View
3
Download
0
Category
Preview:
Citation preview
ICES 1999
Abstract
ICES eM 1999IL:16 Theme Session L Nordic Seas Exchanges
Thermohaline changes in the Irminger Sea
by
John Mortensen and Beainn Valdimarsson Marine Research Institute (MRI), Reykjavik
121 Reykjavik
Skulagata4
Fax: +3545623790
e-mail: johnm@hafro.is
e-mail: hv@hafro.is
The Inninger Sea is part of the Subpolar Gyre in the northwestern North Atlantic and
plays a central role in the large-scale thennohaline overturning of Atlantic Water which is
believed to influence the long-tenn changes of the climate system. In this area thennohaline
changes are observed at almost all depth levels in the nineties. The most pronounced change
is connected to the Modified North Atlantic Water (MNAW) where an overall increase of
temperature and salinity were observed during the nineties. Time series from the Icelandic
continental slope reveal that the recent onset of increasing temperatures took place in the
winter 199511996 and was accompanied by a more pronounced salinity increase in late
summer 1997.
A historical comparison with nearby sections occupied in the early eighties and late
nineties reveals that changes in the distribution of water masses have taken place recently.
The reason is likely to be connected to the densification of the Labrador Sea Water (LSW)
since the late eighties. The change is seen as a downward movement of the LSW core now
occupying in the Inninger Sea the depth range of 1500 to 2100 m instead of 500 to 1200 m in
e.g. 1981. The changes are observed as a freshening and cooling of the LSW.
Keywords: Inninger Sea, thermohaline changes, Labrador Sea Water and Modified North
Atlantic Water.
Introduction
Since the first recorded oceanographic work in the Inninger Sea was published by
Inninger (1853), the area has attracted periodic interests. By the turn of the century the main
hydrographic structure was known. The cold, deep water overflow across the Greenland
Iceland and Iceland-Faroe Ridges was first observed by Knudsen (1898, 1899) and later
discussed by Nansen (1912) who also presented a surface current map of the area. A
summary of all investigations in the area for the period 1853 to 1957 is given by Dietrich
(1957). A similar summary doesn't exist for the period 1958 until today (1999). Some larger
international programmes since 1957 in the area can be mentioned such as the International
Geophysical Year (1958-59) (Dietrich, 1969), NORWESTLANT 1-3, 1963 (Lee, 1968),
Inninger Sea Project 1963 and 1965 (Stefansson, 1968), ICES Overflow 1973 (Anon., 1976),
WOCE 1990-1997, Nordic WOCE 1992-1997 and VEINS 1997-1999. Beside these a number
of special cruises have been conducted in the area with an increasing frequency in the
nineties. During the sixties the water mass definitions and their approximate origin became
well established and one of the first papers regarding this area and its annual variations was
published by Blindheim (1968). The last water mass to be recognised in the Inninger Sea was
the LSW (Lee and Ellett, 1965, 1967). In the seventies and early eighties the overflow was
the main research subject in the area (Anon., 1976). The general tendency in the eighties was
a northward movement of the international research programs leaving the Inninger Sea less
observed. However, the overflow was still monitored (Dickson and Brown, 1994) and
satellite tracked drifters were released for the first time in the area (Krauss, 1995). The
nineties were an introduction to a new era for the Irminger Sea, from being a low research
activity area to becoming a higher one. Large international programs such as WOCE
(including Nordic WOCE) and VEINS contributed to an increased focus on the climatic
aspect of the area.
In this paper thermohaline changes of the LSW and MNA W in the Inninger Sea
(Figure 1) in the nineties are described by analysis of hydrographic data obtained by Icelandic
and international research programs.
Results
LSW changes
The spreading of the LSW from its formation in the Labrador Sea has been described
by Talley and McCartney (1982). They followed the water mass from its origin into the North
Atlantic by its low potential vorticity, utilizing the fact that convective1y-formed water is
2
characterized by a vorticity minimum.
Hydrographic observations in the Labrador Sea of the LSW over the last decades
show that the LSW has undergone significant changes (Talley and McCartney, 1982; Dickson
et aI., 1996). Of special interest are the late property changes of the LSW which started in
1988 (e.g. Sy et al., 1997) where different LSW vintages produced up to 1994 were
distinguishable by their temperatures. From 1988 to 1994 the LSW water mass was observed
with decreased temperature and salinity of the order of O.4°C and 0.02 psu respectively.
