Bayesian palaeoclimate reconstruction

We consider the problem of reconstructing prehistoric climates by using fossil data that have been extracted from lake sediment cores. Such reconstructions promise to provide one of the few ways to validate modern models of climate change. A hierarchical Bayesian modelling approach is presented and its use, inversely, is demonstrated in a relatively small but statistically challenging exercise: the reconstruction of prehistoric climate at Glendalough in Ireland from fossil pollen. This computationally intensive method extends current approaches by explicitly modelling uncertainty and reconstructing entire climate histories. The statistical issues that are raised relate to the use of compositional data (pollen) with covariates (climate) which are available at many modern sites but are missing for the fossil data. The compositional data arise as mixtures and the missing covariates have a temporal structure. Novel aspects of the analysis include a spatial process model for compositional data, local modelling of lattice data, the use, as a prior, of a random walk with long-tailed increments, a two-stage implementation of the Markov chain Monte Carlo approach and a fast approximate procedure for cross-validation in inverse problems. We present some details, contrasting its reconstructions with those which have been generated by a method in use in the palaeoclimatology literature. We suggest that the method provides a basis for resolving important challenging issues in palaeoclimate research. We draw attention to several challenging statistical issues that need to be overcome. Copyright 2006 Royal Statistical Society.

[1]  M. Stuiver,et al.  The GISP2 δ18O Climate Record of the Past 16,500 Years and the Role of the Sun, Ocean, and Volcanoes , 1995, Quaternary Research.

[2]  B Huntley,et al.  July Temperatures in Europe from Pollen Data, 6000 Years Before Present , 1988, Science.

[3]  Patrick J. Bartlein,et al.  VEGETATION AND CLIMATE CHANGE IN EASTERN NORTH AMERICA SINCE THE LAST GLACIAL MAXIMUM , 1991 .

[4]  Sourabh Bhattacharya,et al.  Importance resampling MCMC : a methodology for cross-validation in inverse problems and its applications in model assessment , 2005 .

[5]  Noel A Cressie,et al.  Statistics for Spatial Data. , 1992 .

[6]  An Improved Perturbation Mechanism for Simulated Annealing Simulation , 1998 .

[7]  Hannu Toivonen,et al.  Holocene temperature changes in northern Fennoscandia reconstructed from chironomids using Bayesian modelling , 2002 .

[8]  W. Cramer,et al.  The IIASA database for mean monthly values of temperature , 1991 .

[9]  S. Bhattacharya,et al.  Importance re-sampling {MCMC} for cross-validation in inverse problems , 2007 .

[10]  H. Künsch.  Gaussian Markov random fields , 1979 .

[11]  Patrick J. Bartlein,et al.  Holocene Climatic Change in the Northern Midwest: Pollen-Derived Estimates , 1984, Quaternary Research.

[12]  J. Guiot,et al.  A Method for Palaeoclimatic Reconstruction in Palynology Based on Multivariate Time-Series Analysis , 2007 .

[13]  Håkon Tjelmeland,et al.  Bayesian modelling of spatial compositional data , 2003 .

[14]  Caitlin E. Buck,et al.  Intcal04 Terrestrial Radiocarbon Age Calibration, 0–26 Cal Kyr BP , 2004, Radiocarbon.

[15]  K. Train.  Discrete Choice Methods with Simulation , 2003 .

[16]  Christine Osborne,et al.  Statistical Calibration: A Review , 1991 .

[17]  Sw. Banerjee,et al.  Hierarchical Modeling and Analysis for Spatial Data , 2003 .

[18]  Keith R. Briffa,et al.  The Real Color of Climate Change? , 2004, Science.

[19]  J. Marden.  Analyzing and Modeling Rank Data , 1996 .

[20]  Roger Woodard,et al.  Interpolation of Spatial Data: Some Theory for Kriging , 1999, Technometrics.

[21]  Noel A. C. Cressie,et al.  Statistics for Spatial Data: Cressie/Statistics , 1993 .

[22]  Cohmap Members.  Climatic changes of the last 18,000 years: observations and model simulations. , 1988, Science.

[23]  Andrew R. Millard,et al.  Taking Bayes Beyond Radiocarbon: Bayesian Approaches to Some Other Chronometric Methods , 2004 .

[24]  Robert G. Aykroyd,et al.  A kernel-based Bayesian approach to climatic reconstruction , 1999 .

[25]  J. Guiot,et al.  Methodology of the last climatic cycle reconstruction in France from pollen data , 1990 .

[26]  C. Braak,et al.  Non-linear methods for multivariate statistical calibration and their use in palaeoecology: a comparison of inverse (k-nearest neighbours, partial least squares and weighted averaging partial least squares) and classical approaches , 1995 .

[27]  J. Besag,et al.  On conditional and intrinsic autoregressions , 1995 .

[28]  John Aitchison,et al.  The Statistical Analysis of Compositional Data , 1986 .

[29]  Nalini Ravishanker,et al.  COMPOSITIONAL TIME SERIES ANALYSIS OF MORTALITY PROPORTIONS , 2001 .

[30]  S. Lang,et al.  Bayesian P-Splines , 2004 .

