Changing surface waters and deep-water formation in the eastern
Nordic Seas
In the eastern Nordic Seas, by comparison, planktic
δ18O values of 4.6–4.8 ‰ did not show any substantial
HS-1 meltwater signal over the LGM and most of HS-1, from 18.4–16.3
cal. ka, except for a minor δ18O low
~17.6–17.2 cal. ka (by Δδ18Ο = 0.3 ‰;
Fig. 2b). Figs. 2b and 3, however, display an almost sudden 2-‰ drop in
the planktic δ18O record, that clearly document –
after the first incursion of meltwaters at PS2644 in the west – a
second major ice and meltwater outbreak during the last 1000 yr of HS-1a
starting at 16.3-16.0 cal. ka (as with the onset of HS-1b, precisely
coeval with a second major HS-1 cooling depicted in Greenland ice cores
GRIP2 and NGRIP). The meltwater signal started from the Barents shelf,
spread south down to Faeroe (Sarnthein et al., 2001). The ice breakout
obviously induced a turbulent mixing down to intermediate waters at Site
GIK23074 at 1157 m w. d. (Fig. 3), and lasted until
~14.7 cal. ka. This age estimate on top of HS-1a fits
precisely a δ18O depletion found in Greenland records
NGRIP and GISP2 (e.g., Grootes and Stuiver, 1997).
Over time segments I and II, eastern surface waters showed
fairly low MRA of ~500 to 800 and 1200 yr
(Figs. 2b and 3 that closely resemble MRA values recorded for the very
early LGM by Simon et al. (2023). However, they differed strongly from
the high MRA of 1900-2200 yr obtained from Site PS2644 in the west (Fig.
2a). Starting at 18.4 cal. ka, however, eastern MRA depict a fast rise
to 1730 yr and 2000 yr, rapidly reaching a close match with the MRA
found at western Site PS2644 over time segment III and early segment IV
(Fig. 2b). Accordingly, the differential sense of eastern and western
surface water circulation of time segments I and II ended and was
briefly replaced by a close match during time segments III and IV, when
’old’ surface waters with an Arctic origin similar to those of the EGC
had started also to cover the eastern Nordic Seas. Per analogy, we infer
that this advection dominated until ~16 cal. ka.
During time segment I, extremely low bottom water ventilationages at Site GIK 23074 attest lively intermediate-water convection
somewhere close to the site in the eastern Nordic Sea at least down to
~1200 m w.d. (Figs. 3 and 6; in harmony with Meland et
al., 2008; Thornalley et al., 2015). In contrast, the regime at Site
PS2644 was stratified. Occasionally, extremely young intermediate waters
of Site GIK23074 found their way from the east up to the west, at Site
PS2644 forming DSO mode 1 during time segment II (Fig. 4), where local
deep-water convection continued in the eastern Nordic Seas, though
somewhat attenuated (Fig. 3).
After 18.4 cal. ka, with the start of time segment
III, the circulation geometry of the eastern Nordic Seas was strongly
modified (Fig. 6; modified after Thornalley et al., 2015). Theinflow of NAC stopped . It was r eplaced by surface
waters with high and very high MRA at Site GIK23074 / MD95-2311 (1157
m), that closely equate the MRA at Site PS2644 reflecting the EGC.
Hence, the waters were probably Arctic-sourced. Their arrival was
precisely coeval with the start of meltwater incursion through the
Denmark Strait (Figs. 3, 4). Local intermediate and deep-water
convection (sensu Siedler et al., 2001) then was replaced by
stable stratification and dominantly seasonal, probably fairly modest
volumes of brine-enhanced shelf waters leading to modest deep-water
formation (Bauch and Bauch 2001; Waelbroeck et al., 2011; Thornalley et
al., 2015). This shift is reflected by a local abrupt rise in bottom
water ventilation ages from <1000 yr up to 2100-2200 yr, ages
that slightly exceed the paired, likewise strongly raised MRA of surface
waters reaching 1900-2100 yr (Fig. 3). In turn, the great iceberg and
meltwater outbreak from the Barents Sea after 16.3 cal. ka obviously
resulted in local stirring and an immediate overturning and renewed
local ventilation of intermediate waters in the eastern Nordic Seas
(Fig. 3).
Different from a model of Sessford et al. (2019), the brine
water-induced intermediate waters of the eastern Nordic Seas during time
segments III and IV apparently did never spread up to the westernmost
Nordic Seas to replace DSO waters at Site PS2644. Our benthic
ventilation ages and other proxy data suggest that the circulation
geometry of the west was different from today being dominated by the
North Iceland Jet (sensu Våge et al., 2011) that entrained DSO
mode-3 waters from the east along the northern slope of Iceland and
ultimately, from upper North Atlantic Intermediate Waters to the south
and southeast of Iceland.