Interannual and decadal climate variability in coral and instrumental records, with a focus on the western Indian Ocean

Our understanding of the climate of the western Indian Ocean, and particularly the long-term stability of its climatic patterns, has long been limited by a lack of reliable data. In this dissertation I measured oxygen isotopes (δ 18O) in corals to create a 150 year proxy record of climate of the western Indian Ocean. I then used coral and instrumental records to investigate regional and global climatic relationships.; In Chapter 1, I created the first published monthly resolution coral record spanning more than a century. I used temperature and rainfall data to judge the validity of both instrumental and coral measurements, and to paint a detailed picture of the climate of the western Indian Ocean over the last 150 years. I demonstrate that the western Indian Ocean has responded, without exception, to every El Nino-Southern Oscillation (ENSO) event over at least the last 100 years, and identify a remarkably consistent 12–13 year decadal oscillation as a leading source of variability in the Seychelles coral.; In Chapter 2, I investigate the coral decadal oscillation. This oscillation is part of a tropical decadal teleconnection spanning the Atlantic, Pacific, and Indian Oceans, in which sea-surface temperature (SST) in the central equatorial Pacific oscillates out of phase with SST in the south tropical Atlantic. The decadal oscillation in the Seychelles coral depends on the relationship between Atlantic and Pacific SST: the coral δ18O grows lighter in the “cool Pacific/warm Atlantic” phase, and heavier in the “warm Pacific/cool Atlantic” phase. This pattern, which is distinct from decadal variations in ENSO, has maintained a consistent relationship over the last 150 years.; In Chapter 3, I identify a 9.3 year decadal oscillation in western Indian Ocean SST. I suggest that this oscillation could result from increased tidal mixing of cold sub-surface waters with warmer surface waters as a result of the 9.3 lunar cycle. I compare decadal variability in global SST to patterns of tidal dissipation, and present evidence that the 9.3 year recurrence of extreme tides could influence SST in regions of strong tidal dissipation.