A chronological correlation technique used to date relatively both volcanic and sedimentary sequences. The basic principles consist of Earth’s episodic magnetic field reversals and the correlation of reversals in geological time (Fig.I.9.). Both volcanic and sedimentary sequences hold magnetic minerals reflecting the direction by primary remnant magnetism (Thermoremanent and depositional remanent magnetization) of the earth’s magnetic field at Time 0.
Correlation of the observed polarity sequence can be matched to the radiometrically calibrated geomagnetic polarity time scale (GPTS) developed from well established ocean floor spreading magnetic anomalies (Berggren et al, 1995; Cande and Kent, 1995). Magnetostratigraphy is termed as a relative dating method, it requires an anchor point to match the exact magnetozone in time. Mainly radiometric ages, bio- stratigraphic horizons or orbital tuning are utilised allow us to pinpoint the exact time and fix the polarity sequence.
Problems with this method can arise from magnetic overprinting of samples acquired through chemical and thermal change, periods of long-term reversals such as those within the Cretaceous that lead to a lack of magnetic information in that period. The magnetostratigraphic record is applied to help correlate problems between different environments especially between terrestrial and deep-sea sequences (Lerbekmo, 1995). Chapter III-V will discuss this method in detail, explaining the data acquisition technique and the process of correlating sections from the Western Interior Basin to well defined astronomically tuned European sections.
Presently, one of the most accurate dating methods for the last 35 Ma, whereby astronomers provide valid orbital solutions in Earth’s orbital parameters (Laskar et al, 1999; 2008). Changes in climate due to Milankovitch cyclicity is recorded in the lithology, thus matching patterns of cyclic variations in climate records with patterns of changes in solar radiation.
It is based on cyclical changes in climate proxy records mirroring variations in insolation, which can applied to a particular outcrop allowing high precision time constraints. Currently, orbital solutions by Berger and Loutre (1992) and Laskar et al. (1999) show a concise stable tuning pattern up to ~100M (Laskar et al, 1999), this method has difficulties in older Paleozoic strata due to limitations in tidal dissipation and chaotic diffusion in the inner solar system for precession and obliquity, however the more stable 404 ka eccentricity cycle which is reliant by gravitational interaction between Earth, Jupiter and Venus are recognised as stable through the Phanerozoic (Berger and Loutre, 1992; Laskar et al, 1999). Currently, many GSSP’s in the Neogene have been accurately dated and recorded using this dating method.
Most astronomical dating work has been compiled on deep marine sections as they show the most stable controlled records, however work is now being compiled on fluvial sections in the Western Interior Basin such as the Hell Creek formation in Wyoming USA (Hilgen et al, 2010) and will be discussed further about the possibilities in Canada. Although accurate up to the Cretaceous period, cyclicity can be found in sediments of older age, (Herbert et al, 1999; Olsen and Kent, 1999) thus being utilised as a floating timescale, pinned down by various radiometric ages. Work is also being compiled to use cyclic controlled lithologies +100 Ma to help correct and create possible astronomical tunings for early Earth. Errors for astronomical tuning are hard to infer and climatic feedback is not an instant occurrence, and is thought to have a lag-time of around a few thousand years, still however much more accurate than radiometric methods (Kuiper et al, 2008; Hilgen et al, 2013).
History of 40Ar/39Ar dating technique
Developed from the K-Ar technique, it started from the discovery of the individual isotopes from the two different elements. It forms one of the earliest dating techniques discovered, improving overtime immensely to the current analytical standards of today. Individually the two elements were discovered at different times with Potassium isolated in 1808 by Humphrey Davy and Argon discovered later in 1895 by Rayleigh and Ramsay. The discovery that 40K decays into 40Ar by Aldrich and Nier, (1948) really initiated this as a plausible decay series for radiometric dating. The first K-Ar radiometric ages were published and recognized in 1950 from analysed Sylvite Oligocene evaporate deposits (Smits and Gentner, 1950) developed from using improvements in decay and atmospheric corrections, static vacuum techniques and improved sensitivity in K% quantity.40Ar/39Ar dating originated from principles first developed by John Reynolds in Berkeley, while working on neutron irradiated meteorite samples, recording 39Ar signal derived from neutron bombardment. Primary 40Ar/39Ar ages were obtained by Merrihue and Turner, (1966) and exponentially grew to become one of the most utilised current radiometric dating methods. The concept grew due to the unpopularity of having to measure precise concentrations of K% and the ability to simultaneously measure Argon isotope ratios, increasing the accuracy and precision of ages, thus overtaking K-Ar as the main radioisotope dating method.
