Unidata Users Committee Meeting Summary
15-16 April, 1999
(DRAFT)
Members:
Jennie Moody, Chair
Rich Clark
Dave Dempsey
Steve Koch
Michael Morgan
Clint Rowe
Tony Rockwood
Jim Steenburgh
Doug Yarger Ex-Officio
Members:
Jim Moore, Policy Committee
Tom Whittaker, ATAC
Visitors:
Greg Byrd, COMET
Charlie Murphy, Unidata MetApps
|
UPC Staff:
Sally Bates
Steve Chiswell
Ethan Davis
Ben Domenico
Steve Emmerson
Dave Fulker
Joanne Graham
Linda Miller
Don Murray
Russ Rew
Jeff Weber
Donna West
Mike Wright
Tom Yoksas
|
Administrative Matters:
The proposed date for the next meeting is September 30-October 1, 1999.
A review of the action items:
Action 1:
Fulker will describe Unidata's Java transition in his presentation
to the AMS Head & Chairs
October 8th; there will also be an article on this
in the next issue of the newsletter.
Done
Action 2:
Rockwood will work with Bates to publish an article in the
next newsletter on his transition to
x86.
Done
Action 3:
The Users Committee will keep an eye on NT developments.
(John Caron will provide a briefing on NT platforms and developments
at the next Users
Committee meeting)
Action 4:
Clark and Moore, working with Fulker, will consider how
best to present the GEMPAK/GARP
concerns to the NWS.
Status: Reviewed by Committee
during the meeting.
Action 5:
A demonstration of the FX-net system will be on
the agenda for the next meeting.
Done
Action 1:
John Caron will provide a briefing on NT platforms and developments at the Sept
30-Oct 1, 1999 Users Committee meeting.
Director's Report - Dave Fulker
The following points were made:
- The funding outlook has improved
- documentation impacted in interim
- Case studies continue to be added to CODIAC (17 now available)
- Final OS2 McIdas release made by Don Murray
- SuomiNet proposal submitted for MRI funds to NSF
- New data sources include:
- NOAAPort
- NCEP- CONDUIT
- Level II NEXRAD data via CRAFT
- 6-minute Profiler data from Forecast Systems Laboratory (FSL)
- ACARS data from
FSL
- McIDAS ADDE - efficient remote data access
- Current IDD statistics:
- 600 MB - NWS & WMO data
- 300 MB - GOES images (Univ of Wis)
- 10 MB - Lightning (SUNYA)
- 40 MB - Profiler (FSL)
- 5,000 MB - NCEP high resolution (via Goddard)
- 3,000 MB - NIDS all radars (via WSI)
- 100 MB - WSR-88D full-resolution-one radar (via CRAFT-Univ
of OK, Univ of WA)
- Progress with Java includes:
- two releases of prototypes (using MetApps Task Force input)
- creating Use
Cases
- demonstrating component architecture built on Java Beans
- applicability of Vis-Ad classes for Unidata needs
- use of shape files for background maps
- MetApps Task Force members are testing and providing feedback on applications
- Release of NetCDF in Java
- uses Java only (no C)
- uses Java multiarray abstraction
- support remote data access
- includes change notification
- efficient sub setting and transposition
- utilizes Java Remote Method Invocation (RMI)
- created by Glenn Davis
- PQueue reader for direct access to real-time data and event (new data)
notification
- Vis-AD gaining new users using remote collaboration, incorporating Unidata
advances
- John Caron, Glenn Davis, Steve Emmerson, Russ Rew - developers
- PAGE progress
- Developing Table of Contents and Index for educational materials available
in geosciences
- submitted two proposals for funding from NSF
FSL Demonstrations and Discussions
The Users Committee spent the remainder of the morning visiting FSL at the new
NOAA offices in Boulder. Demonstrations and discussions revolved around
the Java workstation developments of:
FX-Net - provides Internet access to WFO-Advanced workstation products
on a PC workstation. The client application provides an AWIPS-like user
interface with the processing tools required to interact with, e.g., satellite,
radar, and model imagery.
FX-Linux - a pilot effort in preliminary stages of porting FX-Advanced
software to a PC/Linux using Swing, platform. Plans to accommodate changes
in technology and user requirements.
FX-Connect - development that accommodates the interaction of forecasters
located at different locations through a graphical user interface, allowing
some interaction between the FX-Connect and FX-Advanced (AWIPS-like) workstations.
Uses Java's RMI and object serialization.
