Fig. 7: Phase diagram for the pinning transition.

In each panel, our chosen order parameter, the average steady-state pinned interface fraction, is plotted against three potential control parameters for the data from our simulations. These are the average active layer thickness (a), the steady state of the standard deviation of the active layer thickness (b), and the steady state coefficient of variation (i.e. standard deviation/mean) of the active layer thickness (c). Simulation data points are coloured according to whether the simulation is in the unpinned (blue), transiently pinned (red) or pinned (blue) phase. The simulations included in our plot are as in Fig. 4, but the transitional simulations (see Fig. 5) are not included. The simulation with μ_max = 0.4 1/h, S_bulk = 0.005 is also not included, as it had not fully reached the steady state (see Supplementary Fig. 8). To gain resolution in the transition region, we also included additional simulations in the transiently pinned phase, with parameter values of μ_max = 0.3 1/h, S_bulk = 0.007 g/L; μ_max = 0.33 1/h, S_bulk = 0.01 g/L, μ_max = 0.45 1/h, S_bulk = 0.01 g/L; μ_max = 0.25 1/h, S_bulk = 0.005 g/L; μ_max = 0.37 1/h, S_bulk = 0.01 g/L. Error bars are calculated using the standard error of the mean for correlated data (see Methods).