I analysed the last 38 years of JAXA extent date. I took the average values of the 80s, 90s, 00s and 10s (up to now) at each end of a quarter (first table) and in the middle of the quarter (second table) and looked for trends in differences from one season to the next one over these four decades.
Extent [mio km²]
March 31 June 30 Sep 30 Dec 31
80s 15,22 11,14 7,76 13,84
90s 14,83 10,48 7,10 13,48
00s 14,24 9,75 5,94 12,88
10s 13,94 8,95 4,95 12,35
Gains (+) and losses (-) [mio km²]
80s +1,38 -4,08 -3,38 +6,08
90s +1,35 -4,35 -3,38 +6,38
00s +1,36 -4,49 -3,81 +6,94
10s +1,59 -4,99 -4,00 +7,40
Extent [mio km²]
Feb 15 May 15 Aug 15 Nov 15
80s 15,32 13,27 7,88 11,00
90s 14,96 12,87 7,24 10,54
00s 14,43 12,43 6,33 9,91
10s 13,84 11,99 5,40 9,26
Gains (+) and losses (-) [mio km²]
80s +4,32 -2,05 -5,39 +3,12
90s +4,42 -2,09 -5,63 +3,30
00s +4,52 -2,00 -6,10 +3,58
10s +4,58 -1,85 -6,59 +3,86
From these values I see the sharp decrease in the melting over summer (July/Aug) in the latest decade, but also a pronounced increase in winter and early spring, and - surprisingly - a pronounced slower melting in the period between Feb 15 and May 15 in the latest decade. This is in line with the actual observations of a "slightly slower than average" decrease of sea ice extent in the last weeks.
The generally lower maximum (supposedly more in the periphery than in the CAB) could explain this behaviour (when less ice is available in the more peripherical parts which gain sun and warmth at first then the melting starts a little earlier).
The much lower minima in September also lead to a stronger increase towards the winter which explains the higher gains in the latest decade (compared to the earlier ones).
Just some points to discuss...
kind regards from Germany.