Discussion items included:
- above projects are built on NetCDF
- wavelet tranform is the compression technique used for FX-Net
- Product Generator is a unique feature that was of interest to the group
- all FSL-developed code is freely available
A brief discussion of the FSL visit led into the MetApps items. Some issues
were:
- FX-Net had a nice user interface
- the length of development time for Unidata to have the MetApps ready for
the community
- what is the interim situation with GARP, GEMPAK, N-MAP, N-TRANS
- certain functions are currently missing from packages--how will these
functions be addressed?
- how does the community progress while waiting for the Java MetApps
MetApps Task Force Report - Charlie Murphy & Michael Morgan
MetApps currently consists of eight (8) community members who are charged with:
- providing specs for user requirements
- testing and feedback to Unidata developers
- providing a long-term vision of academic needs (5-10 years)
The only meeting of the Task Force took place in July, 1999. Now the Task
Force is a "virtual committee." They are employing "use case" techniques
to provide application recommendations to the UPC developers. Two main use
cases have been created; surface observing viewer, and interactive Skew-T.
These applications are using canned data and there is a general feeling that these
applications are used with real-time data, there will be more enthusiasm and participation
within the Task Force.
Discussion:
Java offers the flexibility of component-oriented software. It also offers
the platform independence feature so important to serve a diverse community
of users. An advantage of creating the applications is that Unidata will
own the applications and be able to modify them according to user needs. Unidata
will create the components, but there will also be opportunities to integrate
contributions from the community.
The level of interaction began with almost everybody on the Task Force participating,
but has gradually decreased. Email discussions are difficult to facilitate,
which has led to the preliminary "prototype" testing of an Internet collaboration
tool recommended and implemented by Mike Wright, who has been rehired in a half-time
position at Unidata. Among the advantages of using the tool is preserving
the discussions taking place among the members. The testing will expand
to the entire Task Force when the final recommended format and templates are
in place.
The collaboration tool, along with additional members for the MetApps task
force should facilitate future activities.
Action 2:
Charlie Murphy will augment the MetApps Task Force with additional community representatives.
AMS Annual Meeting 2000 Special Session - Linda Miller
A brief discussion of the
AMS Annual Meeting
2000 Special Session . This session will be conducted on Wednesday,
January 12, 2000. There will be room for 15 papers to be presented during
the morning session. Unidata university sites are encouraged to send their abstracts
to the AMS-IIPS by July 1, 1999, for consideration in the special AMS IIPS-Unidata
Session.
Russ DeSouza Award - Sally Bates
The Resolution below was approved by the Policy Committee. Sally provided
guidance on the Users Committee responsibility to move forward with their nomination
for the first "Russ DeSouza Award."
Resolution 3 (passed unanimously) by the Users Committee, October 1-2, 1998:
The Users Committee requests the Policy Committee
give the Users Committee full authority
for making the Russell DeSouza Award. Current
Users Committee members will be excluded
from consideration. The Users Committee will
seek nominations from the community, discuss
the nominations in an open meeting, and render
its decision in executive session.
A closed session of Users Committee members-only convened and agreed upon
a candidate for the award. After all participants reconvened, issues such
as:
- method of the delivery of the award
- recognition of recipient, e.g., newsletter article, announcements
- suggestion of having two plaques (one for the UPC with the names of the
recipients, and one for the
recipient was made
Action 3:
Linda Miller will check with the AMS to see if the Russ DeSourza Award presentation
can be made during the Special AMS 2000 Session.
Action 4:
Sally Bates will create a special article for the Unidata Newsletter about
the significance of the award and the first recipient and consider other announcements
Rich Clark extended an invitation to the Users Committee, Policy Committee
and UPC staff to attend a ceremony at Millersville where the dedication of the
new Meteorology building will take place. Many of us contributed toward
the Russ DeSouza Weather Station. The ceremony is scheduled for October 30,
1999.
Data Issues Discussion:
- Dave Fulker summarized NCAR/SCD-Unidata collaborative activities.
There are two activities that the group has focused on. One is the access,
via FTP, to the archive of Unidata data that is being stored at SCD.
Another, very preliminary, long-range activity is building a DODS GRIB server
at NCAR.
- Tom Yoksas reported on the planned removal of non-imagery products from
the Unidata-Wisconsin Channel on July 1, 1999. Announcements have been
sent to the community and reminders will continue to be distributed.
- CIMSS and CIRA have data sets that have been offered to Unidata, including
satellite image data used for AWIPS, with soundings, hourly, 12 km.