Since 1966, improvements within the 40Ar/39Ar system are evolving to include a wide variety of geological problems such as basinal evolution and thermochronology. Currently, work is focused on the age of the standard used as a fluence monitor, decay constants and improvements in mass spectrometer measurements, while J factor determination is constantly pushing the boundary and accuracy of this technique
Concepts the of 40Ar/39Ar dating technique
Potassium naturally occurs in 3 isotopes, 39K (93.2581%), 40K (0.0117%) and 41K (6.7302%). 40K is radioactive (Neir, 1935), decaying with a half life of 1.248 x 10-9 years to both stable ground state 40Ca and 40Ar constituents (Table.II.1). 40Ar derives from in-situ decay of 40K present in most minerals, being inert it does not easily bind with other atoms in the crystal lattice only incorporating once the lattice becomes a closed structure (i.e. when the temperature becomes lower than the closure temperature) entrained through radioactive decay.
Both 40Ar/39Ar and K-Ar dating techniques assume that relative abundances (Fig.II.2) of the isotopes of K are constant in rocks and minerals, essential as 40K is not measured directly and is inferred from isotopic ratios. Constant ratios have been proven throughout, showing change in decay constants no greater than 0.15% (Garner et al, 1975), only in some extra terrestrial samples altered by cosmic radiation (Kendall et al, 1960; Burnett et al, 1966; Verbeek and Schreiner, 1967). Humayun and Clayton, (1994) isotopically analysed K from rocks using secondary ionization mass spectrometry and found variations of just ± 0.5‰ in 39K/41K. thus considered as constant.
Previous Literature of FCT
Currently the main emphasis of the 40Ar/39Ar community seems to focus on the age of the FCT as its main standard whereby Cebula et al. (1986) initially proposed the use of FCT as a possible flux monitor. Since the primary age developed from Steven et al. (1967) there has been vasts amount of published literature (Fig. II.3, Table II.3) attaining ages from a array of methods such as U/Pb and 40Ar/39Ar alongside astronomically tuned ages. Some ages were calculated by fission track dating, but with such large uncertainties they will not be included in this chapter, additionally ages derived from the Rb/Sr method should be cautionally examined due to the relatively large uncertainty of the 87Rb decay constant.
The bulk of ages were derived from inter-calibration with a primary standard (such as MMHB-1 Hornblende, and GA1550 Biotite) whose age was previously determined by K-Ar dating. Over time, aspects of each primary standard such as K content were re-measured against the original ages reported in such articles as Beckinsale and Gale, (1969) and updated to represent the total error. U/Pb ages derived from Oberli.
(1990), Lanphere and Baasgaard, (2001) and Schmitz and Bowring, (2001) show varying ages of older and younger age estimates of FCT due to a variety of factors. Lanphere and Baasgard, (2001) achieved a younger age but was inferred due to possible lead loss and inheritance. While ages derived from Orberli et al, (2002) and Schmitz and Bowring, (2001), exhibit older ages than any other method partly due to prolonged residence time in the magma chamber.
Taylor Creek Rhyolite Sanidine
Formed in the Mogollon Datil volcanic field, Craton County, (New Mexico, USA) the TCR is formed of ~20 lava flows / domes over an area of 100km2. The rhyolite contains 15-35% rounded eu-hedral and sub-heudral quartz, sanidine plagioclase with traces of biotite and hornblende phenocrysts (Duffield et al, 1990). Chemically it is composed of 77.5 ± 0.3% SiO2 content with little varying composition and near constant feldspar phenocryst species, it is homogeneous in its 40Ar*/39Ark ratio. 40Ar/39Ar ages obtained for products of this volcanic field show a growth period of no more than 100,000 years (Duffield and Dalrymple, 1990) making it a relatively instantaneous exposure, erupting from a single magma chamber.
Apparent ages of TCR (Table.II.4) are determined independently from primary standards derived from K-Ar ages (Inter-calibrated to primary GA1550), additionally, ages are calculated relative to FCTs ages. Although similar in composition and age to FCT, a non-uniform distribution is occurent (Renne et al, 2010).
