U23B-01
Using Lava Flows to Constrain Cosmogenic Nuclide Production Rates: Lessons from Hawaii
A potentially major systematic uncertainty in cosmogenic dating is in scaling a calibrated production rate to the altitude of another field site. Despite much recent work aimed at improving scaling models, the scaling remains uncertain because measurements of cosmic-ray intensity, which are used as a proxy to describe the elevation dependence of cosmogenic production rates, have not been checked against measurements of cosmogenic nuclides in geologic samples. Well-preserved, subaerially exposed, Quaternary lava flows on the island of Hawaii have several characteristics that make them ideal for such a comparison: (1) flows cover a large range of elevations, with little change in latitude; (2) many flows are old enough for sufficient cosmogenic nuclide accumulation, yet young enough that significant erosion has not occurred; (3) primary flow surfaces are well-preserved, even in flows that are >60 ka, owing to the semi-arid to arid rain shadows of the volcanoes; (4) extensive measurements of cosmic-ray fluxes have been conducted on Mauna Kea and on the nearby island of Maui. We measured $^{36}$Cl in 40 hawaiite samples collected from an altitude transect on the dry, southwest flank of Mauna Kea, Hawaii ($19.8° N, $155.5° W). The elevation dependence of $^{36}$Cl production is best described by an atmospheric attenuation length of 142±5 g cm$^{-2}$, which is close to the value of 140 g cm$^{-2}$ predicted from cosmic-ray measurements. The good agreement between the $^{36}$Cl elevation profile and the cosmic-ray surveys suggests that cosmic-ray measurements can be used to scale production rates. But difficulties in delineating flow boundaries, together with potential systematic errors due to geological complications, such as erosion or burial, mean that interpreting cosmogenic nuclide data from lava flows can be problematic. Our experience points to the importance of accurate stratigraphic studies and mapping if lava flows are to be used as calibration sites. This reasearch was supported by the National Science Foundation under grants EAR-0001191, EAR-0126209 and ATM-0081403 and by Packard Fellowship in Science and Engineering 95-1832.
U23B-02
Calibration of Cosmogenic $^{36}$Cl Production Rates from Holocene Lava Flows in Iceland
We have measured cosmogenic $^{36}$Cl concentrations in whole rock basalts from twenty-one calibration sites on four radiocarbon-dated postglacial lava flows in Iceland. The $^{36}$Cl data were obtained from splits of rock from the same suite of calibration sites used to determine cosmogenic 3He production rates in recent work (J.M. Licciardi et al., submitted manuscript). These $^{36}$Cl measurements allow a direct co-calibration between these two widely used cosmogenic isotopes. Preliminary results indicate that multiple calibration sites on individual lava flows yield $^{36}$Cl production rates that agree within 1-sigma error, demonstrating strong coherency of the data. Mean $^{36}$Cl production rates among the four lava flows are in reasonable agreement, with the exception of the youngest flow (Lambahraun) which yields notably higher production rates. The higher-than-average $^{36}$Cl production rates from Lambahraun correspond with lower-than-average 3He production rates from this same flow, and may partly reflect the influence of snow cover. Production rates via the dominant Ca spallation pathway, which accounts for 60-70% of total $^{36}$Cl production in the Iceland samples, are within the range of values reported in previous $^{36}$Cl calibration studies. Additional $^{36}$Cl measurements from the calibration sites are expected to allow a more detailed assessment of these preliminary findings. The new $^{36}$Cl production rate calibrations will enable accurate $^{36}$Cl surface exposure dating in Iceland.
