Search for: All records

Abstract Deep exposures of soil profiles on Miocene or Mio-Pliocene alluvial deposits were studied along a 500 km N-S transect in the Atacama Desert. These ancient deposits, with excellent surface preservation, now stand many meters above a broad incised Plio-Pleistocene alluvial terrain. Total geochemical analyses and mass balance calculations allowed the establishment of elemental gains, losses, and redistribution in the soils. From north to south (presently hyperarid to arid), the ancient soils reveal an increase in losses of rock-forming elements (Si, Al, Fe, K, Mg). Additionally, rare earth elements (REE) show losses with increasing southerly latitude and systematic patterns with soil depth. Some REEs appear to be unique chemical tracers of exogenous dust and aerosol additions to the soils. The removal of major elements and REEs is impossible in the present climate (one of salt and dust accumulation), revealing that for a significant period following the deposition of the alluvium, soils were exposed to rainfall, chemical weathering, and mass loss—with a geographical pattern that mirrors the present rainfall gradient in the region. Following the cessation of weathering, the pre-weathered soils have undergone enormous dust and salt accumulations, with the rates and types of salt accumulation consistent with latitude: (1) carbonate in the south and (2) sulfate, chlorides, and nitrates to the north. The quantity, and apparent rates, of salt accumulation have a strong latitudinal trend. Isotopes of sulfate have predictable depth patterns based on isotope fractionation via vertical reaction and transport. The relict hyperarid soils are geochemically similar to buried Miocene soils (ca. 10–9 Ma) in the region, but they differ from older Miocene soils, which formed in more humid conditions. The overall soil record for the Atacama Desert appears to be the product of changes in Pacific Ocean sea surface temperatures over time, and resulting changes in rainfall. The mid-Miocene was relatively humid based on buried soil chemistry and evidence of fluvial activity. The mid to late Miocene cooling (ca. 10–5.5 Ma) appears to have aridified the region based on paleosol soil chemistry. Pliocene to earliest Pleistocene conditions caused weathering of the relict soils examined here, and regional fluvial activity. Since the earliest Pleistocene, the region has largely experienced the accumulation of salts and, except for smaller scale oscillations (glacial-interglacial), has experienced protracted hyperaridity. 

Since the impact ∼50,000 yr ago, surface runoff has entrained and transported sediment from the walls to the floor of Meteor Crater (Arizona, USA). Previous work interpreted this erosion and deposition to be due to predominantly fluvial (i.e., dilute water transport) processes. However, light detection and ranging (LiDAR)−derived topographic data and field observations indicate that debris flows dominated, which were likely generated by runoff that entrained the talus that borders bedrock cliffs high on the crater walls. The low gradient of the crater floor caused debris flows to stop, leaving lobate deposits, while fluvial processes delivered sediment toward the center of the crater. Cosmogenic radionuclide dating of levee deposits suggests that debris-flow activity ceased in the late Pleistocene, synchronous with regional drying. Assuming a rock-to-water ratio of 0.3 at the time of transport by mass flows, it would have taken ∼2 × 106 m3 of water to transport the estimated ∼6.8 × 106 m3 of debris-flow deposits found at the surface of the crater floor. This extensive erosion would require ∼6 m of total runoff over the 0.35 km2 upslope source area of the crater, or ∼18 mm of runoff per debris-flow event. Much more runoff did occur, as evidenced by crater lake deposits, Holocene fluvial activity (which produced little erosion), and contemporary rainfall rates. Rarely on Earth is the total amount of water that creates and runs through a landscape estimated, yet such calculations are commonly done on Mars. Our analysis suggests that erosional and depositional landforms may record only a small fraction of the total runoff.