The westerly winds determine weather patterns and exert an effect on carbon dioxide (CO₂) flux in and out of the Southern Ocean, an important sink for atmospheric CO₂. Our research reconstructs changes in the westerlies over the past 5000 years, using a marine sediment core record collected from the subantarctic Auckland Islands. Our results indicate an intensification of the westerlies around 1600 years ago, contemporaneous with other records from comparable latitudes across the Pacific Ocean.

To explore whether the oxygen-18 to oxygen-16 ratio of atmospheric O₂ is sensitive to the position of the tropical rain belts, we (1) present a record of ice core bubble oxygen isotope measurements from two Antarctic ice cores, and (2) examine the sensitivity of oxygen isotopes in precipitation, weighted by photosynthesis, to the location of oxygen production over the modern-day seasonal cycle. We find a strong modern relationship and discuss implications for past shifts in tropical rainfall.

To improve our understanding of paleoclimate in the northeastern (NE) US, we compiled data from pollen, tree rings, lake levels, testate amoeba from bogs, and other proxies from the last 3000 years. The paleoclimate synthesis supports long-term cooling until the 1800s and reveals an abrupt transition from wet to dry conditions around 550–750  CE. Evidence suggests the region is now becoming warmer and wetter, but more calibrated data are needed, especially to capture multidecadal variability.

We present the first model analysis using a fully coupled dynamic atmosphere–ocean–vegetation GCM over the last 120 kyr that quantifies the net effect of vegetation on climate. This analysis shows that over the whole period the biogeophysical effect (albedo, evapotranspiration) is dominant, and that the biogeochemical impacts may have a lower possible range than typically estimated. This emphasises the temporal reliance of the balance between biogeophysical and biogeochemical effects.

Existing sedimentary palaeoclimate records in Australasia were assessed for suitability for examining the last 2 millennia. A small number of high-quality records were identified, and new Bayesian age models were constructed for each record. Findings suggest that Australasian record chronologies and confidence in proxy–climate relationships are the main factors limiting appropriate data for examining Common Era climate variability. Recommendations for improving data accessibility are provided.