No evidence for attributing nearly four-fold Arctic Amplification to internal variability

The Arctic has been warming nearly four times faster than the rest of the planet over the past decades. Zhou et al. [1] argue that only threefold Arctic amplification is due to external forcing, and the remainder can be attributed to internal variability, in particular an ‘Arctic mode’ of surface temperature changes. Here, we argue that the analysis presented in the paper does not support this conclusion. The analysis period used by Zhou et al. [1] starts with 15-20 years of no observed warming in the Arctic (1965-1980ies). By subtracting a linear trend from the entire timeseries (1965-2022), the authors introduce an early cooling trend and a later warming trend into their ‘detrended’ timeseries. Our analysis suggests that this spurious trend is at the core of what is described as ‘Arctic mode’ internal variability in the main paper. Using synthetic data with only a linear trend in the Arctic starting 20 years later than the analysis timeframe, the method of [1] suggests the existence of an ‘Arctic mode’ in the synthetic data. As in the real data, this spurious ‘Arctic mode’ initially dampens and then increases Arctic warming from an apparently constant ‘forced trend’. We conclude that the statistical analysis in [1] provides no evidence for attributing the recently observed high rates of Arctic amplification to internal variability.