two approaches

Two approachesSediment accumulation rates

Two basic approaches:

-a sediment component that changes at a known rate after deposition (constant, or known, input)Radioactive decay – U series, radiocarbon

- a sediment component with a known, time-dependent input function (constant, or known change, after deposition)excess 230Th , 18O chronostratigraphy

Radiometric datingRadiometric dating

Dating principles – covered in Isotope Geochemistry (Faure) Two “simple” approaches:

Average slopes from age vs. depth plotsAbsolute 14C dates for foraminiferal abundance maxima

Normalization to constant 230Th fluxSediment focusing / winnowing Point-by-point mass accumulation rates

TerminologyTerminology

Radioactive parent daughter + - (electron) or (He nucleus)(or + or emission or e – capture or…)

Isotopes: Same number of protons, differing numbers of neutronschemically similar, different mass (kinetics), different radioactiveproperties

Decay rateRadioactive decay rate proportional to number of atoms present 

N is the number of parent atoms in the sampleλ is the decay constant (units t-1)

λN gives the activity (disintegrations/time)

dNN

dt

Half life

( 0) (0)

( ) (0)

Amount remaining as function of time:

if

with

then

t

tt

dNN

dt

N N

N N e

(Radioactive decay equation)

Half life

Half life

The time for 50% of atoms present to decay.

1 / 2

1 / 2

1 / 2

0.5

ln(0.5)

0.693

tNe

No

t

t

Rule of thumb = radioisotope activity can be measured for about 5 half-lives

isotope (decay const.)

min -1

Half-life

yr238U 2.92 x 10 -16 4.51 x 109

230Th 1.75 x 10 -11 75,200

14C 2.30 x 10 -10 5,730

210Pb 5.92 x 10 -8 22.3

234Th 2.00 x 10 -5 0.066 (24 dy)

values – motivation for atom-counting methods vs decay-counting!

Radioactive parent decaying to stable daughter

ingrowth

decay

230TH cartoonActivity (dpm/g)

0 5 10

Dep

th (

cm)

0

500

1000

234U = supported 230ThTotal 230Th

ln (excess 230Th (dpm/g))

0 1 2 30

500

1000

Cartoon 230Th profile (S = 2 cm/ky)

Regress ln(A) vs. depth (z)

ln(Az) = ln(Ao) – mz

Decay eqn

ln(Az) = ln(Ao) – t

= ln(Ao) – zS)

= ln(Ao) – (Sz

Assumptions for regressions of age vs. depth Accumulation without mixing below the mixed layer The isotope is immobile in the sediment

Constant input activity (reservoir age), or known as a function of time  Recall activity at time = 0 in the decay equation:

( ) ( ) tN t N o e

How well do we know N(o) in the past?

Ku (in Broecker and Peng 1982)

Antarctic sediments – agreement between sed rates over two v. different timescales.

DeMaster et al. 1991

Radiocarbon:

Produced where? How?

Natural variability in production?

Natural variability in atmospheric 14C content?

Human impacts on 14C budgets?

Produced in upper atmosphere,

modulated by solar wind, earth’s magnetic field

Faure

Natural variability in atmospheric 14C content? YES!

Production variations (solar, geomagnetic)

Carbon cycle (partitioning between atmosphere, biosphere, and ocean)

At steady state, global decay = global production

But global C cycle not necessarily at steady state,

and 14C offsets between C reservoirs not constant.

Human impacts on 14C budgets?

Seuss effect (fossil fuel dilution of 14C(atm))

Bomb radiocarbon inputs

14C produced in atmosphere, but most CO2 resides in (and decays in) the ocean

14C-free

Radiocarbon dating of sediments.

Bulk CaCO3, or bulk organic C

standard AMS sample 25 mol C (2.5 mg CaCO3)

Specific phases of known provenance:

Planktic, benthic foraminifera

Specific (biomarker) compounds (5 mol C)

Dating known phases (e.g., foraminifera), at their abundance maxima, improves the reliability of each date.  

No admixture of fossil (14C-free) material. Minimizes age errors caused by particle mixing and faunal abundance variations. But, reduces # of datable intervals.

1414 14 13

14

13 1213

13 12

(2 50) 11000

sample1 1000

0.95 ( )

( / )1 1000

( / )

CC C C

AC

A NBSox

C C samC

C C std

14C data reported as Fraction modern, or age, or Δ14C

t 1/2 ~ 5730 y (half life)

λ= 0.00012097 / yr (decay constant) (about 1% in 83 years)

Activity relative to wood grown in pre-bomb atmosphere

Stable carbon isotopic composition

Account for fractionation, normalize to 13C = -25

Radiocarbon reporting conventions are convoluted!

Peng et al., 1977 bulk carbonate 14C

Regress depth vs. age

14C in the sediment “mixed layer”

3280 m

4675 m

Stuiver et al., 1998

14C of atmosphere, surface ocean, and deep ocean reservoirs in a model.

Ocean mixed layer reservoir age; lower 14C, damped high-frequency variations.

