Dr. A.E. "Sandy" MacDonald, Director
(303-497-6378)
Web Homepage: http://www.fsl.noaa.gov
Dr. Russell B. Chadwick, Chief Engineer, 303-497-6318
Ming Ge, Guest Researcher, 303-497-6021
Dr. Joseph H. Golden, Meteorologist, 303-497-4098
David Himes, Senior Software Engineer/SOS Team Leader, 303-497-5447
Dr. Jin-Luen (Jim) Lee, Meteorologist, 303-497-6097
Dr. Thomas W. Schlatter, Chief Scientist , 303-497-6938
Julie D. Singewald, Secretary Office Automation, 303-497-6818
(The above roster, current when document is published, includes
government, cooperative agreement, and commercial affiliate staff.)
Address: NOAA Forecast Systems Laboratory Mail Code: FSL
David Skaggs Research Center
325 Broadway
Boulder, Colorado 80305-3328
FSL strives for excellence in four major areas: 1) bringing new atmospheric observing systems to maturity, 2) developing and improving regional mesoscale
models for detailed weather predictions, 3) investigating parallel computer architectures as a vehicle for handling the huge computational demands of
environmental models, and 4) developing environmental information systems for a variety of customers, within NOAA and outside.
In the past year, FSL continued to gather atmospheric observations from disparate sources in the service of local forecasting and regional modeling. For example,
the number of Cooperative Agency Profilers, most of them sampling the boundary layer, has grown to over 60. The number of ground-based GPS sites estimating
total column water vapor now exceeds 200. More than 5,000 new surface mesonet observations have been added to the FSL hourly collection. After performing
quality control checks, FSL forwards all these data to the National Centers for Environmental Prediction (NCEP) for use in operational models.
FSL is developing plans for a global observing system with major in situ components that can settle the controversy about long-term tropospheric warming,
monitor natural and anthropogenic atmospheric constituents, and sample the ocean depths. This system would meet requirements for climate monitoring and
prediction and also provide much needed calibration of satellite observations in remote locations. The first phase of this program, called "Pacific Plus," would
employ a combination of unpiloted aircraft, altitude-controlled balloons, and ocean buoys to sound the atmosphere and water across the length and breadth of
the Pacific.
Gauging the contributions of various observing systems to forecast accuracy is an important activity. FSL recently completed the modeling and computing
infrastructure for performing such impact tests. Because commercial airlines, which supply nearly 100,000 temperature and wind reports per day, are financially
strapped, and the continuity of the NOAA Profiling Network is threatened, FSL conducted special impact tests, demonstrating that aircraft and profilers each
contributes substantially to forecast accuracy and that these data are highly complementary.
Several FSL accomplishments in modeling are noteworthy. Three-dimensional variational analysis became operational in the 20-km version of the Rapid Update
Cycle in May 2003. This opens the door to assimilation of many new sources of observations that were previously difficult to accommodate. A Developmental
Test Center is being established in Boulder. FSL, NCAR, and NCEP will contribute staff and share computing resources. The initial focus of the Test Center will
be on the development of the Weather Research and Forecasting Model, destined to become both an operational model and a research vehicle for the larger
modeling community. The Federal Highway Administration supported a collaborative effort between FSL and NCAR to field test a model-based system for
snow plow operators that helps them decide when to plow and what chemicals to spread on the road surface. This Maintenance and Decision Support System
was successfully tested in Iowa in February and March 2003.
In November 2002, FSL accepted a major upgrade to its high-performance computing system,comprising 768 nodes with dual Intel Pentium processors rated at
2.2 Ghz. Also during that month, it was ranked #8 on the Top500 List of the World’s Fastest Computers, and now serves more than two dozen external users
representing most of the line offices within NOAA.
Capitalizing on major development work at FSL, the National Weather Service began installing two Linux-based workstations at each of its Weather Forecast
Offices around the country in January 2002. Eventually these workstations will replace the Hewlett-Packard hardware which inaugurated the Advanced
Weather Interactive Processing System (AWIPS) era. The highly robust Linux was developed in an open-source environment and runs on nonproprietary
hardware.
FSL developed the FX-Net workstation as a low-cost alternative to AWIPS workstations. FX-Net can deliver fairly large datasets primarily because of wavelet
compression techniques developed in-house. FX-Net has become the workstation of choice of the National Interagency Fire Center in Boise, Idaho, and at 11
Geographic Area Coordination Centers throughout the country. Meteorologists used FX-Net close to home last summer in fighting Colorado wildfires during
an unprecedented drought.
For more information on the above items and many more, I hope you will browse through the following pages. We are eager to share with you our
contributions toward NOAA research and technology transfer.
Figure 2. Students at two different locations, Washington, D.C. (top)
and Broomfield, CO (bottom), learning about Science On a Sphere
TM.
(NOAA Photos by Will von Dauster, FSL.)
Nearly any global data can be displayed on Science On a SphereTM, including the weather, climate, geology, images of solar
system bodies, or any type of geographical information that covers a large portion of a planet’s surface. NOAA’s mission to understand and predict changes
in the Earth’s environment can derive special benefit from the ability to present its data with a geometry that corresponds to that chosen by nature. In many
ways, Science On a SphereTM provides an ideal way to educate the public on many important issues, both environmental and
economic, that face NOAA, the United States, and the entire world.
