I don't agree on the covariant character of the velocity field. The velocity field is of the same kind of the displacement field, i.e. it is (naturally) a _contravariant_ vector (the _temporal_ derivative doesn't matter). A natural first example of covariant vector is the "area covector". Usually it is confused (or simply associated) with a vector: in three dimensions it is equal to the cross product of the (vector) sides of the associated parallelogram. The area is a covariant vector: when it acts on a vector (e.g. a velocity) it gives a scalar (e.g. the flux of a physical quantity). In the same way, when the area covector acts on a displacement vector, it gives the volume scalar (the volume of the parallelepiped defined by the surface and the displacement vector)