Modern mixed layer reservoir age corrections, R. Reservoir age = 375 y +/- R. Large range; any reason R should stay constant?

Substantial variation, slope not constant; non-unique 14C ages

Stuiver et al., 1998

Tree ring age

Tree ring decadal

14C

Stuiver et al., 1998

Atmospheric radiocarbon from tree rings, corals, and varves. Calendar ages from dendrochronology, coral dates, varve counting.

Production variations and carbon cycle changes through time

Calendar ages from dendrochronology and Barbados coral U-Th

Bard et al. (’90; ’98) – U-Th on Barbados coral to calibrate 14C beyond the tree ring record. Systematic offset from calendar age.

Reservoir-corrected 14C ages

The product of these radiocarbon approaches is an age-depth plot.Regression gives a sedimentation rate; linearity gives an estimate of sed rate variability.

Typically, sedimentation rates do vary. How many line segments do you fit to your data?How confident are you in each resulting rate estimate?

To estimate mass accumulation rates (MARs) Calculate average sedimentation rates between dated intervals, and multiply by dry bulk density and concentration.

But:Average sed rates can’t be multiplied by point-by-point dry bulk density and concentration to yield time series.

The solution – 230Th-normalized accumulation rates

Two approachesSediment accumulation rates

Two basic approaches:

-a sediment component that changes at a known rate after deposition (constant, or known, input)Radioactive decay – U series, radiocarbon

- a sediment component with a known, time-dependent input function (constant, or known change, after deposition)excess 230Th , 18O chronostratigraphy

Flux estimates using excess 230Th in sediments(M. Bacon; R. Francois) Assume:230Th sinking flux = production from 234U parent in the water column = constant fn. of water depth  (uranium is essentially conservative in seawater) Correct sediment 230Th for detrital 230Th using measured 232Th

and detrital 232Th/238U. Correct sediment 230Th for ingrowth from authigenic U (need approximate age model). Use an age model to correct the remaining, “excess” 230Th for decay since the time of deposition.

Two applications: 1. Integrate the xs230Th between known time points (14C, 18O).

Deviations from the predicted (decay-corrected) xs230Th inventory reflect sediment focusing or winnowing.

 2. Sample by sample, normalize concentrations of sediment constituents

(CaCO3, organic C, etc.) to the xs230Th of that sample. Yields flux

estimates that are not influenced by dissolution, dilution.

Point by point normalization: Activity(230) (dpm g-1) = Flux(230) (dpm m-2 y-1)

Bulk flux (g m-2 y-1) So: Bulk flux (g m-2 y-1) = Flux(230) (dpm m-2 y-1)

Activity(230) (dpm g-1)  = Prod(230) (dpm m-3 y-1) x (water depth)

Activity(230) (dpm g-1)

And:Component i flux (g m-2 y-1) = Bulk flux (g m-2 y-1) x (wt % i) 

Simple examples (without focusing changes): If % C org increases in a sample, but Activity(xs230Th) increases

by the same fraction, then no increase in C org burial – just a decrease in some other sediment component.

 If % C org stays constant relative to samples above and below, but

Activity(xs230Th) decreases, then the C org flux (and the bulk flux) both increased in that sample (despite lack of a concentration signal).

 But: To assess changes in focusing, we’re stuck

integrating between (dated) time points.

Chronostratigraphy based on foraminiferal 18O values

Two premises (observed):

The 18O of seawater responds to changes in global ice volume- high-latitude precipitation is strongly depleted in 18O- more ice on continents => higher seawater 18O

Foraminiferal 18O reflects seawater 18O - (but also temperature)

Foraminiferal 18O provides a global stratigraphy

Dating (radiometric, or orbital tuning) provides timescale

Eccentricity, Tilt, and Precession

Calculated orbital variations

Imbrie et al., 1984

Planktonic foraminiferal 18O (ice volume and water temperature)

Bruhnes-Matuyama

geomagnetic reversal in some cores

Imbrie et al., 1984

18O vs. time

Align control points (“wiggle

matching”) to put cores on same

time-scale.

Assumption – the d18O time series

reflect global signal (ice

volume, and SST)

Normalize, stack, smooth

Result:

A reference 18O stratigraphy

Timescale at base of stack set by

radiometric dating (K/Ar on volcanic

rock) of B/M reversal

Shackleton and Opdyke 1973

K/Ar age of Bruhnes/Matuyama reversal ~730 ky

Age based on tuning to (assumed) orbital forcing is older.

Shackleton et al. (1990): “K/Ar-based timescale underestimates true age by 5 – 8 %”

Age estimates for magnetic reversals (Ma)

K/Ar (1) Milankovich (2) 40Ar/39Ar (3,4)

Bruhnes / 0.73 0.78 > 0.746

Matuyama 0.783

< 0.78

Matuyama / 0.91 0.99 0.99

Jaramillo

(1) Berggren et al. (1985) K/Ar(2) Shackleton et al (1990) orbital prediction(3) Tauxe et al. (1992) Ar/Ar Kenya(4) Baksi et al. (1992) Hawaii