Changes in climate coupled with long-term uplift cause alluvial rivers to aggrade or incise, resulting in the formation of fluvial terrace sequences. Such terrace records provide an opportunity to investigate past interactions between climatic forcings, tectonics, and fluvial systems. Further, they allow us to infer potential channel responses to future climatic change. To accurately decipher such interactions, precise age constraints on fluvial terraces are vital. However, established dating methods such as cosmogenic exposure dating are costly, time intensive, and require direct access to terrace surfaces. Thus, their application is not always possible.

An alternative approach to dating fluvial terraces is based on the temporal evolution of the step (riser) between successive terraces. Downslope sediment transport and elevation change, commonly modelled by hillslope diffusion equations, cause a decrease in gradient with time. If the profile of a riser is obtained in high resolution via GPS or LiDAR elevation measurements, its shape can be related back to the terrace age using inverse modelling schemes.

Here, we apply this theory to a set of fluvial terraces situated along the Río Santa Cruz and Río Shehuen in Patagonia, Argentina. A set of existing ¹⁰Be terrace exposure ages of up to ca. 1 Ma from both rivers is used for age verification and model calibration. GPS elevation transects of terrace risers measured during a field campaign in 2023 are used to establish an age chronology with negligible elevation uncertainties. To explore the feasibility of this method without access to field-quality elevation data, we also test the viability of using riser profiles derived from ca. 12 m resolution TanDEM-X data for dating. Preliminary results from synthetic riser profiles with elevation uncertainties similar to those of TanDEM-X data indicate that such DEM-derived age estimates are theoretically accurate. However, application to the Patagonian river systems and comparison against independent ¹⁰Be ages shows a wide spread in absolute age estimates for single terrace generations. Therefore, in this case, the method appears to be viable only for relative age classification. Post-abandonment riser disturbances and spatially variable sediment transport rates may be key factors hindering our ability to integrate large numbers of DEM-derived profiles into a unified interpretation. Obtaining more robust absolute age estimates may require the use of landscape evolution models that incorporate more complex, non-diffusive processes. Further, the viability of this method may be reduced for very old (i.e., ca. 1 Ma) terraces. Application to younger terrace systems may yield more accurate results.

This technique provides a low-cost, spatially extendable way of dating fluvial terraces and analysing landscape dynamics in fluvial systems. We are currently preparing to release an open-source Python library containing the tools needed to perform these analyses.