Accomplishments
Development of Science On a SphereTM has proceeded in phases at FSL for the past two years. Phase 0 was a proof of concept that
demonstrated the idea of using computer technology, video projectors, and a large white sphere to create the illusion of a planet rotating in space. Phase 1 built
on and improved the underlying concept to create striking, animated images of the Earth’s atmosphere, land, and oceans. Phase 2 development of Science On
a SphereTM is proceeding and will produce a mature, robust visualization system that can be made commercially available for
educational and scientific use across NOAA, K 12 schools, museums, science centers, and other public venues. A patent application for Science On a
SphereTM has been submitted on behalf of NOAA.
The initial Phase 0 prototype was developed in the summer and fall of 2001, using an Apple G4 with four video cards, low-end video projectors, and an acrylic
sphere. The system had limitations in terms of resolution (400 x 400 pixel imagery) and a low video frame rate, along with deficiencies related to mapping and
synchronization of the images on the sphere. However, even with these limitations, the visualizations and overall concept convinced other NOAA organizations
to invest in the project. The Office of Oceanic and Atmospheric Research (OAR); National Weather Service (NWS); National Environmental Satellite, Data, and
Information Service (NESDIS); and National Polar Orbiter Environmental Satellite System (NPOESS) provided resources to support the next development stage
of Science On a SphereTM. NESDIS gave additional help and played an important role by providing a freestanding suspension
structure for the sphere, significant support for dataset creation, and indispensable aid in planning and logistics for conferences and remote events.
The goals during Phase 1 development were to improve the system by 1) creating additional media sets to display on the sphere; 2) increasing the pixel resolution
used by the display system; 3) improving the focus, convergence, and overall quality of the projected imagery; 4) creating a user interface to control the system;
and 5) developing more portable and scalable construction techniques for the sphere itself. All of these goals were achieved to a large extent by December 2002.
Phase 1 development culminated in a series of presentations and remote events, starting with a week-long educational workshop at Broomfield Heights Middle
School (near Boulder) in October 2002 to test the idea: "Can Science On a SphereTM be used as a teaching tool?" The system was set
up in a science classroom at the school, and for one week, nearly 500 students (some future scientists, no doubt) were led through exercises in atmospheric science,
geography, and other earth science disciplines. The reactions from the students during the teaching exercises were fascinating to observe. They voiced their
surprise to see the Earth and other planets at this scale and were very responsive in the classroom exercises. Some students were so influenced by the presentations
and lessons that they were still discussing science at lunchtime in the cafeteria, a rare event according to one teacher. Parents attending parent/teacher conferences
saw the sphere in the evening as enthusiastic students showed them what they were learning in science class.
In December 2002, Science On a SphereTM was at the NOAA Science Center in Silver Spring, Maryland, for presentations to
NOAA staff and management. Various NOAA groups participating in the program were excited with the prospect of displaying new NOAA data from their
own group on the sphere. Science On a SphereTM was exhibited at the Annual Meeting of the American Meteorological Society
in Long Beach, California, where FSL hosted nearly 2,000 participants from the general public and the scientific community during the week-long conference.
Overall comments and audience reactions were very positive, with the more common response that this is a dramatic way of looking at the geographic and
atmospheric processes of our planet and conveying that information to the viewer.
Most recently, Science On a SphereTM was shown at the Department of Commerce’s Hoover Building in Washington, D.C.,
(Figure 3) as part of the national program, "Excellence in Science, Technology, and Mathematics Education Week."
Figure 3. NOAA Administrator Vice Admiral Lautenbacher (top) presenting NOAA's Science On a SphereTM at the Herbert
C. Hoover building in Washington, D.C., and (bottom) Sandy MacDonald leading a demonstration.
(NOAA Photos by Will von Dauster, FSL.)
Projections
The next development phase during 2003 will involve creating an "industry-ready" system that can be made available commercially through a technology
transfer mechanism. Two versions of Science On a SphereTM will be created. The version suitable for use in schools will
comprise a smaller sphere, lower-end computers, and less expensive display technology. The other version, a higher-end system, suitable for museums,
conferences, and scientific institutions will have a corresponding increase in sophistication in terms of overall appearance, computing power, and projection
technology. Additional features planned for implementation during Phase 2 development include:
- Annotation Provide the ability to draw markers and place text over data being displayed on the sphere.
- Simplified System Architecture Investigate and implement more compact computers and technology, along with a lighter and easier
to assemble display structure, to facilitate setup and reduce the cost of transporting and assembling Science On a SphereTM for
conferences, workshops, and schools.
- Documentation Create technical documentation and a user manual describing the operation and use of Science On a
SphereTM.
- Display Preparation Package Develop and provide a software package to convert scientific data to an SOS-ready format. Provide
documentation on SOS data format.
- Real-Time Data Ingest Provide for real-time data ingest for a limited set of NOAA satellite and model data products and develop an
automatic way of distributing Science On a SphereTM ready images from a central facility to remote Science On a
SphereTM sites.
- Unattended Operational Mode Build the capability for Science On a SphereTM to run in an unattended mode for
public display, featuring automatic loading of data with synchronized audio narration.
- Pole Rotation Add the ability to display the Earth or other planets at arbitrary angles of rotation. This feature includes the ability to rotate the
Earth so that the poles can be displayed anywhere on the sphere for easier viewing and data exploration.
The continued success of NOAA Science On a SphereTM is dependent on the support of NOAA and other government
organizations, industry, and the private sector. Early and future development of the system is also supported by a very dedicated, capable team at
FSL under the leadership of Sandy MacDonald.