Ice, wind, and water: Synoptic-scale controls of circulation in the Chukchi Sea

Highlights

• Northeasterly winds reverse the flow in Barrow Canyon.

• Atlantic Water is most commonly upwelled to the head of Barrow Canyon in winter.

• Bering Strait inflow influences freeze-up and melt-back of sea ice in Chukchi Sea.

• The Northeast Chukchi polynya has sensible heat and wind-driven components.

• There are dominant patterns of coherent flow across the eastern Chukchi Sea.

Abstract

A composite dataset of 27 moorings across the Chukchi Sea and Bering Strait in 2013–14, along with satellite sea ice concentration data, weather station data, and atmospheric reanalysis fields, are used to explore the relationship between the circulation, ice cover, and wind forcing. We find a clear relationship between northeasterly winds along the northwest coast of Alaska and reversed flow along the length of Barrow Canyon and at a mooring site ∼ 100 km upstream on the northeast shelf. Atlantic Water is frequently upwelled into the canyon during the fall and winter, but is only able to reach the head of Barrow Canyon after a series of long upwelling events. A pair of empirical orthogonal function (EOF) analyses of ice cover reveal the importance of inflow pathways on the pattern of freeze-up and melt-back, and shed light on the relative influence of sensible heat and wind forcing on polynya formation. An EOF analysis of 25 mooring velocity records reveals a dominant pattern of circulation with coherent flow across the shelf, and a secondary pattern of opposing flow between Barrow Canyon and Bering Strait. These are related to variations in the regional wind field.

Introduction

As one of the most productive areas in the global ocean (Grebmeier et al., 2006) and as part of a region currently experiencing a significant decline in sea ice cover (Frey et al., 2015), a better understanding of physical processes in the Chukchi Sea is critical to our understanding of the changing ecosystem dynamics. Timing of melt-back in sea ice controls the seasonal transition from ice algae to phytoplankton as primary producers. An earlier melt-back could shift the ecosystem from a benthic-dominated to a pelagic-dominated regime (Grebmeier et al., 2006, Moore and Stabeno, 2015). Through brine rejection during ice production, the frequency, extent, and duration of polynyas influence the density of winter waters and the depth to which they ventilate the western Arctic (Itoh et al., 2012, Weingartner et al., 1998). The resulting deep convection homogenizes the water column and can resuspend nutrients from the sea floor (Pacini et al., 2019, Pickart et al., 2016). Circulation patterns then control the distribution of nutrients throughout the Chukchi Sea (Pickart et al., 2016), which in turn influences the location and strength of the phytoplankton blooms.

In recent years, a general understanding of the circulation across the Chukchi Sea and regions of likely exchange with the Arctic Basin has emerged (Fig. 1). Flow through Bering Strait is primarily northward, with a higher transport in the summer months (Woodgate et al., 2005a). Northward flow continues across the Chukchi Sea along three major, topographically steered pathways. The coastal pathway, known as the Alaskan Coastal Current (ACC) in the summer, is the most direct route and follows the Alaska coastline across the shelf, draining through Barrow Canyon. The Central Channel pathway flows northward from Bering Strait and subsequently divides into several branches, each eventually turning eastward towards Barrow Canyon (e.g. Pickart et al., 2016). The western pathway is the most circuitous. A portion of this water flows westward through Long Strait into the East Siberian Sea (Woodgate et al., 2005a, Woodgate et al., 2005b). The rest flows around the west side of Herald Shoal into Herald Canyon. A bifurcation just north of the canyon diverts some of the flow eastward, which joins the Central Channel pathway near Hanna Shoal. Thus, a large portion of the Bering Strait inflow ends up flowing through Barrow Canyon. However, transit times vary greatly by pathway, ranging from as little as 2–3 months in the coastal pathway (Tian et al., 2021, Weingartner et al., 1998) to 6–8 months for the northernmost branch of the Central Channel pathway (Spall, 2007). Additionally, each pathway experiences intermittent flow reversals associated with local winds, which increase the transit times.

Several studies have explored the connection between local wind forcing and flow reversals within Barrow Canyon (e.g., Itoh et al., 2013, Weingartner et al., 2017, Lin et al., 2019a, Pisareva et al., 2019). Northeasterly wind along the northwest coast of Alaska drives offshore Ekman transport, which in turn leads to upwelling in the canyon. Such wind events are frequent, especially in fall and winter. Continental shelf waves are also thought to play a role in upwelling in this region (Aagaard and Roach, 1990, Danielson et al., 2014). Episodes of upwelling have been known to draw warm, salty Atlantic Water from the deep Arctic Basin, sometimes far onto the shelf (Bourke and Paquette, 1976, Ladd et al., 2016).

