Give every direction a frame

A reference frame — open definition supplies a coordinate convention for describing location and motion. An Earth-fixed frame rotates with Earth. A body frame rotates with the spacecraft. An approximately inertial frame can support a different description of motion. None makes the others meaningless; the task is to transform consistently between them.

“Move right” is incomplete without specifying whose right. The same physical displacement may have different component values in two frames. Units, origin, axis orientation and epoch — open definition belong with the number. A precise vector in the wrong frame can be less useful than a rough vector in the correct one.

In an invented two-dimensional example, reference X points right and reference Y points up. Initially the spacecraft’s body x and y axes align with them. Rotate the spacecraft 90 degrees counterclockwise: body x now points along reference +Y, and body y along reference −X.

A displacement of 10 m along reference +X now has body components x = 0 m and y = −10 m. The physical displacement did not reverse or shrink; only its coordinate description changed. Drawing the named arrows is often safer than trying to infer signs from a verbal instruction.

A point fixed on Earth can have constant Earth-fixed coordinates while moving in an inertial description. Similarly, a sensor attached to a turning spacecraft stays fixed in body coordinates. Time-dependent transformations connect these descriptions. Treating a rotating frame as inertial without the required corrections can produce wrong motion predictions.

An Earth image can make longitude and surface geography intuitive, but it does not by itself define the full frame used by a navigation — open definition system. The coordinate convention and time must come from the accompanying data. A familiar picture is not a replacement for metadata.

Same displacement, different components

Before combining two data products, confirm their units and frame definitions. In a simple test, transform known basis directions and check that lengths are preserved by a pure rotation. Then transform back and recover the starting vector within the intended numerical precision.

A realistic interface test also checks timestamps and whether the rotation describes body-to-reference or reference-to-body coordinates. These are inverse operations. Writing the direction of the transformation in words can prevent a silent sign or convention error from becoming a pointing command.