[31]  Murray Gray,et al.  Global Climates since the Last Glacial Maximum , 1994 .

[32]  J. Christen,et al.  Radiocarbon peat chronologies and environmental change , 2005 .

[33]  Sandy P. Harrison,et al.  Monsoon changes for 6000 years ago: Results of 18 simulations from the Paleoclimate Modeling Intercomparison Project (PMIP) , 1999 .

[34]  Jean-Jacques Boreux,et al.  Inverse vegetation modeling by Monte Carlo sampling to reconstruct palaeoclimates under changed precipitation seasonality and CO2 conditions: application to glacial climate in Mediterranean region , 2000 .

[35]  D. Jolly,et al.  Tropical climates at the Last Glacial Maximum: a new synthesis of terrestrial palaeoclimate data. I. Vegetation, lake-levels and geochemistry , 1999 .

[36]  Rachid Cheddadi,et al.  The climate in Western Europe during the last Glacial/Interglacial cycle derived from pollen and insect remains , 1993 .

[37]  Marc B. Parlange,et al.  STATISTICS OF EXTREMES: MODELING ECOLOGICAL DISTURBANCES , 2005 .

[38]  Simon Brewer,et al.  The temperature of Europe during the Holocene reconstructed from pollen data , 2003 .

[39]  Simon P. Wilson,et al.  Parallel algorithms for Markov chain Monte Carlo methods in latent spatial Gaussian models , 2004, Stat. Comput..

[40]  H. J. B. Birks,et al.  Pollen-based reconstructions of late-Quaternary climate in Eurpoe - progress, problems, and pitfalls. , 2005 .

[41]  Sandy P. Harrison,et al.  Reconstruction of Holocene Precipitation Patterns in Europe Using Pollen and Lake-Level Data , 1993, Quaternary Research.

[42]  Andreas Hense,et al.  Transfer Functions for Paleoclimate Reconstructions — Theory and Methods , 2004 .

[43]  Simon Brewer,et al.  A probabilistic approach to the use of pollen indicators for plant attributes and biomes: an application to European vegetation at 0 and 6 ka , 2003 .

[44]  Kanti V. Mardia,et al.  Recent trends in modeling spatio-temporal data , 2005 .

[45]  James D. Annan,et al.  Assimilation of paleo-data in a simple Earth system model , 2002 .

[46]  Brian Huntley,et al.  The use of climate response surfaces to reconstruct palaeoclimate from Quaternary pollen and plant macrofossil data , 1993 .

[47]  Hannu Toivonen,et al.  A Bayesian multinomial Gaussian response model for organism-based environmental reconstruction , 2000 .

[48]  Heikki Mannila,et al.  APPLYING BAYESIAN STATISTICS TO ORGANISM-BASED ENVIRONMENTAL RECONSTRUCTION , 2001 .

[49]  Brian Huntley,et al.  Dissimilarity Mapping Between Fossil and Contemporary Pollen Spectra in Europe for the Past 13,000 Years , 1990, Quaternary Research.

[50]  Ian G. Enting,et al.  Kalman filter analysis of ice core data 1. Method development and testing the statistics , 2002 .

[51]  Colin Prentice,et al.  Records of vegetation in time and space: the principles of pollen analysis , 1988 .

[52]  D J Spiegelhalter,et al.  Approximate cross‐validatory predictive checks in disease mapping models , 2003, Statistics in medicine.

[53]  Scott F. Lamoureux,et al.  Five centuries of interannual sediment yield and rainfall‐induced erosion in the Canadian High Arctic recorded in lacustrine varves , 2000 .

[54]  C E Buck,et al.  Formal Statistical Models for Estimating Radiocarbon Calibration Curves , 2004, Radiocarbon.

[55]  Brian Huntley,et al.  Weichselian palynostratigraphy, palaeovegetation and palaeoenvironment; the record from Lago Grande di Monticchio, southern Italy , 2000 .

[56]  R. Katz.  Techniques for estimating uncertainty in climate change scenarios and impact studies , 2002 .

[57]  L. Tierney.  Markov Chains for Exploring Posterior Distributions , 1994 .

[58]  Carl Wunsch,et al.  Greenland—Antarctic phase relations and millennial time-scale climate fluctuations in the Greenland ice-cores , 2003 .

[59]  Mike Rees,et al.  5. Statistics for Spatial Data , 1993 .

[60]  Kenneth E. Train,et al.  Discrete Choice Methods with Simulation , 2016 .

[61]  Kevin Hayes,et al.  Residuals for the linear model with general covariance structure , 1998 .

[62]  Ian G. Enting,et al.  Kalman filter analysis of ice core data 2. Double deconvolution of CO2 and δ13C measurements , 2002 .

[63]  Anthony O'Hagan,et al.  Outliers and Credence for Location Parameter Inference , 1990 .

[64]  C.J.F. ter Braak,et al.  Effects of atmospheric NH3 on epiphytic lichens in the Netherlands : The pitfalls of biological monitoring , 1998 .

[65]  W. Cramer,et al.  A global biome model based on plant physiology and dominance, soil properties and climate , 1992 .

[66]  P. Guttorp,et al.  Statistical Interpretation of Species Composition , 2001 .