Geomorphic process rates in the central Atacama Desert, Chile: Insights from cosmogenic nuclides and implications for the onset of hyperaridity

Abstract

Water plays a critical role in erosion and sediment transport and this relationship is most evident in the hyperarid Atacama Desert of Northern Chile, a region characterized by erosion rates that fall to near zero and cobbles and boulders with cosmogenic nuclide concentrations indicative of exposure for many millions of years. Cosmogenic nuclide concentrations from the Atacama are used to both determine the age of hyperaridity onset and to place important constraints on rates of geomorphic processes in this uniquely arid environment. Prior determinations of cosmogenic nuclide concentrations from the Atacama Desert focus primarily on exposure ages from boulders/cobbles or erosion rates from bedrock. However, recent determinations of cosmogenic nuclide concentrations from boulders, bedrock, and sediment at the same location suggests that material from diverse sample types have different cosmogenic nuclide concentrations. Therefore, it is critical to determine which concentrations of cosmogenic nuclides are most representative of overall erosion rates from the Atacama.

Here, concentrations of cosmogenic nuclides in more than 100 samples across two east-west transects within the central Atacama Desert (22-26°S) of Northern Chile are considered. Concentrations of ¹⁰Be and ²⁶Al were measured in samples of bedrock, alluvial sediment, active stream sediment, and boulders within several sub-regions of the Atacama Desert, including: the western and eastern Coastal Cordillera, the inner absolute desert (including the Cerro de los Tetas), the Cordillera Domeyko, and the western flank of the Andes. This data allows detailed comparisons of cosmogenic nuclides concentrations both within diverse sample types at a given site and between major geomorphic sub-regions of the Atacama. The general pattern of cosmogenic nuclide concentrations in hyperarid environments is characterized by concentrations that are higher in boulders, moderately high in bedrock, lower in hillslope sediment, and lowest in channels that flow across the desert. At many locations in the central Atacama, boulders and bedrock have erosion rates as much as an order of magnitude slower than that of finer grained sediment at the same location, a relationship that is attributed to the fact that boulders sit above thick gypcrete soils. The hillslope to basin concentrations of cosmogenic nuclides within each sub-region, along with ²⁶Al/¹⁰Be ratios, suggests that concentrations of ¹⁰Be in most bedrock and sediments reflects erosion rates. However, in the western Coastal Cordillera ¹⁰Be concentrations in sediment also reflects transport time.

The overarching trend in this data set is that inferred erosion rates are lower to both the east and the west, corresponding with increases in both precipitation and erosion rates towards both the Andean Flank and in the Coastal Cordillera. This trend is previously noted for the Atacama; however, this large data set allows the first observation of the influence of other variables upon erosion rates. Most notably, another clear influence on erosion rates in the Atacama is slope. In some cases, differences in slope are enough to overcome the influence of aridity. For example, erosion rates on the flanks of the Cordillera Domeyko are faster (>1 m/Ma) than that at the crest (<0.5 m/Ma), despite the fact that the crest of this mountain range receives more moisture. Moreover, erosion rates in boulders are up to an order of magnitude slower than that of finer grained sediment. Taken together, these patterns suggest that in the extreme environment of the central Atacama erosion rates are sensitive not only to direct changes in precipitation but also to variables such as slope and soil cover. The relationship between decreased erosion and either low slope or subaerial exposure in the Atacama is potentially stronger than similar relationships documented elsewhere. These results are part of a growing body of research suggesting that even in extreme environments erosion is a complex process controlled by a multitude of variables, and where erosion rates are strongly limited by one variable, any other variable that acts to increase erosion may also have a significant effect. These new insights invoke reexamination of the global relationship between erosion and precipitation, and it is suggested that precipitation has an increasingly greater impact on erosion rates at lower values, but with some influence beginning at rates of ∼1000 mm/year.

Understanding process rates in the central Atacama is essential to interpretation of cosmogenic nuclide concentrations as the timing of the onset of hyperaridity. Previous cosmogenic nuclide studies in the Atacama have specifically targeted remnant boulders to determine the age of initial aridification and this new data set shows that the boulders do not have the same cosmogenic nuclide concentration as the sediment on which they rest. Thus, the use of boulder ages or hillslope ages for the purpose of topographic reconstruction can be problematic despite the fact that, broadly, exposure of boulders and cobbles over million-year timescales implies lack of erosion over this same period. Several aspects of these results suggest that the onset of hyperaridity in the Atacama Desert predates the Pliocene. First, a number of ages from boulders predate the Pliocene. Second, new Miocene age constraints from a surface previously identified as Pliocene was shown; the age of this surface previously formed a critical component in arguments for decreased incision rates in the late Pliocene. Third, it is shown that a site that is today very sensitive to sediment fluxes from modern storm events is at least 3 Ma in age. Finally, it is shown that most of the hillslopes in the central Atacama were exposed during the Quaternary and do not have complex exposure history; this indicates that the Atacama is not and probably never has been a frozen or static landscape.