Reconstructing Multi-Decadel to Centennial-Scale Variations of the Indian Ocean Dipole During the Last Two Millennia


OCCI Funded Project: 2007


The Indian Ocean Dipole (IOD) is a coupled mode of interannual variability that causes droughts and floods in many tropical regions, with large socio-economic and ecological impacts. Like Pacific interannual variability (the El Niño Southern Oscillation, or ENSO), the typical IOD is phase-locked to the annual cycle. It begins with cooling in the eastern tropical Indian Ocean and evolves and grows through positive feedbacks between wind and upper ocean dynamics. The fully developed IOD is characterized by anomalous cooling of the eastern and anomalous warming of the western tropical Indian Ocean. Several lines of evidence suggest that the IOD is an intrinsic coupled mode, although its frequency, amplitude, and climate effects interact with those of ENSO. The IOD may also modulate the ENSO-Indian Monsoon link. Due to the short instrumental record, and the poor representation of the IOD-ENSO link in climate simulations there is considerable uncertainty about multi-decadal and longer time scale IOD variability, its independence from ENSO, its influence on the ENSO-Indian monsoon link, and its behavior under different climatic regimes.

I request funds from the OCCI to begin research on how the IOD varied on centennial time scales during the last 2,000 years, and how it interacted with variations in other interannual or decadal modes on this time scale. I propose to generate reconstructions of sea surface and thermocline temperature and δ18O of seawater spanning the last two millennia from rapidly accumulating marine sediments (>60 cm/ky) from the eastern Indian Ocean, west of Sumatra, in the heart of the Indian Ocean warm pool. The site has a small seasonal cycle, but its surface temperature (SST), salinity, and thermocline structure respond dramatically to IOD events. Although we will not be able to resolve individual interannual events, modeling studies show that the mean climate state on longer time scales, which we can resolve with sediment records, is influenced by variations in interannual variability (e.g., Clement et al. 1999; Abram et al., 2007).

The questions that will be addressed include:

  • Are today’s warm SST’s in the Indian Ocean warm pool unprecedented during the last two millennia?
  • Are SST changes consistent with a North Atlantic link during the last two millennia?
  • What is the influence of Indian Ocean warm pool SSTs on centennial variations in the amplitude/frequency of the IOD and its interactions on these time scales?
  • Do periods of greater high latitude Northern Hemisphere warmth (e.g, the Medieval Warm Period) correspond to intervals with higher IOD activity (e.g., Kripalani and Kumar, 2004), compared to cooler intervals (e.g., the Little Ice Age)?
  • Is centennial scale IOD variability correlated with reconstructed dry/wet cycles over Africa? Indian monsoon rainfall? With local (Indonesian) rainfall anomalies?
  • Are IOD events more frequent/intense when ENSO variability is weak or absent e.g, around 1000 A.D, and episodically during the last millennium (Moy et al., 2003; Rein et al., 2006)?
  • Does the IOD modulate the ENSO-monsoon link on centennial time scales?The analyses proposed includes a core-top component that would provide proof of concept for thermocline reconstructions in the region. Together with the downcore work, this would improve.
The analyses proposed includes a core-top component that would provide proof of concept for
thermocline reconstructions in the region. Together with the downcore work, this would improve
chances of success for an NSF proposal, including work on abrupt events (e.g., during the last
deglaciation, glacial abrupt events) at the same and additional sites from the eastern Indian

The IOD is an important player in the tropical hydrologic cycle. It causes droughts and floods in
many regions. It interacts with ENSO and modulates its effects. Thus, this proposal addresses
the OCCI theme of how the hydrologic cycle influences climate on decadal to millennial times