Unidata Program Center staff and our Users Committee members have been hard at work planning the upcoming Unidata Users Workshop, which will be held 10-14 July 2006. Not to be confused with our annual software training workshops, the users workshop is always built around a theme that we hope will appeal to a large segment of our community. This year's theme, Expanding the Use of Models as Educational Tools in the Atmospheric & Related Sciences, has attracted an impressive roster of speakers to excite, inspire, inform, and motivate thinking and learning. We look forward to this opportunity to meet our community face to face, share resources, and open a dialog on current issues.

Kicking things off will be keynote speaker, Kelvin Droegemeier of the University of Oklahoma's Center for Analysis and Prediction of Storms. Dr. Droegemeier's presentation on the LEAD project will explore how the project "enables scientists, students, models, and sensors to interact dynamically with weather." Two related labs will further demonstrate the power of LEAD's technology.

We're very excited to tell you that the 2006 workshop is chock-full of hands-on lab presentations. We hope this will give attendees a chance to really experience the way our community is using models in the educational setting. You'll find a listing of our distinguished speakers and lab presenters on the workshop homepage.

Though we do require a $50 registration fee, funding from our sponsor, NSF, has enabled us to provide most guests with lodging, ground transportation, breakfasts, lunches, an opening reception, and an evening barbeque at UCAR's Mesa Lab, nestled above Boulder in the foothills of the Flatirons. Some funding for airfare is even available to a limited number of graduate students to broaden our attendance. (Inquire at support-workshop@unidata.ucar.edu. Please include a brief description of your interest in the workshop.)

The 2006 Users Workshop is right around the corner. The current registration deadline is May 15th, so please register today while there's still space! We're looking forward to sharing Boulder's beautiful summer with you!

DataFed: Federated Data System for Air Quality Management and Science

by Rudolf B. Husar and Stefan R. Falke, Washington University, St. Louis, Missouri

Recent developments in surface and satellite sensing technologies now allow real-time observation of air pollution. Chemical models driven by meteorological transport models can forecast the near-term pollution pattern. The instantaneous "horizontal" diffusion of observational and model data via the Internet permits, in principle, the delivery of the right information to the right people at the right place and time.  Nevertheless, air quality analysts face significant hurdles that hamper the effective use of air quality data. These include finding, accessing, filtering, aggregating, fusing and delivering high grade knowledge to a variety of users. 

The federated data system, DataFed, aims to reduce the above hurdles in dealing with air quality-relevant data. In designing DataFed, we assume that new data sets and data processing services will continue to emerge spontaneously and autonomously on the Internet. The key role of DataFed is to mediate the flow of data between the emerging providers and users. Specifically, it (1) facilitates registration of the distributed data; (2) homogenizes, on the fly, all the data into a physically-based global data model (space and time); (3) supports interoperability through the use of standards-based protocols (OGC, XML-SOAP); (4) provides a set of basic tools for data exploration and analysis. Example data providers include the AIRNOW real-time monitoring project where data are collected by the States, combined by the federal EPA and used for informing the public (figure above, click to enlarge). However, in addition to its primary use, the hourly ozone and PM2.5 data are also placed in the data pool for reuse by the broader community. Since 2004 there are over 50 distributed air quality-relevant data sets accessible through the federated virtual database. The web-based tools include a catalog for data access, data viewers, view editors and data consoles for synchronized data browsing (figure below, click to enlarge).

Currently, the promise of the Internet standards for easy data sharing (e.g. netCDF, OGC Services) and Service-Oriented Architectures (SOA) for building flexible applications from reusable components is only partially fulfilled. Our group is actively collaborating with Unidata and others to develop and to test the new agile Internet technologies and architectures (e.g. the GALEON Interoperability Experiment). We hope that this collaboration will also yield a closer interaction between the meteorological and air quality analysis communities.

Editor's Note: Use of Unidata's Integrated Data Viewer, aka IDV, is growing. When we learned that a long-time community member who's used another visualization and analysis tool for many years was migrating to the IDV, we wondered why and asked him to provide a write-up about the topic. Dave Dempsey's (SFSU) explanation is below.

by Dave Dempsey, San Francisco State University

We decided last year to begin migrating to the IDV for classroom use, based on a long-awaited convergence between the development of the IDV and our ability to use it.

In the early 1990s, we adopted WXP for classroom use, when it was supported by Unidata and the IDV wasn't yet on the radar screen. WXP seemed best suited to the needs of our program at the time. All of the available packages were command-line driven; a GUI wasn't an option.

Command line interfaces are not particularly easy to use. To make WXP programs easier to use for our curricular purposes, I began developing Unix scripts that simplified the WXP command-line interface. By the time GUI interfaces became available for the competing weather-graphics software packages, I'd written (and debugged!) probably 60,000 lines of code and developed a required meteorology major's class that in part taught students how to use the scripts (along with WXP itself).

