Increases in the Pacific inflow to the Arctic from 1990 to 2015, and insights into seasonal trends and driving mechanisms from year-round Bering Strait mooring data

Year-round in sins Bering Strait mooring data (1990-2015) document a long-term increase (0.01 Sv/yr) in the annual mean transport of Pacific waters into the Arctic. Between 2002 and 2015, all annual mean transports (except 2005 and 2012) are greater than the previously accepted climatology (-0.8 Sv). The record-length maximum (2014: 1.2 0.1 Sv) is 70% higher than the record-length minimum (2001: 0.7 0.1 Sv), corresponding to a reduction in the flushing time of the Chukchi Sea (to-4.5 months from-7.5 months). The transport increase results from stronger northward flows (not fewer southward flow events), yielding a 150% increase in kinetic energy, presumably with impacts on bottom suspension, mixing, and erosion. Curiously, we find no significant trends in annual mean flow in the Alaskan Coastal Current (ACC), although note that these data are only available 2002-2015. Record-length trends in annually integrated heat and freshwater fluxes (primarily driven by volume flux trends) are large (0.06 0.05 x 1020 J/yr; 30 20 km3/yr; relative to-1.9 C and 34.8 psu), with heat flux lows in 2001 and 2012 (3 x 1020 J) and highs in 2007 and 2015 (5.5 x 102 J), and a freshwater range of 23001=3 (2001) to-3500 km3 (2014). High-flow year 2015 (volume transport-1.1 Sv) has the highest annual mean temperature recorded,-0.7 C, astoundingly warmer than the record-length mean of 0.0 0.2 C, while low flow year 2012 (0.8 Sv) is also remarkably cold (-0.6 C), likely due to anomalously weak northward flow in January-March, partly driven by anomalously strong southward winds in March. A seasonal decomposition of properties of the main flow shows significant freshening in winter (-0.03 psu/yr, January-March) likely due to sea-ice changes, but no trend (or perhaps salinization) in the rest of the year. A seasonal warming trend in the strait proper in May and June (0.04 C/yr) is reflected in a trend to earlier arrival (0.9 0.8 days/yr) of waters warmer than 0 C. Contrastingly, no significant trend is found in the time of cooling of the strait. The strait's seasonal increasing transport trends (-0.02 Sv/yr) are largest from May-November, likely due to the large wind-driven variability masking the signal in other months. We show that Ekman set-up of waters along the coast in the strait can explain the strong correlation of the water velocity with along-strait winds (as opposed to across-strait winds). We highlight the strong seasonality of this relationship (r-0.8 in winter, but only-0.4 in summer), which reflects the weak influence of the (seasonally weak) winds in summer. By separating the flow into portions driven by (a) the local wind and (b) a far field (Pacific-Arctic pressure-head) forcing, we find the increase in the Bering Strait throughflow is primarily due to a strong increase in the far-field forcing, not changes in the wind. We propose a higher annual mean transport for the strait for the 2000s, (1.0 0.05 Sv) based on recent flow increases, and present estimated seasonal climatologies for properties and fluxes for the strait and for the ACC. Heat and freshwater seasonalities are strongly influenced by the ACC and stratification. Finally we conclude that year-round in situ mooring are still the only currently viable way of obtaining accurate quantifications of the properties of the Pacific input to the Arctic.