The meridional shifts of the Oyashio Extension (OE) and of the Kuroshio Extension (KE), as derived from high-resolution monthly sea surface temperature (SST) anomalies in 1982-2008 and historical temperature profiles in 1979-2007, respectively, are shown based on lagged regression analysis to significantly influence the large-scale atmospheric circulation. The signals are independent from the ENSO teleconnections, which were removed by seasonally varying, asymmetric regression onto the first three principal components of the tropical Pacific SST anomalies. The response to the meridional shifts of the OE front is equivalent barotropic and broadly resembles the North Pacific Oscillation/western Pacific pattern in a positive phase for a northward frontal displacement. The response may reach 35 m at 250 hPa for a typical OE shift, a strong sensitivity since the associated SST anomaly is 0.5 K. However, the amplitude, but not the pattern or statistical significance, strongly depends on the lag and an assumed 2-month atmospheric response time. The response is stronger during fall and winter and when the front is displaced southward. The response to the northward KE shifts primarily consists of a high centered in the northwestern North Pacific and hemispheric teleconnections. The response is also equivalent barotropic, except near Kamchatka, where it tilts slightly westward with height. The typical amplitude is half as large as that associated with OE shifts.

Seven ice mass balance instruments deployed near 83°N on different first‐year and second‐year ice floes, representing variable snow and ice conditions, documented the evolution of snow and ice conditions in the Arctic Ocean north of Svalbard in January–March 2015. Frequent profiles of temperature and thermal diffusivity proxy were recorded to distinguish changes in snow depth and ice thickness with 2 cm vertical resolution. Four instruments documented flooding and snow‐ice formation. Flooding was clearly detectable in the simultaneous changes in thermal diffusivity proxy, increased temperature, and heat propagation through the underlying ice. Slush then progressively transformed into snow‐ice. Flooding resulted from two different processes: (i) after storm‐induced breakup of snow‐loaded floes and (ii) after loss of buoyancy due to basal ice melt. In the case of breakup, when the ice was cold and not permeable, rapid flooding, probably due to lateral intrusion of seawater, led to slush and snow‐ice layers at the ocean freezing temperature (−1.88°C). After the storm, the instruments documented basal sea‐ice melt over warm Atlantic waters and ocean‐to‐ice heat flux peaked at up to 400 W m−2. The warm ice was then permeable and flooding was more gradual probably involving vertical intrusion of brines and led to colder slush and snow‐ice (−3°C). The N‐ICE2015 campaign provided the first documentation of significant flooding and snow‐ice formation in the Arctic ice pack as the slush partially refroze. Snow‐ice formation may become a more frequently observed process in a thinner ice Arctic.