Policy Committee Meeting Summary: 
February 5-6, 1998

Boulder, Colorado

Participants

Members  Representatives  UPC Staff
John Merrill (Chair)  Harriet Barker (UCAR) Sally Bates
Robert Fox  David Carlson (UCAR/ATD) Steve Chiswell
Colleen Leary Harry Edmon (U. Wash./ATAC) Glenn Davis
James Moore Jack Fellows (UOP) Ben Domenico
Charles Murphy David Fulker (UPC) Jo Hansen
Julie Winkler Bernard Grant (NSF/ATM) Linda Miller
Other
 "Fritz" Hasler (NASA) Don Murray
Richard Chinman (UCAR/IITA) Clifford Jacobs (NSF/ATM) Sandra Nilsson
Jon Corbet (UCAR/ATD) Mohan Ramamurthy (U. Ill./Users) Russ Rew
Mary Marlino (UCAR/PAGE) Tim Spangler (UCAR/COMET) Tom Yoksas
Don Middleton (UCAR/SCD) Karen Wiggins (NSF/ATM)
"Stick" Ware (UCAR/UNAVCO)
Paul White (UCAR/F&A)

Administrative Matters

 

Status Reports

Director's Report
Copies of Fulker's Director's Report  were distributed at the meeting; Nilsson's budget report is in the notebook.  Fulker's overall status summary is: Discussion
* Dan Vietor is leaving Purdue and joining Unisys; he's negotiating to take WXP with him. Since WXP has been under community support for some time, there was agreement that this does not  require any action on Unidata's part.
* There was a question about staff morale; Fulker indicated that there was some tension from engaging in development efforts while trying to maintain support at the usual levels.
* In answer to a question about market forces on technical staff in Colorado, Fulker indicated that UCAR seems to be keeping up with the market and not losing staff to industry.
* There was discussion about the exhibits at this year's AMS meeting in Phoenix:  foot traffic through the hall was sparse, there was unhappiness in the meeting's layout generally, a real displeasure at the miserable Internet connectivity. Several people had communicated these concerns to the AMS.
 
Users Committee Report
Ramamurthy reported that the AMS Bulletin article on the summer workshop was near completion. The committee discussed accessing Level II NEXRAD data and, while there was some interest, there was concern that most participants would have little use for it , so NEXRAD should not be delivered at the expense of data of more general interest.  There was no discussion about displaying these data. There was interest in discovering whether expert users might create products for others or teach others about the uses of these data. Ramamurthy expressed some frustration with the lack of participation by some Users Committee members and asked that potential members be polled about their commitment before appointments are made.

NOAA Report
Bob Fox spoke with Doug Sargeant and summarized his conversation as follows:
 

  1. AWIPS
    •  Original deployment to the field was 12 systems; these have since been upgraded to software Build 3.
    •  NWS is in the process of deploying another 21 systems at this time, to be completed by the end of March. These systems are using Build 3, to be upgraded to Build 3.1 asap.
    •  NWS has received Secretary of Commerce permission to deploy another 19 systems contingent upon passage of certain milestones, which look achievable. This deployment is scheduled to start in June. Build 4.0 of the software is scheduled to be completed in May, and these systems could be upgraded to Build 4.0 software prior to or soon after deployment.
    • All systems in the field will have software upgrades as new Builds are released. Builds 4.0, 4.1 and 4.2 are anticipated during 1998.
    • All deployments beyond the 12 + 21 + 19 = 52 cited above require additional higher headquarters approval. The optimists believe that this approval will come yet this month, the pessimists believe it will take longer. About one third of the total systems have been approved to date for deployment.
    • Build 3.1 and 4.0 of the software incorporate WAN capabilities to permit administrative use and interactive forecast preparation; Build 4.0 will provide the local data acquisition prototype capability.
    •  AWIPS software development emphasis is now on the NWS for site segment software and upon PRC for network segment software.

      • Note: the deployments to the field reported above carefully did not use the word "operational" or "commissioned." These systems will be used in the routine functioning of the stations where they are located, but "commissioning" in the NWS jargon has a very specific meaning, and these systems will not be commissioned until many other programmatic milestone have been accomplished.

  2. NOAAport: NWS is actively working to  stabilize the contents of the NOAAport broadcasts and refine their content at this time.
  3. Other: Doug is not aware of any discussion concerning the possible transition of AWIPS to Java at this time. He thinks that this is an interesting route for Unidata to follow, and encourages our discussion/action on this topic.

