Cosmogenic 36c1 and 14C produced in the atmosphere are being used to estimate water residence times in the unsaturated zone at Yucca Mountain. Results thus far show a systematic discordance in that 14C-based ages are generally one to two orders of magnitude younger than 36Cl-based ages. This lack of concordance probably arises from one or more of the following reasons: (1) different transport mechanisms, e.g., vapor transport for I4C; (2) , different magnitudes and timing of bomb-pulse signals; (3) mixing of waters fiom different flow paths; and (4) possibly inadequate methods for correcting for the effect of sample contamination by carbon or chlorine fiom sources other than the infiltrating water. Prelirmnary numerical simulation results usig the FEHMN code suggest that spatial variation in infiltration rates can enhance lateral flow and mixing that leads to discordance in apparent ages depending on the dating technique. Examples are presented to show that disparate radiometric ages are inevitable and to be expected where mixing of waters of markedly different ages occurs.

A chronology of alluvial surfaces on piedmont slopes below the western Ajo Mountains, southern Arizona, has been obtained using cosmogenic36Cl accumulation and AMS radiocarbon dating. The apparent36Cl ages of individual boulders range from 520,000 to 13,000 yr, and the14C ages of organic material in the two young terraces are 2750–2350 and 17,800 cal yr B.P. The sequence of36Cl ages is consistent with the apparent stratigraphic order, but groupings of similar ages for different surfaces appear to result from repeated reworking of older surfaces associated with the deposition of younger ones. The youngest surface gave a distribution of36Cl ages about 30,000 yr older than the14C and soil ages; however, this distribution had36Cl ages that overlapped with36Cl ages from active channels and hillslopes. We attribute the older-than-expected exposure ages of sampled boulders to inheritance of36Cl while residing near the surface during very slow erosion on the mountain front. Our results show that although cosmogenic nuclide accumulation can help establish chronologies for surfaces in piedmont settings, care must be used in evaluating the effects of complex exposure histories.

In this study we compare the stable-isotope composition of late Holocene pedogenic carbonate from a late Holocene soil developed on the pediment below the Ajo Mountains in southern Arizona (USA) with measurements of the δ18O of the soil water throughout the seasonal cycle and with δ18O and δ13C measurements on older, dated soils. The comparison with the soil-water composition shows that the pedogenic carbonate is forming in equilibrium with highly evaporated soil water at the highest summer temperatures. The δ13C of the carbonate apparently reflects equilibrium with CO2 of an isotopic composition consistent with the present mixture of C3 desert shrubs and CAM cactus. Comparison with glacial-period soil carbonates shows that there has been little change in the δ18O, but that the δ13C was about 4.5‰ heavier. Independent evidence indicates that glacial-period summer temperatures were probably much cooler, and that the δ18O of precipitation was also lighter. The increase in the carbonate-water fractionation factor with decreasing temperature probably counteracted the decrease in the δ18O of precipitation, producing little net change in the δ18O of the soil carbonate. We attribute the decrease in the δ13C to replacement of a glacial-period C4 grassland by Holocene C3/CAM desert shrubs and succulents. This replacement was probably in response to an increase in temperature and reduction in summer precipitation at the end of the last glacial period. The δ13C of older soil carbonates indicates that landscape was dominated by grassland for most of the past 700 ka and that the present vegetation likely represents a relatively brief anomaly.

ABSTRACT Chlorine-36, including the natural cosmogenic component and the component produced during atmospheric nuclear testing in the 1950's and 1960's (bomb pulse), is being used as an isotopie tracer for groundwater infiltration studies at Yucca Mountain, a potential nuclear waste repository. Rock samples have been collected systematically in the Exploratory Studies Facility (ESF), and samples were also collected from fractures, faults, and breccia zones. Isotopie ratios indicative of bomb-pulse components in the water ( 36 Cl/Cl values > 1250 × 10 -15 ), signifying less than 40-yr travel times from the surface, have been detected at a few locations within the Topopah Spring Tuff, the candidate host rock for the repository. The specific features associated with the high 36 Cl/Cl values are predominantly cooling joints and syngenetic breccias, but most of the sites are in the general vicinity of faults. The non-bomb pulse samples have 36 Cl/Cl values interpreted to indicate groundwater travel times of at least a few thousand to possibly several hundred thousand years. Preliminary numerical solute-travel experiments using the FEHM (Finite Element Heat and Mass transfer) code demonstrate consistency between these interpreted ages and the observed 36 Cl/Cl values but do not validate the interpretations.

