UPDATE: 11th August 19.37pm (BST)
The NOAA have explained the issue is related to the unprecedented sea ice retreat revealing a numerical issue with their model and are working on a fix. We’re working on deploying alternative data in the interim.
UPDATE: 11th August 12.15pm (BST)
We’re still seeing significant issue on the northern boundary of the model. We’ve taken steps to identify and remove the problem data from your local forecast page but this is an interim fix and won’t fix affected charts. Apologies for the ongoing issue.
We’re seeing what look like erroneous results in the main NWW3 model product we use for our forecasts. A large long period swell entering the model from the northern boundary. We’re investigating the issue.
Throughout the model period we’re seeing forecast swell in the 28 second period range propagating in from the north between Greenland and the Bering Straight. We’re suspicious of this for a couple of reasons: The swell looks well outside the range of normal wind generated waves and crucially the northern boundary of the model is closed at 82 degrees north.
This issue is originating from the NOAA’s distributed NWW3 model data, you can see evidence of the same problem directly on the NOAA’s NWW3 viewer here:
The swell isn’t showing in our in-house model or in other wavewatch products. However as we don’t produce this data ourselves we’re unable to remedy the issue at the moment. We’ve contacted the NOAA for clarification and will update when we have more information.
As we’ve said this result is in error – so you’d be well advised to not make plans based on it – sorry for any inconvenience.