New modeling of storm surge

An AP
wire story mentions that climatologists at Florida State are going to
look into building a better model for storm surge:

Current models for predicting storm surge were created more than
a quarter century ago with outdated technology, leading to inaccurate
predictions, said James O’Brien of the Center for Ocean-Atmospheric
Prediction Studies at Florida State.

“The storm surge is the thing that really causes a lot of problems,” said
O’Brien. “Look at Katrina, look at the other storms that hit, what happened
in Mississippi and so forth. I mean, that huge storm surge is what really,
really caused the devastating damage instead of just the winds.”

O’Brien said the model’s limitations were shown during Hurricane Dennis
earlier this year. The National Hurricane Center predicted 2 to 5 feet of
storm surge south of Tallahassee, but the area was hit with a 10 to 12 foot
surge, causing heavy damage.

“If they had a modern model with higher resolution and better winds and
modern computer capability, I think they could’ve done it (better),” O’Brien
said. “The current model was built 25 or 30 years ago where computers were
smaller and nowadays, there’s no technical barrier to doing it
correct.”

That’s good news. Even better news would be getting word from the cat
modelers that they’ll beef up their work on storm surge and/or incorporate
new research. One of the lessons I learned from this cat season — Katrina
in particular — is just how much uncertainty exists in cat model outputs,
particularly when looking at single- or multi-location accounts, and most
especially when considering potential damageability from storm surge.