The Hydrological Cycle, The Oceans, and Climate

1
Recent Changes in the North Atlantic Bob
Dickson, CEFAS with Ruth Curry, WHOI and Igor
Yashayaev, BIO
Royal Society Meeting on Abrupt Climate Change
Evidence, Mechanisms and Implications, 4-5
February 2003. Session 2- Modern observations
and processes
2
1. What have been the changes in forcing? 2. What
ocean changes are we expecting? 3. What changes
do we find? 4. What observing strategy to adopt?
Royal Society Meeting on Abrupt Climate Change
Evidence, Mechanisms and Implications, 4-5
February 2003. Session 2- Modern observations
and processes
3
1. What Changes in the Forcing?
Royal Society Meeting on Abrupt Climate Change
Evidence, Mechanisms and Implications, 4-5
February 2003. Session 2- Modern observations
and processes
4
When we plot air temperature as a function of
latitude and time, two things are clear 1) the
World is warmer. Including 2002, all ten of the
warmest years since records began in 1861 have
occurred since 1990 Jones and Moberg, 2003. 2)
in the last two decades the distribution of
warming has become global.
Courtesy Tom Delworth, GFDL
5
.and our instrumental and proxy records suggest
that the NAO in the 1990s may have been at a 600
year extreme positive state.
Phil Jones CRU, in press
6
2. What changes of global scale and importance
are we expecting to observe in the ocean?
2i. A slowdown of the MOC, and 2ii. An
acceleration of the water cycle
Royal Society Meeting on Abrupt Climate Change
Evidence, Mechanisms and Implications, 4-5
February 2003. Session 2- Modern observations
and processes
7
IPCC, 2001
Most (but not all) coupled climate models
anticipate a slow down of the Atlantic Meridional
Overturning Circulation (MOC) under greenhouse
gas forcing as a result of freshening and warming
of subpolar seas
8
The Water Cycle Will Accelerate With Global
Warming

• A warmer atmosphere will carry more water vapor,
  because of the exponential increase of vapor
  pressure with temperature.
• An enhanced water cycle will change the
  distribution of salinity in the upper ocean.
• A program for monitoring salinity changes is
  needed.

