45 million dollar proposed budget. These dollars, equaling
150-180 in the FY2001 phase, are accessible only through the proposal process.
(See http://www.ITR.nsf.gov/.) Distribution
of the funds in the 2001 Initiative is threefold: 40% for single-PI projects,
40% for small groups (<$1M per year up to five years), and 20% for large
groups (>$1M per year up to five years). Proposal quality within the three
groups may precipitate some changes in the percentage allocations.
Interdisciplinary proposals will hold priority, but in Geoscience-specific topics, application-distinct proposals will have priority. A preproposal screening process will be utilized in the application process which should facilitate the process considerably.
2001 Budget
A 17% budget request has been made; a 14% increase is anticipated, but
the 2001 budget is still in the Congressional review process. (A different administration
probably would not change the budget picture for 2001.) The bulk of the NSF
budget increase may be earmarked for initiatives; thus base funding will suffer.
While the prospect for a significant NSF budget increase seems positive, uncertainty
exists about how an appropriation might be made. It seems clear that a large
portion will of the funds will be used for new initiatives. This means less
new funding at the directorate level. ATM is concerned that it may receive an
increase earmarked for HIAPER. This is a two-edged sword: HIAPER is recognized
as a compelling need, but as earmarked or encumbered funds it would tie up a
large proportion of the budget and still be insufficient for purchasing an aircraft.
[NOTE: since the meeting NSF Director Rita Colwell sent the following information
to the scientific community: "The agreement passed today recommends $4.424
billion for NSF overall in Fiscal Year 2001 - $526 million over Fiscal Year
2000."]
Grand Challenges in Environmental Sciences
Of the great challenges identified by the NSF in Environmental sciences four
are of specific interest for Unidata:
Other Grand Challenges: Biogeochemical Cycles, Climate Variability, Institutions and Resource Use, and Reinventing the Use of Materials.
Unidata technology integrated with GIS (Geographic Information Systems) Geoscience-specific proposals UCAR/NCAR should be a facilitator for the university member community.
Successful proposals intermingle computer science, mathematics, and geoscience to compete successfully for existing dollars.
Equipment Grants
The RFP for Equipment grants yielded 17 proposals 10 of which were funded for
total funding at the 101K level. One new applicant this year was the University
of the Virgin Islands. Proposal quality was much improved over last year's proposals.
Discussion
Committee members agreed that the Equipment Grant Process represents a
successful model for UPC and NSF synergy which serves the NSF, the UPC, and
the university community as well. The grants encourage university participation
in the UPC's processes. At this time, approximately 50-60 universities have
received Equipment Grant funds resulting in a strengthened community.
Director's Report
Fulker summarized the program's overall status as follows:User Relations: Green
Data Sources: Green
Data Distribution: Green
Current Software: Green
Documentation: Red
Java Software: Yellow
Scheduling: Green
Finances: Yellow
Staffing: Yellow
Documentation and other communications. Characterized as "red" or being a cause for concern. Sally's retirement exacerbates what already was an understaffed area. Dave posited three options for improving the situation: One, rebuild the group to its two-year ago status; two, restructure Unidata's communications function so that both its and DLESE's communications can be met; three, restructure and elevate communications group's role to meet other program needs such as proposal development and broadened community engagement. (Potential resolutions to this problem were presented in the "Staffing and Strategic Planning" portions of the agenda.)
Java Software, characterized as "yellow" or requiring attention because the project is in a transition period where it must move from creating prototypes to creating products capable of being widely-used.
Finances. (Covered more fully in Graham's budget report. Description below.) Finances are cause for concern. Unidata operated below total budget in FY2000 but went over its annual allocation. The program has carryover for FY2001, but concern exists because of the NSF Funding Picture and additional staffing needs. (See Jacobs report.)
Staffing. In addition to concern about the Documentation area noted above, other elements of the program are stretched a little thin--primarily because of Unidata's expanded mission which has slowed development and stretched staff resources. Fulker prefers to invest FY2000 carryover in staff expansion while continuing to pursue new funding sources.
While the above comprise areas of concern, the majority of the Unidata program is moving along nicely.