Utilizing temperature, salinity and chlorofluorocarbon concentration data Sy et al. (1997)
found surprisingly rapid spreading rates of newly formed LSW. They estimated that it
reached the southern parts of the Inninger Sea in about 6 months. Yielding a mean spreading
speed of 4.5 cm/s. It is therefore of interest to see if these velocities are maintained in the rest
of the Irminger Sea and how the density increase of .the LSW influences the water mass
distribution.
Figure 2 shows the temporal changes of the thermohaline characteristics of the LSW
in the northern part of the Irminger Sea as potential temperature versus potential density
relative to 1500 dbar for the LSW salinity minimum or core. The temporal development of
the LSW core reveals a cooling in the northern part of the Irminger Sea of the order of0.45°C
during the period 1988 to 1996 and then a slight increase during 1997 to 1998. The salinity
changes were of the order of -0.03 psu and +0.01 psu during the same periods. The depth of
the LSW core was continuously increasing during the entire period, being ca. 1300 m in 1988
increasing to 1600 m during the years 1992 to 1994 ending up at approximately 1850 m in
the period 1996 to 1998. It is important to note that the density increase observed during 1997
and 1998 is linked to the salinity increase mentioned above, suggesting a mixing between the
LSW and Iceland-Scotland Overflow Water (ISOW).
With support in results reported by Sy et aI. (1997) the present data suggests that the
transit time of the LSW from the WOCE AlE section situated at 600 N to the northern part of
the Irminger Sea is in the range of 6 to 12 months. With an assumed length of the pathway of
444 km, this is equivalent to a mean speed of about 1.4 - 2.9 cm/s.
A historical comparison of two nearby sections occupied in the early eighties (1981)
and late nineties (1998), reveals changes in the distribution of water masses (Figure 3) that
have taken place recently in this period. The changes are seen as a downward movement of
the LSW core in the late nineties occupying the depth range 1500 to 2100 m instead of 500 to
1200 m in 1981.
3
MNA W changes
During its northward path from the North Atlantic Current the MNA W experiences a
gradual cooling and freshening due to atmospheric interaction and mixing with the Subpolar
Gyre and East Greenland Current. After passing the Reykjanes Ridge from the Iceland Basin
it is generally accepted that the MNA W participates in the cyclonic circulation of the
Irminger Sea as the Irminger Current. After branching in the Denmark Strait the western
branch of the MNAW finally joins the East Greenland Current on its southward path as a
band of variable width. The onshore side is characterised by cold and fresh waters of polar
origin, whereas the central Irminger Sea is occupied by relatively cold and fresh Irminger Sea
Water (IW), also referred to as Subarctic Water (SAW) (Dickson et aI., 1998).
A station from the MRI Faxafl6i section, Faxafl6i 9 (fx9), is used to describe the
thennohaline variability in the northern part of the Irminger Sea through the nineties. This
station (Figure 1) is located near the 1000 m isopath on the Icelandic continental slope in the
outer core of the MNA W water mass. Therefore this station can be considered as a good
proxy for the MNA W in the area.
Isopleths of potential temperature, salinity and potential density for fx9 (Figure 4) all
reveal both interannual and seasonal variations in the entire water column. The most
pronounced seasonal signal is found in the surface layer mainly connected to air-sea
exchange. During early summer a seasonal thennocline develops in the surface layer which is
followed by a freshening due to increased stability and precipitation. During winter the
surface layer is broken down by winter mixing .which can reach depths greater than 500 to
600 m. This cycle can to some degree be disturbed by advection. An example of this is seen
in the isopleth of salinity after 1996 as a less distinguished fresh water layer than before.
Below the surface layer the seasonal signal is mainly connected to the winter mixing and the
advection cycle of the MNA W.
The isopleths (Figure 4) indicate warming during the winter 95/96 of the upper layers,
followed by increasing salinities especially since late 1997. During 1998 temperatures and
salinities reached levels (T>7.5°C; S>35.15) which are comparable to those observed during
the warm and saline period before the "Great Salinity Anomaly" in the late sixties (Lee,
1968). Heat and salt content of the water column between 5 and 950 m for fx9 (Figure 5)
support these tendencies and give a clearer picture of the development. The salt content
reveals a clear seasonal signal but this is less distinguishable in the heat content. Beside the
seasonal signal, four pronounced non-seasonal signals are evident. The first change was
observed during the winter 92/93 as a cooling and freshening of the water column. The
second change was observed during the period November 1995 to August 1996 where the
temperature increased by -0.7°C (2.8 GJ/m2) and the salinity by 0.01 psu. Since August 1996
the heat content has stayed at nearly the same level and has only been interrupted by a
4
cooling and freshening event (third change) in spring 1997. This change was followed by
increasing heat and salinity (fourth change) which peaked in May 1998 raising the salinity of
the water column by about 0.095 psu.
