North Atlantic sea surface temperatures (SSTs) are crucial for Northern Hemisphere climate. Understanding the drivers of SST changes, particularly in the context of anthropogenic climate change, is vital. Recent decades have witnessed a notable warming trend across much of the North Atlantic, a consequence of rising greenhouse gas concentrations. However, climate models also predict a counterintuitive phenomenon: significant future cooling at mid-latitudes, known as the North Atlantic warming hole. This pattern has been observed in recent decades, raising the question of its origin: is it a product of human activity or simply a fluctuation within the climate system's natural variability? This research directly addresses this question using advanced climate models and comprehensive observational datasets to formally detect and attribute the observed North Atlantic warming hole to anthropogenic forcing. The study's significance lies in its ability to isolate the human impact on this complex climate pattern, separating it from inherent internal climate variability. This separation is critical for refining climate projections and for understanding the influence of human actions on regional and global climate dynamics.

Previous research has extensively investigated the link between SST patterns in the North Atlantic and changes in oceanic circulation. Several studies have explored the relationship between the Atlantic Meridional Overturning Circulation (AMOC) and the warming hole, proposing a weakening AMOC as a contributing factor. Others have focused on the role of atmospheric processes and their interaction with ocean currents. However, these studies often lacked a formal detection-attribution analysis to definitively determine the extent to which the observed warming hole pattern is a result of anthropogenic influences as opposed to natural variability. This paper builds upon this existing literature by providing a rigorous detection-attribution analysis to isolate the human fingerprint in the North Atlantic SST trends.

The study employs a combination of observational datasets and state-of-the-art climate model simulations. Two high-resolution SST datasets—NOAA SST and HadISST—combining satellite and in-situ measurements, were used to depict recent (1982–2017) North Atlantic SST trends. These observations were compared against simulations from two large ensembles of climate models: the Community Earth System Model 40-member Large Ensemble (CESM-LE) and the Max Planck Institute Earth System Model 100-member Grand Ensemble (MPI-GE). Both ensembles were run under historical and RCP8.5 forcing scenarios. To account for potential model biases, simulations from the Coupled Model Intercomparison Project Phase 5 (CMIP5) were also included. The analysis employed a fingerprint detection method, calculating the signal-to-noise ratio (SNR) to formally identify the external forced signal within the observed SST data. The leading empirical orthogonal function (EOF) of the 1982–2017 North Atlantic SST anomalies from each model ensemble was used to define the external forced fingerprint. Trends of increasing lengths of the projection of the observed SST anomalies onto these fingerprints were calculated to represent the signal. The noise was determined using SST data from pre-industrial control runs. The forced fingerprint was considered statistically detectable when the signal exceeded a 5% significance threshold relative to the noise. To attribute the fingerprint to specific forcing agents, four additional CESM-LE ensembles were used, each with one forcing agent (greenhouse gases, aerosols, biomass burning, land use/land change) held constant. Finally, the study investigated the underlying mechanisms of the warming hole by analyzing the CESM-LE mixed-layer temperature tendency equation, examining the roles of air-sea heat fluxes, zonal and meridional advection, and vertical heat transfers. This involved decomposing the mixed-layer temperature tendency into its contributing terms and analyzing their temporal evolution.

The analysis revealed a clear warming hole pattern in both observational datasets and model ensembles, characterized by stronger warming at high and low latitudes and weaker warming (or cooling) at mid-latitudes. The forced response in CESM-LE and MPI-GE was found to be similar to the observed pattern, although the cooling in the model means was stronger. The CMIP5 multi-model mean also displayed a similar, though weaker and poleward-shifted, warming hole pattern. Fingerprint detection analysis showed a statistically significant anthropogenic signal in the observed SSTs, emerging from the internal variability around the year 2000. The signal was detectable in nearly all individual members of the CESM-LE and MPI-GE after 2000. Attribution analysis using the CESM-LE ensembles with fixed forcing agents indicated that the observed fingerprint was primarily attributable to greenhouse gas emissions. The warming hole pattern was found to persist through the end of the 21st century in all three ensembles. Analysis of the mixed-layer temperature tendency equation revealed that the main balance in the CESM-LE simulations was between air-sea heat fluxes and meridional heat advection. A decline in meridional heat advection, primarily associated with greenhouse gas emissions, was identified as a major factor in creating the warming hole. This decline in meridional heat advection was also largely attributable to greenhouse gas emissions, aligning with previous findings on the weakening ocean circulation.

The findings demonstrate that the North Atlantic warming hole, a seemingly paradoxical feature given the overall warming trend, is indeed a consequence of anthropogenic greenhouse gas emissions. This is a critical finding, as it highlights the complex and sometimes counterintuitive ways in which human activities are affecting the climate system. The results emphasize the importance of considering regional variations in climate change alongside the global average temperature increase. The study's attribution analysis firmly establishes the link between greenhouse gas emissions and the observed SST pattern. The identified mechanism—a decline in meridional heat advection—confirms previous studies linking this pattern to changes in ocean circulation. The persistence of the warming hole pattern into the future, as projected by the models, underscores the continued significance of this phenomenon and its potential impact on regional climate. This research contributes to a more comprehensive understanding of anthropogenic climate change, providing robust evidence of its influence on regional climate patterns and highlighting the intricate interplay between atmospheric and oceanic processes.

This study provides strong evidence that the observed North Atlantic warming hole is of anthropogenic origin, primarily driven by greenhouse gas emissions and associated with a decline in northward oceanic heat flux. The formal detection-attribution analysis definitively links this complex regional climate pattern to human activity. Future research should focus on improving the representation of ocean circulation and air-sea interactions in climate models to better project the evolution of the warming hole and its potential consequences for regional climate and weather patterns. Further investigation into the potential feedback mechanisms and broader impacts of the warming hole is also warranted.

The study relies heavily on climate model simulations. While the use of large ensembles helps reduce uncertainty associated with internal climate variability, potential biases in model parameterizations and physical representations could still influence the results. The analysis of the mixed-layer temperature tendency equation is based solely on the CESM-LE model, limiting the generalizability of the mechanism identification to other models. The observational data used has a limited time span (1982–2017), potentially restricting the ability to identify longer-term trends and natural variability.