Fulker also reported that an NSDL proposal on which he is PI has been funded at 600K by NSF-EHR and GEO for one year. Eight Unidata staff are authors or co-authors on papers submitted for the AMS. A new concept for distributed thematic data servers (THREDDS) is an agenda item for Ben Domenico.
DiscussionUnidata Budget Report
Joanne Graham profiled Unidata's anticipated FY2000 year-end information and presented scenarios for the outyears. Those can be seen at . It is clear that fairly substantial increases to the Unidata program are necessary in the next year or so even in the worst case scenario (one in which the UPC staff level remains static). Financial benefits realized from not enlarging the staff would, unfortunately, be short term, because increased staffing costs and overhead rates for existing staff are a part of the problem.
IDD Report
Ben Domenico reported that the UCAR server, motherlode (motherlode.ucar.edu), replaces the Unidata server, Thelma (thelma.ucar.edu), as the main IDD source machine. The motherlode machine is a Sun E450 with four 450 Mhz processors, 4 GB memory, redundant power supplies, and 500 GB disk.
July data losses and related problems resulted from power and network outages at NCAR. To mitigate effects of such outages in the future, Unidata implemented external sites for all top level sites to serve as failovers.
The IDD and CONDUIT stats were reported in one report, but the report was not comprehensive because the statistics were separated only by a week or ten days.
Discussion
Discussion yielded suggested changes to reporting methodology. Specifically:
Strategic Planning
Fulker introduced the Strategic Planning agenda item. He stated that this portion of the meeting would be devoted to converting the goals, articulated at the May Policy Committee meeting, into specific steps to guide the program office in the realization of its expanded mission. The "Unidata Strategic Plan: A Strawman" document served as an outline and guide for the discussion.
Before embarking on specifics Fulker noted that the "Vision" articulated in the document represents an expression of community desire. He urged that the "Goals" be articulated (insofar as possible) linked to the "Grand Challenges" articulated by the NSF and reiterated by Cliff in his presentation.
Participation Policy
The committee reviewed a draft Participation Policy (found in the notebook) under Tab Section A. Following discussion they made the following recommendations:
Jacobs suggested that Unidata representatives meet with members of the MM5 group at NCAR/MMM (whose experience, while somewhat different, is extensive and relevant) as they refine the Unidata's Policy.
The committee voted to accept the Participation Policy in principle pending inclusion of suggested editorial adjustments and other additions in the form of caveats.
Action Item I: Unidata should complete articulation of Participation Policy and circulate the revision to the Policy Committee for review.
Action Item 2: Upon completion of Action Item 1 members of the Unidata staff should meet with NSF representatives to to present and discuss the participation policy and its broader more discipline-independent position.
Real-time, self-managing data flows
Linda Miller detailed real-time data flows currently available provide upper air, lightning, FSL, NIDS, ACARS point-to-point, CONDUIT-NCEP, and global model data. Data flows could include a NOAAPort subset, CRAFT-Level II WSR-88D data; SuomiNet-GPS; Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS); GODAE, and Mesonets.
One flow, CONDUIT (Cooperative Opportunity for NCEP Data Using IDD Technology), presently has 15 participating sites scattered over a wide geographical range.
Yet another promising data stream, CRAFT (Collaborative Radar Acquisition and Field Test), will soon be ready to receive and test Level II data on “Motherlode” server at NCAR. Twelve radars already on line and funding is in place to activate 18 additional sites.
LDM
Russ Rew's presentation focussed on the LDM 5.1.2 which was released in August.
Its most prominent features are its enhanced efficiency: 1) it is five times
faster than 4.0 in search, insertion, and deletion capabilities. While the previous
algorithm depended on the number of products in the queue, the newer version
does not vary the number of products; and 2) Its scalability is enhanced. Also,
a previous bottleneck, pqexpire, has been eliminated. Some ongoing problems:
static routing, non-extensible feed types, configuration difficulty, and non-use
of multicast .
Future development plans are to include alternative architectures and location transparencies in applications.
Discussion
Committee members expressed that the increased access to real-time independent
data sets which provide enhanced access control is pivotal in serving the community.
This amplified capability can facilitate realization of NSF's major research
themes.