Discussion and conclusions
During the nineties the Irrninger Sea was subjected to pronounced thennohaline
changes. At intennediate depth the water mass distribution underwent pronounced changes
due the arrival of the so called "1988 LSW cascade" (Sy et aI., 1997) of newly fonned, cold
and dense LSW. The arrival of the cascade was seen as a pronounced downward movement
of the LSW. Changes in the MNA W layer show a complex nature and observations from
Faxafl6i 9 station suggest that at least four larger changes have taken place during the period
1992 to 1999. Together they resulted in an overall increase in temperature and salinity of the
MNA W water column of the order of 0.5°C and 0.03 psu during this period. The reason for
the seasonal signal (observed particularly in the salt content) has not yet been established but
is likely linked to the advection cycle ofMNAW in the Irrninger Sea.
With respect to the 1988 LSW cascade, it can be added, that there is evidence of
transit times of the LSW in the Irrninger Sea in the range of 6 to 12 months which are
equivalent to mean speed of about 1.4-2.9 cm/s. It is important also to note that the LSW
density increase observed during 1997 and 1998 is linked to the salinity increase, suggesting
a mixing between the LSW and ISOW.
The question is now, are the changes observed in the MNA W at Faxafloi 9 caused by
local or (large-scale) advective processes. Among local processes can for example be
mentioned ocean-to-atmosphere heat fluxes, wind stress curl influencing the doming of the
Subpolar Gyre, frontal movements, eddies and changed mixing conditions between the
MNA W and the interior of the Subpolar Gyre.
Answering this question is not straight forward from the data available. The question
is therefore turned around to become: Which signals or causes mentioned above can be seen
in the Faxafl6i 9 time series?
Estimates of heat fluxes produced by the National Centers for Environmental
Prediction (NCEP) reanalyses and by the European Centre for Medium Range Weather
Forecasts (ECMWF) analyses in the study area has been reported by Reverdin et al. (1999).
They arrive at results, which suggest positive heat flux anomalies larger than 100 W 1m2
during the winter 199511996. Further they report strong negative heat fluxes during the winter
199211993 and a negative anomaly of short duration in late winter 1996/1997. It is interesting
to notice that all three anomalies are well described by the heat content time series observed
at Faxafl6i 9 (Figure 5).
5
Not explained by heat fluxes given by Reverdin et al. (1999) is the heat and salt
content increase observed during the period June 1997 to May 1998. During this period
Icelandic MRI hydrographic cruises observed increasing salinities both south and west of
Iceland. For the first time since the early sixties (e.g. Lee, 1968) salinities above 35.20 were
observed at the Reykjanes Ridge. These observations suggest that advection plays a
significant role in the observed change of the MNAW characteristics between 1997 and 1998.
References
Anon., 1976, Overflow '73 inventory, ICES oceanographic data lists and inventories, no. 29,
203 pp.
Blindheim, J. 1968. Hydrographic investigations in the Irminger Sea in the years 1954-1964,
FiskDir. Skr. Ser. HavUnderk., 14, 72-97.
Dickson, RR, and J. Brown 1994. The production of North Atlantic Deep Water: Sources,
rates, and pathways, J. Geophys. Res., 99, C6, 12319-12341.
Dickson, RR, J. Lazier, J. Meincke, and P. Rhines 1996. Long-term coordinated changes in
the convective activity ofthe North Atlantic, NATO ASI Series, I 44, 211-261.
Dietrich, G. 1957a. Schichtung und Zirkulation der Irminger-See im Juni 1955, Ber. Dtsch.
Wiss. Komm. Meeresforsch., 14(4),255-312.
Dietrich, G. 1969. Atlas of the hydrography of the Northern Atlantic Ocean based on the Polar Front Survey of the International Geophysical Year, winter and summer 1958, International Council for the Exploration ofthe Seas, Hydrographic service, Copenhagen, 140 pp.
Irminger, C. 1853. Om Havets Stnmminger m. m., Nyt Archiv for Sf3Vresenet, Anden Rrekke,
8 bd., 115-137, Ki0benhavn.
Knudsen, M. 1898. Hydrografi, Den danskelngolf-Expedition, Bd. I, No.2, 23-154, K0benhavn.
Knudsen, M. 1899. Hydrography, The Danish Ingolf-Expedition, Vol. I, NO.2, 23-161,
Copenhagen.
