The problem here is that when we examine how violent flood events carve things, they don't tend to curve very much. Certainly not the hairpin turns seen in the Colorado River. I'm not sure if you espouse the view that all sediment was laid down in the flood, but if you do, these hairpins form a serious problem for that hypothesis, as while erosion was clearly the culprit, a single flood event simply could not have done that.
Right, but I'm looking for a
scientific defense of Noah's Ark. Some people claim that this is scientifically possible, and science excludes supernatural causation (for reasons detailed
here).
Of course it is, but how the water moves determines how things erode. Flood channels are typically fairly straight and shallow. They don't leave winding channels such as those seen in the grand canyon. Those are most typically left by a river flowing across an uplifting plateau - as the plateau uplifts, the river dams, backs up, and cuts through the softest surrounding strata, leaving a lazy, meandering path.