Résumé
Determining the P-T conditions at which metasomatism occurs provides insight into the physical conditions at which fluid-rock interaction occurs in the crust. However, application of thermodynamic modeling to metasomatized rocks is not without pitfalls. As with "normal" metamorphic rocks, the main difficulty is to select mineral compositions that were in equilibrium during their crystallization. This essential task is particularly difficult in metasomatized rocks because it is often difficult to distinguish textures produced by changes in P-T conditions from those caused by fluid-rock interactions and associated changes in bulk composition. Furthermore, the selection of minerals in equilibrium in metasomatized rocks is made difficult by the great variability of scale of mass transfer (see Chapters 3 and 4 of this volume), and therefore equilibrium, which varies from micrometer- to hand-sample or larger scale, depending on the amount of fluid involved and the fluid transport mechanisms (e.g. pervasive or focused). Finally, another major limitation that is discussed in detail in Chapter 4 of this volume, is that fluid composition coming in or out of the rock is unknown. Since fluid is a major phase component of the system, neglecting its impact on the phase relations might be problematic for thermobarometry. Despite these pitfalls, we describe in this contribution examples where thermobarometry has been apparently successfully applied. We emphasize that pseudosection thermobarometry is particularly suitable for metasomatized rocks because the effects of mass transfer can be explored through P-T-X phase diagrams. Application of thermodynamic modeling to metasomatized rocks requires (1) detailed mineralogical and textural investigation to select appropriate mineral compositions, (2) essential geochemical analyses to define the relative and absolute mass changes involved during the metasomatic event(s), and (3) forward modeling of the effects of mass transfer on phase relations.