@ seaicesailor & oren
Please do not read into my comment something that was simply not there - and certainly was never intended.
I have always considered longer-term averages to be more meaningful - and more statistically robust - than their short term equivalents. Way back in August (this thread #2941) I mentioned that the IJIS/ADS rolling annual average extent was already in very interesting territory, and that the regrowth rate following the Sept min would be instrumental in determining just how low this year's rolling annual average would become.
In my recent post (#3372), I merely mentioned that viddaloo had also been pointing this out for some time - both here and on the ASIB. The fact that he has recently started trying to extrapolate the current rate of change in the annual average extent is, I feel, a completely untenable position.
Several years ago, Neven posted an article of mine on the ASIB. In this, I argued that trying to extrapolate monthly average extents even a few months ahead was extremely problematic.
http://neven1.typepad.com/blog/2013/06/problematic-predictions.htmlRob Dekker followed this up shortly afterwards by his own take on the difficulties involved in determining time series correlations.
http://neven1.typepad.com/blog/2013/07/problematic-predictions-2.htmlGiven my views then, I am hardly going to subscribe to the notion that one can simply extrapolate a specific variable (in this case, a rate of change) several years into the future.
The projection (prediction?) made by viddaloo of a totally ice free Arctic in 2023 does not stand up to any serious scrutiny. Currently, the IJIS/ADS rolling annual average extent is dropping by around 3,000 sq kms each day. That comes about because today's value is approximately 1.1 million sq kms lower that the equivalent day's value in 2015. For as long as the differential between the 2016 extent and the same day's extent last year remains at about 1.1 million sq kms, the daily rate of change in the annual average extent will remain the same - i.e. dropping by about 3,000 sq kms each day.
However, there is absolutely no physical reason to suppose that, for the next 7 years, the daily extent will continue to be clocked up at 1.1 million sq kms below the value recorded on the same day of the previous year.
Even worse, the situation is mathematically impossible. Should, by some miracle, the 1.1 million offset from the previous year continue for the next 45 months, then, by late August 2020, the daily value will hit zero. As it is somewhat difficult to have a negative value for extent, the projection simply cannot continue in an unmodified form beyond that point. It simply does not make mathematical - never mind physical - sense.