Recent determination of FCT age
As discussed previously, many studies involving different dating techniques have been devoted to the FCT. The age of 28.02 Ma (Renne et al, 1998) have been used extensively for the last decade, but it was recently challenged by two recent studies proposing a significantly older value. From 40Ar/39Ar analyses performed at Vrije Universiteit Amsterdam (VU) and Berkeley Geochronology center (BGC) of tuffs intercalated within well constrained sedimentary sections using cyclostratigraphy, an age of 28.201 ± 0.023 Ma was proposed by Kuiper et al. (2008). More recently, from an approach based on comparison with 238U/206Pb dating of zircon and newly proposed decay 40K decay constant values, Renne et al. (2010) proposed an even older age of 28.305 ± 0.036 Ma.
These two recent FCT standard ages differ slightly from each-other. Inter-laboratory experiments show both the BCG and the VU laboratories are within 2σ analytical error, therefore demonstrate that the difference is not due to analytical systematics. In addition, both laboratories utilise the same error propagation technique and programs, and similar MAP mass spectrometers. The only noticeable difference between the two laboratories comes down the packing of samples for irradiation and the calculation of the J-factor.
While the VU uses a similar set up as University Paris Sud (see previous section), the BGC laboratory irradiates samples using a cylindrical vessel that places standards around the unknowns, the J value is then calculated as an 3D array, determining the neutron flux change both along the Y and X axis.
Uncertainties in Kuiper et al. (2008) age.
To obtain this FCT age of 28.201 ± 0.023 Ma (Kuiper et al, 2008), it was assumed that the orbital forcing and sedimentary expression develops no lag time and is determined as zero in the calculation. Work is now in place to determine this by monitoring sapropel layers in the Mediterranean basin (Hilgen et al, 2008; Kuiper et al, 2008), with, currently, a proposed value of ± 10 kyr. This approach also requires that the sedimentation rate for the astronomically tuned section is constant for the whole succession and that no hiatus occurred. Astronomical tuning is also dependent on the accuracy of the solution presented (Laskar et al, 2004), and is calculated with variables such as tidal dissipation. Presently, these uncertainties are not accounted for and might significantly increase the FCT age proposed with an uncertainty of only 0.023 Ma (1σ).
Uncertainties in Renne et al. (2010) age.
To utilise the new Renne et al. (2010) age of 28.305 ± 0.036 Ma, one must also include the new decay constants published in the Renne et al (2010) paper. These ne decay constants pushed the 40Ar/39Ar ages for several major events significantly older than either U-Pb or astronomical estimates. For example, newly re-calculated Bishop Tuff (BT) age of 778 ± 4 ka using this FCT age leads to an age of 793 ± 4 ka for the Matuama-Brunhes reverse to normal polarity geomagnetic transition (MBT), when the 15 kyr well-constrained stratigraphic offset between BT and MBT is considered (Sarna-Wojcicki et al, 2000). Such MBT age is significantly older than ages derived from orbital tuning, which lies between 773 and 781 ka (Bassinot et al, 1993; Channell et al, 2004; Lourens et al, 2004). Finally, it should be noticed that the new 40K decay values proposed here have been challenged by Schwarz et al. (2011), who suggest that important issues remain to be clarified before they can be used in 40Ar/39Ar or K-Ar age calculations.
Table of contents :
I.1. Project Aim
I.2. Introduction to the Cretaceous
I.3. Introduction to Geochronology
II.2. Fluence Monitors
II.4. Chosen Monitor
III.2. Western Interior Basin
III.3. Introduction to Foreland Basins
III.5. Sequence Stratigraphy
III.6. Previous Work
III.7. Geology of Canada
III.8. Horseshoe Canyon Formation
IV. The Cretaceous/Paleogene Boundary
IV.2. Geological Setting
IV.3. Samples and Methods
V. The Campanian Maastrichtian Boundary
V.2. Geological Setting
V.3. Samples and Methods
VI.2. Geological Setting
VI.3. Previous Work
VI.4. Sample Localities
VI.5. Analytical Techniques
Appendices A (Chapter II)
Appendices B (Chapter IV)
Appendices C (Chapter V)
Appendices D (Chapter VI)