USTropics wrote:Below is a chart for average positional error by nautical mile ONLY for Irma (includes all model runs since Irma initialization). ECMWF data only goes out to 72 hours:
data source - http://tropicalatlantic.com/models/models.cgi?basin=al&year=2017&storm=11&display=model_error&error_type=average&latestrun=1&type=table
This is good stuff. What is concerning is how accurate the UKMET and TVCN projections have been. Only 100 miles off at 120 houra and they are both taking it further West. Another 100 mile shift west error from them at 120 hours puts it right on Florida's door step. Kudos as well to GFS 120 accuracy. Taking into account all 3. It's becoming clear there will be a hit.