Barrow Canyon is an important choke point of the Chukchi circulation that influences the export into the western Arctic. Roughly half of the annual Bering Strait inflow of Pacific Water drains through the canyon (Itoh et al., 2013). It is thus important to understand the factors controlling the circulation and its variability there. For example, what conditions determine which water masses get upwelled through Barrow Canyon? What portion of the Chukchi shelf is influenced by these upwelling events? Pisareva et al. (2019) found that upwelling sometimes delivers denser water to the head of Barrow Canyon and other times delivers lighter water. They noted that much of the difference was due to strong seasonality of the water masses present on the Chukchi Shelf, i.e. the initial water mass. Unlike the findings of Lin et al. (2019b) for the Alaskan Beaufort Slope, Pisareva et al. (2019) found that upwelling of Atlantic Water to the head of Barrow Canyon occurs only infrequently. However, Itoh et al., (2013) report that Atlantic Water is maintained below 150 m at the mouth of Barrow Canyon year round.

While strong northeasterly wind can force a flow reversal (upwelling) in Barrow Canyon, flow is down-canyon under all other wind directions (Lin et al., 2019a). A sea surface height gradient (pressure head) between the Pacific and Arctic Oceans is the primary driver of northward flow through Bering Strait (Coachman and Aagaard, 1966, Woodgate et al., 2005b). The signal of northward flow is largely coherent across the eastern Chukchi Sea (Woodgate et al., 2005b). However, the northward flow is opposed by the mean wind in the region, i.e. northeasterly wind along the northwest coast of Alaska (Pisareva et al., 2019) and northerly wind in Bering Strait (Woodgate et al., 2005b). Woodgate (2018) found that the northward transport through Bering Strait has been increasing in recent years and attributes that trend to an increase in the pressure head. Danielson et al. (2014) propose that the increase in sea surface height on the Bering Sea end of the Strait is due to an eastward shift in the mean position of the Aleutian Low. Danielson et al. (2014) also present evidence of northward propagating shelf waves, which follow the Alaska coastline, and can increase or decrease Bering Strait transport on synoptic time scales, with the sign of the velocity signal depending on the direction of wind that sets up the shelf wave. If flow through Bering Strait and across the Chukchi Sea is primarily driven by the pressure head, but strongly modulated by local winds, do local winds near Bering Strait and near Barrow Canyon always act in concert? What sort of dynamical response might there be if/when they do not?

Although there has only been a slight warming in Bering Strait inflow waters, because of the increased volume transport there has been a significant increase in heat transport through the strait (Woodgate, 2018). Serreze et al. (2016) found heat transport through Bering Strait to be the strongest predictor in timing of both spring ice retreat and fall ice advance. This additional heat flux can also promote sea ice thinning across much of the western Arctic Ocean (Woodgate et al., 2015). How apparent is the effect of seasonal variations in Bering Strait inflow on the spatial pattern of sea ice?

Polynyas are a common occurrence in the Arctic. As stated previously, northeasterly winds in the region of Barrow Canyon drive offshore Ekman transport that results in upwelling. Similar wind conditions can also drive offshore ice transport along the northwest coast of Alaska. Such regions of ice divergence along a coastline are known as wind-driven, or latent heat, coastal polynyas (Morales Maqueda et al., 2004). Because the water column in a wind-driven polynya remains at the freezing temperature, new ice is readily formed at the surface. Continued offshore transport of ice makes this type of polynya an ice production zone and contributes to the densification of the water column. Alternatively, sensible heat polynyas are formed when warm ocean waters are introduced to an ice-covered region, melting the existing ice and preventing new ice from forming. The location, extent, and duration of such polynyas are dependent upon the same characteristics of the warm water mass (Morales Maqueda et al., 2004). Hirano et al., 2016, Hirano et al., 2018) provide evidence that the recurring polynya in the vicinity of Barrow Canyon is a hybrid latent and sensible heat polynya, influenced by both wind-driven ice divergence and upwelling of warm waters through Barrow Canyon. However, their analysis shows the influence of sensible heat on the polynya is limited to localized areas very near the coastline. Other investigators (e.g. Ladd et al., 2016) have suggested a much larger extent of warm water influence.

In this study we use data from an extensive set of moorings deployed across the Chukchi Sea in 2013–14, from Bering Strait to the western Beaufort Sea, to address some of the above questions. This allows us to explore the coupled nature of the flow across the shelf in relation to the wind forcing and the sea ice concentration. We begin with a description of the data sources utilized and the method used for identifying wind events in Section 2. In Section 3, we give a brief description of the mean flow over the study year. In Section 4, we explore the flow and water mass response in Barrow Canyon to northeasterly wind events. Section 5 investigates spatial patterns in sea ice on regional and local scales and their relationships with potential forcing mechanisms. Shelf-wide circulation patterns and their relationship to regional wind patterns are explored in Section 6. A summary of our results is presented in Section 7.