Two measures of the 2026 Arctic sea-ice minimum, one depleted baseline
The 2026 Arctic sea-ice minimum has two different headline numbers. The National Snow and Ice Data Center (NSIDC) estimates 4.60 million square kilometres on September 12, tied for the tenth-lowest minimum in its satellite record. The UK Met Office, using the OSI SAF Sea Ice Index, reports 5.06 million square kilometres on the same date and ranks it fourteenth-lowest in that series.
The difference is not evidence that one agency saw an ice-free sea and the other saw a recovery. They use different satellite-derived records. Small differences in estimating ice cover can change the measured extent, and a cluster of years with similar minima can change the rank. The Met Office explicitly notes this sensitivity when comparing its series with NSIDC's. The two institutions agree on the day of the apparent minimum and on the larger finding: Arctic sea ice remains far below the extent typical of earlier decades.
A rank needs a reference point
The Met Office puts its 2026 minimum 1.51 million square kilometres below the 1981–2010 average minimum in its series. It also reports below-average extent across nearly every part of the Arctic Ocean. NSIDC calls its 4.60-million-square-kilometre estimate preliminary and says wind-driven compaction could still change the season's final minimum.
NASA, reporting the joint NASA–NSIDC analysis, notes that the 20 years from 2007 through 2026 produced the 20 lowest annual Arctic minima in the continuous satellite record. That finding gives context to a year ranked tenth or fourteenth: being less extreme than the very lowest year is not the same as being close to the older baseline.
The satellite figure is extent, not a measurement of all the ice's volume or thickness. It describes the ocean area counted as ice-covered under a dataset's concentration rules. It can move with weather, winds, and the way thin or broken ice is classified. The summer's rank is therefore a useful comparison within a series, but it cannot by itself describe the whole physical condition of the ice.
What the measurement can support
The two records should be quoted with their names attached. “4.60 million square kilometres, NSIDC Sea Ice Index” and “5.06 million square kilometres, OSI SAF Sea Ice Index” are distinct measurements; silently combining one figure with the other's rank would mislead readers. Nor should a single year's minimum be treated as a forecast of the next year. Seasonal weather can move an annual result around a much longer trend.
For people planning around Arctic change, the more consequential comparison is with the system that used to exist. Both monitoring groups place the 2026 minimum substantially below their historical averages. The exact number and rank may be revised as the season closes. The depleted baseline is already visible in either series.