We examine the spatiotemporal variability of shelf water properties in the Prydz Bay and possible mechanisms that control the dense shelf water production, using CTD data in figure 2 and mooring measurements at the Prydz Channel near the shelf break. The mooring CTD measurements from March 2013 to March 2015 show a clear seasonality in the properties of dense shelf water. Salinity, consequently density, peaks in October and November, which lags sea ice production season by approximately a month. In the Prydz Bay shelf region, the sea ice formation is the only source for the formation of high salinity water. Both modified Circumpolar Deep Water (mCDW) and ice shelf water reduce salinity in the lower water column where remnant of winter dense shelf water resides. The dense shelf water observed at the mooring site at the Prydz Channel is likely transported to the area from the Amery Depression where the densest shelf water was found. The mean current at the Prydz Channel supports this assumption. At the mooring site, salinity tends to be lower in austral winter. Color coded mooring T/S diagram suggests that ice shelf water frequently appears at this location from July to December, which reflects a seasonal density minimum in July and August. It suggests that sea ice formation does not dictate the seasonality of local shelf water density near the shelf break. Regardless, in this two-year mooring record, the potential density of dense shelf water reaches higher than 27.82 for approximately 3 months in each austral spring, which is dense enough to sink down to the continental slope.

Water properties also exhibit clear interannual variability at the density layers between 27.70 and 27.78. This density layer contains ice shelf water near the coast and mCDW intrusions at the shelf break. The water at this layer across the shelf is colder and fresher in 2015 than in 2012 and 2013 (figure 2). The layer is also thicker in 2015 than in 2012 and 2013, particularly near the coast (not shown). This distinct year to year variation extends to the water properties at the bottom of CTD casts across the shelf, suggesting that the densest water mass on the shelf also has the same year to year variation. The color-coded T/S diagram of ship based CTD data show that ice shelf water is colder and fresher (less diluted) in the summer of 2015 than in summers of 2012 and 2013. The highest density observed at the mooring site near the shelf break reaches 27.875 in the spring (September, October, November) of 2013. All evidence points to that more ice shelf water outflow occurred in 2015 than in the other two years, resulting in less dense shelf water in 2015.

To investigate whether ice shelf water outflow is the key factor that dictates the shelf water density, we also examine sea ice production rate in these years, using reanalysis surface fluxes. Here we assume that ocean heat flux is negligible and all heat loss to the atmosphere goes toward to sea ice formation (Nihashi and Ohshima, 2015). The results show that the maximum sea ice production rates are relatively the same in the winter prior to our hydrographic surveys of 2012 and 2013 but slightly higher in the winter prior to the 2015 survey. However, the higher maximum winter ice production rate and more sea ice in following summer did not overcome the freshening effect of ice shelf water outflow in 2015. It indicates that the sea ice formation plays a minor role in determining shelf water density. We further examine the surface wind patterns in these three summers to isolate mechanisms controlling this interannual variability of the shelf water. In 2015, divergent winds prevail in the outer shelf region and weak convergent winds occur near the Amery Ice Shelf front. In contrast, convergent winds prevail in the entire Prydz Bay shelf area except near Cape Darnley in the summers of 2012 and 2013. We surmise that divergent winds at the sea surface encourage ice shelf water outflow at the subsurface.

We further examine the surface wind patterns in these three summers to isolate mechanisms controlling this interannual variability of the shelf water. In 2015, divergent winds prevail in the outer shelf region and weak convergent winds occur near the Amery Ice Shelf front. In contrast, convergent winds prevail in the entire Prydz Bay shelf area except near Cape Darnley in the summers of 2012 and 2013. We surmise that divergent winds at the sea surface encourage ice shelf water outflow at the subsurface. A manuscript is prepared to report these results.

We have four conclusions. (1) Density of dense shelf water near the shelf break exhibits clear seasonal cycle with a maximum in austral spring and minimum in austral winter. (2) There is clear year-to-year variability in the shelf water properties from mid depth to the bottom of water column. Shelf water is lighter, fresher and colder in the summer of 2015 than the summers of 2012 and 2013. (3) Sea ice formation is not a key factor for the seasonality of dense shelf water properties near the shelf break, nor a key factor for the interannual variation of the shelf water properties. (4) Ice shelf water dictates the shelf water properties at seasonal and interannual time scales. Surface wind patterns likely influence the volume of ice shelf water out flow.

Figure 2. The mean temperature (top) and salinity (bottom) in the density layer of 27.70-27.78 in 2012 (a, d), 2013 (b, e) and 2015 (c, f).

Data

CTD Data

2015 - in progress

Publications/Presentations

Yuan,. X, M. Kaplan, M. Cane, The interconnected global climate system - a review of tropical-polar teleconnections. Revised for Journal of Climate.

Yuan, X. L. Gao, Y. Sun, et al., Spatiotemporal variability of the shelf water in Prydz Bay. In preparation.