Boundary normal coordinates, as their name implies, are defined relative to some boundary such as the magnetopause or the bow shock. They allow the data to be ordered in a way which is related to that boundary. The L and M axes, equivalent to x and y, lie in a plane tangential to the boundary - and the N axis, equivalent to z, is normal to the boundary.
There is no universal convention to assign the positive sense of the N axis. In magnetopause studies, N is usually defined as positive away from the Earth.
Similarly, there is no universal convention to resolve the L and M axes. In work on the magnetopause, L is usually defined as the projection of the GSM Z axis on to the boundary tangent plane (Russell and Elphic, 1978). Thus the unperturbed geomagnetic field in the outer magnetosphere should lie close to the L direction.
The transformation matrices between boundary normal coordinates and other systems such as GSE maybe determined by a number of methods:
Last updated 29 July 1997 by Mike Hapgood (Email: M.Hapgood@rl.ac.uk)