Unidata Users Committee Meeting Summary
17-18 May 2005
|Jim Steenburgh, Univ. of Utah, Chair|
|Elen Cutrim, Western Michigan Univ.||Steve Chiswell|
|Chris Herbster, Embry Riddle Aeronautical Univ.||Ethan Davis|
|Kevin Knupp, Univ. of Alabama - Huntsville||Ben Domenico|
|Anton Kruger, Univ. of Iowa||Emily Doremire|
|Gary Lackmann, North Carolina State Univ.||Steve Emmerson|
|Donna Tucker, Univ. of Kansas||Joanne Graham|
|Michael Voss, San Jose State Univ.||Jo Hansen|
|Tom Whittaker, SSEC - Univ. of Wisconsin-Madison||Jeff McWhirter|
Please Note: Proposed Meeting Date: 13-14 October 2005
The first half day of the Users Committee meeting was combined with the Unidata Policy Committee meeting. This was the first time the two committees had met jointly.
Introductions/Stage Setting (Mohan)
Mohan welcomed all committee members and attending staff to this historic meeting. He noted that a late-season snowstorm forced a change in the Users Committee meeting plans, and that discussion among staff and committee members confirmed that a joint half-day meeting would be desirable. Mohan believes that if the meeting proves to be successful, it could be a format we should consider repeating each spring.Software and Support (Rew with input from Yoksas)
The presentation and discussion grew from an action item taken at the October 2004 Policy Committee meeting. Increased use of Unidata software, growth of the Unidata community, and new data streams have increased the need for support over the years. Providing excellent support is a core Unidata value, and this is why the UPC is exploring ways to maintain support at a high level.
Users experience support in several ways, most frequently in the form of e-mail exchanges with staff. Others are via topical e-mail lists, occasional drop in and telephone consultation and workshops. The latter is less preferable than e-mail forms, since they (e-mail) leave a paper trail which often prove useful to other community members with similar questions. Another form support takes is in making new data available. Support provides developers with benefits while at the same time competing with actual time spent on development.
From a largely internally-developed system supporting only a few users, Unidata support has grown to 30 queries a day or 140 a week, and indications are that the load will grow. All developers participate in the support effort. Proactive support in the form of training (and other) workshops is important.
Several solutions for improving the situation are under consideration. Two packages: eSupport (www.kayako.com) and Request Tracker (www.bestpractical.com/rt) are being evaluated by Unidata's Support Task Force. One of these packages will most likely be implemented.
V2 of the Unidata portal will feature a roll-out of eSupport, users will be alerted to the fact that it's an experiment and if proven successful additional features will be added: FAQs and knowledge bases from support archives.
Planning to transition the McIDAS package to a version--McIDAS-V-- that can easily visualize data from the next generation of weather satellite instruments is underway. At thirty-plus years of age the program, the program originally written in Fortran 77 and C has limited extensibility possibilities, new visualization concepts cannot be incorporated, and, as noted, environmental satellite data cannot be used in the best way (GOES-R & NPOESS operational systems).
The plan is to enhance the functionality of the package while retaining its current capabilities, including user-developed code. To do so several options were considered, with the decision to build upon an innovative rapidly evolving system design to move existing functionality to an environment that would allow considerable future growth in both capacity and functionality and the selected environment is Unidata's IDV. The IDV is based on SSEC's VisAD package and is designed to replace both McIDAS-X and GEMPAK in due course.
The IDV can integrate several kinds of data that are and will be crucial for McIDAS users and some of those are geo-referenced netCDF data, GIS-ESRI shapefiles, ADDE, among others. ADDE is the key data access tool for McIDAS -V, since the emphasis is on satellite data. OpenADDE is an open source version of the ADDE servers and includes required McIDAS-X routines (about 30% of the total code). Open ADDE is already being used to allow "smart subsecting" when reading local AREA (image) files in the IDV.