U23B-03
Geological Context of Cosmogenic-Nuclide Sampling Sites for the CRONUS-Earth Project on the Shoreline of Paleo-Lake Bonneville, Utah
One of the objectives of the NSF-funded CRONUS-Earth Project is to obtain geologically and geochronologically well-constrained samples in which the production rates of a large number of cosmogenic nuclides can be compared. One area sampled for this purpose, in July 2005, was associated with the shoreline of Lake Bonneville, a very large Pleistocene pluvial lake that occupied the current Great Salt Lake basin. The Bonneville highstand represented the all-time high lake level in the basin. The timing of this event is well constrained by 14C dating. The lake approached the Bonneville highstand by ~18.9 ka. The highstand probably lasted ~1.7 ka and ended at 17.4±0.2 ka by downcutting of the lake sill at Red Rock Pass. We sampled two localities associated with the shoreline: Tabernacle Hill and Promontory Point. The Tabernacle Hill basalt flow was erupted during the Bonneville highstand and was partially covered by the lake. We sampled the flow at seven locations approximately 500 m south of the northeast margin of the flow. The location of the Bonneville shoreline, marked by wave erosion and tufa deposition, was verified to be to the northeast of, and below, the sampling sites. Samples were collected from the tops of push-Up ridges using a rock saw. Promontory Point was sampled from a pronounced wave-cut bench on the west side. At the sample site, the depth of wave erosion was ~140 m. The bench was cut into quartzite. Samples were collected by rock saw and hammer and chisel from wave-planed bedrock outcrops. Based on geological mapping, the sample sites were 18 to 27 m below the lake surface elevation (depending on sample location), with an estimated uncertainty of ±5 m. The geometry of the pre-Bonneville topography can be reconstructed with a fair degree of confidence, enabling correction of the measured nuclide concentrations for pre-Bonneville production by muons.
U23B-04 INVITED
The rapidly advancing field of applications of nuclides produced in terrestrial solids in earth and planetary sciences
The field of cosmogenic in situ nuclides in terrestrial solids is about 2 decades old, beginning with the realization that the commonly occurring mineral quartz allowed convenient study of in situ produced nuclides, 10Be and 26Al (Lal and Arnold, 1985; Nishiizumi et al., 1985)), which in turn provide estimates of cosmic ray irradiation history of quartz. Since then the field has continually developed, providing time scales in several disciplines of paleoclimatology, geomorphology, tectonics, meteorology, and volcanology. Realizing the importance of determining accurate time scales in earth sciences, several scientists joined in to implement a project: Cosmic-Ray-Originated Nuclide Systematics on Earth Project (CRONUS), for improving rates of production of cosmogenic nuclides in targets exposed to cosmic radiation under different conditions. These efforts will undoubtedly immensely increase the potentials of the cosmogenic in situ method. In modeling of the cosmogenic nuclide data, one generally faces a dilemma in most cases, namely that one does not have sufficient amount of useful geological data about the samples under study. For instance, one of the principal obstacles constraining exposure histories is the information on its erosion and exfoliation history during exposure. A new approach is feasible to tackle the above mentioned shortcoming utilizing the unique feature of the cosmogenic interaction, namely that the cosmic ray flux at a given point within an object depends on the angular distribution of path lengths of cosmic ray particles arriving at the point from different directions. Interestingly, this distribution continuously changes as the surface of the object evolves with erosion and/or exfoliation. Herein lies the potential of determining both the exposure history and the evolutionary history of the solid object under investigation from multiple determinations of nuclide concentration at different points within a target. We present as examples, exposure histories of boulders, soils and beach terraces, where one can take advantage of the fact that the magnitude of cosmic ray interactions in a solid is very sensitive to the geometry of irradiation, which allows constraining its evolutionary history.
U23B-05
A Monte Carlo Approach to Uncertainty Analysis in Cosmogenic Nuclide Dating
There are a number of sources of uncertainty in cosmogenic nuclide dating, including analytical uncertainties in the measurement of cosmogenic nuclides, erosion rates, topographic shielding, effects of latitude and elevation, the history of the earth's magnetic field, and in reaction cross sections. The mathematical models used in cosmogenic nuclide dating are quite complicated with nonlinearities that make linearized uncertainty analysis difficult. In this presentation, we describe a Monte Carlo approach to uncertainty analysis that incorporates these sources of uncertainty into a framework that can be used to compute ages for samples together with associated uncertainties. Further analysis of the model provides information on which sources of uncertainty make the largest contributions to uncertainty in the ages obtained.