Krauss, W. 1995. Currents and mixing in the Irminger Sea and in the Iceland Basin, J.
6
Geophys. Res., 100, C6, 10851-10871.
Lee, A.J. 1968. NORTHWESTLANT Surveys: Physical oceanography, ICNAF Special Publ.
No.7, Part 1,31-54.
Lee, AJ., and D.l. Ellert 1965. On the contribution of overflow water from the Norwegian
Sea to the hydrographic structure of the North Atlantic Ocean, Deep-Sea Res., 12, 129-142.
Lee, A.l., and D.l. Ellert 1967. On the water masses of the northwest Atlantic Ocean, Deep
Sea Res., 14, 183-190.
Nansen, F. 1912. Das Bodenwasser und die Abkiihlung des Meeres, Internationale Revue der
Gesamten Hydrobiologie und Hydrographie, 5, (1), 1-42.
Reverdin, G., N. Verbrugge, and H. Valdimarsson 1999, Upper ocean variability between
Iceland and Newfoundland 1993-1998, (Submitted to J. Geophys. Res.)
Stefansson, U. 1968. Dissolved nutrients, oxygen and water masses in the Northern Inninger
Sea, Deep-Sea Res., 15,541-575.
Sy, A., M. Rhein, 1. Lazier, K.P. Koltennann, J. Meincke, A. Putzka, and M. Bersch 1997.
Surprisingly rapid spreading of newly fonned intennediate waters across the North Atlantic
Ocean, Nature, 386, 675-679.
Talley, L.D., and M.S. McCartney 1982. Distribution and circulation of Labrador Sea Water,
J. Phys. Oceanogr., 12, 1189-1205.
7
Figure 1. Location of the Faxafl6i section and station Faxafl6i 9 (fx9). In addition are shown the location of repeated VEINS CTD sections and moorings deployed by Icelandic research vessels.
3.3 I
• May 88
3.2
G 3.1 .July92 t:iJ Q)
:8-w Sep93 _
• Feb 93 0::: 3.0 ::>
Aug 94 .. May 93 ~ 0::: W c.. 2.9 :2 w ..... May 98 ....= • 0 2.8 Nov 96- • May 1997 c..
2.7
2.6 +----,----r--..-------,---,--.,.--__,_---,--....--------,-----,---j
34.65 34.66 34.67 34.68 34.69 34.70 34.71 SIGMA-t.5(kglm**3)
Figure 2. Potential temperature versus potential density relative to 1500 dbar for the LSW salinity minimum.
8
";
TIO Lc~ 5 11·1" "",,,,,,, I~I
Distance (km)
Figure 3. The distribution of water masses in the Irminger Sea, illutrated by the Transient Tracers in the Ocean (TTO) in August1981 (top) and VEINS (Faxaf16i) in May 1998 (bottom) salinity sections. Shading indicates different water masses: Labrador Sea Water (LSW), Denmark Strait Overflow Water (DSOW) and Iceland-Scotland Overflow Water (ISOW) respectively (the TTO section is adapted from Dickson and Brown, 1994).
9
1992 1993 1994 1995 1996 1997 1998 1999
~ ~ 6! ~:5 ~ ~~ S; ~ ~i ~ ~ ~lH:E~!~ fJ ~ !! ~ ~
lime (months after 01/01/1992)
1992 1993 1994 1995 1996 1997 1998 1999
~ ~ a~ i:m ~!i ~ ~!i ~! SH!~~!5~~~ ~ ~ ~~ ~ ~
lime (months after 01/01/1992)
1992 1993 1994 1995 1996 1997 1998 1999
lime (months after 01/0111992)
Figure 4. Isopleths of (top) potential temperature, (middle) salinity, and (bottom) potential density for Faxafl6i 9 in the period 1992 to 1999.
10
1.140 33.25 .-... N
.-... iC iC
N E iC iC" -E 0') - ~ J 33.20 ('I) r- I , , , iC "-" • , , I
iC
r- I 0
~ T""
Z , , "-" W 1.136 III ~ r-r- z z , .. , ,
LU • 0 I ill, I r-() ... /~ ~ I 33.15 Z
~ I • , . 0 \ I ... ' . • I , , , () ,
W ~ ~, I , , I r-:::c \ , 'S ..J I I I , ' I , J. « I •• ,I
~ CJ)
• 1.132 III 33.10
92 93 94 95 96 97 98 99 00 Time (years)
Figure 5. Temporal variability of heat and salt content of the 5-950 m depth range for Faxafl6i 9 in the period 1992 to 1999.
11
Recommended