Coriolis deflection depends on relative velocity

Take a horizontal platform rotating counterclockwise as seen from above, so its angular velocity points along positive z. Let x point right and y point upward on the platform diagram, making a right-handed Cartesian basis. An object moving along positive x relative to the platform has angular velocity crossed into relative velocity along positive y. The minus sign and factor two in the Coriolis expression therefore give an acceleration along negative y. Reversing either the platform rotation or the relative velocity reverses this acceleration. Reversing both preserves it.

At a given instant, the Coriolis direction is perpendicular to the relative velocity and rotation axis. A velocity parallel to the rotation axis has zero Coriolis acceleration, even though the frame rotates. The factor two arises because a moving object's displacement and the rotating basis both contribute when differentiating acceleration; it is not a convention that can be dropped. When resolving components, distinguish angular speed in radians per second from rotation frequency in cycles per second. Replacing angular speed by frequency without the factor of two pi creates a quantitative error even when the direction is correct.

The dot product of relative velocity with its Coriolis acceleration vanishes identically. Consequently the Coriolis term alone does no instantaneous work in the rotating-frame kinetic-energy balance. This does not imply that every object's speed is constant in a rotating frame. Gravity, contact forces, the centrifugal term, or a changing frame can alter relative kinetic energy. Energy accounting must include the full equation and any suitable effective potential. A bent path is not by itself evidence that a sideways coordinate term supplies or removes mechanical energy.

For a short interval, hold the leading relative velocity approximately constant and estimate the Coriolis contribution to transverse displacement using constant acceleration. The approximation requires the accumulated direction change to be small, typically angular speed times duration much less than one. If the launch is near the rotation axis, the centrifugal term starts small, but it develops as the object moves away. Its leading displacement correction can then occur at higher order in this short-time expansion. At longer times, integrate the full rotating equation or transform an inertial trajectory; do not extend the quadratic transverse estimate indefinitely.

For local horizontal motion on Earth, resolve its rotation vector into local vertical and horizontal parts. The vertical component is angular speed times the sine of latitude. Keeping only this component gives the familiar horizontal Coriolis acceleration perpendicular to horizontal velocity. It deflects toward the right of motion in the Northern Hemisphere and left in the Southern Hemisphere. Those statements require the local horizontal approximation; the full three-dimensional effect also contains vertical couplings. At the equator the simplified horizontal deflection vanishes, but the complete Coriolis vector does not vanish for every possible velocity.

Compare the relevant time and length scales before including or omitting Earth's rotation. A ball crossing a classroom in a fraction of a second usually accumulates a tiny rotation-induced deflection compared with launch uncertainty and air resistance. A large-scale flow evolving for many hours can be strongly affected. Observing a curved path alone does not isolate Coriolis acceleration: pressure gradients, contact constraints, or aerodynamic forces may dominate. A controlled platform experiment can reverse rotation while preserving launch conditions, providing a sign-reversal test that helps separate a rotation-dependent effect from a fixed laboratory bias.