A New Northwest Passage: Tracing Atlantification through the Amerasian Arctic

Abstract

The Arctic Ocean has undergone rapid warming, resulting in sea ice decline, displacement ofpolar species, and shifts in the seasonal cycles of Arctic primary producers. A major driver of these changes is ”Atlantification”, a process where increased amounts of heat and/or volumes of water from the Atlantic Ocean progress deeper into the Arctic Ocean. However, sparse observations make it difficult to assess how Atlantification has progressed over time, particularly in under observed regions like the Amerasian Basin. This work attempts to address this issue by utilizing new mesoscalepermitting ocean state estimates (reanalyses), which combine observations with numerical models, to address questions about how the transport of heat into the Amerasian Arctic has changed over time, what the drivers of this change are, and the impact it is having on the Arctic surface mixed layer. The first section of the work analyzes the consistency of four new reanalyses with observations of Atlantic Water (AW) structure, finding that only two of the reanalyses adequately reconstruct the observed AW layer both upstream of the Amerasian Basin and within the Amerasian Basin. From the reanalyses which are consistent with observations, we derive a 40-year time series of heat transport into the Amerasian Basin, finding that heat transport has increased by 0.36 TW/year, leading to a 25% increase in total Amerasian Basin heat content. Overlaid on this long-term trend is significant decadal and sub-decadal variability. The decadal variability is consistent with changes in the strength of the Arctic Dipole Anomaly (an atmospheric pattern associated with an enhanced Beaufort High and Laptev Sea low). A strengthening of the Arctic Dipole since the late 2000s has increased the amount of heat transport into the Amerasian Basin but has insulated the Beaufort Gyre region from this additional heat. We also find that accurately resolving the upstream heat loss in the Barents Sea is crucial for accurately modeling AW structure downstream in the Amerasian Basin. In the second part of the work, we further assess the upstream impacts of the Barents Sea using the SODA4 reanalysis. In doing so, we find that heat transport into the Barents Sea has increased by 0.23 TW/year since 1980, but that the heat transport out of the Barents Sea and into the Eurasian Basin has only increased in 0.11 TW/year over the same period. This indicates that a ”Barents Sea Cooling Machine” is responsible for removing a significant portion of the excess heat from the inflowing AW. However, the cooling machine is unable to completely keep pace. We also observe that the increase in heat transport is entirely the result of increases in temperature rather than in the volume transport, and that downstream changes in the heat transport into the Amerasian Basin result from the increases in heat transport out of the Barents Sea rather than from the Fram Strait, which has no significant trend in heat transport. We also show the first evidence of an ”ocean feedback” component of the Barents Sea Cooling Machine, which contributes to heat loss from AW in the Barents Sea. In the final part of the work, we examine changes in the Arctic surface mixed layer. Prior studies have found conflicting trends in the Mixed Layer (ML) depth making it unclear how the mixed layer is changing and what this means for the entrainment of heat from AW into the surface mixed layer. We find that overall, the winter ML has shoaled since the 1980s while the summer ML has deepened, but changes are most significant in the winter months. We also derive a time series of winter ML depth anomalies from 1980-2024 which can be found to reconcile nearly all of the prior observed trends. This is a result of significant interdecadal variability in the winter ML depth. This variability appears to have its source in the thinning of sea ice over the Canadian Basin and Central Amerasian Basin, which allows for stronger wind stresses on the ocean surface and potential for increased entrainment of heat from the AW layer into the surface mixed layer.