The worst of Laura’s surge narrowly missed Lake Charles LA (hence this busted fcst) thanks to a few factors including a last-minute E wobble of the center.

We still have a lot to learn abt surge forecasting.

Plus we probably should work more on probabilistic surge communication https://twitter.com/billkarins/status/1298897541954760709

It often feels like the only numbers discussed in surge forecasts are "reasonable worst case scenarios". That's what the NHC inundation graphics are based on, it's what gets on TV, etc.

There is a reason for that: we want you to prepare for that level of flooding.

But there's a downside when folks hear those big numbers and perhaps may be less inclined to take subsequent threats seriously when they don't verify (which by design they shouldn't most of the time!)

What's the solution? Probably to focus more on the full range of psbl outcomes