Radiocarbon Calibration

1
Radiocarbon Calibration

• By Jonathan Chacon
• ESS 433
• 14November2007

2
Fairbanks et. al. 2005

• Paired radiocarbon and 230Th/ 234U/ 238U ages
  determined on coral samples help extend
  radiocarbon calibration curve (RCCC) from 12k-50k
  yr BP
• Tree-ring dating calibration 0-12k yr BP (Reimer
  et. al. 2004)

3
Reimer et. al. 2004

• Discusses calibration data for Northern
  Hemisphere C14 using tree-ring measurements 0-12k
  BP
• www.radiocarbon.org

4
Fairbanks

• Extended TRC by using coral reef cores from
  Barbados (Atlantic), Kiritimati Atoll (CenPac),
  and Araki Is. (WesPac)
• Accurate calibration essential for measuring time
  and rates of change for numerous scientific fields

5
(No Transcript)
6
(No Transcript)
7
Fairbanks cont.

• (CO2)14 produced by cosmic rays in atmosphere,
  mixes in troposphere and exchanges w/ reactive C
  reservoirs of ocean and biosphere, where it
  decays
• THUS the use of coral core samples!

8
Fairbanks cont.

• Existing RCC data sets infer calendar ages based
  on interpolations and correlations of local
  climate proxies in deep sea cores to chronologies
  of ice core proxies, or assumptions about
  sedimentation rates.

9
Fairbanks Methodology

• Require that each data point has measured
  calendar age (230Th/ 234U/ 238U) and RC age w/
  known errors independent of each other
• Other studies not included b/c they use samples
  w/ 1-5 calcite, which corrupts RC data
• Fairbanks -0.2 calcite samples

10
RC Age Calibration

• All RC ages are readily subject to contamination
  by modern C during sample handling and processing
• 230Th/ 234U/ 238U and 14C ages have independent
  errors measured, computed, and accounted for

11
Accuracy and Precision

• Tested by
• a.) measuring samples that overlap TRC
• b.) making paired 230Th/ 234U/ 238U and
  231Pa/235Uages on select samples
• c.) through 230Th/ 234U/ 238U, 231 Pa/235U, and
  14C measurements in laboratory calibration samples

12
230Th/ 234U/ 238U

• Abundance determined by Thermal Ionization Mass
  Spectrometry (TIMS) opened wide range of dating
  applications
• New half-life estimates for 230Th and 234U
• Improvements made to sample pretreatment quality
  control

13
Advantages to Fairbanks

• Quality of coral samples combined w/ high
  precision 230Th/ 234U/ 238U and 14C measurements
• Goal 3 yrs make 400 new 230Th/ 234U/ 238U and
  14C to fill in gaps of RCC 0-50k BP w/ samples of
  100 yr resolution or better
• Rigorous error estimation!

14
Quality Control

• lt0.2 calcite most important screening
  criteria, uses X-ray Diffraction
• 230Th/ 234U/ 238U and 231Pa/235U ages determined
  by Multi-Collector Magnetic Sector double
  focusing Inductively Coupled Mass Spectrometer
  (MC-MS-ICPMS) w/ age uncertainty of 0.5!

15
AAAAGGGGHHH!!!
16
More Methodology

• No benefit to amalgamating subsets of calibration
  pts. from other data sets b/c dramatic increase
  in scatter
• 10-29k BP all pristine coral samples
  exclusively from marine environments
• Consistency of screening, handling, analysis, and
  precision of measurements justify stand-alone
  CC

17
Fairbanks to Reimer

• Fairbanks CC overlaps w/ existing TRCC and extend
  calibration to 50k BP
• Paired 230Th/ 234U/ 238U and 14C age
  determinations fr. pristine corals provide, high
  precision/accuracy RCCC for use beyond superior
  TRCC

18
Reimer

• Tree Ring Data Sets
• Random Walk Model (RWM) account of uncertainty
  in calendar and 14C age to calculate underlying CC

19
IntCal04

• Group that established criteria for acceptance of
  data into calibration set
• Tree ring measurements used 0-12.4k BP, marine
  data 12.4-26k BP
• www.radiocarbon.org

20
Tree Ring Data Sets (0-12.4k BP)

• Dendrochronological dating and cross-checking
  tree rings is required
• Rigorous testing by internal replication of many
  overlapping sections
• Heidelberg Data Set
• Belfast Data Set
• Waikato Data Set
• Seattle Data Set
• Pretoria and Groningen Data Sets
• Arizona Data Set

21
Regional Offsets

• Trees assumed to have sampled well-mixed
  atmosphere
• -lat-dependent diffs in ocean area, surface water
  ages, and wind speeds cause 14C offsets of 1 (8
  14C yr) for most of the Northern Hemisphere with
  the exception of the high Arctic
• - growing season diffs, altitudinal effects,
  melting permafrost, proximity to areas of intense
  ocean upwelling, and volcanic emissions -gt
  significant in certain localities and time periods

22
Sources of Uncertainty

• Counting statistics -gt standard deviation in 14C
  age difference of replicate samples (mathematical
  process)
• Others may be the result of differences in
  sampling, sample pretreatment, or laboratory
  operation, which are difficult to quantify for
  individual samples

23
(No Transcript)
24
(No Transcript)
25
Data Collection

• For comparison purposes, calculated a weighted
  avg of all updated tree-ring C14 measurements
  w/in a 10 year window
• RWM assumes equal rise/fall in atmospheric 14C
  levels

26
Reimer Conclusions

• -b/c RWM accounts for scatter around true
  calibration curve, IntCal04 curve is smoother
  than the tree-ring-derived portion of IntCal98
  (0-11440k BP)

27
(No Transcript)