The goal of PROMS (Pacific Rainfall on Millennial Scales) was to bring together climate modelling and palaeoclimate research to understand changes in rainfall over the South Pacific. In paticular we studied the South Pacific Convergence Zone (SPCZ), a large band of rainfall which stretches diagonally across the South Pacific. State-of-the-art coupled climate models struggle to represent the SPCZ well due to biases in tropical Pacific Ocean temperatures. We get around this issue by using a simpler atmosphere-only model.
This work was a collaboration primarily between the University of East Anglia and the University of Southampton. I led the running of our climate model simulations, supervised by Profs Manoj Joshi, Tim Osborn, and Adrian Matthews.
Variability over the past 1,000 years
The project team travelled to a number of islands across the South Pacific to collect lake sediment cores, before analysing these samples in the lab at Southampton. Through a combination of carbon dating and isotope analysis, the team were able to reconstruct records of rainfall across the islands over the past few thousand years. We found that the SPCZ has shifted eastward over the past 1,000 years, with more eastern islands like Nuku Hiva getting wetter, and more western islands like Samoa getting drier.
Furthermore, by using IGCM4 (Intermediate-complexity General Circulation Model, version 4) climate model simulations, we were able to show that this eastward shift in rainfall is consistent with a change in ocean temperatures over the same period. The difference in temperature between the warm tropical west Pacific, and relatively cooler tropical east Pacific, weakened over the past 1,000 years. The weakening of this temperature gradient caused a subsequent weakening of the Pacific Walker circulation, leading to the eastward shift in rainfall.
You can read more in our paper, Ocean variability drives a millennial-scale shift in South Pacific hydroclimate, published in Communications Earth & Environment.
Drivers of decadal-to-millennial scale SPCZ variability
Through further climate model simulations with IGCM4, we sought to better understand the mechanisms that drive changes in the SPCZ over long timescales. To do this, we ran IGCM4 with a number of different sea surface temperature (SST) patters applied. These SST patters corresponded to three key modes of climate variability: the Interdecadal Pacific Oscillation (IPO), Atlantic Multidecadal Variability (AMV), and shifts in the Southern Hemisphere polar jet (SHPJ).
We found that the IPO dominates changes in the SPCZ, with a more positive IPO (warm waters in the tropical east Pacific, north-western Pacific, and south-western Pacific) causing the SPCZ to shift north-eastward. The AMV has very little effect on the SPCZ, however we note that our atmosphere-only model may not capture oceanic teleconnection pathways. Finally, shifts in the SHPJ can cause rainfall changes across the South Pacific by altering Hadley and Ferrel cell, however these changes do not correspond to large-scale shifts in the SPCZ.
You can read more in our paper, Decadal-scale drivers of South Pacific hydroclimate variability, published in the Quarterly Journal of the Royal Meteorological Society.
South Pacific rainfall during the last Ice Age
Whilst the majority of our sediment cores contain information about the last 1,000-5,000 years, one particular core, taken from the island of Nuku Hiva, French Polynesia, stretches back 60,000 years. This means that we can analyse rainfall in the region during the last ice age (Last Glacial Maximum) around 20,000-25,000 years ago.
We found that the weather on Nuku Hiva would have been cooler and drier in the last ice age than it would have been today, which contradicts the predictions of coupled climate models. This difference is, however, likely due to biases in the climate models. Moreover, pollen samples also present in the sediment core show an absence of plants usual found in “cloud forrests”, which again points to a drier climate.
You can read more in our paper, Cooler and drier climate in the South-Central Pacific during the last glacial period, published in Communications Earth & Environment.
The SPCZ and human migration
Having shown that rainfall has shifted eastward over the past 1,000 years, we then went on to link this rainfall change to human migration patterns across the South Pacific. Having migrated from South East Asia into western Polynesia around 3,000 years ago, humans then ceased their eastward migration in what is known as the “long pause”. Around 1,000 years ago migration resumed, reaching eastern Polynesia before spreading to Hawaii and New Zealand.
We showed that this resumption in migration coincides with the shift in rainfall, suggesting that this climatic shift may have been a driving factor in the migration. As western Polynesia got drier, conditions became less optimal for agriculture, whilst east Polynesia got wetter and therefore better suited to supporting a society.
You can read more in our paper, How did a changing climate in the tropical South Pacific contribute to the eastward migration and settlement of Polynesia?, published in the Journal of Pacific Archaeology.