Transcript Document

A decomposition of the Faroe-Shetland Channel water masses
using POMP analysis
Mckenna, C.1, Berx, B.2, and Austin, W.1,3
1 School of Geography and Geosciences, University of St Andrews, St Andrews, Fife, KY16 9AL, Scotland
2 Marine Scotland Science, Scottish Government, Marine Laboratory, 375
Victoria Road, Aberdeen, AB11 9DB, Scotland
3 Scottish Marine Institute, Scottish Association for Marine Science, Oban, PA37 1QA, Scotland
1. Background
The poleward flow of warm and saline Atlantic water through the FaroeShetland Channel (FSC) accounts for a large fraction of the total Atlantic
inflow into the Nordic Seas1,12. Therefore, the FSC is an important conduit
for the poleward transport of salt, heat and nutrients, which creates
favourable conditions for the economically important Nordic fish stocks10.
This transport of salt also enhances the formation of intermediate and
deep waters in the Arctic6. These waters then flow back towards the
south, transporting a total ~5.6Sv of water into the North Atlantic13, of
which
which ~2.1Sv passes through the FSC7.
The FSC is thus an integral gateway in the
present operation of the thermohaline
circulation (see video) and, as such,
research into the nature of mixing and
circulation in the channel is important.
Figure 2: The distribution of the FSC water masses across the
NOL section, established from temperature and salinity
measurements.
NAW – North Atlantic Water; MNAW – Modified North Atlantic Water; MEIW – Modified East Icelandic Water; NSAIW – Norwegian Sea Arctic
Intermediate Water; NSDW – Norwegian Sea Deep Water.5
Video showing the important role that the FSC (highlighted in box) plays in the North Atlantic thermohaline circulation.
Source: https://www.youtube.com/watch?v=c-GOFHPkf6Q
2.
Current state of the art
Since the 1970s, water properties have been monitored regularly across 2
sections of the FSC (Fig. 1), which has established that 5 water masses flow
through the channel (Fig. 2). Their contrasting origins gives them distinct
temperature and salinity signatures, which have been used in the past to detect
their presence in the FSC11,15. But, while an absent signature implies the absence of
a water mass, this could also arise from intense mixing. Thus, the nature and
extentnd mixingextent
Figure 1: Map of the Faroe-Shetland Channel and the 2 standard hydrographic sections: Fair-Isle-Munken (FIM) and
Nolso-Flugga (NOL).
and extent of mixing between the FSC water masses is currently uncertain.
Simple empirical mixing models have been used to calculate water mass
ratios in the FSC6,8, but due to a limited number of known variables they can
only model up to 3 of the FSC water masses with statistical significance. We
therefore propose using the novel method of Parametric Optimum MultiParameter (POMP) analysis2, which can model all 5 FSC water masses while
maintaining statistical significance.
1
3.
4.
POMP analysis
POMP analysis assumes that water properties at each point in the water column can be decomposed into
contributions from a number (𝑛) of source water masses. Temperature (𝑇), salinity and nutrients
(nitrate, phosphate and silicate) were the properties used here. Mixing equations of the form below are
then solved simultaneously for each property, say 𝑇, at each point:
𝑇 = 𝑥1 𝑇1 + 𝑥2 𝑇2 + ⋯ + 𝑥𝑛 𝑇𝑛 + 𝑒𝑇 ,
where 𝑇 is the measured temperature (known), 𝑇𝑛 is the source water temperature (known – see Fig. 3
for how this is found), 𝑥𝑛 is the fraction of the source water (unknown) and 𝑒𝑇 is random noise. POMP is
particularly powerful as it takes into account that if a water mass is in high proportions at a point, it isby
is likely to be in high
proportions at adjacent points.
This places an extra constraint
on the problem, therefore
reducing the number of
unknown
variables
and
improving the statistical
significance of the solution.
Figure 3: Plot of temperature vs.
salinity using data collected from the
FSC over the period May 2009-2013. The
source water properties were identified
from prominent end members e.g. NAW –
most saline point; NSAIW – least saline
point in the region -0.5°C to 0.5°C; NSDW –
densest point. Dotted lines are density
contours.
