Table 1.
Overview of published sensitivities of the September sea-ice cover to global warming
| Sensitivity | Ice-free at | Sea-ice data | Temperature data | Time period | Ref |
|---|---|---|---|---|---|
| [106 km2/K] | |||||
| Based on sea-ice extent | |||||
| − 5.7+ | + 0.6 ∘C | NSIDC sea-ice index | GISTEMP | 1979–2013 | [13] |
| − 4.4 | + 0.8 ∘C | NSIDC sea-ice index | GISTEMP | 1979–2014 | [40] |
| Based on sea-ice area | |||||
| − 2.6 | + 0.9 ∘C | HadISST 1∗ | GISTEMP | 1979–2007 | [12] |
| − 4.1 | + 0.6 ∘C | Sea-ice index & HadISST 1∗ | HadCRUT4 | 1953–2016 | [36] |
| − 3.3 | + 0.7 ∘C | HadISST 2.2 | GISTEMP | 1953–2016 | [36] |
The first column is the estimated observed sea-ice loss per degree of global warming. The second column is the additional warming above present levels needed to obtain an Arctic sea-ice coverage of less than 1 million km2 (based on average September sea-ice extent of past ten years of 4.7 million km2 and average September sea-ice area of 3.3 million km2 [1]). ∗Note that [41] identified an inconsistency in the HadISST 1 sea-ice area, and recommended that it should only be used when merged with a consistent satellite record. +This estimate is based on the gradient ratio of the temporal trend of global-mean warming and of the temporal trend of sea-ice coverage. All other estimates are based on ordinary regression of sea-ice coverage on global-mean temperature (see [11] for details)