improve documentation on volatile dependence

[chrishavlin]

The current behavior is fairly confusing -- it's hard to know when and where VBR.in.SV.Ch2o gets used.

Also, with #92 , we'll be getting separate Ch2o fields for water in melt and solid phases.

So, would be good to add a page to documentation about water (and generally volatiles) describing in detail how it will impact results.

[chrishavlin]

Copying from a slack answer for future reference:

for the pre_melt scaling, water will only have an affect through the solidus that you provide -- no direct effect of water on maxwell time.

For andrade_psp: andrade_psp will not depend on water content.

For eburgers_psp, the default behavior will not depend on water content. But if you set the paramter VBR.in.anelastic.eburgers_psp.useJF10visc=0 and provide a viscous method, VBR.in.viscous.methods_list = {'HK2003';}, then the maxwell time for the anelastic calculation will use the diffusion creep viscosity from HK2003, which will have a dependence on water if you set Ch2o.

For xfit_mxw, this method just directly pulls the diffusion creep viscosity of the viscosity calculation in VBR.in.viscous.methods_list, so if you set HK2003 and a nonzero VBR.in.SV.Ch2o field then you should see some dependence on water.

Ch2o units are ppm.