1. Two reasons not to take the Telegraph piece at face value:

a. If does not provide a link to the underlying research paper.

b. Not a single researcher is identified by name.

(c. No offense to the U of Bordeaux, but I have not heard them mentioned as a hotbed of cutting-edge research.)

It’s not clear whether the inadequacy is in the research or the reporting. Please excuse my lousy eyesight if I’ve overlooked something.

2. That said, the work could be important if it is correct. Hurricane prediction involves computer models that are cumbersome even on supercomputers. The major models do not always agree. As hurricanes make their way up the US East Coast for example, a few tens of miles can be the difference between heavy rain & high wind and a devastating impact.

If the reported research shows a way of demonstrating that a lot of the complexity in those models is irrelevant, that would be important, both to make the computations faster & more responsive to data updates, and to guide attention to understanding the truly significant factors.