FEMA to expand disaster prediction

The Federal Emergency Management Agency and the National Institute of Building Sciences (NIBS) last week agreed to issue a request for proposals to expand a system that predicts the number of deaths and economic losses from natural disasters.

In four to six weeks NIBS a congressionally authorized nonprofit organization that is developing the system to predict losses from natural disasters plans to release the RFP for a new component to the Hazards US (HAZUS) system which will predict the damage caused by hurricanes tornadoes hail storms and other high-wind weather systems.

"We identified more models that may serve as the basis for the development of the wind loss estimation model and we determined how we are going to conduct the procurement process " said Phil Schneider director of the multi-hazard loss estimation program at NIBS. He said last week's meeting defined the way forward.

Currently state and local governments use HAZUS to forecast the number of deaths and economic losses from earthquakes and use those predictions to plan relief efforts. By importing HAZUS' database which contains the size and types of buildings in an area and demographic information state and local governments can use their geographic information systems to estimate the number of deaths and economic losses from earthquakes. HAZUS predicts how the ground will shake the number of buildings damaged the number of casualties the amount of damage to transportation systems disruption to the electrical and water utilities the number of people displaced from their homes and the estimated cost of repairing projected damage and other effects.

The wind component as well as one for floods will be completed in three to four years and together with the earthquake prediction component will provide more detail on natural disasters than FEMA's Consequence Assessment Tool System (CATS) which FEMA now uses to forecast losses from natural disasters.

"HAZUS is clearly an expansion in scope as well as a refinement in detail " said Joseph Minor chairman of the FEMA/NIBS wind committee. "CATS was able to project certain things for certain hazards at a certain level of detail. HAZUS will extend the scope of that to a greater level of detail."

NIBS has a cooperative agreement with FEMA to produce the HAZUS earthquake wind and flood models and has distributed the HAZUS earthquake modeling software to emergency management agencies in 50 states. The NIBS' flood committee is looking into user requirements for the flood modeling software which will forecast losses from floods.

HAZUS will completely replace CATS for predicting losses from natural disasters in about three to four years - the amount of time it will take to develop HAZUS' wind and flood components Minor said.Gil Jamieson chief of risk assessment within FEMA's mitigation directorate said CATS still will be applicable for certain disasters. "We're always going to use CATS for chemical [and] biological high-computing end-modeling capabilities " Jamieson said. "If you're ever in a situation where you're talking about a toxic release or a chemical spill of one sort or another you want the experts in one place running the model on their high-end computers and then passing along the results to state and local governments."

HAZUS also gives states and local governments equipped with desktop PCs the ability to make quicker decisions in the event of disasters. FEMA provides state and local governments with a compact disc so that officials can make up their own scenarios and input information relevant to their own communities as they plan for disasters. CATS which was built on a Unix workstation and requires high-end computing capabilities simply sent the forecast information to state and local governments.

Scott McAfee a GIS analyst with the earthquake program at the Office of Emergency Services in the California governor's office said HAZUS is easier to integrate with the state's GIS system but one of the drawbacks is that results are produced at a slower pace.

"If I were to run a scenario it would take longer on a desktop Intel-based system than it would on a high-end Unix [system] " McAfee said. He also said it could take "overnight or a day and a half" to get predictions for a complex geological area with many buildings and covering such as five counties in one metropolitan area.