The Gulf of Maine, a semi-enclosed sea on the east coast of North America, has recently experienced significant warming, exceeding the global average ocean warming rate. This rapid warming, marked by increased sea surface temperatures (SSTs), extreme marine heatwaves, and subsurface temperature increases, has severely impacted the region's ecosystems and fisheries. The decline of several species, including right whales and cod, and changes in lobster landings, highlight the profound ecological and economic consequences. Multiple factors likely contribute to this warming, including increased atmospheric temperatures and pressures, changes in Gulf Stream warm core rings and western destabilization point, and internal climate variability. The Gulf Stream's position relative to the Gulf of Maine is implicated as a key driver, with a more northerly Gulf Stream leading to warmer water throughout the water column. The strength of the Atlantic Meridional Overturning Circulation (AMOC) is also linked to the Gulf Stream's position; a weaker AMOC allows the Gulf Stream to shift closer to the Gulf of Maine. Previous studies, using limited data (short instrumental records or discontinuous proxy data from few specimens), have indicated long-term cooling over the last millennium. However, the lack of continuous, high-resolution data has made understanding the onset and uniqueness of recent warming challenging. This research aims to address this gap using multiple geochemical proxies to reconstruct the water mass history of the Gulf of Maine over the last 300 years, placing this variability within the context of the past 1000 years using climate model simulations, ultimately pinpointing the start of recent warming and its potential triggers.

Existing literature points to a complex interplay of factors driving temperature changes in the Gulf of Maine. Studies such as Pershing et al. (2015) have linked rapid warming to the collapse of the cod fishery, while others (Pershing et al., 2018; Chen et al., 2014; Record et al., 2019; Thomas et al., 2017) have explored the impact on various marine species and ecosystems. The role of the Gulf Stream and its relationship to the AMOC has been extensively discussed (Zhang & Vallis, 2007; Thibodeau et al., 2018; Karmalkar & Horton, 2021; Caesar et al., 2018; Saba et al., 2016; Neto et al., 2021), with analyses showing links between Gulf Stream position and the regional warming. A significant portion of the previous research relied on relatively short instrumental records (<40 years), creating limitations in understanding long-term trends. The longest existing instrumental SST record from Boothbay Harbor, Maine (since 1905), shows warming since its inception, but provides limited insights into pre-1905 conditions. Previous paleoceanographic work (Wanamaker et al., 2008) utilizing oxygen isotopes in *Arctica islandica* shells indicated millennial-scale cooling in the Gulf of Maine, consistent with other western North Atlantic records, but lacked continuous long-term data. This study builds upon these previous findings, aiming for a more comprehensive and detailed understanding of the region's hydrographic history.

This study employs a multi-proxy approach using geochemical data from *Arctica islandica* shells collected from the western Gulf of Maine. The shells, known for their longevity and annual growth increments, provide high-resolution records of past environmental conditions. Three geochemical proxies were analyzed: oxygen isotopes (δ¹⁸O), reflecting seawater temperature and salinity; nitrogen isotopes (δ¹⁵N), indicating water mass source; and previously published radiocarbon isotopes (Δ¹⁴C), further characterizing water mass origins. These geochemical records were absolutely dated using cross-dating techniques, extending from 1694 to 2013 CE for δ¹⁸O (annually resolved with some gaps), 1751 to 2008 CE for δ¹⁵N (decadally resolved), and 1685 to 1935 CE for Δ¹⁴C (decadally resolved). The oxygen isotope data was validated by comparing it to various instrumental temperature records (Boothbay Harbor SST record and gridded data products like ERSST, HadISST, OISST, and EN4) and significant correlations were observed. To contextualize these geochemical data, the study utilized climate model simulations from the Community Earth System Model-Last Millennium Ensemble (CESM-LME). The CESM-LME includes fully-coupled simulations with 13 ensemble members under full forcing conditions (all external climate forcings), as well as individual forcings (greenhouse gas, volcanic, solar, and ozone/aerosol). The model output, specifically annual average temperature at 35 m depth in the Gulf of Maine, was analyzed to compare with geochemical data. Segmented regression analysis was employed to identify breakpoints indicating shifts in trends for both geochemical and climate model data. Monte Carlo simulations were performed on the climate model output to determine the statistical significance of the recent warming trend compared to other 100-year periods in the past millennium. Finally, the study analyzed broader-scale North Atlantic conditions during extreme temperature years (top and bottom 10% of annual temperatures) in the Gulf of Maine, to investigate links between regional conditions and large-scale ocean dynamics. The AMOC strength, also derived from the CESM-LME simulations, was analyzed to explore its potential role in Gulf of Maine temperature variability.

