As Fig. 3, but for a ‘variable’ dust-to-gas ratio in the disk; same grain sizes and colour bar. The (lower) ‘variable’ dust scale height results in much lower dust densities in the wind than in the ‘fixed’ model, even for very small grains. The slightly wiggly appearance of the in-disk densities at R ≃ 150 AU is caused by the underlying gas disk; it can be ignored as most dust entrainment occurs for smaller R (in respect to both total mass and radius).