Results in May 2013
The interesting shape of
MNAW implies its
recirculation at FIM, as is
thought in the
literature14,16. The POMP
analysis suggests this also
causes increased vertical
mixing of NAW, MEIW and
NSAIW in the centre of FIM,
~300m.
POMP confuses MEIW and
MNAW in surface waters
on the Faroese side of the
channel, especially at FIM.
This probably reflects that
MNAW contributes to the
MEIW formation process5.
Internal solitary waves are thought to
propagate up the Shetland side of the FSC,
causing mixing between deeper and upper
water layers4,9. The mixing fractions manage
to capture this process and suggest this
mixing may be significant on NOL .
Generally, the water masses enter the channel as
well defined cores, but spread out and sink as
they flow through the channel. The POMP analysis
captures these mixing processes well, as well as
correctly predicting the general distribution of
the water masses (Fig. 2).
2
5.
Acknowledgements
Conclusions
This work received funding from the MASTS pooling initiative (The Marine Alliance for Science and Technology for Scotland) and their
support is gratefully acknowledged. MASTS is funded by the Scottish Funding Council (grant reference HR09011) and contributing
institutions. Salinity analyses were conducted by members of the Oceanography Group at Marine Scotland Science: we would like to
thank David Lee, Matthew Geldart, Dougal Lichtman and George Slesser for their efforts. Nutrient analyses were performed by the
Analytical Chemistry Group at Marine Scotland Science: we are grateful to Pamela Walsham, Alison Taylor and Lynda Webster. Oxygen
isotope analyses were conducted by WENA and Angus Calder at the University of St Andrews; we acknowledge helpful discussions of
these data with Lauren Gillespie. The POMP analysis computer code was kindly given to us by Anouk de Brauwere, for which we are very
grateful.
 POMP currently works well in the FSC with temperature, salinity and nutrient data.
 Even though the model is empirical, it tell us a lot about which mixing processes and relationships are
significant on each section of the channel.
6. Future perspectives
Table 1: A comparison of the cost function (CF) for different data combinations. CF = distance between
the modelled fractions and measurements. DF = degrees of freedom; if the CF is larger in magnitude than DF,
then there are potentially model errors present. Case 1 used the full resolution nutrient dataset, whereas
cases 2 and 3 only used nutrient data where there was δ18O data.
 Our May 2013 results highlight the mixing processes between the water masses.
However, we need to investigate different seasons and years to explore the
temporal variability of these processes. Changes in the properties of one water
mass could propagate into another, with wider implications for the thermohaline
circulation. Indeed, mixing in the FSC between freshening Atlantic waters and
intermediate waters may have helped wide-scale freshening of the northern North
Atlantic in the 1960s, 70s and 80s, which is thought to have inhibited convective
overturning3.
References
1.
2.
FIM
Case
(1) No δ18O, all nutrients
(2) No δ18O, limited
nutrients
(3) With δ18O, limited
nutrients
 The results could be more accurate if we used an additional variable in the model:
stable oxygen isotopes (δ18O). Even though δ18O is currently sampled at a relatively
low resolution in the FSC, it still improves the results (compare cases 2 and 3 in
Tables 1 and 2). Samples will be collected at a higher resolution in October 2014.
NOL
3.
CF
DF
CF
DF
142.74
405
455.09
352
5.
6.
91.05
88
-
-
7.
8.
102.71
112
-
-
9.
4.
10.
11.
12.
Table 2: A comparison of the uncertainties associated with the mixing fractions (in %) for each case
tested (calculated from 1000 Montecarlo simulations in each case).
13.
14.
15.
 POMP may help us to identify NAW in the FSC
and, so, improve estimates of the volume
transport of the Atlantic inflow into the Nordic
Seas (i.e. the strength of the thermohaline
circulation).
Case
(1) No δ18O, all
nutrients
(2) No δ18O, limited
limited nutrients
(3) With δ18O, limited
limited nutrients
FIM
NOL
NAW
MNAW
MEIW
NSAIW
NSDW
NAW
MNAW
MEIW
NSAIW
NSDW
5.2
7.0
3.9
5.0
4.6
5.3
7.1
3.1
4.2
3.6
8.5
11.0
3.9
5.6
5.9
-
-
-
-
-
8.1
10.6
3.9
5.6
5.9
-
-
-
-
-
16.