Ray Schmitt,WHOI pers comm
9
3. What changes do we observe?
Royal Society Meeting on Abrupt Climate Change
Evidence, Mechanisms and Implications, 4-5
February 2003. Session 2- Modern observations
and processes
10
We cant measure change in the MOC directly
But we can measure a range of its known or
supposed associates e.g. 3a) an increase in the
freshwater fluxes from the Arctic that are
supposed to slow it down. 3b) slowing or
density-change in the overflows that drive
it. 3c) changes in the trans-ocean gradients of
steric height that will reflect a change in
overturning rate.
11
3a. Changes in Freshwater Flux from the Arctic.
Royal Society Meeting on Abrupt Climate Change
Evidence, Mechanisms and Implications, 4-5
February 2003. Session 2- Modern observations
and processes
12
Schematic of the northern loop of the oceans
Great Conveyor. McCartney et al 1996
The elements of an ASOF freshwater flux array is
funded and is partly in place. Few results
yet.....but we do have one proxy measure of fw
flux from the AR7W Line
13
The offshore density gradient in the 0-150m layer
from Labrador Shelf to the Central Lab Sea is our
only (proxy) measure of the changing fw flux to
the NW Atlantic. This gradient has progressively
steepened with the NAO over the past 4 decades
(equivalent to a 20 incr. In Sgoing transport)
largely thro freshening of shelf upper-Slope
waters. Data from AR7W, 0-150m.
Density gradient
Density
14
3b. Changes in the Overflows i) transport ii)
temperature (see paper) iii) salinity
Royal Society Meeting on Abrupt Climate Change
Evidence, Mechanisms and Implications, 4-5
February 2003. Session 2- Modern observations
and processes
15
The main indication we have of a change in
overflow transport is proxy evidence, and applies
to the eastern overflow. There, the depth of the
?t28.0 isopycnal in the upstream reservoir of
the Norwegian Sea controls the pressure head that
drives overflow through the Faroe-Bank Channel
(Hansen et al 2001).
16
Over the past 4 decades, for a variety of reasons
associated with the amplifying NAO the fresh
water accession to the Nordic Seas has increased
steadily
17
Salinity at OWS M, Norwegian Sea, 1950-97,
Courtesy Svein Østerhus, UiB
The 50-year record at OWS M in the Norwegian Sea
is our best benchmark of this broadscale change,
showing a long term freshening reaching to depths
of gt1 km.
18
Depth of the ?t28.0 isopycnal at OWS M
As a result, the ?t 28.0 isopycnal has itself
steadily deepened, by ?70m in 50 years, (Hansen
et al 2001).
19
Hansen et al (2001) use this finding to suggest
that the coldest, deepest part of the eastern
overflow may have decreased by 20 since 1950.
20
There is no such evidence of decreasing flow
speeds from the western overflow through Denmark
Strait. There, the 187 monthly means of speed
collected from the overflow core since 1986 (z
?2000m) show little sign of seasonal
variability..
21
Neither has there been any obvious change in the
transport of the overflow core.
22
3biii. Changes in overflow salinity
Royal Society Meeting on Abrupt Climate Change
Evidence, Mechanisms and Implications, 4-5
February 2003. Session 2- Modern observations
and processes
23
Salinity at OWS M, Norwegian Sea, 1950-97,
Courtesy Svein Østerhus, UiB
The broadscale freshening of the Nordic Seas over
the past 4 decades has reached depths of gt1 km,
so is accessible to the two main overflows which
cross the Greenland-Scotland Ridge.
24
Tapping-off this layer, the two dense overflows
that renew and ventilate the deep ocean have also
freshened over the past 4 decades. Dickson et
al, Nature 2002.
25
so that if we construct salinity time series at
intervals along the spreading pathways of both
overflows from their sills to the deep Labrador
Sea
26
we find that the entire system of overflow and
entrainment that ventilates the deep Atlantic has
undergone a remarkably rapid and remarkably
steady freshening over the past four decades. A
change in the ocean-climate of sub-arctic seas
has thus been transferred to the deep and
abyssal ocean at the headwaters of the Great
Conveyor Dickson et al 2002
27
NAO-
NAO
The resulting full-depth change in the Labrador
Sea salinity is believed to be the largest change
in the instrumented oceanographic record. By
1992, equivalent to adding an extra 6 m of fresh
water at the sea surface.
28
Reported with the usual objectivity!
29
Fresh water gain by layers

• ?2
  Principal Accumulated Range
  Resident
  Fresh Water
• 36.884 - 36.966 Labrador Sea Water 2.52 m
• 36.968 - 37.060 Northeast Atlantic Deep 0.63
  m
• 37.062 - 37.182 Denmark Strait Overflow 0.31
  m
• Total 3.5 m

..Or more generally over the NW Atlantic,
equivalent to mixing in an extra 3.5 m of
freshwater, unevenly distributed over the
watercolumn.
30
The result has been a dramatic shift in the ? -S
relation for waters of the NW Atlantic. Igor
Yashayaev, unpublished
31
2ii. A change in the Atlantic hydrologic cycle?
32
As we follow the deep freshening south through
the western Atlantic to the W Line .we
encounter something else
Curry, Dickson and Yashayaev, in press
33
This is Salinity Maximum Water (SMW), formed in
the subtropical gyre at the Atlantic E-P maximum,
where density layers in the range ?0 25.50-26.50
are ventilated.
Curry, Dickson Yashayaev, in press
34
Over the same 40-year period that salinities of
high latitude water masses have freshened,
salinities at the E-P maximum have been
increasing . a shift in the entire Atlantic
hydrological cycle?
Calculated for surface to?n 26.50 (waters
outcropping S of 30?N in the E-P max) or 27.0 (S
of 40?N)
Curry, Dickson and Yashayaev, in press
35
This change in salinity is not confined to the
NAO or even the Atlantic. We compare changes late
50s-early 60s to 1990s from 30S to 60N through
the W Basins, through main centres of SSS, E-P
and steric height.
36
ATLANTIC WATER MASSES 1990-99 minus 1956-64
SALTIER
FRESHER

• NEADW DSOW
• LSW Surface Labrador Sea
• AAIW UCDW Vent.Therm

• Surface Tropics / Subtropics
• MOW UNADW (S)

