Climate in northern Eurasia 6000 years ago reconstructed from pollen data

https://doi.org/10.1016/S0012-821X(99)00171-5Get rights and content

Abstract

Using a climatic calibration based on the scores of the plant functional types (PFTs) calculated for 1245 surface pollen spectra, the climate at 6 ka BP has been reconstructed for a set of 116 pollen spectra from the former Soviet Union and Mongolia. The results are presented as maps of climatic anomalies and maps of probability classes showing the significance of these differences from the modern climate. The reconstructed patterns are spatially coherent, but have confidence levels that vary from region to region, due to the often-large error ranges. At 6 ka, the winters were more than 2°C warmer than today north of 50°N, with a high significance east of the Urals. Summers were also more than 2°C warmer than today with a high level of confidence north of the Polar Circle and in central Mongolia. In the mid-latitudes of Siberia, in northern Kazakhstan and around the Black and the Caspian seas, 6 ka summers were significantly cooler than today. The reconstructed moisture availability (ratio of actual to equilibrium evapotranspiration) was more than 10% higher than today in the Ukraine, southern Russia and northern Mongolia, and more than 10% lower than today in central Mongolia. This pattern corresponds partly with that of the water budget (annual precipitation minus evaporation) reconstructed from lake level records.

Introduction

In order to accurately predict future changes in the climate, robust models are required. As the past has experienced major climatic changes, one of the best ways to test and improve the forecasting power of climate models is to compare model-simulated conditions for a particular past time period, with the conditions estimated from geological data. The last 20 ka cover a wide variety of climates against which atmospheric general circulation models (AGCMs) can be tested [1]. Within this time period, the mid-Holocene (6 ka) represents a key time-slice. It has been used by modellers to test whether solar forcing has a direct effect on the Earth's climate system [2]. At that time, ice sheets were reduced to their modern extent, the mid- and high-latitude summer insolation were notably higher than today and the winter insolation lower than today.

Here we present a quantitative climatic reconstruction for northern Eurasia at 6 ka based on pollen data. The aim of this paper is three fold: (1) to test the new `PFT method' [3]on the latest pollen dataset from northern Eurasia; (2) to compare the results with earlier continental-scale climate reconstructions based on different statistical and non-statistical methods 4, 5; and (3) to discuss the reconstructed climate patterns in comparison with some AGCM simulations.

The method of climate reconstruction [3]used in the present study uses the concept of `biomization' [6], which attributes the appropriate biome to any given pollen spectra, by grouping the pollen taxa into plant functional types (PFTs — the broad classes of plants selected by their ecology, leaf morphology and bioclimatic tolerance [7]). The method [3]uses PFT scores instead of the pollen taxa abundances used in the traditional best modern analogue method [8]. The `PFT method' has already been successfully applied to the scarce pollen data of the Last Glacial Maximum from southern Europe [3]and northern Eurasia [9]. These studies demonstrate robust and homogeneous results of climate reconstruction, proving the hypothesis [7]that groups of taxa have a better-defined response to climate than individual taxa, whose responses to climatic change may be obscured by competitive and migrational processes.

Section snippets

Pollen and climate data

A set of 1245 modern surface pollen spectra covering northern Eurasia and western North America has already been used by Tarasov et al. [9]to establish the relationships between the numerical scores of the PFTs and the modern climate at the sampling sites. The present-day climate was calculated at each site by weighted distance interpolation [10]of values taken from the present-day climate database [11].

In the present study we have reconstructed four bioclimatic variables considered as

Results

The biome distribution in northern Eurasia at 6 ka was previously reconstructed by Tarasov et al. [12], using the standard method of biome reconstruction [6]. The improvement here was the ability to define steppe biome as cool steppe and desert as cool desert (Fig. 1a), using a modification of the biomization method [14]. Cool and warm steppes (and deserts) require different climatic conditions and the ability to determine them from pollen data is a valuable contribution to the validation of

Discussion and conclusions

Reconstructed values of 6 ka climate show spatially coherent and generally significant patterns. The large error bars obtained with the applied `PFT method' make the reconstructions partly non-significant (Fig. 3), suggesting that the method is less precise than the best modern analogues method. However, this is not true, as the small error ranges of the analogues method are often underestimated [5].

The results of Tc, Tw, GDD5 and α reconstruction are consistent with the qualitative

Acknowledgements

PT thanks the Délégation aux Relations Internationales et à la Coopération (Ministère de l'Education Nationale, de l'Enseignement Supérieur, et de la Recherche) for the financial support. The present work is a contribution to the international PMIP project and has been funded by EEC contract ENV4-CT95-0075 and by the French PNEDC (projet `Climat glaciaire: extrême et variabilité; approche conjointe modèles-données'). PT compiled the 6 ka pollen data used in this study within the framework of

References (21)

  • O Peyron et al.

    Climatic reconstruction in Europe from pollen data, 18,000 years before present

    Quat. Res.

    (1998)
  • J Guiot

    Methodology of the last climatic cycle reconstruction from pollen data

    Palaeogeogr., Palaeoclimatol., Palaeoecol.

    (1990)
  • S.P Harrison et al.

    Late Quaternary lake-level record from northern Eurasia

    Quat. Res.

    (1996)
  • H.E. Wright, J.E. Kutzbach, T. Webb III, W.F. Ruddiman, F.A. Street-Perrott, P.J. Bartlein (Eds.), Global Climates...
  • S. Joussaume, D. Taylor, Status of the palaeoclimate modelling intercomparison project (PMIP), Proc. First Int. AMIP...
  • B. Frenzel, M. Pecsi, A.A. Velichko, Atlas of Paleoclimates and Paleoenvironments of the Northern Hemisphere, Gustav...
  • R Cheddadi et al.

    The climate of Europe 6000 years ago

    Climate Dyn.

    (1997)
  • I.C Prentice et al.

    Reconstructing biomes from palaeoecological data: a general method and its application to European pollen data at 0 and 6 ka

    Climate Dyn.

    (1996)
  • I.C Prentice et al.

    A global biome model based on plant physiology and dominance, soil properties and climate

    J. Biogeogr.

    (1992)
  • P.E Tarasov et al.

    Last Glacial Maximum climate of the former Soviet Union and Mongolia reconstructed from pollen and plant macrofossil data

    Climate Dyn.

    (1999)
There are more references available in the full text version of this article.

Cited by (76)

  • Assessing the importance of climate variables for the spatial distribution of modern pollen data in China

    2015, Quaternary Research (United States)
    Citation Excerpt :

    However, in general, the regional differences in the importance of climate variables are an underappreciated issue in continental-scale climate reconstructions. In many studies the large-scale modern pollen datasets have been applied to a high number of fossil records for reconstructing the same climate variables over continents or globally (Tarasov et al., 1999; Davis et al., 2003; Viau et al., 2006; Bartlein et al., 2011). It is likely that in such studies the results may be less reliable in regions where the variables of interest are not among the most significant regional variables.

View all citing articles on Scopus
View full text