Poster #
Estimates of the seasonal variability of
volume, heat, and freshwater fluxes
in the eastern subpolar North Atlantic
T1-39
Stefan F. Gary1*, Stuart A. Cunningham1, Clare Johnson1, N. Penny Holliday2, Loic Houpert1
SAMS, Oban, UK 2NOC, Southampton, UK
1
1) Motivation
2) Our Goals
The Extended Ellett Line (EEL) is a repeat
hydrographic section with a 40 year history.
Understanding of the variability in the area
may help climate predictions1. Rough winter
weather results in few winter occupations.
Previous work has resolved interannual
variability2 but not seasons3.
Contribute to the search for a seasonal cycle of
volume, heat and freshwater fluxes with winter glider
occupations.
Merge glider data with ship and float data. Sampling
on a single section may be impacted by spatial and
temporal aliasing of the eddy variability4.
EEL
3) Where would we expect geostrophic transports to reflect the seasonal cycle?
Model-based conclusions:
●
IB
●
SAMS Glider Missions
in this region
●
●
●
●
●
●
●
Talisker EEL, '09-'10,
784 dives, 3020 km
Talisker EEL, 2011
836 dives, 2292 km
Talisker FASTNEt, '13-'14,
1857 dives, 2851 km
Ardbeg
FASTNEt, '13-'14,
1803 dives, 2835 km
Ardbeg
EEL, 2015,
1200+ dives, 3100+ km
Laphroig FASTNEt, 2014
1515 dives, 1743 km
Jura
OSNAP, '14-'15
1844 dives, 3103 km
Knockando FASTNEt, 2014
1316 dives, 1919 km
●
Scapa
OSNAP, '14-'15
1593 dives, 2987 km
●
Bowmore OSNAP, 2015
946 dives, 2607 km
●
Bellatrix 5, EEL, 2014,
651 dives, 1423 km
●
+3 more in the water
right now! Follow:
@samsglider
velocity.sams.ac.uk/gliders
*
SAMS, Scottish Marine Institute,
Oban, Argyll, PA37 1QA
(+44) (0)1631 559419
(+44) (0)7474 387218
[email protected]
References:
1) Hermanson et al., 2014, GRL,
doi:10.1002/2014GL060420
2) Holliday et al., in review, JGR
3) Holliday et al., 2000, DSRI,
doi:10.1016/S0967-0637(99)00109-0
RHP
RT
●
Figure 1: r2 for co-located depthintegrated absolute and
geostrophic transports along
sections in the FLAME 1/12o model5
over Rockall Trough (RT), Rockall
Hatton Plateau (RHP), and Iceland
Basin (IB). 15-year time series, at
3-day resolution, were smoothed
with a 3-month boxcar filter.
●
Good correlations in the RT, RHP
and southern IB are not isolated to
a particular section.
Difficult to resolve the northern IB.
There is strong, bottom-intensified
flow over sloping topography here
due to the overflow water.
There is a seasonal cycle in the
model absolute velocity field which
is consistent with the geostrophic
velocity field.
However, the uncertainties in the
seasonal cycle can be impacted by
strong interannual variability.
4) Data sources and Quality Control (QC)
5) Volume, heat, and freshwater fluxes
Figure 4:
Seasonal RT
fluxes using
geostrophic
transports
referenced to,
and summed
above, 1200 m
in a smoothed,
¼o, 3D, QC'ed,
seasonal
climatology.
The black line
is the EEL,
others are the
dotted sections
shown in Fig. 1.
← Glider missions. Ship & Argo data span 1950-20126.
Challenges: Assumptions are needed for processing
SeaGlider data7 but QC'ing glider data based on ship &
float data may remove variability observed by gliders.
Gliders profile every ~5 km but ships & floats profile
every ~10-100 km. The density of glider profiles can
bias climatologies and the QC process.
Solution: all profiles that fall in each year, season, and
¼o x ¼o bin are isopycnally averaged. The statistical
envelope of the binned profiles is used to QC ship,
float, and glider data together8. Here, we using 1.2x105
profiles with 7.9x106 S,T,P triplets, which are 83% and
90% of the input into the QC filter, respectively.
Figure 3: Example of T-S
space statistical data QC
filter in a 1o x 1o square in
RT. Accepted points are
shaded by pressure, 02000 dbar, rejected points
are black + symbols and
the mean and ± 2.3 std of
the binned profiles at each
density level are shown in
green & red, respectively.
Conclusions:
●
●
4) Sherwin et al., 2015, OS,
doi:10.5194/os-11-343-2015
5) Böning et al., 2006, GRL,
doi:10.1029/2006GL026906
6) NODC, 2012, www.nodc.noaa.gov/OC5/
SELECT/dbsearch/dbsearch.html
7) UW SeaGlider QC manual, 2012
8) Lozier et al., 1995, Prog. Oce.
www.whoi.edu/science/PO/hydrobase/
php/index.php
9) Johnson et al., 2015, D22.31,
http://www.naclim.eu
Here, glider data are retained at 92%, similar to ship
& float data. Compare to 77% for separate QC9.
Figure 2: (top) Transport in
the upper EEL RT in FLAME.
(bottom) Seasonal cycle of the
time series above with ± 1 std.
●
No clear seasonal cycle yet, but
transport consistent with previous work3.
Geostrophic fluxes are sensitive to
smoothing since they are based on
density gradients. More data (update
ship and float database, more glider
missions this winter) will help reduce the
smoothing. Also, revisit the data
pipeline to minimize smoothing.
Estimate error bars with a “leave-someout” analysis.
Acknowledgements: We are grateful for NERC funding. The research leading to these results has received funding
from NACLIM, a project of the European Union 7th Framework Programme (FP7 2007-2013) under grant agreement
n.308299. We are also grateful for Claus Böning (GEOMAR) sharing the FLAME model output. Thanks to Estelle
Dumont, Toby Sherwin, Mark Inall, Marie Porter, Karen Wilson, Colin Griffiths, John Beaton, Lovro Valcic, the Captains
and crews of the RRS James Clark Ross, NLB Pole Star, FRV Magnus Heinason and Bogi Hansen and Brian King.