In 1995 I got a small SFSU grant to set up what I dubbed the California Regional Weather Server (CRWS) (http://virga.sfsu.edu).  The idea was to generate GIF versions of some of the maps and images that I'd been developing for in-house curricular use and make them accessible to the general public, updating them automatically via cron jobs. There weren't a lot of weather maps and images specific to California available at that  time, so I hoped to help fill that niche. WXP made this possible, and it remains essential to the CRWS's ongoing success.

Classroom use is another story. Many of today’s students have grown up with the Web and don't have much patience for command-line interfaces. It's been just a matter of time before we switched to software with a GUI. However, to switch to the Java-based IDV, several conditions had to be met:

1. We needed to acquire enough computers capable of running not only the IDV but also the variety of other software we use in the classroom.
2. The IDV had to become both comprehensively and computationally efficient enough for classroom use.
3. We had to learn how to use the IDV well enough to incorporate it into our teaching.
4. We had to have confidence that Unidata would develop and support the IDV over the long term.

In 2002 we were able to buy 18 laptop computers to equip a 36-seat classroom that we renovated to permit any combination of traditional lecturing, networked computing, and small-group collaborative work.  For historical reasons most of our Department of Geosciences faculty (2/3  geologists, 1/3 meteorologists and oceanographers) have used Apple Macs,  and Macs running Apple's Unix-based OS X proved friendly enough for students more familiar with Windows PCs, so we bought then state-of-the-art, 15", 800 MHz, Mac Titanium G4 laptops.

By early 2005, Apple finally had an implementation of Java that could  support nearly all of the IDV's features. In summer 2005, we upgraded the memory in our laptops to 1GB of RAM, which is plenty for the IDV. Moore's Law means that our laptops are now about 6-8 times slower than machines available today, but the IDV's efficiency had improved sufficiently that our laptops were just fast enough to make using the it in our classroom feasible.

In summer 2005, I attended an IDV training workshop at Unidata.  Without that training, I don't think I could have taught our students and my colleagues to use the IDV adequately. With the training, I felt I could. Moreover, although I'd known something about the IDV beforehand, I hadn't fully appreciated how much more it had become than just a weather graphics package with a GUI.

For example, the IDV can access data residing not only on the local machine but also on remote servers, whereas WXP can access only locally available data. This means that students can use the IDV to complete part of their class work at home if need be, something that is much harder to do with WXP.

The IDV has some analysis and display capabilities that WXP lacks, some of which are spectacular. The IDV's ability to create (and animate and rotate) 3-D visualizations of model data, radar data, aircraft track data, and the like is particularly eye grabbing. It can display GIS shapefiles, which opens up a vast world of GIS-like possibilities. The User's Guide reveals a number of other distinctive capabilities.

The IDV has its shortcomings, of course. However, Unidata has dedicated staff to the IDV's ongoing development, and because Unidata and the IDV's developers are sensitive to the needs and concerns of academic users like us, many of the IDV's remaining shortcomings will likely be addressed. To me this makes a shift to the IDV feel like a safe bet over the long haul.

Emboldened by that Unidata training workshop, I decided to begin our switch to the IDV. We were finally ready to make the leap.

Last fall I modified a required majors course, in which I used to teach students to use WXP, to use the IDV instead. I relied heavily on Unidata's training workshop materials as well as the Users Guide. I'd found the IDV's learning curve a bit steep in the beginning (though no more so than other packages), but the students seemed to scale it more easily than I, which was a bit disconcerting!

We discovered some Mac-specific and other bugs, but the IDV developers worked with us to fix them quickly. Dealing with those bugs detracted from the IDV learning experience but it also made us feel like contributing members of the community of IDV developers.  Don and Jeff paid us a half-day site visit during their attendance at the annual AGU meeting in San Francisco last December, and they boosted our understanding of the IDV and noted carefully our wish- list of features and alternative user interface designs. Their responsiveness was welcome.

It'll be a while before we build up the experience and the body of course-specific IDV "bundles" needed to integrate the IDV into our curriculum more fully and effectively, but we're in for the long haul and that will eventually happen.

Meanwhile, I just love making animated, global 3-D jet stream visualizations with a 3-D temperature surface plotted below it. It does just what I'd hoped it would do when I wrote a "Use Case" for it back in the early stages of the IDV's development

by Linda Miller, Unidata Program Center

By now many of you have probably received an on-line survey from Nelson Consulting concerning the impact on people, ideas, and tools that the Unidata Program has had on your teaching and research.  The survey is only one component of a multi-part, independent assessment of the Unidata Program.  As part of that assessment, Nelson Consulting will also be conducting Focus Groups during the Unidata Users Workshop, Expanding the Use of Models as Educational Tools in the Atmospheric and Related Sciences, scheduled for the week of July 10, 2006.   In addition, some of you who have had a long association with Unidata will be contacted by the consultants for telephone interviews on the transformational aspects of Unidata on research and educational activities through the years. 

The results of the assessment will be used to identify future directions and inform Unidata on important themes to consider for the future.  The results will inform planning activities and help Unidata to continue to fulfill its mission of providing data, tools and community leadership for enhanced Earth system education and research. 

If you haven't already completed the survey, please do so by May 15th.

Thank you for your participation in these important assessment activities on behalf of Unidata!

Recent news item on the assessment.