2. Coral records of ENSO through the late Holocene Kim Cobb Hussein Sayani Georgia Inst. of Technology Chris Charles Niko Westphal Scripps Inst. of Oceanography Larry Edwards, Hai Cheng University of Minnesota with thanks to ACS-PRF, NOAA, NCL, PARC, Cobb lab undergrads

3. Motivation: How is the tropical Pacific climate system responding to anthropogenic forcing? Approach: Use well-dated, high-resolution paleoclimate records from the tropical Pacific to assess its response to known natural and anthropogenic climate forcings. Example: El Nino-Southern Oscillation response to insolation forcing during the current interglacial period (the “Holocene”)

4. Dai and Wigley, 2000 El Niño Temperature El Niño Precipitation “El Niño-Southern Oscillation” (ENSO) ENSO is a climate phenomenon in the tropical Pacific which arises from coupled interactions between the atmosphere and ocean ENSO impacts global climate every 2-7 years tropical Pacific climate variability over decades to centuries to millennia poorly constrained; 20 th century trends uncertain Why tropical Pacific climate?

5. Vecchi et al, 2008 - climate models project widely divergent scenarios for tropical Pacific climate under greenhouse forcing Climate model response to greenhouse forcing

6. Vecchi et al, 2008 - instrumental SST datasets contain trends of different signs  no help Instrumental SST trends

7. Ice age cycles are timed to “insolation” forcing

8. What happened to ENSO during the mid-Holocene? Model: Clement et al., 2000; Lake P: Moy et al., 2002; - climate models generally show reduced ENSO activity in response to insolation forcing… BUT are they correct? - primary evidence of reduction comes from 2 lake and 1 marine sediment records (Moy et al., 2002; Conroy et al., 2008; Koutavas et al., 2006) but these do not agree!

9. A composite of available Holocene coral ENSO activity Model: Clement et al., 2000; Lake P: Moy et al., 2002; Corals: Cobb et al., 2003; Tudhope et al., 2001; Woodruffe et al., 2003; McGregor and Gagan 2004; Correge et al., 2001 - scant fossil coral ENSO reconstructions during Holocene - no statistically significant reduction in ENSO variance from fossil corals - primary evidence of reduction comes from 2 lake and 1 marine sediment records (Moy et al., 2002; Conroy et al., 2008; Koutavas et al., 2006) but these do not agree!

10. 150°E180°150°W120°W b SST (°C) AVERAGE CONDITIONS EL NIÑO SST, hydrology, and coral δ 18 O in the Line Islands Coral oxygen isotopes (δ 18 O) decreases when warm (thermodynamics) Coral δ 18 O decreases when rainy (lower seawater δ 18 O)

11. Reminder: Palmyra coral δ 18 O is a sensitive proxy for ENSO Cobb et al., 2003

12. Overlapping fossil coral δ 18 O records during last millennium Cobb et al., in prep

13. Palmyra 50 cores U/Th dated 18 cores undated Christmas 25 cores U/Th dated 51 cores undated Fanning 17 cores U/Th dated 19 cores undated The Line Islands Coral Collection modern cores from three islands splice overlapping cores in last millennium many cores in mid-Holocene 1 2 3

14. Strengths of Line Islands Fossil Corals 1.Modern coral δ 18 O has high correlations with NIÑO3.4 SST index 2.Overlapping fossil coral δ 18 O records highly correlated during last millennium; provide uncertainty estimates for single fossil coral reconstructions 3.New collection of >100 fossil corals extends back 7000 years (20-100yrs each)

15. New fossil coral δ 18 O records from Fanning: 6-7kybp What do we see in the mid-Holocene fossil corals?

16. Scanning electron microscope (SEM) photos of 6ky fossil coral - surface coatings represent no more than ~10% by mass - original coral geochemical signature likely preserved, not “altered”

17. A modern-young fossil coral comparison Crespo et al., in prep

18. New Line Islands Fossil Coral Data  no change in ENSO - added 650 years of monthly-resolved data (red dots = our stuff) - we have tripled the amount of fossil coral data available from the mid- to late- Holocene Model: Clement et al., 2000; Lake P: Moy et al., 2002; Single-foram (purple): Koutavas et al., 2006 Corals: Cobb et al., 2003; Tudhope et al., 2001; Woodruffe et al., 2003; McGregor and Gagan, 2004; Correge et al., 2001 What is the take-home for climate scientists?

19. Conclusions -No fossil coral support for reduced mid-Holocene ENSO activity climate model response questionable, so 21 st century ENSO projections of same models questionable - Reduced ENSO activity throughout late Holocene, combined with evidence for cooler and drier conditions - explanation? - need more data from the tropical Pacific!

20. What’s next? How much does seawater δ 18 O sw vary today across Line Islands? Postdoc Jess Conroy Careful diagenetic investigations on fossil corals & Micro-scale Sr/Ca analyses of pristine sections of fossil coral skeletons Student Hussein Sayani We probably need many more corals! Koutavas-like approach applied to 5-10yr-long sequences?