Introduction: Integrating high‐resolution past climate records for future prediction in the Australasian region

Recognising the challenge of human-induced climate change, the World Meteorological Organization and the United Nations Environment Program established the Intergovernmental Panel on Climate Change (IPCC) in 1988 to assess our understanding of the scientific basis of risk of anthropogenic climate change and opportunities for adaptation and mitigation. Unfortunately, relatively widespread instrumental measurements only extend back to the mid-19th century. As a result, it has not yet proved possible to develop seasonspecific histories for different climate variables at a regional scale to unambiguously find a ‘fingerprint’ of climate response to increasing greenhouse gas emissions (Jones et al., 2001). An alternative approach has been to focus efforts on large-scale changes in the average climate. There exists a wealth of indicators of past climate that show rapid changes have taken place. Annually resolved climate proxies, such as tree-rings, corals, annually laminated lake sediments and historical climate datasets, have formed major contributors to such studies; typically, past surface temperatures have been inferred by calibrating to instrumental temperature data using statistical relationships. The best known is the so-called ‘hockey stick’ temperature reconstruction for the Northern Hemisphere (Mann et al., 1998, 1999); the resulting curve sloped gently downward for most of the past millennium (the handle of the hockey stick), then rose sharply into the 20th century (the blade) until it topped the relative warmth of 800 to 1000 years ago. This reconstruction featured prominently in the 2001 IPCC report. Although some criticism has been levelled at the reconstruction, the results have recently been supported by a National Research Council panel of the US National Academy of Sciences, but it was noted that more must be done to reduce uncertainties in the early part of the curve (Brumfiel, 2006) and in the Southern Hemisphere (Kerr, 2006). Some attempts have been made by the international scientific community to address the concerns highlighted by the US National Academy of Sciences. For instance, the dependence on annually and decadally resolved datasets has not always captured variability on the multi-centennial timescale. By combining lowand