The scientific content of the workshop could be divided into two general themes. The application and use of palaeodatabases and the organisation, type and quality of data that is appropriate for research purposes. Several examples of application and use of databases were presented in some detail with palaeoclimate and land surface reconstruction forming central foci. The recognition that Holocene tree distributions are largely in equilibrium with prevailing climatic conditions means that past climates can be reconstructed from pollen data by inversion of dynamic vegetation models. Pollen assemblages with no modern analogues can be used to reconstruct climates with no modern analogues and these were widespread at the onset of the Holocene. Data-model comparisons, which are a central theme of the EuroCLIM DECVEG project, are testing palaeoclimatic reconstructions derived from an atmospheric general circulation model (AGCM) output with some success (Fig.1).
Figure 1. Pollen data compared with modelled vegetation driven by AGCM output from a site in central Scandinavia. The relatively good match confirms that vegetation dynamics are largely driven by climate change and that the modelled palaeoclimate is close to reality (Miller et al. submitted).
The science of databases themselves has developed with technical computational developments. New user-friendly interfaces conceal complex relational databases that can handle far larger databases than was feasible when the EPD was first established. As a consequence, databases can be less restrictive about the data they accept and the user can select data of appropriate quality for the task in hand with greater ease. Standardised but appropriate age-depth models are needed for typical database queries that users wish to make for e.g. time-slice studies or dynamic modelling.