30S EQ
60N
Curry, Dickson Yashayaev, in press
37
The same symmetrical pattern of change,
freshening intermediate waters from high N S
latitudes and more-saline upper ocean at low lats
has been shown for the Pacific and Indian oceans
by Wong et al 1999, 2001 and Bindoff McDougall
2000.
38
It is therefore tantalizing to speculate that
the observed symmetrical freshening of NPIW and
AAIW/SAMW represents a trend in the climate
system that could be induced by anthropogenic
sources. A natural extension to this work would
be to investigate whether the freshening of
AAIW/SAMW is circumpolar in extent Wong,
Bindoff and Church, 2001
39
3c. A change in the trans-ocean steric gradient?
40
THC Overturning vs Atlantic Meridonal Steric
Gradient in HADCM 3(Thorpe et al 2001)
We assume that change in the MOC will be
associated with some measurable change in the
trans-ocean density gradient, e.g. HadCM3
suggests a close correlation between Atlantic
overturning rate and the S-N gradient of steric
height from 30S - 60N through the W Atlantic.
41
Atlantic Water Mass Variability and the
Meridional Density Gradient
42
Over the 40-year observational record, the
meridional steric height gradient has varied only
by about 20 cm, largely driven by changes at its
northern end.
43
a) T contribution
c) Steric Height change
The change in steric height in the Labrador Sea,
1958 to 2002, been dominated by the effect of
cooling rather than freshening ? a net lowering
of steric height. Courtesy of Igor Yashayaev,
BIO, Canada
b) S contribution
44
THC Overturning vs Atlantic Meridonal Steric
Gradient in HADCM 3(Thorpe et al 2001)
1970
1995
20 cm
40 cm
Mapping the observed steric gradient changes (20
cm) onto the Thorpe et al results, we find that
the Atlantic MOC strength has varied little over
the past 40 years.
. But also that a
50 reduction from the present MOC strength could
result from just a 40 cm drop in the gradient
below the 1995 gradient.
45
Change in the zonal trans-ocean density gradient
along 25N, ---the Marotzke, Bryden and
Cunningham proposal in NERC-RAPID
46
Since mass transport between 2 points depends
only on the pressure difference between them,
their moored array along 26.5N will measure the
overturning rate by continuous observation of
density at the Wern Eern boundaries.
47
A 1st look at changing gradients of steric height
along 24N suggest two problems. High-amplitude
variability in the west and small variability
elsewhere cf the W Atlantic line.
48
So in summary, the observed changes are
49

• As accompaniments to the extreme climatic forcing
  of recent decades, the following obs. seem
  relevant
• 40 year increase in the sgoing fw flux
  circuiting around the Lab Sea margins.
• 50 year decrease in deep, dense overflow from
  FSC
• 40 year freshening of both GIN-Sea its
  overflows
• 40 year increase in E-P upper ocean salinity
  in the subtropics
• some global evidence of the expected
  multi-decadal increase in the water cycle.
• Though no evidence yet of any sustained change in
  the MOC, the above explains justifies the
  current 120M, 4-5 year focus on the oceans role
  in climate (EC, UK, Norway, USA, Canada).

50
4. What observing strategy?
51
The first task is to measure all of the
Arctic-Subarctic Ocean Fluxes that connect the
Arctic to the N Atlantic. ASOF is already
proceeding to implementation (http//asof.npolar.n
o)
52
Royal Society Meeting on Abrupt Climate Change
Evidence, Mechanisms and Implications, 4-5
February 2003. Session 2- Modern observations
and processes
53
3bii. Changes in overflow temperature
Royal Society Meeting on Abrupt Climate Change
Evidence, Mechanisms and Implications, 4-5
February 2003. Session 2- Modern observations
and processes
54
The current meter thermistor data from 2000m in
the overflow core do reveal a dramatic decadal
variation in mean temperatures. These values are
not smoothed. Each 30-day mean is simply the
average of 720 hourly values.
55
and the temperature and salinity of the overflow
core at 2000m off Angmagssalik clearly determine
the density of the DSOW-derived layer of the
abyssal Labrador Sea a further one year later.
56
Potential Temperature
NAO-
NAO