for years i try to start a solution oriented deeper discussion about how to resolve this and any approach is mostly discarded with some "what can we do" or "they are great/sufficient" arguments.
again propose to seriously discuss whether the 15% threshold for extent is still contemporary, considering the fact that we are not looking at a homogeneous ice sheet but highly fractured and somehow perforated ice cover that often remains above 15% but way below the 100% as what they are counted as far as extent is concerned.
sooner or later one has to at least consider to change methods/parameters of measurement to keep up with real world scenarios, or do we really not care if once the ice mass is cut in half an well spread across the arctic to see the ice extent as the same like when it was double the ice but in one single homogenous sheet?
my point is that i believe that many discussions are caused by different interpretations and prioritizing of input and to adapt to new facts could get us rid of many times seeing the same thing differently from different "angles" so to say.
IMO extent as it has always been (measured) should not be a holy cow.
....contributing ideas, i would ask to consider to raise the extent threshold to something between 30 and 50% because if there is less than half of an area covered with ice it's hard to understand why this should count as 100% extent.
this is just one example, there are other factors that undergo heavy changes in the last few years, one is humidity and cloud cover as compared to dry cloudless conditions over vast periods in the past. etc.
People always seem to jump on the "its an observable metric, and changing it would make comparisons with previous years difficult." Bandwagon when we try to suggest that its not a good measure. The comparing apples with apples argument. But its like saying we have the similar number of barrels of apples as we did ten years ago. Which may be fine if you are intending to compost them. But if you want to ship them to the supermarket to be sold, then ignoring that every year there is more rotten apples in those barrels, and now most of them are is a big problem.
Personally I think area is failing in the same way. Say to wash them you put your apples in a turbulent stream that flows into a pond, and measure the areas of the clumps that accumulate around the shore to estimate the size of the harvest. If the apples are getting smaller you have less barrels of apples, but they may cover the same extent and area. If the apples are rotten and the stream is carrying more water and more turbulent, then you'd be a right twat to estimate the revenue you will get from selling them by just measuring the area or extent of the mats of disintegrated rotten apples and bashed and broken bruised ones on your pond.
The fractured rubble fields between smaller and thinner floes, than we used to have, should should not be viewed as equivalent in area or extent to vast areas of large thick floes with deep pressure ridges and keels, and clearly defined large leads easily resolved by kms wide satellite sensor pixel resolution.
If cloud effects could be removed then Albedo would be a far more valid comparison. Perhaps some combination of microwave and ultraviolet frequency albedo maps would be a good way to go. But comparing with past years would be difficult.
In terms of melt susceptibility and prediction for where the season might end up compared to previous years SMOS archive going back to 2010 might be the best thing to work with. It does suffer from high daily variation with changing temperatures and precipitation events and sogginess and salinity of the ice resulting. So running 7 day 14 day, and month trailing means would give decent results I would think. It would be better if they plotted the whole dataset rather than clipping at max 50cm thickness where their error percentage starts to rise on dry ice for thickness measurement purposes. Still it would be very easy to gif a weeks worth for same date in different years for comparison purposes.
Here's 22june 12,16,17,18