Middle to Late Pleistocene alluvial surface ages recorded by their spectral reflectance in Patagonia, Argentina.

Alluvial surfaces like fluvial terraces can act as markers for past changes in sediment and water flux along alluvial rivers. By precisely dating terrace surfaces, we can begin to quantify how sedimentary signals propagate though river channels and better infer the paleo-climatic and tectonic conditions during their formation. However, collecting and processing geochronological samples along larger river systems can be costly and time consuming. Here, we extend alluvial landform age control along the Río Santa Cruz of southern Patagonia by applying a spectral surface characteristic model calibrated from a limited set of field samples and cosmogenic nuclide derived exposure ages. This quantitative method leverages the spectral response of surface weathering to ultimately improve age control of the region's fluvial landforms while reducing the time and cost associated with traditional field dating methods.

Ages of alluvial surfaces may correlate with time-dependent geochemical weathering processes, such as clay mineral formation. Although previous surface-weathering studies have mostly focused on surface ages since the last or penultimate glaciation, we analyzed the change in weathering state for southern Patagonian fluvial terraces up to a million years old consisting of quarzitic and granitic source lithologies. We find that multispectral Landsat 8 data show a 20% increase in the band 6 to band 2 ratio with terrace elevation (and inferred age), highlighting the higher reflectance in the shortwave infrared band often associated with clay mineral formation. It is likely that weathering rates in the dry and cold Patagonian environment are slower compared to regions with less stable and warmer climate conditions or lithological sources, where age dependent weathering signals in multispectral data tend to saturate on much shorter time scales. Our new ¹⁰Be results from surface cobbles and amalgamated pebbles yield exposure ages roughly between 45 and 1000 ka for these surfaces, and the calibrated spectral model allows us to interpolate ages of additional 9 alluvial surface generations based on 11 dated surfaces in the region.

Planned in-situ spectral surface measurements will provide robust ground-truthing to the satellite-based observations and allow for further investigation of the mineral changes driving the age-dependent spectral signal. Furthermore, additional terraces will be dated downstream to provide (1) a better understanding of how the weathering process may differ with downstream distance, and (2) a more reliable correlation of surfaces over long distances (> 100 km), enabling us to reconstruct the details of past climate forcing on alluvial-channel evolution.