Workshop on Observing, Modeling, and Understanding the Circulation of the Arctic Ocean and Sub-Arctic Seas

        I am grateful to US CLIVAR for organizing the Workshop on Observing, Modeling, and Understanding the Circulation of the Arctic Ocean and Sub-Arctic Seas. Next summer, we look forward to meeting in person and getting the perspectives of many oceanographers and atmospheric scientists, including early career scientists and students.

        For now I am excited for the chance to start things off, albeit in this online way necessitated by the pandemic. Since before the beginning of the Study of Environmental Arctic Change (SEARCH) in the late 1990s, through the development of all the new ocean observing components of the Arctic Observing Network (AON), and the emergence of an amazing array of remote sensing instruments over the last two decades, it has been a great time to study the Arctic Ocean environment. Our community has learned a tremendous amount about changes in the Arctic Ocean and how these relate to global change. The Workshop is an opportunity to assess what we have learned, refine our questions, and talk about new directions forward. Consequently the Workshop will be divided into three parts assessing: 1) State of Knowledge, 2) State of the Observing System, and 3) Overcoming Challenges to Observation. We anticipate vigorous debate in all three parts in person at the Workshop next summer, but we will kick things off with a series of blog posts of various opinions on the subjects of the workshop. Here is my take on 1) through 3):

1) I think that since the early 1990s, Arctic Ocean near surface circulation has been in a relatively cyclonic state. The mean state of surface circulation dominated, by the Beaufort Gyre (BG), tends to be anticyclonic, but in situ and more recently remote sensing observations show variations in circulation to be dominated not by variations in the strength of the Beaufort Gyre, but by a trough in sea surface height and associated cyclonic circulation along the Russian side of the Arctic Ocean. In its positive phase set against the mean circulation, this cyclonic mode presents a dipole pattern with an intensified but smaller BG opposite an extensive cyclonic pattern. Notable examples of this shift to the cyclonic mode in the early 1990s [1] and in 2007-08 [2] have followed increases in the Arctic Oscillation (AO) index, the principal component of the primary EOF of variation of atmospheric pressure north of 20°N. Since 1989 the AO has averaged about one standard deviation above the average before 1989 [2, 3], and as a result, Arctic Ocean circulation has tended to be in the cyclonic mode since that time. This is important to climate because: a) the AO is a hemispheric climate index that arguably increases with global warming [4, 5], b) the cyclonic mode deflects Eurasian runoff to the Canada Basin [2] and thus weakens the cold halocline layer that isolates sea ice from Atlantic Water heat [6, 7], and c) a positive AO and cyclonic circulation reduce ice extent in the following summer [3, 8] ,  enhance the export of ice and near-surface freshwater [9], which increase the stratification of the sub-Arctic seas and potentially throttle convective overturning [10].

2) The present in situ observing system is nearly blind to the cyclonic mode of circulation change. Figure 1 from Kwok and Morison [2011] [11] illustrates this and the cyclonic mode pattern. Dynamic heights from IPY hydrographic sampling a year after a positive shift in the AO (Figure 1a) reveal an intense Beaufort Gyre that dominates the region sampled by IPY. However, dynamic ocean topography (DOT) from ICESat (Figure 1b) shows that virtually the whole rest of the Arctic Ocean without in situ observations is a cyclonic trough spread along the length of the Russian side of the Arctic Ocean. Comparisons among hydrography, ICESat DOT, and GRACE ocean bottom pressure (OBP) reveal fresh water increase in the Beaufort Gyre is almost completely balanced by freshwater decrease in the rest of the Arctic Ocean [2].

 A lack of in situ data outside the Beaufort Sea remains a critical shortcoming. Arctic Ocean hydrographic observations in the first part of 2019 (Figure 2) including moorings, aircraft sections, Ice Tethered Profiler and UpTempO drifting buoys were all concentrated in the Beaufort Sea. Meanwhile, ICESat-2 sea surface height relative to CryoSat-2 mean sea surface height, 2011-2015 (Figure 3) show the strong depression on the Russian side of the ocean characteristic of the cyclonic mode, completely unseen by the in situ observations.

3) The challenges to measuring the cyclonic mode are formidable. The center of action of the cyclonic mode along the Russian margins, unlike the BG, is difficult to reach. Buoys drifting with the ice tend to converge in an anticyclonic gyre like the BG and diverge out of a cyclonic feature like that shown over most of the Arctic Ocean in Figure 2b. And the shortest most convenient cruise tracks to the Pole tend to line up with the Transpolar Drift and Front and bypass the region of cyclonic circulation. Integrated remote sensing and in situ observing approaches, and international cooperation will be required to overcome these problems. 

 

References

1.    Morison, J.H., K. Aagaard, and M. Steele, Recent environmental changes in the Arctic: A review. Arctic, 2000.53(4): p. 359-371.

2.    Morison, J.H., et al., Changing Arctic Ocean freshwater pathways. Nature, 2012. 481(7379): p. 66-70.

3.    Williams, J., et al., Dynamic Preconditioning of the Minimum September Sea-Ice Extent. Journal of Climate, 2016. 29(16): p. 5879-5891.

4.    Fyfe, J.C., G.J. Boer, and G.M. Flato, The Arctic and Antarctic Oscillations and their projected changes under global warming,. Geophysical. Research Letters, 1999. 26: p. 1601–1604.

5.    Gillett, N.P., M.R. Allen, and K.D. Williams, The role of stratospheric resolution in simulating the Arctic Oscillation response to greenhouse gases. Geophysical Research Letters, 2002. 29(10).

6.    Steele, M. and T. Boyd, Retreat of the cold halocline layer in the Arctic Ocean. Journal of Geophysical Research-Oceans, 1998. 103(C5): p. 10419-10435.

7.    Polyakov, I.V., et al., Greater role for Atlantic inflows on sea-ice loss in the Eurasian Basin of the Arctic Ocean.Science, 2017. 356(6335): p. 285.

8.    Rigor, I.G., J.M. Wallace, and R.L. Colony, Response of sea ice to the Arctic oscillation. Journal of Climate, 2002. 15(18): p. 2648-2663.

9.    Hilmer, M. and T. Jung, Evidence for a recent change in the link between the North Atlantic Oscillation and Arctic Sea ice export. Geophysical Research Letters, 2000. 27(7): p. 989-992.

10.  Dickson, R.R., et al., The Great Salinity Anomaly in the Northern North-Atlantic 1968-1982. Progress in Oceanography, 1988. 20(2): p. 103-151.

11.  Kwok, R. and J. Morison, Dynamic topography of the ice-covered Arctic Ocean from ICESat. Geophysical Research Letters, 2011. 38(L02501): p. L02501.

Figures

Figure 1. Spring 2008 dynamic height from Arctic Ocean hydrography (a) and dynamic ocean topography from ICESat (b). The hydrographic stations include individual CTD (NPEO, Switchyard, and BGEP) and Aircraft eXpendable CTD (AXCTD) profiles, and 10-day averaged Ice-Tethered Profiler (ITP) data.  From Kwok and Morison [2011])

Figure 2. Arctic Ocean IABP drifting buoy observations in March 2019 plus sections and moorings of the Bering Strait mooring program, Beaufort Gyre Exploration Project (BGEP) and Seasonal Ice Zone Reconnaissance Surveys (SIZRS). Figure courtesy of Ignatius Rigor, 2019. 

Figure 3. Monthly ICESat-2 sea surface height anomaly December 2018 through March 2019 relative to the CryoSat-2 mean sea surface.