“…SedSimple simulates sedimentary deposition, erosion and transport over geological timescales given a base topography and relative sea level curve, to create a three‐dimensional geological conceptual model simulation of the subsurface. Compared to other, often commercially available process models, such as SLB's GPM (Courtade et al, 2021; Otoo & Hodgetts, 2021), DionisosFlow™ (Al‐Wazzan et al, 2021; Borgomano et al, 2020; Hamon et al, 2021), SedsimX (Snieder et al, 2021), or CarboCAT (Masiero et al, 2021), SedSimple requires less computational resource making it possible to run a large number of simulations, but at the cost of reduced complexity in modelled processes and hence in the produced simulations. We train neural networks to represent the information in a large set of geometries obtained from SedSimple simulations, to produce networks for shallow marine environments that mirror the fluvial networks of Laloy et al (2018).…”