It's that time of year again, when we see that extent hasn't dropped as low as we feel it ought to have, and so we try to change the definition...

I think instead of 15% concentration Extent we should focus on assessing 50% concentration Extent. 15% takes a lot of areas that are basically 85% ice- free into account.

[... snip ... Discussion of posts where Ned W compares the standard extent data to what it would look like with a 30% threshold instead of 15% ...]

further, while it is great that you calculated that, perhaps you'd post the math behind it for verification and then nobody ever said that it must be 30%, perhaps the critical percentage is not 30% but another threshold. to follow this up it would take a few runs with different threshold to compare and the numbers to verify, after all, where are 30% numbers available, i did not find them which is why i was not able to do my own calculations?

[...] if we have a huge area of 50% ice and 50% water no-one can tell that the result do not differ by at least the 35% that are usually counted and now are not, after al 35% is about a third, hence cannot be insignificant to my understanding

Personally, I think basic logic dictates that if an area is more water than ice, then it should be considered water. 15% as an arbitrary threshold is bizarre. If you look at my map prediction for Sep 1, a 50% threshold will be almost exactly as I predicted.

This comes up every year. 15% is what the community uses, all of the data (except older DMI) are based on it. If we start making predictions about a 30 or 40 or 50% threshold, all it does is ensure that it will be prohibitively difficult to compare them to anything else, quantitively.

If this year's "50%-extent" is X million km2 ... Is that low, high, or average? Nobody will know! Because for every previous year, we have "15%-extent" data.

However ... As noted above, I did go back and reprocess all the NSIDC concentration maps to calculate a "30%-extent", and posted the results in the other thread here:

https://forum.arctic-sea-ice.net/index.php/topic,2223.msg164660.html#msg164660On the next page following that one there's a post by me with a csv file with the results.

Magnamentis wants that redone for other thresholds, and wants me to post "the math" for verification purposes. The "math" involved is merely (a) reprojecting all the ice concentration maps into an equal-area projection, and then (b) counting up the number of grid cells that exceed whatever threshold you want. (A further complication is the "pole hole" whose size changes...) If there's a lot of interest I'd be willing to rerun the processor for different thresholds.

The thing is, though, so far at least it doesn't have any non-trivial effect on the long-term trend in extent. As Neven predicted... So the payoff for doing it is pretty small.