New indicators helpful in predicting surge levels

Hurricane Katrina didn’t follow the rules. Neither did hurricanes Ike or Isaac.

These storms all produced storm surge much larger than could be accounted for by the wind speeds the storms maintained at landfall.

Hurricane Katrina had Category 3 wind speeds when it hit land, but it brought with it a 28-foot storm surge, the largest ever in the U.S., said Hal Needham, a research associate at LSU.

He and state climatologist Barry Keim have been working on several research papers that use high-water marks from around the Gulf of Mexico and the country to determine if there is a better correlation to be made between wind speed and storm surge.

For years, the rule of thumb was the Saffir-Simpson scale, which correlates wind speeds of a storm at landfall with the amount of damage it is expected to cause, including from storm surge.

A Category 1 storm could be expected to cause a certain level of damage, while a Category 5 storm could be expected to result in much greater damage.

In the past decade and especially the past few years, however, there has been a move away from this correlation, as storm after storm hit land with wind speeds that belied the large amount of storm surge associated with them.

During their work, the researchers found that wind speeds 18 hours before landfall are a much better predictor of the scale of storm surge at landfall than just clocking the winds when a storm’s center reaches shore.

At 18 hours before landfall, Hurricane Katrina was a Category 5 hurricane, which much better explains the record-breaking storm surge that came with the storm, which had been downgraded to a Category 3 by landfall.

What the researchers found is that after the 18-hour point, it doesn’t matter if a storm holds its strength or weakens: The storm surge levels have been set.

Their research paper has been accepted for publication by the journal Earth Interactions.

“The approach of our research here is really data-driven,” Needham said. Much storm surge research relies on computer modeling, but Needham and Keim assembled a large historical database called SURGEDAT with information that stretches back to 1880.

This information came from various sources, including details on 600 hurricanes worldwide, newspaper articles, data from various federal agencies and more than 7,500 high-water marks in the U.S.

A number of factors influence the size of storm surge: wind speed, the size of the storm, the depth of the water offshore and the shoreline geography of bays or harbors.

It is possible to look at the storm’s wind speed 18 hours prior to landfall, determine its location and use data from previous storms to get an idea of the storm surge potential, Needham said.

But the usefulness for purposes of evacuations is limited.

“Most evacuations have to be initiated at least 48 hours before landfall,” Keim said.

In a related study, Needham and Keim looked at the physical size of tropical storms and how that might correlate with storm surge. They again found that a storm’s size 18 hours before landfall is the best fit for determining how large its surge may become.