Quantifying the internal variability in Arctic Amplification and identifying the associated spatial trend pattern using machine learning

Recent studies have shown that the magnitude of Arctic Amplification (AA) has increased since the beginning of the 21^st century based on observational data. AA is a robust feature in state-of-the-art climate models, but model simulations rarely reproduce the observed levels of AA for 1980–2022. The ratio of polar and global surface warming yields an AA index of 3.9–4.6, which exceeds the magnitude of AA in over 90% of model ensemble members (the mean from each large ensemble ranges from 2.13 to 3.58). We found that natural fluctuations in the climate system that arise from internal variability have reduced global warming and accelerated Arctic warming which results in enhanced AA that exceeds what model simulations typically produce. After quantifying and removing the effects of internal variability using a machine learning algorithm, the mean AA value is 3.02 based on 4 observational data sets, which is consistent with multi-model ensemble means (2.13 to 3.58). A further study identified a unique spatial pattern associated with the1980–2022 internal variability that manifested as Global Cooling and Arctic Warming (i-GCAW). The i-GCAW pattern exhibit a unique and robust surface air temperature (SAT) trend pattern, which features enhanced warming in the Barents and Kara Sea and cooling in the Tropical Eastern Pacific and Southern Ocean. This pattern is imprinted on observed surface trends and indicates that internal variability is a non-negligible contributor to differences in the pattern of simulated and observed warming.