ABSTRACTS OF TALKS AT MEETINGS

AGE-DEPTH MODELLING: INCORPORATING UNCERTAINTY

Richard J. Telford, Bjerknes Centre for Climate Research

27 February 2003

Absolute chronology is the basis for all comparisons and correlations of Holocene stratigraphical proxy records. It is essential to develop independent absolute chronologies for such stratigraphic sequences before attempting comparisons, correlations, and syntheses.

Correlations and comparisons of stratigraphic data are hypotheses that can be tested using statistical tests, but this cannot be done if the uncertainties surrounding the age-depth model are ignored.

Creating an age-depth model is a multi-stage process

1) Are the dates reliable?
Reservoir effects and contamination can seriously bias dates. Any age-depth model using such data will by unreliable.
2) Make an appropriate calibration curve
The true radiocarbon calibration curve is a smooth function. How should the measured calibration curve be smoothed to reflect this? Additionally, should the curve be smoothed to for samples with long deposition times?
3) Chose an appropriate central point
The intercept method of estimating the central point of a calibrated date is highly sensitive to the mean of the radiocarbon date and the adjustments to the calibration curve. Weighted averages of the probability distribution provide the most robust point estimate.
4) Use Bayesian statistics to reduce uncertainty
The stratigraphic order of dates can be used to constrain their uncertainty
5) Age-depth modelling
Without discarding the uncertainty.
6) Model evaluation and revision

Teleconnections and the nature of global and regional responses

Trond Dokken, Bjerknes Centre for Climate Research

26 March 2003

Global and regional processes are closely linked in the atmospheric system, and together with ocean circulation patterns, influence the earth's climate. Records of palaeoclimate changes can be compared in different parts of the earth and past relationships can be discerned. Of special interest are regions that are in anti-phase, such as the North Atlantic region (including Greenland) and the Cariaco Basin near the coast of Venezuela.

The EU PACLIVA project

Øyvind Lie, Bjerknes Centre for Climate Research

2 April, 2003

Øyvind Lie, together with Atle Nesje and Svein Olaf Dahl, is working on a comparison of Holocene glacier records in Norway and the western French Alps near Grenoble. Proglacial lakes have been selected and cored in the St Sorlin area, and are ready for analysis this spring. The timing of Holocene glacial advances and retreats will be compared with the Norwegian record to see if they were out of phase in the past, in the way that the mass-balance figures are today between the two areas. The likely forcing function is the North Atlantic Oscillation, and the past record may provide information of how the NAO has behaved during the Holocene.

North Norway NORPEC mini-symposium

Jostein Bakke, Anne Bjune, Reidar Løvlie, and Øyvind Paasche

30 April 2003

Jostein Bakke presented an overview of the late Quaternary geology of the Lyngen peninsula, from the deglacial phases to a phase of neoglaciation starting about 3800 BP. Aspvatn contains a sedimentary record of glacial activity, reflected in the LOI sequence. Periods of low LOI in the Early Holocene and around 4000 BP reflect flood events, and LOI becomes stable at low values after the development of a glacier in the catchment.

Reidar Løvlie discussed the magnetic stratigraphy at Aspvatn. The concentration and grain size of magnetic minerals can be estimated by measurements of susceptibility, ARM, and SIRM. Ratios can detect the presence of bacterial magnetite. The magnetic records traced the development of the basin since its isolation from the sea, the phases of minerogenic inwash due to flood events and glacial input, and the formation of bacterial magnetite in the late Holocene.

Anne Bjune presented pollen stratigraphies from a site in Skibotn to the east of Lyngen and Barheivatnet, a lake with a small unglaciated catchment on the Lyngen peninsula. She showed a pollen-based climate reconstruction from the Skibotn site.

Øyvind Paasche discussed the magnetic properties of the Fiskevatn sediments, and speculated that there was 500 years between the Younger Dryas - Holocene transition and the end of permafrost in the catchment. He demonstrated an apparent negative covariance between pollen-based mean July temperatures and the magnetic S ratio. He suggested that the Neoglacial had two phases, an initial phase of dry winters and cold summers followed by a phase of wet winters and cool summers.

Jostein Bakke followed this up by presenting winter precipitation reconstructions based on ELA changes and six different summer temperature records from North Norway based on pollen, chironomids, and tree-lines. Although different in detail, the patterns of changes were similar. A consensus winter precipitation reconstruction suggested an unstable early Holocene with fluctuating precipitation periods, a non-glaciated mid-Holocene, a major expansion of glaciers after about 2300 cal BP, with the maximum extent of the Little Ice Age in the last 200 years.

Report of the ESF HOLIVAR workshop (24-27 April, 2003), on Holocene dating, chronologies, and age modelling

John Birks

21 May 2003

John Birks presented an account of the European Science Foundation HOLIVAR (Holocene Climate Variability) workshop held at Zeist, The Netherlands, 24-27 April, which Einar Heegaard and Richard Telford also attended. After outlining the aims of HOLIVAR and the programme of the HOLIVAR workshop, he summarised the main scientific conclusions from the workshop in terms of:
(1) radiocarbon dating problems of lake sediments and the lake reservoir effect changing with time,
(2) 210Pb and 14C chronologies and age-depth models rarely matching,
(3) the limited value of a single central point estimate of a calibrated 14C age and the importance of presenting the full density probability function of every calibrated age,
(4) developments in age-depth modelling using the full probability density function,
(5) error estimation for speleothems, tree-ring, varve, and ice-core chronologies,
(6) the adoption of the Bayesian paradigm in the calibration of 14C dates,
(7) the value of tephra in constructing late Holocene chronologies,
(8) the importance of careful lithological examination of the sediments prior to sampling and dating,
(9) the value of numerically based wiggle matching for 14C calibration,
(10) the importance of estimating 95% confidence intervals for age-depth models, not only for lake sequences but also for marine profiles, as shown by Oliver Heiri in his comparison of chironomid temperature records and North Atlantic 'Bond events', and
(11) the much greater concern in Holocene research about dating uncertainties and error estimation today than about 5 years ago.