 |
No more canned weather data. For the real thing, check out Weather Visualizer. |
Maybe you can't control the weather, but now you can
control the weather map.
A new visualization tool called Weather Visualizer lets you create customized weather maps on the Web with real-time or archived data. You can follow an evolving storm anywhere in the U.S. or find out if itÕs raining in Des Moines. It's like having a weather station on your desktop.
"Weather Visualizer is one of the very few sites on the Internet where you can tailor a product to your needs and then interact with that product," says Mohan Ramamurthy, UIUC associate professor of atmospheric sciences. He and NCSA's Robert Wilhelmson, who is also a professor of atmospheric sciences, worked with atmospheric sciences students and programmers to develop the tool.
At the heart of Weather Visualizer is point-and-click technology for creating the six most common categories of weather maps: surface observations, upper air observations, upper air soundings, satellite imagery, radar summaries, and forecast data. Click on the icon for one of these maps and a form pops up displaying the available options. The form linked to the satellite image icon, for instance, offers a selection of image types -- visible, infrared, water vapor, or color-enhanced infrared -- as well as a yes-no option for a radar summary overlay. Choose among 10 geographic regions and recent or archived observations. When done click Submit Query, which sends a request to the server to collect the necessary data and compute the map. To change the map, select different options and resubmit the query.
Anyone puzzled by a term or symbol can click on Helper Pages for an explanation. The more inquisitive can explore the extensive An Online Guide to Meteorology. Discover the differences between the four types of severe storms as well as the one most likely to erupt into a tornado. These informational extras divulge Weather Visualizer's genesis as a tool for education. It was developed with NSF support through the Collaborative Visualization (CoVis) Project, a joint endeavor among Northwestern University, UIUC, and the Exploratorium Museum in San Francisco to create K-12 science curriculum for the Web. At UIUC it is coordinated by Steve Hall.
CoVis methods are modeled after scientific collaboration practices and draw upon many of the same tools and data. For Weather Visualizer, data streams in from sources such as the National Weather Service (NWS). Every hour NWS compiles the observations from approximately 1,700 observing stations and transmits them via commercial and nonprofit data vendors to UIUC. Eventually some of the vast amount of data compiled daily by NASA's Earth Observing System will be accessible through the Visualizer. Easy access to this impressive array of data has convinced more than 40 schools to use Weather Visualizer and has won fans among meteorologists, airline pilots, farmers, and weather junkies of every ilk.
Support for this effort also comes from NASA's Project Horizon (centered at NCSA) because that project is also interested in Web scalability issues like those faced by Weather Visualizer. The popularity of the Visualizer has been pushing the server into overload as it struggles to keep pace with requests for the compute-intensive maps. Java may relieve the strain by transmitting small chunks of program, called applets, along with the data so that maps are compiled on the user's computer rather than on the server. The clientÕs computer does not interact with the server again until the client downloads new data.
The snazzy new Java features include graphically displayed pop-up temperature forecast data for point- and-click map locations selected by the user as well as customized map animation. Users can select the specific frames they want to animate, then advance or rewind the maps step-by-step or select an animation speed. More applets are in the works for zooming and map analysis. According to Hall, the CoVis coordinator, "It is all part of our goal of giving the user total control over the weather map."
Return to the Table of Contents.
NCSA: The National Center for Supercomputing
Applications
access / Summer 1996 issue
Email comments to NCSA Publications Group:
pubs@ncsa.uiuc.edu