First I would like to understand the question. The Watts-Williams equation is the so-called stretched exponential which has an analytical expression in the time domain but does not have an analytical Foureier transfrom to the frequency domain, see e.g. http://journals.aps.org/prb/abstract/10.1103/PhysRevB.44.7306 or n.wikipedia.org/wiki/Stretched_exponential_function. In order to fit the frequency domain spectrum you have to assume a certain beta and tau(zero) value (the limiting moduli can be well assessed from the frequency domain data) than you have to perform the Fourier transform. Novoconrol offers softare for such purposes (http://www.novocontrol.de/html/winfit.htm). Other references:
By the Bruggeman equation you mean the effective medium approach of heterogeneous media? It has nothong to do with the form of relaxation. Please explain your question.
Article Investigation on the transformation of time domain spectrosc...
The parameters Fmax a d Mmax can be determined directly from M'' vs. frequency and the parameter beta by sumulation of theorical equation of M'' vs frequency.