Berx, B., Hansen, B., Østerhus, S., Larsen, K.M., Sherwin, T., & Jochumsen, K. (2013). Combining in-situ measurements and
altimetry to estimate volume, heat and salt transport variability through the Faroe Shetland Channel. Ocean Science
Discussions, 10(1), pp.153-195.
de Brauwere, A., Jacquet, S.H., De Ridder, F., Dehairs, F., Pintelon, R., Schoukens, J., & Baeyens, W. (2007). Water mass
distributions in the Southern Ocean derived from a parametric analysis of mixing water masses. Journal of Geophysical
Research: Oceans (1978-2012), 112(C2).
Dickson, R.R., Meincke, J., Malmberg, S.A., & Lee, A.J. (1988). The “great salinity anomaly” in the
northern North Atlantic 1968–1982. Progress in Oceanography, 20(2), pp.103-151.
Hall, R.A., Huthnance, J.M., & Williams, R.G. (2011). Internal tides, nonlinear internal wave trains, and mixing in the Faroe-Shetland
Channel. Journal of Geophysical Research: Oceans (1978-2012), 116(C3).
Hansen, B. & Østerhus, S. (2000). North Atlantic-Nordic Seas exchanges. Progress in Oceanography, 45(2), pp.109-208.
Hansen, B., Østerhus, S., Hátún, H., Kristiansen, R., & Larsen, K.M.H. (2003). The Iceland-Faroe inflow of Atlantic water to the
Nordic Seas. Progress in Oceanography, 59(4), pp.443-474.
Hansen, B., & Østerhus, S. (2007). Faroe Bank Channel overflow 1995-2005. Progress in Oceanography, 75(4), pp.817-856.
Hermann, F. (1967). The TS Diagram Analysis of the Water Masses over the Iceland-Faroe Ridge and in the Faroe Bank Channel
(Overflow ‘60). Rapports et Procès-Verbaux des Réunions du Conseil International pour l’Exploration de la Mer, 157, pp.139-149.
Hosegood, P., van Haren, H., & Veth, C. (2005). Mixing within the interior of the Faeroe-Shetland Channel. Journal of Marine
Research, 63(3), pp.529-561.
Larsen, K.M.H., Hátún, H., Hansen, B., & Kristiansen, R. (2012). Atlantic water in the Faroe area: sources and variability. ICES
Journal of Marine Science: Journal du Conseil, 69(5), pp.802-808.
Martin, J.H.A (1993). Norwegian Sea intermediate water in the Faroe-Shetland Channel. ICES Journal of Marine Science: Journal
du Conseil, 50(2), pp.195-201.
Østerhus, S., Turrell, W.R., Jónsson, S., & Hansen, B. (2005). Measured volume, heat, and salt fluxes from the Atlantic to the
Arctic Mediterranean. Geophysical Research Letters, 32(7).
Sherwin, T.J., Griffiths, C.R., Inall, M.E., & Turrell, W.R. (2008). Quantifying the overflow across the Wyville Thomson Ridge into
the Rockall Trough. Deep Sea Research Part I: Oceanographic Research Papers, 55(4), pp.396-404.
Sherwin, T.J., Hughes, S.L., Turrell, W.R., Hansen, B., & Østerhus, S. (2008). Wind‐driven monthly variations in transport and the
flow field in the Faroe–Shetland Channel. Polar Research, 27(1), pp.7-22.
Turrell, W.R., Slesser, G., Adams, R.D., Payne, R., & Gillibrand, P.A. (1999). Decadal variability in the composition of Faroe
Shetland Channel bottom water. Deep-Sea Research Part I: Oceanographic Research Papers, 46(1), pp.1-25.
van Aken, H.M. (1988). Transports of water masses through the Faroese Channels determined by an inverse method. Deep Sea
Research Part A. Oceanographic Research Papers, 35(4), pp.595-617.
3