%0 Journal Article %1 kolstad2024causal %A Kolstad, Erik W. %A O'Reilly, Christopher H. %D 2024 %J Climate Dynamics %K natl-var %N 5 %P 4223--4236 %R 10.1007/s00382-024-07128-y %T Causal oceanic feedbacks onto the winter NAO %U https://doi.org/10.1007/s00382-024-07128-y %V 62 %X Of the climate variability patterns that influence the weather in the North Atlantic region in winter, the North Atlantic Oscillation (NAO) is the most dominant. The effects of the NAO span from cold air outbreaks to unseasonably warm conditions and unusual precipitation, with significant impacts on human activities and ecosystems. While a connection between the NAO and antecedent sea surface temperature (SST) conditions has been recognised for decades, the precise causal interaction between the ocean and the atmosphere remains enigmatic. In this study we uncover a robust statistical relationship between North Atlantic SSTs in November and the NAO throughout the subsequent winter in the extended ERA5 reanalysis back to 1940. We apply a well-established causal inference technique called mediation analysis, commonly used in social science and now adopted in climate research. This analysis highlights the roles of low-level baroclinicity, latent heat fluxes, and latent heat release in mediating the effect of November SSTs on the NAO in January and February. It is important to recognise that these mediators are interrelated. Moreover, our analysis reveals bidirectional relationships, where the NAO reciprocally mediates the effects of the November SSTs on these variables. This is evidence of a complex web of feedback mechanisms which collectively contribute to the response of the winter NAO to late autumn/early winter SSTs.

@article{kolstad2024causal, abstract = {Of the climate variability patterns that influence the weather in the North Atlantic region in winter, the North Atlantic Oscillation (NAO) is the most dominant. The effects of the NAO span from cold air outbreaks to unseasonably warm conditions and unusual precipitation, with significant impacts on human activities and ecosystems. While a connection between the NAO and antecedent sea surface temperature (SST) conditions has been recognised for decades, the precise causal interaction between the ocean and the atmosphere remains enigmatic. In this study we uncover a robust statistical relationship between North Atlantic SSTs in November and the NAO throughout the subsequent winter in the extended ERA5 reanalysis back to 1940. We apply a well-established causal inference technique called mediation analysis, commonly used in social science and now adopted in climate research. This analysis highlights the roles of low-level baroclinicity, latent heat fluxes, and latent heat release in mediating the effect of November SSTs on the NAO in January and February. It is important to recognise that these mediators are interrelated. Moreover, our analysis reveals bidirectional relationships, where the NAO reciprocally mediates the effects of the November SSTs on these variables. This is evidence of a complex web of feedback mechanisms which collectively contribute to the response of the winter NAO to late autumn/early winter SSTs.}, added-at = {2024-06-04T10:42:48.000+0200}, author = {Kolstad, Erik W. and O'Reilly, Christopher H.}, biburl = {https://www.bibsonomy.org/bibtex/24c0ef33eb9e6564501d29fc925b62190/laura.jackson}, day = 01, description = {Causal oceanic feedbacks onto the winter NAO | Climate Dynamics}, doi = {10.1007/s00382-024-07128-y}, interhash = {1284a52597342ac459a603269b576ce3}, intrahash = {4c0ef33eb9e6564501d29fc925b62190}, issn = {1432-0894}, journal = {Climate Dynamics}, keywords = {natl-var}, month = may, number = 5, pages = {4223--4236}, timestamp = {2024-06-04T10:42:48.000+0200}, title = {Causal oceanic feedbacks onto the winter NAO}, url = {https://doi.org/10.1007/s00382-024-07128-y}, volume = 62, year = 2024 }