relating power spectrum to fractal dimension for a surface

Hi
I'm reading up on how to generate fractal surfaces and have become
confused about something I thought I already understood.
I'm generating these surfaces by fft-ing some normally distributed
noise and scaling the coefficents so that the power spectrum has a
certain slope.

For a 2D surface, I thought that power spectum slope (-B) is related to
fractal dimension (D) by: D=(7-B)/2. However, I've come across some
litriture which implies that the B here refers to the power spectral
slope of 1D transsects through the surface, and that the 2D surface
needs a power spectral slope of -(B+1) for these relationships to hold.

Maybe this is just a question of nomenclature and different people
using the same variables for different things?

Shane

.