Testing For Regional Fluid Flow in Subducted Crust, Tinos Island, Cyclades (Greece)

Yale University, New Haven CT

Investigators

Abstract

Ague EAR-0105927 Profound questions regarding the nature and extent of fluid migration in subducted crust have arisen in recent years. Strong evidence for regional fluid flow and large fluid fluxes can be found in some settings (e.g., California), but is notably lacking in others (e.g., Alps, Cyclades). The latter case is particularly thought-provoking because clays, serpentine, and carbonate minerals in the precursor sediments and hydrothermally altered igneous rocks should have released abundant volatiles during prograde high-pressure/low-temperature (HP/LT) metamorphism. In an effort to resolve this apparent paradox, it is proposed to test for the presence of large-scale, regional fluid flow during both prograde subduction and retrograde exhumation and tectonic imbrication in the Cycladic Archipelago (Greece) by integrating high-accuracy digital field mapping, chemical analyses of rocks and minerals, and numerical models of coupled fluid flow and reaction. The specific focus is on the HP/LT rocks of Tinos island, which to date have yielded no clear evidence for regional flow despite prograde metamorphism during the Alpine orogeny to epidote-blueschist/eclogite facies. Futhermore, the potential for regional flow is not limited to the prograde stage, because many of the HP/LT rocks of the Cyclades underwent fluid infiltration and greenschist facies retrogression during exhumation. However, the processes and length scales of flow remain unclear. The field and laboratory work will focus on the following straightforward criteria, applicable to both prograde and retrograde settings, to test for limited or no regional flow. (1) Local or limited devolatilization (prograde case) or volatile uptake (retrograde case); (2) derivation of vein mass from local surroundings as opposed to precipitation from through-going fluids; (3) limited mass transfer of non-volatile elements by flow processes; (4) small estimated fluid fluxes and; (5) absence of regional conduits for volatiles. Regional flow, on the other hand, requires that one or more of these criteria be violated. The numerical modeling will establish what type of flow regimes, whether local or regional in scope, are compatible with observed net transfer reactions, amounts and spatial distribution of reaction progress and element transport, volume changes of reaction, fluid compositions, and distribution of any flow conduits. The major goal will be direct determination of processes of volatile generation and pathways of volatile escape from downgoing slabs. The results will constrain models of the major and trace element composition of subduction zone fluids, critical for our understanding of mantle metasomatism and arc magma chemistry. Furthermore, the general directions of fluid motion, whether subvertical or slab-parallel will be estimated. Slab-parallel flow confined to regional tectonic units, in strong contrast to subvertical flow across layers, may deprive the mantle hanging wall of volatiles and transported non-volatile elements, therefore limiting mantle metasomatism and, ultimately, arc magma generation. Tectonic juxtaposition in active subduction zones allows downgoing rock units undergoing prograde metamorphism to release fluids that retrograde HP/LT terranes undergoing exhumation. What remains to be resolved is if the fluids can migrate regionally, or if the HP/LT packages absorb volatiles locally, preventing large-scale outwards fluid migration to the surface.