Discussion
* COMET will be bringing AWIPS into UCAR and will incorporate it into its classroom.
* AWIPS ran into receiver problems and changed to C-band; they're now using a new demodulator

NASA Report
Hasler gave a presentation on El Niño that employs TRIMM satellite images (with the first active orbiting radar); the clip was produced for showing in mini-theaters at places like the Smithsonian. He also presented  "Earth on line,"  an interactive model that is Web-based. Copies of his slides were distributed at the meeting.

NSF Report
Jacobs reported on the current state of NSF's FY99 budget request, NSF's priorities and directions, the impending review of Unidata (by a panel meeting in Boulder at the end of March), and on the new Knowledge and Distributed Intelligence program that was just funded. With a total of $65M to spend on KDI awards, NSF is planning to hold $15M to fund proposals already submitted to NSF.  An internal committee with members from across NSF (Jacobs will be representing ATM) will decide how to allocate the funds to the "extant" proposals.  He is hopeful that some funding for Unidata from this pool may be possible.  He also encouraged UCAR to submit a coordinated proposal. Copies of Jacob's slides were distributed at the meeting.

Discussion
 * There is a Research Fund that cross-cuts agencies and which is broadly supported by Congress.  NSF funding seems to be broadly supported as well.
*  FY99 budget request includes $78M for KDI; all KDI monies must be committed each year. The short lead time on this year's KDI program will strain NSF's review process.  Since KDI awards can be up to $1M, NSF expects to be inundated with proposals.  NSF has not decided on the review process, but Jacobs predicts that only 5% of the proposals will be funded.  All proposals must be submitted by NSF's on-line system, which may swamp the system.  Proposers should submit earlier than the 8 May deadline to ensure that their proposals make it through the system.
* Commercial entities may not apply for KDI funds.
 

PAGE (Marlino)

Marlino summarized PAGE's activities, which included an intense month of focus-group meetings across the country. The preliminary findings suggest that the highest priority needs for the community are:  (1) a Web-accessible clearing house of geoscience educational resources, (2) Web  access to multimedia materials from the COMET modules, (3) workshops on pedagogy and technology, and (4) assistance with assessment and evaluation. Copies of Marlino's slides are in the notebook.

Discussion
* The relative roles of PAGE and Unidata were questioned.  Marlino noted that PAGE was committed to the Unidata model of sharing resources.  While Domenico is on loan to PAGE for the immediate future, the programs are otherwise distinct;
* One challenge facing Unidata is the heterogeneity among campuses and Marlino was questioned about PAGE's approach to this problem; Marlino noted that the have/have-not situation in the PAGE community is not equivalent to the Unidata situation (which requires participant commitment and resources) since PAGE is dealing with Web-based resources.  Clearly, however, some campuses will need more help than others in applying technology to the classroom.
* An interesting difference between PAGE and Unidata is that campuses that "have" the technology still may not be able to apply it in education.  The PAGE challenge is to combine what we know about learning with state-of-the-art science.  The Users workshop was a good example.
* UCAR's involvement in broad geoscience education was questioned; Jacobs noted that UCAR's mission is to support atmospheric "and related" research and education.
 

SuomiNet (Fulker)

Fulker reported that Unidata and UNAVCO delayed submitting the SuomiNet proposal for MRI funds on the belief that the idea needed further clarification. There were indications that people needed to know more about the scientific value of the observations, which may be addressed by hosting a workshop this summer. Randolph "Stick" Ware, director of UNAVCO, mentioned that the need for real-time data in the solid earth community was limited to a few disciplines at present, and therefore he needed more input on direction from his community.

Unidata and UNAVCO intend to pursue the resolution of current issues and submit for MRI funds next year.

Discussion
* Although some MRI funds would be used by Unidata and UNAVCO for the administration of SuomiNet during the period of the award, Jacobs questioned how the network would be sustained in the period after the duration of the MRI grant.
* Chiswell asked for recommendations as to the level of support that universities should be expected to provide toward cost sharing for such a proposal in light of the recognition that smaller universities should not be overlooked.
* Harry Edmond pointed out that institutions such as U. Washington had several continually operating GPS receivers already, and so the possibility of adding the meteorological packages to the existing network for precipitable water estimates rather than purchasing complete systems was desirable from their point of view.
 * Fulker was questioned whether SuomiNet represented a new paradigm for Unidata--as a data generator. Fulker noted that UNAVCO is already creating these data and would be doing the processing; Unidata's role is to disseminate those data, as is done with data from FSL.  Unidata would also be the broker of the funds and would manage the data-distribution topology.
* The proposal might better present SuomiNet as a "proof of concept" (the concept being that these data will improve forecasting) with a limited term.  This would sidestep issues of instrument maintenance over time and would represent valid research.  The proposal would need to reference the simulation work that indicates these data would enhance forecasting, however.
* Unidata is interested in SuomiNet as a test of IDD capabilities and to further Unidata's goal of involving other disciplines.
* When questioned why Unidata was investigating GPS data instead of radar data, Fulker noted that the use of radar data (everyone wanting data from a particular radar when weather in that area was interesting) would require developing automatic rerouting capabilities and there are no funds available for developing and testing these.
 

