Northridge disaster tests GIS workstations

The Los Angeles earthquake of January 17, 1994 measured 6.8 on the Richter scale, breaking water and gas lines, collapsing freeways and destroying buildings and lives. In the following days, the staff of the Los Angeles County Emergency Operations Center (EOC) received thousands of frantic calls from residents wanting to know if their water was safe to drink or if there were alternative routes to avoid damaged roads.

The day after the quake, EOC asked Wayne Bannister, manager of Urban Research Section in Los Angeles, to help it translate field damage reports into maps. “They were getting five or six pages of faxes at a time from the highway patrol — pretty cryptic stuff about stretches of road that were damaged or closed,” Bannister says.

Before the earthquake, the EOC had been planning to install its own workstation-based GIS, and the performance of the Urban Research team was proof under fire of its value. “The workstation gave us superior graphics, and because it doesn’t spread its processing power over many users the way a mainframe does, we were able to get these maps out very quickly.”, Bannister said.

Fire department workers brought in maps indicating areas without water, from which the team constructed a large map showing areas of contaminated water, water distribution sites and water companies involved in each area of damage. They gave the maps to the EOC phone staff, the Highway Patrol and Caltrans, the state transportation department. “The maps were much easier for the emergency staff to work with than flipping through pages of faxes to see if a particular bridge was damaged or to find the location of the closest water distribution truck,” Bannister says.

The 20 members of the URS typically do not respond to emergencies. The mathematicians, sociologists, statisticians and geographers provide county departments with information, such as the best locations for new trauma centers or how large the population of a certain neighborhood will be in 20 years. Much of their work involves huge databases.

Until recently, the work was performed on a mainframe, but the lack of graphics software and terminals had become a serious limitation. Slow response time and high user costs were other problems in centralized computing.

Although the $20-perminute fee was passed on to customers, the URS struggled to keep its rates competitive. “Even though the power of the mainframe was very great, it took a long time to get answers, said William Montgomery, senior analyst, says. “A job might require five minutes of CPU time, but we wouldn’t get our results for more than an hour because we were competing with 50 or 60 other users.”

With the GIS workstations, researchers could produce clearer and more explicit charts, graphs and maps. Faster output has increased overall productivity while lowering workstation leasing and maintaining costs. According to Montgomery, the average monthly cost of a quality workstation is less than two hours of CPU time, a savings that can be passed on to customers.