Unidata Users Committee Meeting Summary

25-26 September 2006
Boulder , Colorado

 

Members UPC Staff
Gary Lackmann, North Carolina State Univ, Chair Tom Baltzer
Eugene Cordero, San Jose State Univ Steve Chiswell
Elen Cutrim, Western Michigan Univ Ethan Davis
Chris Herbster, Embry Riddle Aeronautical Univ Ben Domenico
Kevin Knupp, Univ of Alabama, Huntsville Emily Doremire
Anton Kruger, Univ of Iowa Steve Emmerson
Leigh Orf, Central Michigan Univ Jo Hansen
Scott Rochette, Suny College at Brockport Yuan Ho
Michael Voss, San Jose State Univ Linda Miller
Tom Whittaker, SSEC - Univ. of Wisconsin-Madison Terry Mitchell
  Jeff McWhirter
  Don Murray
  Jennifer Oxelson
  Sandra Petrie
  Mohan Ramamurthy
  Russ Rew
  Jeff Weber
 

Tom Yoksas

Please Note: Proposed Spring Meeting Date: 17-18 May 2007

Meeting Agenda

Review of Action Items from 10-11 April 2006

Monday, 25 September 2006

Administrative:

Discussion:

Report Jim made to the Policy Committee at the September meeting centered on the summer workshop. Points made were:

  1. grad student representative for the Users Committee would provide a good form of professional development and would facilitate use of Unidata data and tools.
  2. momentum from workshop should keep going-summer visitor program to have faculty come to Unidata and COMET-unsure of mechanism and funding
  3. THREDDS and IDV important and will be included in spring meeting for Policy Committee discussion.

Mohan indicated 60-70 % of PolComm meeting was spent discussing strategic planning and International possibilities related to Unidata's future.  White paper on Unidata’s international activities.  Why is it important to Unidata? The Policy Committee worked on defining strategies for the next proposal.

By the spring Users Committee meeting, we should have a new strategic plan for review and comments. The Policy Committee will meet three times, due to next funding proposal.
Overarching comments from the Policy Committee indicated that Unidata was not being bold enough, and should be taking a much longer term view of proposed goals that will benefit the community. One suggestion was to consider connecting Unidata to Web-2.0 and the use of web services to create new services that could be connected in creative ways.

Director's Report-Mohan Ramamurthy (Presentation)

Highlights include:

Discussion:

GIS integration-The Policy Committee had a tour of the lab at the Univ of Hawaii site-they leveraged Unidata funds along with matching funds from the Dean which made a huge impact for their lab with funding received from the 2005 awards.

A discussion of Web2.0 at PolComm suggested that there is a need to make sure Unidata creates tools that connect to others, e.g., Google Earth. The IDD can import KML files into GE in a lightweight integration.

The question of community understanding of THREDDS was discussed. The first generation of users are data developers, but the goal is to have additional local data assets via THREDDS Data Server. Data is not getting published and shared, and the goal is to help to make that data available via TDS.

Will the TDS replace the IDD? As the data volume grows, we will need more reliance on "just in time" data.  A large number of users will need data, particularly for labs, they will need case study data - more on demand. "Save this file" will be added to THREDDS. Remote access will be important. 

THREDDS and THREDDS servers – should another equipment award round be based on THREDDS? Perhaps it's time to revisit THREDDS for the next Equipment Awards-maybe combining that with thematic case studies-provide quasi-persistent storage-NCDC, NCAR, et al is important to have THREDDS

ECMWF restrictions make IDD impossible-TIGGE, for example, ECMWF will require user accounts and the data from ECMWF will be embargoed for 48 hours. TIGGE data will be available from NCEP however.

ACTION: The relevant UPC staff will provide a THREDDs demo/presentation at the next Usercomm meeting.

Unidata Status Reports

Discussion:

GEMPAK use has expanded due to the MeteoForum and international efforts.   Some support responses are now being taken care of by international users.  Brazil is using GEMPAK exclusively-it's their favorite visualization and analysis package.

CPTEC has adopted the Unidata vision – they are sending mesonet data into IDD Brazil and running local models. Unidata IDD participants can get the data in their IDD - there are several types of data now available in the experimental IDD feed.  CPTEC is a data provider for THORPEX-TIGGE and will provide their global model.

LEAD has a large database on the Unidata testbed that holds seven LEAD datastreams, so the community might want to take advantage of that. The data are archived for 120 days.  It’s a valuable resource for the community.  It’s not widely advertised due to the fact that it is a testbed, but it could be useful for backfilling.  It is a 40TB system, and LEAD is creating a table on what data exists and how its accessible. There was a question of workshop evaluations of LEAD-Seige interface to the LEAD system.  A large scale evaluation has not been published yet.  Details on architecture, and it should be ready by spring semester.