Understanding soil water movement is important for water resource management and for analyzing pollutant transport in the vadose zone. Seasonal variations in shallow soil profiles may have profound effects on deep soil water and solutes. We have investigated seasonal soil water movement in the top meter of undisturbed desert soil in the southern Arizona using 2H, 18O, Cl, and bomb-36C1. Six soil profiles were sampled from a young terrace (Holocene) during 1991 to 1992 on a seasonal basis. One soil profile was sampled from an older terrace (latest Pleistocene) in October 1992, and one from a yet older fan surface (late Pleistocene) in March 1992. The results indicate that repeated seasonal cyclic movement of soil water in the top 60–80 cm active zone tends to produce a consistent stable isotope composition for the soil water below the active zone in different seasons. The deep δ 18O on the oldest surface is about 6‰. heavier than on the younger surfaces, probably resulting from greater evaporative loss due to the larger proportion of fines in the older soil. This deep isotopic composition is closely related to the composition of the local average annual precipitation. An annual average evaporation rate of 35 mm year−1 was estimated using a steady-state diffusion model, whereas the actual regional evapotranspiration rate is 175–250 mm year−1. This suggests that about 80% of the soil water is lost through transpiration and first-stage evaporation before quasi steady-state is reached. The long-term average infiltration rate below the active zone is about 4 mm year−1 for the younger terrace, based on both chloride mass balance and bomb 36Cl, and 0.02 mm year−1 for the older terrace and the fan surface, based on chloride mass balance. The correlation between the deep δ 18O, the reduction of downward flux and the surface age suggests that increasing soil development has resulted in a significant decrease of soil water infiltration and hence an increase in evaporation and surface runoff.

Indurated pedogenic carbonate layers (calcretes) are common in soils on stable surfaces in arid to semiarid climates. The morphology and composition of calcretes provide important information on the geomorphic and climatic histories of the regions where they are formed, but they have proved difficult to date with conventional radiometric methods. We report cosmogenic Cl-36-buildup ages from three fractions (leachable Cl, carbonate, silicate) of a calcrete from the surface of an alluvial slope below the Ajo Mountains in southern Arizona. All three fractions give reasonably concordant ages, ranging from 700 ka at the base of the calcrete horizon to 200 ka at its top. These ages are in good agreement both with estimates of age based on correlation with similar, independently dates, soils in the region and with Cl-36-buildup ages on surficial boulders. These results support the ideas that calcretes accumulate upward with time and that water movement through the carbonate matrix is very limited after induration.

Cosmogenic nuclides produced in situ within minerals at the surface of the Earth are proving to be an effective means of assessing geomorphic histories. The use of multiple cosmogenic nuclides permits both exposure times and erosion rates to be determined. However, if two nuclides are produced only by spallation reactions, the systematic differences in their accumulation rates depend only on the differences in their production rates and half‐lives. The relatively small differences that result require a high degree of analytical precision to yield useful results. In contrast to other spallogenic nuclides, 36Cl is also produced by low‐energy neutron absorption, which creates a different pattern of production as a function of depth. We have measured the thermal flux with depth in a concrete block using 3He‐filled neutron detectors. The measured thermal neutron profile agrees well with predictions from a simple diffusion‐ based thermal neutron distribution model. Calculations of 36Cl production using the model suggest that the use of 36Cl along with a purely spallogenic nuclide to determine erosion rates and exposure times should be less sensitive to analytical error than are determinations from two purely spallogenic nuclides.