PIOMAS is calibrated to NSIDC sea ice concentration, which has coarse resolution (25x25km). Perhaps NSIDC shows this area with some ice?
That it uses NSIDC concentration is good information.
I've been comparing the latest modeled ice thickness against concentration maps. I have some observations but unfortunately not much insight.
POIMAS definitely shows ice where NSIDC does not, but this could be explained by the 15% threshold. This is not conclusive at all, because posters have pointed out the area around Barrow as inconsistent.
However, it also ignores ice that NSIDC clearly shows, such as shore ice in the Bering Sea. The granularity of the PIOMAS chart is lower, so it could get lost in some sort of averaging, but I am not convinced of that.
As others have said, it seems to have problems with overestimating thinner ice, but I'm seeing it go both ways. It may have similar issues with thick ice, but that's tougher to tell using NSIDC concentration. There are short lived low concentrations (NSIDC) that move across the interior of the ice pack that are not reflected in the morphology of the PIOMAS map thicknesses. I am wondering if some sort of box car averaging is being used.
I am not trying to discredit PIOMAS, but it's important to understand it's limitations and I'm not there yet.
We are obviously moving towards thinner ice, so these anomalies will be more widespread, so if anyone has more insight, I am all ears.