Resolution 1   (passed with one opposed):
The Policy Committee recommends that Unidata work with UNAVCO to submit a proposal to NSF for MRI funding to start, operate, and distribute data from a SuomiNet for a fixed period of time.
 

Resolution 2 (passed with one opposed and one abstaining):
The Policy Committee recommends that it be given time to review the revised SuomiNet proposal prior to submittal.

 Action 1:
UPC staff should pull together an annotated bibliography of the science involved in the SuomiNet proposal and add links on the SuomiNet Web page to current SuomiNet-like products.

NOAAport

Fulker noted that NOAAport receivers are composed of off-the-shelf components, with the exception of one piece.  The belief is that, with the advent of NOAAport, there will be a drop to zero of FOS subscribers.  The fate of DIFAX is still unresolved, but this does not affect Unidata; sites currently subscribe to DIFAX through Alden.  The cost of a PRC NOAAport receiver is about $65k; SSEC is now offering a NOAAport receiver system for about $15k (for one channel only), with additional channels costing a bit more than $10k per channel. The letter detailing the SSEC offer is in the notebook.  Unidata was originally planning to receive NOAAport data from COMET, which is scheduled to receive a PRC receiver later this spring.  With the SSEC offer, Unidata may reconsider its options.

Discussion
* Receivers  may eventually be cheaper than even the SSEC prices.
* SSEC is receiving NOAAport data already, but has not compared the NOAAport stream with FOS
* DIFAX availability will be an issue for the community; one suggestion is for Unidata to organize the creation and dissemination  of DIFAX-like products by a university.
* If several universities install NOAAport receivers, then this redundancy on the IDD will replicate Alden's 7X24 operational data flows.  Perhaps Unidata should consider applying for MRI funds to fund equipment for 4-5 universities to capture NOAAport.
* NOAAport imagery is all remapped and may not be adequate for all research purposes.

Action 2:
The Users Committee will discuss NOAAport issues at its next meeting.
 

NEXRAD (Fulker)

Prior to this Policy Committee meeting, Fulker held a meeting to consider the issues posed by distributing NEXRAD data.  Participants included Woody Roberts, Rit Carbone, Harry Edmon, Richard Chinman, and Linda Miller.  Edmon is currently  repackaging Level II data,  compressing it, and sending it to his university via the commodity Internet using the LDM. He finds the data volumes "reasonable" (about 9.8 MB/6 minute) and uses Zebra to look at the data.

Discussion
* The NWS is not interested in connecting directly to universities since the current configuration for doing so would require running a Sun workstation as an interface; in the future (about 18 months), NEXRAD sites will have 4 direct ports each and connections will be possible using open RPG .
* USWRP is interested in having access to NEXRAD base data; Unidata may cooperate with them and with NCAR's RAP, which is incorporating radar data into hydrological models.
* NCDC is archiving all the radar data.
* Current arrangement with NIDS vendors probably will not continue beyond September, 1999.
 

Outstanding Participation Award (Merrill)

The award subcommittee recommends that the award be Jacobs reported that no NSF funds can be used for this award, suggesting that the UPC ask UCAR for money from its General Fund.

Resolution 3 (passed unanimously):
The Policy Committee accepts the purpose and procedures drafted by its subcommittee to create a Russell  L. DeSouza Outstanding Participation Award.
 
Action 3:
The UPC will write a letter to the DeSouza family asking permission to use his name.

Action 4:
 The UPC will hold discussions with UCAR management about the possibility of using UCAR funds for a cash award.

Action 5:
Bates will contact the AMS about the possibility of their sponsoring  the Russell DeSouza award.

Action 6:
The final form of the award, which depends on the results of Actions 3-5, will be a topic on the next agenda.

JAVA Transition (Fulker and Rew)

The UPC finds it difficult to describe what users can expect of the new Java software.  In general, Fulker envisions components--small pieces that can be rearranged by users and that can contain contributions by others in the community.  Rew described the methodology for developing the components, which involves building prototypes in increments using the "scrum" method (daily meetings, short-term goals, rapid prototyping). There will still be a need for a framework, which must be initially determined through the development of a limited application then subsequently readjusted as needed.  The development group has settled on using the "three-example framework" (i.e., they will choose three related components) but has not identified the specific components.

