Abstract

Fossil pollen records are well-established indicators of past vegetation changes. The prevalence of pollen across environmental settings including lakes, wetlands, and marine sediments, has made palynology one of the most ubiquitous and valuable tools for studying past environmental and climatic change globally for decades. A complementary research focus has been the development of statistical techniques to derive quantitative estimates of climatic conditions from pollen assemblages. This paper reviews the most commonly used statistical techniques and their rationale and seeks to provide a resource to facilitate their inclusion in more palaeoclimatic research. To this end, we first address the fundamental aspects of fossil pollen data that should be considered when undertaking pollen-based climate reconstructions. We then introduce the range of techniques currently available, the history of their development, and the situations in which they can be best employed. We review the literature on how to define robust calibration datasets, produce high-quality reconstructions, and evaluate climate reconstructions, and suggest methods and products that could be developed to facilitate accessibility and global usability. To continue to foster the development and inclusion of pollen climate reconstruction methods, we promote the development of reporting standards. When established, such standards should 1) enable broader application of climate reconstruction techniques, especially in regions where such methods are currently underused, and 2) enable the evaluation and reproduction of individual reconstructions, structuring them for the evolving open-science era, and optimising the use of fossil pollen data as a vital means for the study of past environmental and climatic variability. We also strongly encourage developers and users of palaeoclimate reconstruction methodologies to make associated programming code publicly available, which will further help disseminate these techniques to interested communities.