High resolution from low resolution; a new method for time-series pollen-climate reconstruction and its application in investigating abrupt Holocene climate change in Southern Europe/Mediterranean
by
Basil Davis
IMEP-CNRS, Centre Universitaire d'Arles, Arles, 13200, France; Department of Geography, Newcastle University, Newcastle upon Tyne, NE1 7RU, UK
Coauthors: Simon Brewer (IMEP-CNRS), Tony Stevenson (Newcastle University), Steve Juggins (Newcastle University)

There is an increasing need for palaeo-climate records with high temporal resolution in order to investigate short-term periodic and abrupt changes in climate. In particular, the possibility of a future collapse of the North Atlantic thermohaline circulation (THC) as a result of anthropogenic forcing has highlighted the need to study the terrestrial/atmospheric impact of past changes in the THC identified in marine cores. Any development of ocean-atmosphere models that can predict changes in the THC will require validation against comparable paleoclimatic data. This ideally means long term (Holocene), high resolution (decadal) quantitative palaeoclimate records of a range of climate parameters (temperature, precipitation etc) at a continental scale (Europe).

Pollen data has been succesfully shown in the past to fulfill all these criteria except one, that of high temporal resolution. The traditional 'blurred snapshot' approach (eg 6000±500 BP) is inappropriate for identifying short-term dynamic climate change of the kind associated with THC events. However, whilst few individual pollen cores have been sampled at this kind of resolution, this poster shows how the considerable number of records now available in the European Pollen Database (EPD) offers a new approach to the problem.

A major feature of pollen data is its regional cohesiveness, with pollen records from one site often being used as a basis for correlation with others to form a regional vegetation history. This regional cohesiveness is further improved when using quantitative pollen-climate reconstruction methods that reduce complex taxa to simple climatic parameters. A large number of cores from a region can therefore be expected to record similar climate changes, albeit at variable time intervals and over a variable climatic (eg lapse rate) gradient. By assimilating clusters of cores from a region and interpolating through time and space to a constant (virtual) location, a new record (a 'cluster-core') can be created based on thousands of samples with considerably improved resolution.

This technique has been applied to the Southern European/Mediterraean area using 133 cores with extended Holocene records from the EPD and the authors. Pollen-climate reconstruction was based on the improved modern analogue technique using PFT scores, together with revised and calibrated chronologies. Data is presented for the area as a whole, as well as for a series of regional cluster-cores. Results indicate that the technique is succesful in capturing both long-term solar driven change, as well as short term events shown in the Greenland dO18 record and marine core records. Inter regional comparison of cluster-core records indicate important spatial variations between events, as well as changes in this relationship over time. For Holocene sub-Century palaeoclimate reconstruction based on a uniform technique applied over a wide spatial area, the cluster-core technique shows considerable promise. The technique should also be applicable to other proxies.

Acknowledgements The authors would like to thank Joel Guiot and Rachid Cheddadi for help in supplying pollen and modern climate data, and particularly all contributors to the EPD, without whom this study would not have been possible.

Date received: May 15, 2001

Copyright © 2001 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cahi-93.