Discussion
* The iterative approach is particularly powerful when learning a new language.
* Java performance on GUIs is not an issue
* The implementation of most candidate applications should be capable of being run by anyone in the community; more complex applications (3-D) may need to wait for implementation in Java extensions.
* The same people who maintain the current system are also the developers and so are well aware of user needs; however,  UPC still needs explicit interactions with users to evaluate the prototypes.  Need to select testers carefully.
* UPC may want to select applications with an eye toward KDI funds: e.g., applications where there are problems of scales and structure or involve the interplay between computation and different kinds of data,
* NSF review panel will be looking at the program as a whole, not just the plans for the Java transition; reviewers may have data or support issues and the UPC must be prepared to answer these.  The reviewers are users not developers.
* The future of Java was questioned.  Object-oriented methodology is useful in itself; once designed, the applications can be written in any object-oriented language.
 

Response to KDI Announcement

Fulker identified seven activities that might form the basis of a coordinated UCAR proposal or set of proposals to NSF for KDI funds.  (Copies of Fulker's "Thoughts on KDI" were distributed at the meeting.) They are as follows:
  1. Community-Oriented, Interactive Earth-Observing Network (basically an expanded IDD).
  2. Object Components for Visualizing Earth Systems (a set of software components for data analysis and display).
  3. Simulators of Earth Systems and Fundamentals (a range of models, from the CCMs to mini-models that can be used in the classroom).
  4. Object Modeling of Earth-System Concepts (a multidisciplinary classification framework for organizing scientific principles, data, hypotheses, learning objectives).
  5. Community-Wide Resource Discovery System (for helping people find data, simulations, and any other need resources on the Internet).
  6. Hyperlinked "Story Board" for Geoscience Course Creation (that allows users to assemble a course using community resources but tailored to individual classes or learners).
  7. Peer Comments on Data, Simulations, and Explanations (a peer-review process for the resources created or made available in the above activities).
Discussion
* KDI represents not a Unidata effort, but a UCAR effort that would incorporate Unidata, PAGE, and several NCAR divisions.
* What is the Policy Committee's role vis-a-vis activities in the rest of UCAR? The KDI activities will involve groups that have their own advisory committees.  The potential for Unidata playing a leadership role in UCAR activities is occurring more frequently; the UPC and the Policy Committee should address this.
* KDI poses some fundamental organizational issues for the UPC and UCAR.  Unidata users have been pushing for the integration of science research and education (the fusion of data, dissemination, teaching, and research).  Unidata is the spokesperson for the community, and this voice should not be fragmented.  Fulker prefers to see Unidata as a contributor to a UCAR-wide team rather than expanding to incorporate new initiatives.
* There are numerous examples of collaborations in the university community (e.g., UIUC); in a KDI proposal UPC should engage one or more of these universities as well.
 

Resolution 4 (passed; one opposed):
The Policy Committee expresses its support for UPC participation leading to a UCAR-wide KDI proposal, along the lines summarized in the white paper presented by Fulker.  The Policy Committee wishes to stress the importance of including community participation.
 

Summary of Resolutions and Action Items

 
Resolution 1   (passed with one opposed):
The Policy Committee recommends that Unidata work with UNAVCO to submit a proposal to NSF for MRI funding to start, operate, and distribute data from a SuomiNet for a fixed period of time.
 
Resolution 2 (passed with one opposed and one abstaining):
The Policy Committee recommends that it be given time to review the revised SuomiNet proposal prior to submittal.

 Resolution 3 (passed unanimously):
The Policy Committee accepts the purpose and procedures drafted by its subcommittee to create a Russell  L. DeSouza Outstanding Participation Award.

Resolution 4 (passed; one opposed):
The Policy Committee expresses its support for UPC participation leading to a UCAR-wide KDI proposal, along the lines summarized in the white paper presented by Fulker.  The Policy Committee wishes to stress the importance of including community participation.

Action 1:
UPC staff should pull together an annotated bibliography of the science involved in the SuomiNet proposal and add links on the SuomiNet Web page to current SuomiNet-like products.
 
 Action 2:
The Users Committee will discuss NOAAport issues at its next meeting.

Action 3:
The UPC will write a letter to the DeSouza family asking permission to use his name.

Action 4:
 The UPC will hold discussions with UCAR management about the possibility of using UCAR funds for a cash award.

Action 5:
Bates will contact the AMS about the possibility of their sponsoring  the Russell DeSouza award.

Action 6:
The final form of the award, which depends on the results of Actions 3-5, will be a topic on the next agenda.
 

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 .