The geochemical records reveal a multi-century cooling trend in the Gulf of Maine before the mid-1800s, followed by a rapid warming trend beginning in the late 1800s. The δ¹⁸O record shows a breakpoint in 1875, marking the shift from cooling to warming. The δ¹⁵N and Δ¹⁴C records support this observation. Climate model simulations (CESM-LME) corroborate this shift, showing a long-term cooling trend reversed around 1904 (ensemble mean). This reversal of a multi-century cooling trend is largely consistent between geochemical records and climate model simulations. The most recent century (1901-2000) shows the fastest warming rate (0.44 °C century⁻¹) compared to almost any other 100-year period in the last 1000 years. Analysis of single-forcing climate model runs indicates that volcanic forcings primarily drove the pre-industrial cooling trend, while the post-1850 warming is largely attributable to greenhouse gas forcings, although partially offset by ozone and aerosol cooling. The multiproxy geochemical approach (δ¹⁸O, δ¹⁵N, Δ¹⁴C) helps track water mass changes. Before the 1870s, an increase in Labrador Slope Water (LSW) was observed, contributing to colder conditions, potentially amplified by the negative NAO in the late 1800s. After the 1870s, an increase in warmer and less saline Warm Slope Water (WSW) signals changes in the Gulf Stream position and potentially a northward shift, consistent with a weakening AMOC before recent strengthening. Analysis of climate model simulations during extreme warm and cold years shows changes in upper ocean temperature, salinity, and current velocity patterns in the Gulf Stream region, which likely influenced WSW influx and hence Gulf of Maine temperature. However, it is important to note that the AMOC simulations in this study did not support a recent significant weakening of the AMOC, instead indicating strengthening in the mid-1800s prior to subsequent weakening.

The results demonstrate the successful use of multiproxy geochemical records to reconstruct the hydrographic variability in the Gulf of Maine over the last three centuries. The combined data reveal a significant climatic shift in the late 19th century, marking the end of a long-term cooling phase and the beginning of a warming trend consistent with climate model simulations. The findings highlight the crucial role of volcanic forcings in pre-industrial climate variability and the dominance of greenhouse gas forcing in the recent warming. The observed changes in water mass proportions indicate significant shifts in the North Atlantic circulation, with an increase in WSW influx potentially linked to a northward shift of the Gulf Stream. While the AMOC shows some relation to Gulf of Maine temperatures over the last millennium, suggesting a possible role in amplifying volcanically forced changes, the results do not clearly indicate that AMOC weakening is the primary driver of the recent warming. Instead, increased greenhouse gas concentrations, likely influencing other factors such as the Gulf Stream's position, seem to play a more significant role. The discrepancies between the model simulations and some historical events (e.g., tilefish die-off) might point to the need for higher-resolution model simulations to fully capture the complexity of the interaction between multiple driving forces in the region.

This study provides a robust, long-term perspective on Gulf of Maine hydrographic variability using a multi-proxy approach combined with climate model simulations. The findings highlight the influence of both volcanic and anthropogenic forcings on regional climate, with greenhouse gas forcing being the dominant factor in recent warming. The results support the understanding that the recent warming trend is unprecedented in the last millennium. Future research could focus on improving the resolution of climate model simulations to better capture the detailed interplay of different factors, particularly the interactions between the Gulf Stream, AMOC, and atmospheric circulation patterns. Further work is also needed to fully characterize the impacts of anthropogenic nitrogen deposition on the nitrogen isotope records.

The study's limitations primarily involve the limited temporal extent of some data. The radiocarbon record, crucial for tracking water mass changes, does not extend beyond the 1930s due to the effects of nuclear bomb testing. Furthermore, while the study correlates oxygen isotopes with instrumental temperature data, variations in the correlation coefficients across different datasets highlight potential uncertainties and the need for careful interpretation of the oxygen isotopes as temperature proxies. The resolution of the climate model used also impacts the level of detail in interpreting finer-scale changes in circulation patterns within the Gulf of Maine itself.