U23B-06
Extracting in situ cosmogenic 14C from olivine: significance for the CRONUS-Earth project
One of the main goals of the Cosmic-Ray-prOduced NUclide Systematics on Earth (CRONUS-Earth) project is to compare production rates of in situ cosmogenic nuclides (CNs) at several well-dated locations in various rock types. Quartz is the most commonly used target mineral for several CNs (e.g., 10Be, $^{26}$Al, 21Ne, 14C), but is generally absent in mafic volcanic terrains, where flows of different ages can constrain temporal variations in CN production at a given location. Because of its short half-life (5.73 ka), in situ cosmogenic 14C (in situ 14C) can be particularly useful for elucidating temporal variations in CN production over much shorter time scales than other CNs. While CNs such as $^{36}$Cl and 21Ne can be measured in both mafic and felsic rocks, clearly it would be advantageous to measure in situ 14C in mafic rocks as well. As such, we have worked to develop reliable protocols to extract in situ 14C from olivine. We conducted numerous stepped combustion experiments testing the efficacy of various chemical pretreatments. We were able to extract a stable and reproducible in situ 14C component from olivine using a LiBO2 flux, following pretreatment with dilute HNO3. However, measured concentrations in olivine (normalized to SiO2 composition) from two known-age basalt flows, the Tabernacle Hill flow (17.3ñ0.4 ka in age) in central Utah and the McCarty's flow (3.0ñ0.2 ka in age) in western New Mexico, were 3 to 5 times lower than predicted in situ 14C concentrations based on measurements in quartz. This discrepancy appears to arise from (1) a synthetic spinel-like mineral formed during our extraction process by the chemical interaction of the Al2O3 sample boat and olivine dissolved within the LiBO2 flux, and (2) undissolved pyroxene phenocrysts (difficult to separate in quantity from olivines). Although we do not fully understand how the formation of the synthetic mineral may affect carbon atoms liberated from olivine, the concentration of in situ 14C atoms that we measured is directly proportional to the Fe-to-total-cation (Fe:TC) ratio of each sample. After applying simple correction factors based on the Fe:TC ratio and the percentage of pyroxene in the sample, measured in situ 14C concentrations were indistinguishable from predicted values at both calibration sites. Because the mineralogical composition (~30% fayalite, 70% forsterite) of the olivines studied here is common in basalt flows elsewhere and the Fe:TC correction factor appears to be predictable, in situ 14C can now be applied to CN research in basaltic terrains, complementing other CN measurements made for CRONUS-Earth.
U23B-07
Boulder, Pavement, Pit; Sample Selection for Cosmogenic Nuclide Dating on Alluvial Fans
Selecting sample targets and methods that minimize the effects of erosion, inheritance, and material movement is one of the largest issues facing cosmogenic isotope dating. Since sample material availability often sharply limits methods, establishing error estimates in less-than-ideal sampling situations is also important. Our study of alluvial fan offsets in the Eastern California Shear Zone (Mojave Desert) takes advantage of multiple sample targets to develop method inter-comparisons. Alluvial fans along the Calico fault had abundant large basalt and quartz monzonite boulders, targets for 3He and 10Be respectively. In addition to the boulders, the incipient desert pavement on the youngest alluvial fan surface was sampled. The basalt boulders had higher relative cosmogenic nuclide concentrations, and a much higher scatter, due to very high inheritance and also likely boulder recycling into younger terraces. The quartz monzonite had lower, consistent concentrations on the youngest terrace, but presented erosional concerns due to the spalling and grussification common to granitic material. The incipient desert pavement had a comparable 10Be concentration to the quartz monzonite boulders; the highest concentration in each was the same, and the average pavement concentration lagged the boulders only slightly. The Lenwood fault alluvial fans had two targets. Small clasts formed weakly developed to mature pavement on the surfaces. The fan deposits were also thick enough to dig pits (1.5m). On the younger fan surface, both sediment and clasts (gravel to cobble) were collected in the pit; concentrations were comparable between the two. The profiles indicated high inheritance (comparable to modern wash samples), and some mixing at the near surface. Due to high inheritance, the surface sample alone would far overestimate the age. Using wash samples to estimate inheritance slightly overcorrects, due to mixing in the near-surface. Two types of clasts were sampled in the older surface, mixed granitoids and hydrothermally altered, cryptocrystalline metavolcanics. Successive leach steps and comparable concentrations to the granitoids indicate that the metavolcanics provide a good sample target. Obvious migration of the clasts, due to ped formation and surface dust inflation, make age interpretations more challenging. Surface samples become a relatively better target, since the contribution from inheritance becomes proportionally less through time.