The question of two data formats for netCDF was posed? Not all of netCDF can be represented in HDF4 or 5.  The issues are complex.  NetCDF is a simpler data model and has shared dimensions for grid.  There is no good way of doing this in HDF5 yet.  The Common Data Model (CDM) is an attempt to make it easier to have a common data model with HDFand netCDF, and data can be accessed through both of these interfaces through the CDM.  Through the GEO-IDE effort (NOAA support), GOES area files will look the same, and the data looks the same through this common interface.  HDF5 data is also available through the same interface.  All versions of netCDF are backwards compatible.  HDF4 is not used much for new datasets, but there is much valuable data in HDF-EOS archives based on HDF4.  HDF5 provides the advantage for high performance computing.

Case Study Project-Persistent data storage needs to be provided for the case study project and for the LEAD project. It is envisaged that a drag and drop from one repository to a longer term data repository when needed might be doable. The UPC is not involved in long term repository. The policy has been a conscious decision made that theUPC will not become a long term repository. If there is a need, it would become a policy issue and discussed at that level.

Users Committee site phone calls - It is important to continue the site contacts. The list was circulated for sign ups. Try to get technical contact names as well as department heads and users.

The Users Committee discussed the term length for a student representative for the Users Committee. This idea, spawned from the summer workshop, was discussed at the Policy Committee and now the Users Committee needs to create the process to move forward with a solicitation to the student community. It was agreed that the student representative would be treated as an equal member with voting privileges. The student needs to be a graduate student and the term length would be two-years. A nomination process would be followed and would include name, email, and short abstract with background and interest in Unidata. The student would be encouraged to represent Unidata, would view this as an opportunity to engage in Unidata activities. It would also be considered professional development, and the student would help spread the word about Unidata's mission, etc.

Action: Gary Lackmann will develop an invitation for nominations for grad student rep on UserComm Conditions:
            Must be nominated, graduate student, 2nd year or later, and use Unidata software
            Must hold an interest in Unidata (and explain why they want to be involved)

There was a unanimous Users Committee vote in favor of this resolution (that a grad student rep would be selected based on the above conditions).

An action taken from the Panel Discussion on the last day of the workshop was a recommendation to move forward and research the idea of curriculum materials to accompany presentations made during the workshop. The idea was to invite interested faculty members to work with an instructional designer with summer salary (stipend) to be included, and to provide modules to accompany presentations, such as those made during the summer workshop.  

After extensive discussion, it was determined that this activity should be aligned within COMET's purview and not Unidata's. COMET does a great job with module development and might consider such an activity.

Engaging the Community in IDV development -Don Murray and Jeff McWhirter

Don provided a brief IDV overview, beginning as the MetApps project 1998-2003. Even then, the Users Committee was deeply involved through the creation of the MetApps Task Force, a sub-committee of the Users Committee. The committee provided use cases on which to develop and test the software.
MetApps was born out of the desire to have a platform independent application that would ultimately have many of the characteristics of GEMPAK and McIDAS, and include a gridded data viewer that would provide both 2-D and 3-D versions. Several prototypes were built, including the gridded data viewer, sounding applications, image viewer and surface observations viewer. In 2003 the name was changed to Integrated Data Viewer with 1.0 released.

The current release of the IDV is Version 2.0 released in August 2006. The development is now driven by the IDV Steering committee. Speed issues were the focus of this release which provides greater performance. The user interface has changed, and menus are more user friendly. The IDV dashboard aggregates windows into one main window. New icons have been integrated and caching mechanism has been implemented.

Discussion:

There will be a "developer" training workshop in 2007. There will be continued coordination with THREDDS. The community should be assisting in defining the next datasets for the IDV. Support for modeling efforts like TIGGE and other ensembles will be completed.

Some users just want a tool that they can "walk up and use". The response to that is "IDV Lite" which has been discussed, but not yet developed. The IDV developers were encouraged to go forward with that idea. Some said that "simple is good!" It was also suggested that an IDV web site be created for user contributed plugins and bundles.
           
Newsletter

Action 1 from the April Users Committee meeting pertained to engaging the community in reading the e-letter. Jo Hansen provided a graphic of newsletter readers and the graph indicated substantial number of readers-average of 300 per day, with the most popular being April 2006 workshop sign up issue. The newsletter is now available via RSS feed, and there were 454 who subscribed to that. This number does include crawlers.

Discussion:

It was suggested that more visual content and less text be used. It was also suggested that a PDF version be created for printing purposes. This would allow a cut and paste from the newsletter for talks and presentations by Users Committee members if so inclined. The question of how many undergraduates embraced the newsletter is not resolved, because we cannot identify who the students are.

ACTION: A printer-friendly version of E-letter will be made available for future releases.

A student UserComm representative should be able to help us advertise. In fact, the idea of creating a <studentsonly> email list was mentioned.

 

Unidata assessment process

The assessment process is moving along on schedule. The Focus Groups from the Unidata workshop and the COMET workshop were very well attended. There were five groups which broke into one group grad students, three groups of faculty and one group of "others." At this point, the survey, software projects interviews and Focus groups have been completed, leaving the phone interviews and alignment table remaining. The case studies and narrative will be the final actions to be completed before 31 December 2006 when the process will be completed.