Also fog products being produced at CIRA will eventually be offered.
- ACARS data will be provided by FSL via LDM, through a subscription service,
similar to the SUNYA lightning data. This service should be in place
during May, 1999, after FSL is settled in their new building.
- NIDS Floater management was discussed. The committee agreed that
the management was going well considering it is not an "operational" function,
and Millersville is coordinating coverage with students during off-hours of
the day, as well as handling it during normal working hours.
Other additional products were discussed in a general manner. If there are
specific products Unidata should research, the Committee should contact the UPC
to check the availability of the data.
GEMPAK Discussion
A question of bridging the gap between GEMPAK and GARP to Java was discussed.
There are functions in the current GEMPAK and GARP applications that were described
as desirable for the community. A discussion about the functions led to
the following action item.
Action 5:
A list of current application capabilities of GEMPAK will be created by Steve
Koch. The list will be sent to the Users Committee and the UPC for review.
Action 6:
Steve Koch, as the lead, in conjunction with the Users Committee, will create
a specific list of requirements for GEMPAK development to be considered by the
UPC during the interim period leading to the Java MetApps.
Action 7:
The UPC will clarify strategy on the GEMPAK issue and report back to the Committee
after receiving the information noted in Action 5-6.
Unidata Summer 2000 Workshop
The theme, coordination and organization of the summer workshop was discussed.
This workshop is intended to be a co-sponsored workshop by Unidata (NSF) and COMET
(NWS).
Objectives:
- educate the user community on data/information available
- engage users in how they are using the data
- include substantive education, pedagogy and research through techniques
and tools being used in the community
- foster development of ideas/materials to enhance application of data in
instruction/student research (through a set of lectures and labs)
Discussion:
To attract broader participation for the workshop, consider the broader scope
discussed in "Shaping the Future", and the "Geo 2000" reports by the National
Research Council. With a broader representation, there could be several
tracks being followed during the course of the one-week workshop.
Case studies could be used to create new case studies and/or to augment current
case studies. Use of CONDUIT and
CRAFT
data could be considered for creating case studies, along with other data sets.
There is a need to de-emphasize lectures and emphasize labs. Activity or lab
should already be created prior to the workshop. Need to identify
skills needed for the workshop; if people do not have the skills, provide a
half-day session on the Saturday or Sunday prior to the workshop for training.
"Key" people would be identified to meet in Boulder for one week prior to the
workshop for preparation. It was agreed that "the process is the product"
that participants would leave with.
A suggestion for the workshop title was: "Shaping the Future: Unidata
Users as Leaders". No decision was made concerning the title.
Summary of Workshop Discussion:
- Tony Rockwood and Doug Yarger agreed to co-chair the workshop
- focus of the workshop will be directed toward Unidata applications
- have prepared materials before the workshop using "key" people prior to
the start of workshop
- community user-oriented workshop model (hands on, lab-oriented)
- thematic approach - "Geo 2000" or "Shaping the Future" (earth systems approach),
not exclusive to meteorology
- examples of community use of Unidata data and applications
Logistics include:
- dates of the workshop will be June 18-23, 2000
- maximum of 81 participants
- lodging will likely be at the University of Colorado dorms
- conference rooms, auditorium, and the COMET classroom will be reserved
at FL2
- buses will be provided for transportation
- night session was suggested for discussion of MetApps - including user
requirements
Wrap Up - Miscellaneous
- An informal survey for the training workshop scheduling was conducted by
Tony Rockwood. The majority of the respondents preferred summer workshops.
The summer workshop is schedule to begin on July 28-August 7, 1999.
Don Murray is the contact.
- The use of Users Committee notebooks was discussed briefly. There
was no concensus by the Committee on this issue. Suggestions included:
- placing all reference materials and URLs (links) in one document for
printing at the university site
- continue providing the information in the notebooks without sending the
3-ring notebook
- An LDM special workshop will be provided for NOAA groups using the LDM.
The objective is to train one person at each site using the LDM. That
person will train others who use the LDM at the site. This is a one
time situation, based on the CONDUIT project and access to high resolution
data sets.
Action 8:
The Users Committee recommended that Unidata send the letter regarding GEMPAK
(N-AWIPS) to the NWS acknowledging its usefulness and value to the academic
community.
Action 9: Tony Rockwood and Doug Yarger will be responsible for creating
and coordinating the community workshop 2000 announcement and selecting members
for the workshop sub-committee.