EMFT hand book？ can you send it to me? or can you tell me how he do it?

You could measure this by diluting finely ground luminescent material (i.e. without large aggregates) into a transparent binder, which is made in the form of a disk, for instance by using a two-component epoxy-like material, polystyrene,.... This binder should be transparent for the excitation wavelength (no absorption, no fluorescence). Then you can measure in transmission mode the amount of excitation light being absorbed by the luminescent material relative to a phosphor-free disk. Preferably this is done in an integrating sphere, because the phosphor particles can also induce some scattering of the excitation light (without absorption).

By making this phosphor-binder under controlled conditions (such that you know the 'thickness' of the phosphor layer by knowing the  amount of phosphor material, its density, the binder volume and the thickness of the final binder-phosphor disk), you can make estimates on the absorption and thus the penetration length. Preferably you perform this for a few different phosphor concentrations in disks with similar thickness, thus varying the pathlength 'within' the phosphor particles.

Multiple effects are at play, such as scattering of the excitation light, so experiments should be carefully performed. It should be favorable to look for a binder material which has similar refractive index to the phosphor material, in order to reduce scattering.

 Philippe Smet  thanks