problem about mindt and time3d

[zhifengmuyun]

I'm puzzled about the influence of the parameter time3d and mindt. I'm modeling the river bends that can form on gentle slopes under the influence of a strike-slip fault. I set the slope to 0.1, the size of the simulation area is 4km*4km and the location of the fault is shown in the figure.

When I simulate with the two sets of parameters mindt=10, time3d=10 and mindt=10, time3d=20 respectively, there is a very noticeable difference.
this is mindt=10, time3d=10

and this is mindt=10, time3d=20

I didn't change any other parameters, so I don't now the reason why there are such obvious differences between these results.