Catastrophe modelers have capitalized on one of the few positive things to come from Hurricane Katrina—a wealth of claims data and other information they are using to improve their products and help insurers and risk managers mitigate windstorm losses.

Katrina and the spate of Gulf Coast hurricanes that preceded it in 2004 produced an abundance of loss information that has helped shape models, said Jayanta Guin, senior vp of research and modeling at Boston-based AIR Worldwide Corp.

“Insurers did a very good job of collecting that information,” Mr. Guin said. “We have never had that much data to validate modeling with such precision.”

Katrina's impact on modeling can be seen in improvements that allow exposures to be examined in finer detail, sources say. The storm's widespread destruction also led to models that now include some previously unexamined risks. In addition, esoteric influences on claims costs such as the political pressure on insurers that mounted after Katrina now are included in some models, according to experts.

“What happened definitely influenced changes,” said Robert Muir-Wood, London-based chief researcher for Risk Management Solutions Inc. in Newark, Calif. “Not just to modeling, but to science.”

Modeling has benefited from the same data from Katrina that engineering scientists have used to examine the potential failure of flood defenses under storm surge, Mr. Muir-Wood said in citing an example. “Catastrophe modeling had to learn these broad lessons as well.”

Modeling since Katrina has become a “game changer,” said Al Tobin, New York-based national property practice leader for Aon Risk Services. While previous models largely were in the domain used by insurers and reinsurers, more risk managers are using them today, he said.

Insurance buyers are “very familiar with the benefits of having quality data” from models that are used as tools to help determine exposures and the amount of insurance needed to cover them, Mr. Tobin said.

Katrina influenced modelers to include some exposures in newer versions of their models.

The storm caused a significant contraction in the offshore energy market and led the companies to model that exposure in their updated products.

“We've included offshore energy,” said David Smith, senior vp at Oakland, Calif.-based EQECAT Inc. “There's been an increased demand for that type of modeling” since Katrina, he said.

Revised models treat some risks differently from how they did before Katrina because the storm revealed features about those exposures that were previously unknown to the modelers.

Casinos that floated just offshore in Mississippi were destroyed by the hurricane. “Although they were accounted for in the models, they were quoted as buildings on solid ground,” said Mr. Guin. Those risks are now properly accounted for in AIR's model, he said.

An RMS report cited several unmodeled lines of business that escaped major storm losses until Katrina hit, some of which are now covered by the modeling process and others that may be included one day. Among those lines, RMS listed cargo, inland marine and recreational watercraft, floating casinos, onshore energy and automobiles.

Modelers are modeling some of these exposures, and are hoping to include others in updated versions.

“We're certainly doing better in capturing the casinos,” said Mr. Muir-Wood. “Cargo is still a challenge. It's not insured by location,” he said. It is an asset that moves around and can be difficult to track. “Inland marine is slightly easier than cargo,” he added, but remains a complicated risk to model.

Katrina was a storm that had a “high political profile” that added never-before-seen complexities to the modeling process, Mr. Muir-Wood said. “We saw a lot of insurers paying for things that were not covered,” he said, in some cases because of “extreme political pressure” to make policyholders whole even though their losses were not insured.

Such influences are now considered in RMS' modeling, Mr. Muir-Wood said.

EQECAT's Mr. Smith said there has been a “renewed emphasis on storm surge” in models. “That is a big part of what happened in Katrina,” he said. Much of Katrina's destruction along the coast was attributed to the powerful tidal surge that damaged and overtopped levees in New Orleans and washed away structures along other coastal areas.

Demand surge, or the increase in the costs of materials and services to recover from a storm, is another component of modeling that has gotten new attention since Katrina.

“After Katrina, we saw four separate phenomena pushing up costs,” said Mr. Muir-Wood.

“Classic demand surge,” which pushes up prices because of a scarcity of building materials and other resources, was evident post-Katrina, Mr. Muir-Wood said. The effects were worsened by a similar surge created by several hurricanes that hit the coast in 2004 and powerful storms after Katrina.

Repairs often took an unusually long time to complete after Katrina because contractors and materials were scarce, Mr. Muir-Wood said. Repair costs rose as rain caused further water damage and mold took hold.

Katrina produced an unusually high number of claims, which meant overwhelmed insurers also received a larger-than-usual amount of fraudulent submissions, Mr. Muir-Wood said.

Insurers also experienced “coverage expansion” as they felt obligated by political pressure to pay some claims for damage they had not covered, he said.

As modelers improve their products, they monitor their effectiveness by simulating losses. When they consider a storm like Katrina today, insured losses are not significantly different from five years ago.

That's partly because reconstruction after Katrina has been slower than other hurricanes and parts of the devastated area remain blighted, said Mr. Guin. “There are certain areas of New Orleans and Mississippi that haven't been rebuilt to pre-Katrina days,” he said.

Property values have risen since Katrina, Mr. Guin acknowledged, but those increases are at least partly offset by the absence of so much property that was destroyed by Katrina and not rebuilt.

Mr. Muir-Wood said flood protection programs in New Orleans since Katrina would help hold down losses from a similar storm. RMS models, he said, “show a very significant reduction in areas that would flood.”