A four-phase approach to the transition has been initiated. (Phase One includes plans to run the heritage McIDAS within the McIDAS-V environment.) The entire transition will take 3-4 years. At this time, SSEC is seeking partners in proposals to fund this effort. One proposal has already been submitted, but since the overall amount required is probably ~$1.3M it is certain that more than one proposal will have to be submitted.
A resolution to propose creating a new journal by either the AGU or the AMS, or both, was presented. Creating the resolution grew from discussion during the committee's first day concerning the need for a peer-reviewed journal whose focus would be on cyberinfrastructure. Existing journals have a cyberinfrastructure component, but little in this area appears in their contents. The resolution will be refined via e-mail and subsequently a vote of all members of both committees will be taken.
There has been a steady increase in interest in the application of information systems in earth and atmospheric sciences, as evidenced by the number of submissions to the AMS IIPS Conference and the formation within AGU of a Focus Group on "Informatics for Earth and Space Sciences". The National Science Foundation is making a substantial investment in cyberinfrastructure development across all the sciences, at a level of approximately $800 million per year in its current budget. Considerable intellectual innovation is occurring across traditional disciplinary boundaries and the field of informatics in the earth and atmospheric sciences has matured to the point where a new professional society journal is justified, as current AGU and AMS journals are focused on science advancement.
The Policy Committee recognizes the need for a new journal on earth and atmospheric science informatics and resolves to promote this idea through a letter addressed to AGU and AMS Presidents, asking the two societies to consider this as they develop their future publication plans.Unidata Metrics/Evaluation
This discussion was a continuation of one begun during Monday's meeting while NSF representatives were present. Monday's discussion underscored the importance of focusing on NSF's three-point: People, Tools, and Ideas.
All agreed that defining Unidata's community is of primary concern. It is also critical that outside independent professionals have to guide the evaluation and structure any surveys.
Many metrics for Unidata include ill-defined and intangible elements, unconventional uses of Unidata software and tools. A term applicable to these sorts of uses might be "cross-pollination". Quantifying such efforts, while complex, would be well worth while.
Action items developed from this discussion are:
3. UPC will prepare a fleshed-out schedule detailing milestone dates for the assessment project which will be provided to committee members by the fall meeting, specifically something more detailed than "survey to be completed by 2007."
4. The UPC will select an independent, outside consultant to provide guidance in preparing a metric for the next funding proposal, after which a working group comprising UPC staff, Policy and Users Committee representatives will be appointed to provide input to the process.Joint meeting
Melanie and Jim as committee chairs will make a decision about future joint meetings.
Review of Action Items
Discussion: can a site do an informal workshop without UPC assistance? Tom Whittaker has already done it. There is a question of expertise and use of software, but it can certainly be done. There are tutorials on the Unidata Web site for sites to follow.
Policy Committee meeting summary from October 2004 meeting – Jim Steenburgh
Director's Report – Mohan Ramamurthy
Review of Status ReportsUnidata Status Reports:
Floater - Chiswell, Yoksas
In response to Action Item #2 from the October 2004 User Committee
meeting which contained:
"The Users Committee also requests that the UPC implement a Level II floater based on Unidata automated selection scheme".
A method for objective determination of a floater region based on precipitation forecast for a 24 hour period was presented. The selected floater region could be used to provide coordinated radar and satellite imagery as well as additional high resolution data sets such as mesoscale model output, to replace the current floater radar and satellite floater sectors which are independenly chosen by the
community. The revised floater data feed would provide an integrated data set.
Could generate a "bundle" for the IDV and make it available.
Concept is fine, but there's something missing at the education level for the suggested case study approach. Users Committee suggests putting the task aside in favor of other priorities at this time.
Discussion of lessons learned from the 2003 workshop
The topic of use of models was discussed. Discussion points included:
Send community email to provide a short survey to all community about workshop topic to gauge interest. This method could help to reshape the workshop focus if needed. (see Action below)
Users Committee should provide a list of speakers as soon as possible, due to the importance of the speakers for promotion of the workshop, and the development of a funding proposal.