Software to visualize/analyze geoscience
Met-Apps
The team of Murphy and Murray presented demonstrations of two of the MetApps
prototypes (Gridded Data Viewer and Sounding Application). During the demonstrations,
features of the prototypes were hightlighted to illustrate the lessons learned
during the development. These included some of the advances made in developing
fundamentals such as image navigation, coordinate transforms and interactive
selection of data. A summary of these features and other lessons learned is
available at http://www.unidata.ucar.edu/staff/donm/MetAppsStrategy.html.
The discussion then focused on clarifying the next steps in the MetApps development--near-
and long-term. The near term objectives are basically the same as stated in
the original goals of MetApps (see: http://www.unidata.ucar.edu/community/committees/metapps/goals.html)
with additional focus on data-type extensibility over function and linkages
with DLESE as practical. The incorporation of the A server in the Interactive
Sounding prototype is a significant step in moving toward a network enabled
methodology for data access.
Distributed, organized collections of digital material
THREDDS
Ben presented an outline for a new project for helping the community to
build a set of distributed thematic data servers to be "acronymized"
as THREDDS or Thematic Realtime Earth Data Distributed Servers. He provided
a schematic of the system which would combine push and pull access technology
and enable users to find and use historical as well as realtime data. The major
components of the system were briefly described along with a list of possible
collaborators which might include SCD, DLESE, Lamont-Doherty, SSEC, CRAFT, GODAE,
IRIS, and NCDC.
Discussion
The committee suggested that a Woods Hole group should also be considered as
a possible collaborator as well as the Open GIS community, possibly working
through John Snow should also be considered for collaboration. To develop a
vision of how this system of of thematic servers might serve Earth Systems faculty
in its teaching and research, three committee members (Ackerman, Ramamurthy,
and Murphy) will work with Ben and Linda Miller on an ad-hoc committee.
When the plans for THREDDS are more firm, the group will make a presentation
to several other groups. ESIG should be contacted in the near future. THREDDS
group members will ultimately present its vision to biology, sociology, and
environmental science groups at NSF.
Action Item 3: Contact the Social and Behavioral Group at NCAR, (ESIG) and invite Bob Harriss, Director, to the next meeting, February 1-2 2001 meeting.
Discussion of the following goals: Computer augmented discourse and discovery and Support and engagement of diverse tool users was deferred until the next Policy Committee meeting 1-2 February 2001.
Updates on Digital Library Efforts
Fulker updated the committee on his and Unidata's interactions with the Digital Library Community generally and his direct involvement in the DLESE group. He noted that DLESE is in some senses evolving in ways similar to those in which Unidata evolved as it grew. For example, the UMADA community which Mike Wright, (former Unidata staff member now with the DLESE group) helped to develop is a valuable model for DLESE in developing its community. An important challenge from Unidata's perspective will occur when DLESE is ready to establish links to atmospheric data. Fulker is PI on a successful SMETE (Science Math Engineering Technology Education) proposal funded at the 600K level. Its goals of developing leadership and community building mechanisms are also commensurate with Unidata's experience. Other co-PIs on the proposal are from the University of Missouri, Eastern Michigan, Columbia, and UC-Berkeley. Funding for the proposal is from the NSF-EHR group.
Staffing Issues
Fulker addressed the committee on staffing changes that will occur within the Unidata Program in the near future. (The committee can make recommendations relating to personnel decisions within Unidata although its mission does not include making personnel-related policy.) The main change: because of his increasing involvement in digital library issues and the community development, Dave proposes to take a "sabbatical" to focus effort on partnerships and proposals needed for Unidata to be a player in ITR and Education and digital library initiatives. He presented two alternative staff restructuring possibilities:
Discussion
Committee concerns for Option One: funding sources for the position; a concern
that the option does not appear, at this point, to include a science component
for the position. In the second option the committee discussed the possibility
of the Director's permanently leaving the Unidata Program.
The committee will indicate its support for the staffing decision in a letter to Rick Anthes.
Action 4: The Committee will compose and send a letter to UCAR Director, Rick Anthes, endorsing Unidata's expansion and changing personnel structure.