Could relying on ensembles and historical models backfire in hurricane forecasting?

[SeekerOfTruth007]

Ensembles are where you make like 50 or even 1000 different "close guesses" of the initial conditions, and then try to run a group average of what each initial condition adds up to.

Historical models just try to average what has always happened in the past (allegedly).

Problem: One in so many hurricanes disobeys Historical Averages and does something totally unexpected, so that makes historical modeling functionally USELESS.

Ensembles theoretically waste most of the processing power on the WRONG output.

Theoreticlaly, if you knew the real starting conditions and knew the real formula for weather, you should put all your processing power on one run of the model.

[Anonymous]

Nope

[SeekerOfTruth007]

You're wrong. Category 5 Hurricane Otis hit Acapulco as a Category 5 hurricane.

the day before landfall, it was still a Tropical Storm, and the computer models and the NHC were forecasting it to stay weak and move out to sea and never make landfall.

The hurricane immediately intensified close to world record pace and made a direct hit on the city and destroyed all the resort structures in the city, especially right up to the waterfront property.

Computer models are not supposed to replace human skill in forecasting, plus I learned decades ago when it comes to forecasting weather, you should always err on the side of CAUTION, and above all else, a forecaster should NEVER say that something CANNOT or WILL NOT happen.