Summer Workshop Discussion

The Users Committee expressed the desire to feature the workshop link on the Unidata home page, especially after populating it with all of the presentations that were made during the workshop. The Committee said that the presentations could be useful to a broader audience.  

BAMS article

It is important to include the focus of the workshop (remember that the focus changed from the original proposal for the workshop).  The intent was that panel discussion would address the three issues, which are

  1. Reflect on workshop content
  2. Consider ways to advance teaching with models
  3. Materials/results: ways to keep up the momentum
Some ideas:

ACTION: BAMS workshop paper- Leigh will draft, send to Usercomm by Monday, 10/2.

ACTION: Link Summer 2006 Workshop page more prominently from main Unidata pages. Prepare follow up message to presenters to share web links to codes. He will work with Sandra Petrie to send it.
            Prepare “follow-up” message to presenters to share www links to codes
            [We still need to finish thank-you letters!]

ACTION:  Sandra Petrie will create a participants workshop list for use by the committee.

ACTION: Edit the summer workshop page to include all workshop information and develop an area for contributed workshop presentations to be added. Provide the workshop page in a prominent area on the Unidata Web site (Petrie).

ACTION:  create a forum for communication and outcomes of summer workshop.

The committee needs to finish the thank you letters to workshop presenters.

Tuesday, September 26, 2006

Equipment Awards-Sandra Petrie and Mohan Ramamurthy

Mohan and Sandra provided an overall status of the Equipment Awards. The Users Committee should set the theme. Most of the solicitations have been thematic. The timeline for the formal announcement will be in December-January timeframe. Fliers will be provided at the AMS meeting in San Antonio.  There was a lot of discussion surrounding the next theme, which included Case Study data sets used with THREDDS, but it was decided that it should wait until next year when additional capabilities are completed and available.

Discussion:

Modernizing labs to be used in atmos sci courses to make use of new tools, LDM, IDV for example, is important, but generally costs more. Another consideration in the past has been how will the case benefit the Unidata community? If a university is unable to modernize labs, it benefits the UPC to get the lab up to speed.
Example: benefiting the community via a server for the IDD. The question of how much it costs to get a university into the system was raised. It's about $5K. Cost matching is not required from a university, however many participants approach their department head for the cost matching and it works. The next theme will be the dual theme of:

Mohan, Terry, and Sandra will craft the RFP.  They will create priorities and be sure to provide those in the RFP. Some Users Committee members will be asked to participate in the review panel. Obviously, the review panelists cannot be in contention for the Equipment Award. The next Panel Review will be held on 17 May 2007.

ACTION: Equipment award themes:  (i) Upgrade labs to IDV-compatible systems, and (ii) community expansion.  [Next time:  THREDDS servers?]

Case Study-STORM fund - Jeff Weber and Mohan Ramamurthy

The STORM fund (an internal UCAR/UOP fund) was granted to Unidata to work with a module expert (previously affiliated with COMET), Dr. Brian Etherton, and his student from NC-Charlotte, and THREDDS catalogs utilizing OPeNDAP and ADDE server technologies and visualized using the IDV. The THREDDS server and THREDDS Data Repository will allow upload of datasets to create case studies and pave the way to add other data, such as societal impacts, etc. THREDDS repository will become richer by sharing curriculum, lab exercises, and unique views will all be available. The particular prototype case study is
looking at climatological aspects of hurricanes. The STORM funding is meant for seed money. There will be presentations about this project at both AGU and AMS. The hope is to create a community where they can contribute aspects of the data set and enhance the data set. It becomes a living case study.

Data persistence issue-how are educational materials packaged with it?  The hope is that this will become ubiquitous and others will serve case studies.

NOTE: Next year’s equipment award-consider the case studies with THREDDS.

The first prototype case study is scheduled for spring of 2007. 

Telephone survey calls

Several committee members missed their survey calls for this meeting. With new members, a new Chair, and the summer workshop, it is understandable that some were unable to make the calls, but we hope committee members will provide summaries of the calls made for the spring 2007 meeting.

ACTION: provide phone calls to university members —try to talk to people besides the sysadmin.  Initiate contact via email so that they will be aware
            Unidata needs to go through the list to see who is using Unidata

ACTION: post the list and have committee members post their calls

ACTION: Try to do better with phone calls to community universities.  Jo Hansen will send out a reminder email to the committee with their contact info listed.  Also, the web page link that contains contact information will be changed to link to the .xls file.

DeSouza Community Award

ACTION:  DeSouza discussion: end of October-nominations due.  Send announcement to community
Send a note to UserComm with possible candidates with a short description.

Lightning Data-USPLN-Chris Herbster

Chris Herbster is still working with USPLN to make the data available to the Unidata community. Unidata will provide an announcement regarding the service when it becomes available.  The coverage includes South America.

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