A Method for Recovering Fault Kinematics and Long-Wavelength Surface Uplift from the Inversion of Landscape Features and its Application in the Eastern European Alps

A large body of geomorphological research has shown that topography records the history of both uplift and horizontal motions. Numerical landscape evolution models can be used to test the effects of various kinematic scenarios on landscape development. This enables the possibility of using landscape evolution models to solve an inverse problem and quantify tectonic motions based on observed features of a landscape. Various approaches of inverse landscape evolution modelling have been employed in recent years, with most previous work focusing on purely vertical motions, and/or the inversion of longitudinal stream channels in 1D rather than 2D landscapes. In this study, we introduce an approach using Ensemble Kalman Inversion – an efficient, ensemble-based data inversion method – to recover both vertical and horizontal kinematics from the present-day topography. Our approach is capable of handling large numbers of free parameters and quantifying uncertainty in the result. We use the average elevation and average normalised river steepness index (K_sn) calculated in a moving window along a profile across-strike of the orogen, to which we fit a large number of landscape evolution models. In this way, the models target first-order geomorphological features avoiding second-order variations characteristic in natural settings. Given the high data density and knowledge of the upper crustal structural evolution in the European Alps, we first demonstrate our method using a synthetic model set-up that emulates the structural geometry beneath the Tauern Window along the TRANSALP transect. We specifically include the possibility of long-wavelength surface uplift in our models, which may be derived from various mantle processes or isostatic responses. Our novel inversion approach can recover magnitudes and changes in surface uplift and horizontal advection rates in space and time. Not surprisingly, the ability of the method to determine past rates of deformation decreases the farther back in time they occur. However, initial results suggest that this effect is less pronounced for horizontal advection rates than for surface uplift rates, indicating potential limitations in modelling studies that do not consider horizontal advection. Our novel approach is now being applied to the topography along TRANSALP. We will use our inversion results to assess to what extent orogen-scale geomorphological features are able to record the tectonic and geodynamic evolution of Cenozoic mountain ranges.