Local winds drive interannual variability of the Gulf Stream North Wall: Results from an adjoint sensitivity analysis

The Gulf Stream (GS) is bounded to the north by a strong temperature front known as the North Wall (NW). The NW’s dominant mode of interannual variability is a longitudinally coherent north–south migration. Fluctuations in NW position have important consequences for regional climate because the GS transports considerable heat from the ocean at low latitudes to the atmosphere at high latitudes. Numerous studies have shown that the NW’s latitude is correlated with the North Atlantic Oscillation (NAO), with the NW north of its mean position when the NAO is anomalously strong. Baroclinic Rossby waves are commonly invoked as the mechanism linking the NAO to the NW. The long crossing time of baroclinic Rossby waves (4–9 years) compared to the reported lag between NAO anomalies and NW shifts (0–2 years), suggests that this explanation will only work if the Rossby waves originate very close to the GS.

The sensitivity of NW latitude to atmospheric forcing is examined in the ECCOv4(r2) ocean state estimate using adjoint and forward perturbation experiments. While the seasonal cycle of NW latitude is largely driven by buoyancy forcing (primarily heat flux) over and upstream of the GS Extension, interannual variability is primarily sensitive to wind stress curl anomalies due to zonal wind perturbations. Consistent with previous ideas about Rossby waves linking the NAO and the NW, the sensitivity pathways are organized into beams of westward-propagating Rossby waves. However, a dynamical reconstruction of NW variability reveals that the NAO typically projects incoherently onto the sensitivity pathways over most of the Atlantic; that is, different regions contribute opposing tendencies to NW latitude. Any small net tendency remaining propagates toward the GS so slowly that it is often canceled by opposing NAO variability in subsequent years. The area southeast of Newfoundland, near the extreme eastern edge of the GS, is a notable exception—wind stress curl anomalies produced there propagate rapidly up the GS, affecting the entire NW within 0–2 years. While most previous studies overlooked this region due to the relatively small projection of the NAO loading pattern, we show that anomalies located there are the only ones capable of producing a coherent and statistically significant response in the latitude of the NW.