In mid June I posted a blog post asking about whether we were facing a crash this year. This was based on the US Navy's ACNFS Model, which I have previously called HYCOM. Shortly after the post the model version I had used was put behind a password wall, the US Navy had realised that it wasn't signed off for public release.
Now it seems there is a problem with ACNFS, as found out by Peter Ellis at the Sea Ice Forum. Basically the massive melt ACNFS showed this summer was due to a glitch in the model. These things happen, the only people who don't make mistakes are those who never try anything, and the more complex one's endeavour the more likely mistakes are.
So my question in mid June about us facing a crash this year was based on flawed data.
This doesn't affect any of my recent blog posts because the password walling of ACNFS made me suspect something was up, on that I was wrong in the end, it was walled for another reason. But after the PIOMAS data I suspected something was off with ACNFS, by my June Status post I used ACNFS but came to the conclusion there would not be a crash this year.
I will be using ACNFS for ice movement, but for other metrics will probably rest use of it for the remainder of the summer season, perhaps longer, to give them time to get things sorted.