John Graham of SDSU created this visualization showing the before and after of the Tsunami.

Viz Center researchers acquired data from a number of sources. The majority of the imagery data sets were taken by the National Oceanic and Atmospheric Administration (NOAA) with a specially equipped plane that shot high-resolution color photographs over the damaged area. Before photos were acquired from the United States Geological Survey Earth Resources Observation and Science (USGS EROS) Data Center and the Army Corps of Engineers. NASA provided before and after satellite imagery that recorded the widespread flooding and destruction wrought by Katrina. The high-resolution photographs were acquired repeatedly as the water drained to provide insights into the changing conditions over time as the levees broke and the water flooded and receded.

A team of image processors and computer scientists from a number of different institutions worked together on a task that normally would have taken weeks or months to accomplish. These experts included John Graham, the Visualization Center’s senior research scientist, who led the processing effort and built the social network of specialists; Norm Vine, an independent contractor working with the University of New Hampshire and the Wood Hole Oceanographic Institution; Chuck Stein from GeoFusion; and numerous others. Tom DeFanti at the University of California San Diego allowed the CPUs of their SGI Prism system to be networked into the SGI Prism at San Diego State in order to speed processing of the immense data sets.

Watch a 3D visualization of the progression of the water level in New Orleans caused by Katrina, in an accurate detailed representation of how the water inundated the area created by Frank Weed. All 3DNature images © 2005 3DNature.com.

Because of an inherent aspect of the low-altitude photographs that causes the scale to be different in the center than at the outside edge because the plane is closest to the center — like taking a very close picture of someone where the nose looks bigger than the rest of the face — the team of experts needed to color balance and geo-rectify the aerial shots so that the pixels were placed where they actually are on the Earth. Once the photos were geo-rectified, various data sets were added as to where roads, hospitals, schools, refineries, gas stations, police and fire stations and hazardous waste were located. Besides providing a photographic record of the destructive effects of the storm, addresses could be geo-located on top of the photos to determine whether residents had a house to return to. The imagery was also used with data gathered on the ground, such as soil samples that measured lead leached by the salt water from a pre-colonial landfill. The soil sampling data was located on an overlay to show the pattern of toxin dispersal created by the floodwaters.

The capability to respond to disasters would be greatly improved with some inexpensive investments in infrastructure. Frost strongly advocates the need to continue innovation in order to provide faster access to original data, through networks such as the National Lambda Rail and the addition of 10GB Ethernet connections, especially at government facilities and research centers. Frost makes an analogy of the difference improved bandwidth would make in the event of another disaster, “If you loaded up a moving van in San Diego and where going to race it off to New Orleans for relief and you went one mile an hour, immediately every one would say, ‘Get going! This is crazy to go one mile and hour.’ That’s largely the response because of the network.”

Having a faster network in place before the next disaster strikes would enable all major SGI facilities, data sources, visualization centers and command-and-control hubs to be connected. This would allow experts throughout the country to immediately ingest, process and serve up data sets, speeding the creation of 3D visualizations and greatly impacting the ability to respond to natural disasters in the future.