3D lithological structure in a steady state model drives divide migration

Predicting future relief is a longstanding challenge in the field of geomorphology. Past denudation and incision rates can be reconstructed and modelled from field data such as thermochronometers, cosmogenic nuclides or optically stimulated luminescence, whereas future rates are then, by definition, fully unknown. Predicting future landscape evolution is further complicated by the dynamic nature of drainage networks, as well as the necessity of constraining properties such as erodibility in order to make sensible predictions. One of the few constraints available for future landscape properties is the underground stratigraphy imaged by wells or geophysical methods. The 3D rock structure will eventually be exhumed and can be utilised to constrain the future states of model simulations.

In this contribution, we present a landscape evolution model capable of ingesting 3D lithologic information and adapting to alternative channel networks, and demonstrate it using a study area in the Swiss Jura Mountains. The model calculates local relief using steady state solutions of the stream power incision model, and also quantifies hillslope relief using a very simple critical slope gradient where hillslope angles are set to a critical value on pixels that have a small drainage area. Further, drainage divides are allowed to migrate to minimize sharp breaks in relief across drainage divides.

We calibrate the values of erodibility, K, for each lithological unit by extracting ranges of apparent K value from the present-day landscape based on drainage area and gradient along the drainage network. This is then further refined by i) using a Monte Carlo approach to create combinations of K based on these ranges, and ii) comparing the real and model landscape for each combination with the aim to minimise the difference between the two. We then run selected model simulations of future base level fall and potential drainage reorganisation events, highlighting the effects of i.) spatially variable erodibility and ii.) lateral changes of the main channel axis on divide migration.