Gulf Stream Temperature, Salinity, and Transport during the Last Millennium

Transcription

1 Gulf Stream Temperature, Salinity, and Transport during the Last Millennium David Lund - University of Michigan Part I - Geostrophic estimation of Gulf Stream flow - Gulf Stream strength varied by ~10% - Weaker vertical shear and transport during Little Ice Age - Consistent with timing of North Atlantic cooling Part II - Sea-surface temperature and salinity - Gulf Stream salinity increased during LIA - Likely due to southward migration of the Inter-Tropical Convergence Zone Part III - Linking the oceanic circulation and ITCZ

2 North Atlantic surface circulation wind-driven gyre flow NADW compensation Schmitz, 1996

3 Gulf Stream volume and heat transports 31 Sv (1Sv = 1x10 6 m 3 s -1 ) 1.3 PW (1PW = W) 25-40% of heat transport due to shallow gyre overturning Talley, 1999 Schmitz, 1996 Larsen, 1992

4 Location of core sites in Florida Straits

5 Core retrieval and sampling

6 Benthic foraminiferal δ 18 O and density Cibicidoides floridanus 200 µm increasing S and decreasing T causes sea water density AND foraminiferal δ 18 O to increase after Lynch-Stieglitz et al., 1999

7 Foraminiferal density estimates match modern observations Water Depth (m) density ( t ) Lund et al., 2006

8 Estimating Gulf Stream transport Modern techniques - current meters ~ Sv Pillsbury, 1890; Schott et al., 1988; Leaman et al., submarine cable estimates of 29 Sv (Jan) to 33 Sv (July) Wertheim, 1954; Baringer and Larsen, geostrophic estimates ~ Sv Montgomery, 1941; Schmitz and Richardson, 1968 Geostrophic estimation (using thermal wind equations) v z u z = - = g ρ o f ρ x g ρ o f ρ y ---> vertical current shear is proportional to horizontal density gradient * transport can be calculated only if velocity at a given depth is known

9 Dry Tortugas δ 18 O Bahamas δ 18 O

10 Florida Current density cross-sections 200 yr BP 1100 yr BP

11 Lower Gulf Stream transport during Little Ice Age modern annual average L I A what do estimates of salinity tell us about LIA conditions? Lund et al., 2006

12 Dry Tortugas reflects tropical Atlantic salinity Dry Tortugas salinity= ggc and 62mc Levitus, 1994

13 North Atlantic surface salinity controlled by evaporation-precipitation rate I T C Z I T C Z evaporation-precipitation rate dasilva et al., 1994

14 Calculating surface salinity using planktonic foraminifera Globeriginoides ruber δ 18 O calcite = δ 18 O water (SST) Kim and OʼNeil, 1997 Lynch-Stieglitz et al., 1999 δ 18 O water = δ 18 O calcite (SST) 200 µm δ 18 O water ( ) Salinity (psu) Mg/Ca (mmol/mol) Anand et al., Temperature ( C)

15 Dry Tortugas δ 18 O calcite increased during LIA G. ruber δ 18 O ( ) 95% confidence interval 100yr running mean Lund and Curry, 2006

16 Dry Tortugas sea surface temperature increased during LIA Lund and Curry, 2006

17 Dry Tortugas δ 18 O w increased during LIA salinity Lund and Curry, 2006

18 Dry Tortugas δ 18 O w record is replicable δ 18 O water anomaly ( ) salinity

19 The magnitude of δ 18 O w variability is due to either: A) influence of thermocline water or B) incorrect Mg/Ca calibration! 18 O c (, PDB) Calendar Age (yr BP) SST (ºC) Schmidt, 1999 Craig & Gordon, 1965! 18 O w (, SMOW) * * * Salinity based on multivariate equations of demenocal et al., 2007

20 Higher LIA salinity driven by southward ITCZ migration I T C Z I T C Z evaporation-precipitation rate dasilva et al., 1994

21 Coherent change in northern Venezuela precipitation saltier northern tropical Atlantic drier northern Venezuela δ 18 O water anomaly ( ) LIA Haug et al., 2001

22 Gulf Stream characterized by low transport and high surface salinity during the Little Ice Age

23 A function of reduced windstress curl? Annual mean Sverdrup streamfunction My 1 β Tx y Johns et al., 2002

24 nomalies exhibit a zonal band of increased aridity jus precipita orthmodel of theresults equator, although precipitasuggest curl opposing increases positive when ITCZ on migrates anomalies to the south of this band are more subtle southward! - SST anomalies scaled to simulate - forcedsignificant NCAR CAM3 using SST igure 2. mean annual precipitation (contours (contours, anomalies from hosing experiment LIA observations!1!1 wind stress (vectors, dyn cm ) anomalies mm day(zhang) &and Delworth, 2005) or (a) NA1, (b) NA2, (c) XTA1, (d) XTA2, (e) TNA1, and ) TNA2. Shading highlights positive (dark/green) and Saenger et al., 2009

25 A role for the MOC? warm ` cold HadCM3 surface temp. (ºC) anomalies wet ` dry HadCM3 precipitation (cm/yr) anomalies Vellinga and Wu, 2004

26 Conclusions Gulf Stream transport varied by ~10% during the last millennium, but was 3±1 Sv lower during Little Ice Age Surface Gulf Stream salinity increased during the LIA, most likely due to southward ITCZ migration Simultaneous transport and salinity variability implies tight linkage between oceanic circulation and hydrologic cycle on centennial time scales Southward migration of wind-field would likely enhance flow, implying MOC was primary driver of LIA transport anomaly

27 Holocene variability in Gulf Stream transport a) Transport (Sv) Transport increased ~ 4 Sv during Holocene b) Calendar Age (yr BP) Lynch-Steiglitz, et al., in press ITCZ migrated southward during Holocene Haug et al., 2001

28 Dry Tortugas δ 18 O water record mimics Δ 14 Catm Lund and Curry, 2006

29 North Atlantic region cooled by ~1 C during LIA Greenland Temp. ( o C) Dahl-Jensen et al., 1998 Lamb, 1995 Marchitto and demenocal, 2003 Bond et al., 2001 Keigwin, 1996 demenocal et al., 2000

30