Chris Herbster provided an overview of the new lightning network, USPLN.– would like the data to be injected into the IDD via peer-to-peer, similar to NLDN. There are other subtle differences from NLDN. Embry Riddle has offered to be middle-group, e.g., SUNY-A is to NLDN.
Some data archiving is being performed. There was the question of decoders; it would look similar to NLDN data. USPLN is willing to provide data in NetCDF or ASCII
It was brought to the committee’s attention that there is an issue of validation of the data (QA), which created some concern to the committee. Linda Miller offered to contact the NWS to research the network and any testing that they are aware of. Anton Kruger is very interested in the data and offered to test the network.
DeSouza Award-Jim Steenburgh
The request for nominations for the DeSouza Award will be sent to the community in late July and again in early August. The presentation will be made at AMS meeting.
RICO Field Program-Don Murray
Don Murray presented a briefing on the RICO. RICO took place in Barbuda, Antiga during November, 2004-January 2005. The project is a highly collaborative project involving multi-agencies and groups within UCAR. EOL solicited Unidata’s presence vis-a-vis the IDV.
Table of Priorities
Mohan set the stage for discussion. This exercise grew out of discussion in the Policy Committee meeting in Oct 04 regarding appropriate priorities for Unidata. Unidata’s funding was less than requested, and it is necessary to identify the priorities from those proposed in the 2008 NSF proposal. . Another issue that comes into play is that NSF funding is one of several proposals funded at UPC. The UPC needs to fold activities from all proposals into the highest priorities to proceed successfully.
The Users Committee reviewed the table and made recommendations that will be shared with the Policy Committee. Jim Steenburgh offered to write up Users Committee recommendations and forward them to the Policy Committee. (see Action below)
Ken Howard, NSSL-NMQ-QPE and short term QPF (need presentation from Ken)
NSSL has been working collaboratively to coordinate a community wide research and development of hydrometeorological applications. With the advent of CRAFT and 88Ds, set up infrastructure to process the data in RT.
TDWR radar ingestion along with WSR-88Ds and Canadian network which will also provide level III data. The Users Committee was in favor of collaboration with NSSL on this project to make the data available to the community. Unidata will work with NSSL on format issues, etc., with the time frame goal of fall, 2005.
Discussion of Users Committee phone call surveys:
Recommendation that the Unidata homepage provide more information on IDV, apprising the community that IDV works on Windows with painless installation. The Users Committee will contact individuals they’ve already talked to by email to request contacts related to education at their institutions (see action below).
Action Item (joint with Policy Committee): Selected UPC staff, polcomm members, usercomm members, and members of the at-large Unidata community will work with external consultants to develop a plan to independently evaluate the Unidata program. The plan and an implementation timetable will be provided to Polcomm and Usercomm at their fall meetings.
Action Item: Unidata Support will provide announcement to community about cessation of floater
Action Item: The workshop co-chairs will provide a short community announcement regarding the workshop topic to ascertain the interest of the workshop focus.
Action Item: Gary Lackmann and and Chris Herbster, in collaboration with Joanne Graham and the UPC, will prepare an NSF proposal to fund the summer workshop. Users Committee members will work with Gary and Chris this summer to identify potential speakers and contributors for the workshop, with planning continuing at the Fall meeting.
Action Item: Users Committee will issue a a request for nominations for the Russell L. DeSouza award in late July and then again in early September (nominations due by late September) and then select an award winner at the Fall meeting.
Action Item: UPC will work with Ken Howard's group to begin distributing his radar products via the IDD for community use.
Action Item: Users Committee members will have phone conversations with two Unidata sites prior to the next meeting and contact prior sites to obtain lists of academic faculty, research faculty, technical support staff, and teaching support staff that are using Unidata products.
Linda Miller, Community Services, Unidata
University Corporation for Atmospheric Research
P.O. Box 3000
Boulder, CO 80307-3000
303-497-8646 fax: 303-497-8690