Topics, Tools and Techniques in Paleoclimate Research

Th/U geochronology
Speaker: Warren Beck,  Physics/AMS
March 10th 2004

Summary provided by Jessica Rowland

Warren Beck gave a lecture regarding the "nuts and bolts" of, or rather, "everything you ever wanted to know" about U-series dating. He presented the ²³⁸U-²³⁰Th and the ²³⁵U-²³¹Pa decay chains, but focused mainly on the ²³⁸U-²³⁰Th decay chain, which has a longer dating interval and can give precise ages up to 600 ka. Warren also explained how carbonate samples are pretreated and spiked, and how the isolated U and Th are measured using either TIMS or ICPMS. He then derived the radioactive decay equation in terms of activity (A), and showed that with no initial daughter present the equation can be further simplified to A₂₃₀/A₂₃₈ = 1-e^-(λ230t). However, ²³⁸U and ²³⁴U are not in secular equilibrium, and this disequilibrium is due to crystal lattice damage caused when ²³⁸U decays to ²³⁴U. This alpha recoil effect causes ²³⁴U to preferentially go into solution under conditions of aqueous weathering. So, A₂₃₄/A₂₃₈> 1 in all aqueous solutions, with then oceans having an activity ratio of 1.15. δ²³⁴U describes the deviation from secular equilibrium, and can even be used as an independent age determinant. More commonly, t is solved iteratively from the modified age equation, which assumes ²³⁴U-²³⁸U disequilibrium and additionally adds a term to correct for initial ²³⁰Th.

Warren went on to explain paleoclimatic applications of the U-series dating method. He began with marine corals, which are a good material to date because they are composed of aragonite that has U/Ca similar to that of seawater, and is low in initial Th. Coral species that originally grew near the sea surface can be sampled and used as a proxy for sea level. During the LGM sea level was 120m lower than today, so corals can either be drilled from the ocean bottom, or sampled from tectonically uplifted islands. Sea level change based on U-series dating of these corals shows much more detail than ¹⁴C sea level curves. Two melt-water pulses are recorded, as well as a slowed melting of icecaps during the Younger Dryas. The hemispheric insolation gradient (both 65^oN summer insolation and 15^oN winter insolation) is seen to be an important factor in influencing sea level change. U-series dated corals can also be used to calibrate the radiocarbon curve, by identifying changes in atmospheric ¹⁴C production. Lastly, climate can be reconstructed in terms of evaporation/precipitation and salinity, when the Sr/Ca and δ¹⁸O derived temperature curves are compared. Decadal oscillations as well as seasonal variations are observed.

Lastly, Warren spoke about detecting diagenetic alteration in corals. δ²³⁴U can be used to test for diagenesis. Because δ²³⁴U_{modern seawater} = 150 permil, and δ²³⁴U values should be stable ±10-20% over 100ka, unaltered corals should have values ~150 permil ±10-20%. Diagenetic alteration can also be observed if samples do not fall on a concordia plot of ²³¹Pa/²³⁵U vs. ²³⁰Th/²³⁴U. A popular method of testing for contamination is to use isochron plots. Samples from the same growth band with differing amounts of detrital or hydrogenous Th contamination (²³⁰Th/²³²Th) are plotted in order to attain the true age and the initial ²³⁰Th/²³²Th value. Initial ²³⁰Th is present in deep sea corals, deep sea sediments, lake sediments, soil carbonates and speleothems. For example, the Bahamaian speleothem that he analyzed was first tested for detrital contamination (none was found), then corrected for hydrogenous Th, and finally this corrected value was compared to the ²³¹Pa age (they agreed).

Warren finished his presentation with numerous examples of paleoclimate applications of U-series dated speleothems. He mentioned how the Hulu Cave speleothem has been used to calibrate the GISP2 chronology. He also spoke about how speleothems record changes in monsoons, sources of moisture, the ratio of C₃:C₄ plant biomass, the carbon cycle, ocean circulation, and possibly even show "annual" banding (controversial).

Back to TTT04 syllabus.