The Community for Technology Leaders
RSS Icon
Subscribe
Issue No.05 - September/October (2004 vol.24)
pp: 4-5
Published by the IEEE Computer Society
In May and June of 1991, floods destroyed 24 million hectares of farmland across 19 provinces in eastern and central China. Around 5,000 people were killed. Pak Chung Wong, Harlan Foote, David Kao, Ruby Leung, and Jim Thomas—working at the Pacific Northwest National Laboratory (PNNL) in Richland, Washington—used fusion-based visualization technology to study the typhoon. 1
Wong explains that this issue's cover image "depicts the direction and speed of the wind patterns in the typhoon and surrounding areas." The image is a visualization that helped the scientists understand the storm's dynamics, movement, and intensity. "The visualization combines a number of characteristics and features of the typhoon and the weather patterns surrounding it," Wong elaborated. "Some of the characteristics depicted in various patterns of colors and layers are wind velocity, direction, and magnitude; curvature and critical points of the flow field such as a curl; and vortex centers and convergent flow points (that is, saddles). These features are critical to understanding the dynamics of wind flow patterns and are related to predicting the course of such a storm."


Cover image.

Conventional visualization usually draws upon a single image. Fusion-based visualization, on the other hand, uses multiple layers superimposed on top of each other, letting scientists explore parameters not available with conventional methods. Although the scientists originally implemented their strategies to analyze a storm, their visualization methodologies are fully applicable for other scenarios as well.
"The visualization techniques can augment traditional techniques used by creative professionals ranging from magnetic fields on the sun to the dynamics of currency flow," Wong explained. "The concepts of these visualization techniques have been applied repeatedly in a variety of applications at PNNL. Among them is In-Spire (see http://in-spire.pnl.gov and the article by Hetzler and Turner in this issue), which has the ability to uncover how the relationships, trends, and themes hidden within data can lead to new knowledge and new insights that could be used to assess terrorist threats, determine how to treat a medical condition, or to gather market research on the competition."
Managed by the US Department of Energy's Office of Science, PNNL is one of nine DOE multiprogram national laboratories. Wong began studying data visualization as a graduate student at the University of New Hampshire and joined PNNL soon after he received his PhD in computer science.
"I've had the fortune to team up with some of the best physicists and meteorologists and support them in their work," he explained. Among these he lists Harlan Foote, who has studied flow visualization in subsurface media for decades, and Jim Thomas.
Global Climate Change
Figures 1 and 2 are part of the same project. In each instance, the image is a fusion of more than one visualization. Techniques that Wong and the other scientists used to create the images include texture-based visualization, color-based visualization, and/or intensity-based visualization. According to the group, the visualizations are fused together via alpha channel manipulation, filigreed graphics, and/or elevation terrain mapping.


Figure 1. The color spectrum is used to indicate the flow direction instead of using arrows. Streamlines provide the texture and flow detail.

"[In Figure 1], the color spectrum indicates the flow direction instead of using arrows," Foote explained. "Streamlines provide the texture and flow detail. [In Figure 2], the velocity magnitude is used for relief shading, color for direction, and streamlines for structure."


Figure 2. The velocity magnitude is used for relief shading, color for direction, and streamlines for structure.

Ted Tanasse, a graphic illustrator and user interface designer at PNNL, supplied Figure 3 along with a background map of Asia to provide a geographical reference and a glimpse into the climate modeling data set and the wind speed. The big swirl at the lower right is the typhoon.


Figure 3. An overlay map gives scale and location. The large typhoon is the big swirl at the lower right.

Wong et al. wrote in a research paper that
the ultimate goal is to loosely combine multiple single-purpose techniques into a fast and effective multipurpose visualization tool that capitalizes on the best of all worlds. The concept of data fusion has been explored intensively in the remote sensing and imagery industry for decades. It is our intension to take advantage of its dynamic integration concept to visualize a multivariate dataset. 1
Foote noted that once in a while they'll create an Adobe Photoshop stack to fuse the visualizations, but the majority of the work is done with their own software: "To generate the shaded topography of the magnitude of velocity, and curvature, and different kinds of streamline texturing, and stuff like that, [we use] our own programs."
And the visualization of the East Asian typhoon data set is only part of the overall endeavor, one that Foote and his colleagues said is an essential component of their ongoing visualization effort supporting a grand-scale global climate change program at PNNL. The typhoon study was just one component of their grand plan.
"This was just one example," said Foote. "We had a lot of similar data—weather, especially velocity fields—and different depths of the atmosphere for different areas. [Some are] for the US, and this one is around China and Japan."
Analyzing Terrorist Threats
Fusion-based visualization research at PNNL has expanded into other areas as well. The researchers there are using visuals to upgrade the analysis of intelligence to counter future terrorist attacks.
Wong explains that although the field of visualization is still young, it's growing steadily into a more diversified visual-based exploration concept known as visual analytics (the theme topic of this issue). Wong's colleague and co-guest editor of this issue, Jim Thomas, is director of the National Visual Analytics Center (NVAC), a department of PNNL that is funded by the Department of Homeland Security's Science and Technology Directorate. According to NVAC's Web site (see http://nvac.pnl.gov/), the center is a national resource that provides strategic direction and coordination of activities to discover, develop, and implement innovative visual information analysis methods.
Established just this year, NVAC looks to set a national agenda on visual analytics and hopes to play a pivotal role in countering future terrorist attacks around the globe. Using research conducted at PNNL for the last decade, NVAC is providing new analytic methodologies for the intelligence community.

Reference

6 ms
(Ver 2.0)

Marketing Automation Platform Marketing Automation Tool