Abrupt climate change during the last glacial cycle based on Gulf of Mexico sediments
Investigators: Benjamin P. Flower, David W. Hastings (Eckerd College), Heather W. Hill, and Terrence M. Quinn
1. Testing Hypotheses
on Abrupt Climate Change
• Greenhouse gas forcing
• Modulation by thermohaline circulation
• Routing of Laurentide ice sheet meltwater
2. Why the Gulf of
Mexico?
• Relation to tropical Atlantic
• Laurentide meltwater history
• Laminated Orca Basin
3. Paired data
• d18O data on Gs. ruber (Leventer et al.,
1982; this study)
• Mg/Ca SST: Mg/Ca =0.38 exp[0.09 • SST (°C)] (Dekens et al.,
2002)
• d18Oseawater: T (°C) = 14.9 - 4.8 (d18Oc
-d18Osw) (Bemis et al., 1998; Thunell et al., 1999;
Spero et al., 2003)
4. Comparison to
GISP2 and Byrd
• DSST ~4°C 17-15 ka during Heinrich 1
• asynchronous with GISP2
• major d18Oseawater decreases during B/A and “precursor”
during H1
5. Deglacial Phasing
of SST
• Synchronous with GISP2
• Not synchronous with GISP2
• Similar to Tobago Basin (Ruhlemann et al., 1999) but not Cariaco Basin
(Lea et al., 2003)
6. Estimating salinity
from d18Oseawater
• Correct d18Oseawater for ice volume effect
based on U/Th dated sea level history (Fairbanks, 1989; Bard et al., 1992
• Assume -7 ‰ low-salinity end-member for Holocene and -25 to -35
‰ end-member for deglaciation
• re-routing of LIS meltwater during H1 and onset of YD?
7. Deglacial Phasing
of SST
• similar relation during H1 and H4
• but no Heinrich event at “cessation event” at ca. 35 ka
• PP42A-0855 Hill, Flower, and Quinn, 2003
8. Conclusions
• Deglacial GOM SST warmed >3°C ca.17-15.2 ka (during H1, prior
to Bølling/Allerød)
• Asynchronous relation to GISP2 is consistent with Atlantic THC modulation
of deglacial climate, but must be tested using other cores
• Separation of SST and d18Osw in GOM (for the first time) indicates
• Laurentide ice sheet meltwater input during H1 and the Bølling/Allerød
• Salinity estimates using reasonable low-salinity end-members indicate
modest changes (<4 psu) in MIS 2/1 and MIS 3 relative to early Holocene
• LIS meltwater routing shows a similar pattern during the last deglaciation
and MIS 3, except no H event during “cessation event” ca. 35 ka
9. Acknowledgements
Ethan Goddard (USF), David Hollander (USF), Jenna LoDico (USF), Jess Adkins
(CalTech), Gerald Haug (Potsdam University), Danny Sigman (Princeton University),
Tom Guilderson (LLNL), David Lea (UC Santa Barbara), Howie Spero (UC Davis),
IMAGES Program, The 2002 “Geochemical and Oceanographic Tracers”
class Eckerd College/ College of Marine Scienc, American Chemical Society PRF
Grant 38056-AC2
