Evidence for Eocene aridification of the Atacama Desert’s hyperarid core

Article https://doi.org/10.1038/s41467-026-73422-4 Evidence for Eocene aridification of the Atacama Desert’s hyperarid core Received: 10 June 2025 Accepted: 8 May 2026 1234567890():,; 1234567890():,; Check for updates Benedikt Ritter-Prinz 1 , Steven A. Binnie 1, Finlay M. Stuart2, Derek Fabel Richard Albert 3, Volker Wennrich 1 & Tibor J. Dunai 1 2 , The Atacama Desert is the most arid non-polar region on Earth, yet the timing and drivers of its hyperaridity remain debated. The earliest record of extreme aridification is preserved in the Coastal Cordillera of Northern Chile at the Oligocene-Miocene boundary. However, clast exposure ages on low-relief surfaces and supergene mineralisation ages suggest that low precipitation, and thus limited surface activity and weathering, may have been established earlier. To test the Miocene hyperaridity hypothesis, we have established a record of surface activity based on cosmogenic 21Ne concentrations in 135 locally-derived quartz clasts from low-relief surfaces in the desert’s core. Thirty-two clasts have modelled exposure durations of Oligocene age or older. Their long-term surface preservation suggests exceptionally low landscape evolution rates and implies that aridification initiated earlier than the development of the Humboldt Current and major Andean uplift. We hypothesize that global cooling following the Early Eocene Climatic Optimum was likely a key driver of regional aridification. Transitions to extreme climatic states are linked to interaction between astronomical forcing, global climate, oceanic circulation and regional tectonics1–8. The Atacama Desert in Northern Chile is one of the oldest and driest deserts on Earth (Fig. 1), it is an extreme habitat for life on Earth serving as a commonly-used analogue for the Martian surface9,10, and is characterised by exceptionally low rates of erosion and sedimentation11–13. Easterly-derived moisture is the primary source of precipitation at the western Andean foothills, where precipitation declines from more than 300 mm/yr at 5000 m altitude to less than 20 mm/yr at 2300 m14 (Fig. 1C). Extreme hyperarid conditions prevail below 2300 m, with an mean annual precipitation in the region of less than 2 mm/yr14, especially in the Chilean Coastal Cordillera (Fig. 1C). When and why the region attained hyperaridity is not well established15–20. The paucity of datable paleoclimate archives in the sedimentary record has hampered the development of a complete understanding of the Cenozoic evolution of the region. The sediment record from the Atacama Desert points to arid conditions since 150 Ma21. The exceptionally high concentration of cosmogenic 21Ne in locally-derived quartz clasts from low-relief surfaces in the Coastal Cordillera of northern Chile15,16,22 preserve the earliest record of extreme aridification (Early Miocene) that appears to coincide with the establishment of the Humboldt Current some-time during the late Oligocene-Early Miocene and the uplift of the Andes during the Miocene15–17,22–24. 40Ar/39Ar ages of supergene mineralisation from the arid/hyperarid Andean Precordillera record the decrease of precipitation rates below the threshold (100-120 mm/year) necessary for deep leaching of metals from porphyry copper deposits around the same time24–26. Paleosols from the Precordillera and sedimentological archives from the Andean foothills, however, record a switch to hyperarid conditions during the Late Miocene19,27 that coincides with the development of the high altitude Altiplano and implies that the development of an intensely arid climate was not synchronous across the region16,24. A small number of quartz clasts from widely dispersed very low relief surfaces in the Coastal Cordillera record Late Oligocene exposure ages determined from cosmogenic nuclide concentrations15,16,22. This implies that landscape stagnation occurred significantly earlier than the prevailing models predict, and implies that the transition to 1 Institute of Geology & Mineralogy, University of Cologne, Cologne, Germany. 2Scottish Universities Environmental Research Centre, East Kilbride, UK. Frankfurt Isotope and Element Research Center (FIERCE), Goethe-Universität Frankfurt, Frankfurt, Germany. e-mail: 3 Nature Communications | (2026)17:4520 1 Article https://doi.org/10.1038/s41467-026-73422-4 Fig. 1 | Overview of the study area. Satellite imagery (Earthstar Geographics SIO, ©2023 MAXAR) of the study area in northern Chile. (A) Overview of the study area located in the Coastal Cordillera of the northern Atacama Desert south of Arica. The white line marks the drainage catchment ( ~ 34 km²) of the sampled depositional surface. The sampled surface is bounded to the north by the deeply incised Quebrada de Tiliviche and protected from influence from the east by an uplifted fault scarp. The grey circles are sampling sites published by Dunai, et al.15. The black dots indicate the new sampling sites in this study. The Quebrada de Jazpampa incised into the surface and caused Site A to be isolated from the depositional system. PI17-004, to the south of the main study area, is located in the higherelevated area of the Coastal Cordillera and may contain a clast population that has not been affected by erosion, such as samples PI06-1, 2 and 4. The orange dots mark outcropping tephra deposits studied by Mortimer, et al.43 and Hoke, et al.77 and sample sites TIL22-02, 03. (B) Close-up of the sampling locations. Water and entrained sediment move along the stippled arrows (based on satellite imagery and verified by field observations). Yellow stippled lines indicate the transport of vein quartz clasts from potential source areas. The white dashed lines indicate the basal contact (bedrock) of the Azapa Formation sediments15. The Quebrada Tiliviche and Quebrada de Jazpampa dissected these sediments following regional uplift. (C) Map of South America indicating the study site in red. Topographic west-east profile through the Atacama Desert created using ArcPro – WorldElevation3D/ Terrain3D data. Indicated is the division of the Atacama Desert into its geographic units. Blue lines mark extrapolated recent mean annual precipitation modified from ref. 14. hyperaridity may have been earlier. Occasionally pre–Miocene 40 Ar/39Ar ages have been recorded for secondary minerals from supergene copper deposits from the Precordillera24,25,28–31. These hints that tectonic events, such as the uplift of high-altitude Andes and growth of the Altiplano, may have intensified regional aridification24, but they may not have been responsible for the initiation of the intense aridification of the hyperarid core of the Atacama Desert. The relict low-relief surfaces of the Coastal Cordillera in Northern Chile are ideally suited for studying the long-term climate history of the Atacama Desert15–17,22,32. Here we use cosmogenic 21Ne concentrations to determine the exposure duration of 122 locally-derived quartz clasts from a suite (...truncated)