Abstract
The quality and availability of climate data suitable for palaeoclimatic calibration and verification for the Australian region are discussed and documented. Details of the various datasets, including problems with the data, are presented. High‐quality datasets, where such problems are reduced or even eliminated, are discussed. Many climate datasets are now analysed onto grids, facilitating the preparation of regional‐average time series. Work is under way to produce such high‐quality, gridded datasets for a variety of hitherto unavailable climate data, including surface humidity, pan evaporation, wind, and cloud. An experiment suggests that only a relatively small number of palaeoclimatic time series could provide a useful estimate of long‐term changes in Australian annual average temperature. Copyright © 2006 John Wiley & Sons, Ltd.