This is a reasonable question. At the scale of a waterspout, the inertial forces of fast-moving air should be large compared to the viscous forces (i.e., very large Reynolds number). Yet the inflow along the surface of the water is laminar, where we would ordinarily expect boundary-layer vorticity (i.e., turbulence). A detailed description of the expected properties of suction vortexes (including the boundary-layer turbulence) can be found here:
Thermodynamic Tornadoes?
An hypothesis that directly addresses the anomalous behavior is here:
Tornadic Inflow
This work includes plenty of references in case you want to research it further. For example, you could read up on two-fluid simulations, which accurately describe the fast-moving, laminar flow along the surface (but beg the question of how two-fluid behaviors are instantiated in a well-mixed fluid such as the air).