Typically in climate observations, averages are referenced without paying attention to the high degree of component variability from year to year, and over longer time periods. Mid May is when the Spring melt is well underway, but with the Arctic core still frozen solid. Yet the animation above shows on day 135 over the last 15 years, there are considerable differences as to how much ice is in which regions.
On the bottom left is Bering Sea which had ice extents on this day ranging from a high of 682k km2 (2012) to a low of 38k km2 (2018). The day 135 average for Bering is 293k km2, but with a standard deviation of 192k (65%). Okhotsk center left is the next most variable, from 290k (2012) to 99k (2019), averaging 188k with std. deviation of 63K (33%). Barents Sea center top has a large variability from 568k km2 (2009) to 223k (2012), averaging 422k km2 +/- 111 k km2. Other Arctic regions vary little on this day from year to year. For example, Hudson Bay is close to 1.2M km2 every year on day 135.
The effect on NH total ice extents is presented in the graph below for the period mid April to mid May, comparing the 14-year average with 2021 MASIE and SII, and some other years of interest.
Note on average this period shows an ice loss of 1.5M km2. MASIE 2021 is about 200k km2 below average, 1.6% down, or having the same total extent 3 days ahead of average. Interestingly, SII shows about 200k higher, matching the MASIE average for day 135.
The table below shows the distribution of sea ice across the Arctic regions.
|Region||2021135||Day 135 Average||2021-Ave.||2007135||2021-2007|
Overall NH extent March 31 was below average by 200k km2, equivalent to the deficit in Baffin Bay. Elsewhere smaller deficits were offset with surpluses. The onset of spring melt is as usual in most regions.