Carbon–oxygen isotope and trace element constraints on how fluids percolate faulted limestones from the San Andreas Fault system: partitioning of fluid sources and pathways

Understanding the way fluids flow in fault zones is of prime importance to develop correct models of earthquake mechanics, especially in the case of the abnormally weak San Andreas Fault (SAF) system. Because fluid flow can leave detectable signatures in rocks, geochemistry is essential to bring light on this topic. The present detailed study combines, for the first time, C–O isotope analyses with a comprehensive trace element data set to examine the geometry of fluid flow within a significant fault system hosted by a carbonate sequence, from a single locality across the Little Pine Fault–SAF system. Such a fault zone contains veins, deformation zones, and their host rocks. Stable isotope geochemistry is used to establish a relative scale of integrated fluid–rock ratios. Carbonate δ¹⁸O varies between 28‰ and 15‰ and δ¹³C between 5‰ and −7‰. From highest to lowest delta values, thus from least to most infiltrated, are the host rocks, the vein fillings, and the deformation zone fillings, respectively. Infiltration increases toward fault core. The fluids are H₂O–CO₂ mixtures. Two fluid sources, one internal and the other external, are found. The external fluid is inferred to come essentially from metamorphism of the Franciscan formation underneath. The internal (local) fluid is provided by a 30% volume reduction of the host limestones resulting from pressure solution and pore size reduction. Most trace elements, including the lanthanides, show enrichment at the 100-m scale in host carbonate rocks as fluid–rock ratios increase toward the fault core. In contrast, the same trace element concentrations are low, relative to host rocks, in veins and deformation zone carbonate fillings, and this difference in concentration increases as fluid–rock ratio increases toward the fault core. We suggest that the fluid trace elements are scavenged by complexation with organic matter in the host rocks. Elemental complexation is especially illustrated by large fractionation of Y–Ho and Nb–Ta geochemical pairs. Complexation associated with external fluid flow has a significant effect on trace element enrichment (up to 700% relative enrichment) while concentration by pressure solution associated with volume decrease of host rocks has a more limited effect (up to 40% relative enrichment). Our observations from the millimeter to the kilometer scale call for the partitioning of fluid sources and pathways, and for a mixed focused–pervasive fluid flow mechanism. The fluid mainly flows within veins and deformation zones and, simultaneously, within at least 10 cm from these channels, part of the fluid flows pervasively in the host rock, which controls the fluid composition. Scavenging of the fluid rare earth elements (REE) by host rocks is responsible for the formation of REE-depleted vein and deformation zone carbonate fillings. Fluid flow is not only restricted to veins or deformation zones as commonly believed. An important part of fluid flow takes place in host rocks near fault zones. Hence, the nature of the lithologies hosting fault zones must be considered in order to take into account the role of fluids in the seismic cycle.     

Seismic-reflection profiles of the central part of the Clarendon–Linden fault system of western New York in relation to regional seismicity

Geological and geophysical research in upstate New York, with few exceptions, has not definitively associated seismicity with specific Proterozoic basement or Paleozoic bedrock structures. The central part of the Clarendon–Linden fault system (CLFS) between Batavia and Dale, NY is one of those exceptions where seismicity has been studied and has been spatially associated with structure. The CLFS is either a complex system of long faults with associated shorter branches and parallel segments, or a region of many short faults aligned north–south from the Lake Ontario shore southward to Allegany County, NY. Interpretation of 38 km of Vibroseis and approximately 56 km of conventional seismic-reflection data along 13 lines suggests that the CLFS is a broad zone of small faults with small displacements in the lower Paleozoic bedrock section that is at least 77 km long and 7–17 km wide and spatially coincident with a north-trending geophysical (combined aeromagnetic and gravity) lineament within the basement. The relative offset across the faults of the system is more than 91 m near Attica, NY. The CLFS is the expression of tectonic crustal adjustments within the Paleozoic rock above the boundary of two basement megablocks of differing petrologic provinces and differing earthquake characteristics that forms the eastern side of the Elzevir–Frontenac boundary zone. Deep seismic-reflection profiles display concave-eastward listric faults that probably merge at depth near the mid-crustal boundary layer. An interpretive vertical section provides the setting for refined definitions of the CLFS, its extensions at depth and its relation to seismicity. Most modern seismicity in western New York and the Niagara Peninsula of Ontario occurs in apparent patterns of randomly dispersed activity. The sole exception is a line of seven epicenters of small earthquakes that trend east from Attica, NY into the Rochester basement megablock. Earthquakes may be triggered at the intersections of north- and east-trending brittle faults within the Niagara basement megablock. Current interpretations of the mechanisms for earthquake generation in western New York and the Niagara Peninsula of Ontario require conservative estimates of seismic hazards that assume that an earthquake the size of the 1929 Attica, NY, event (M_b=5.2) or larger could occur anywhere in the Eastern Great Lakes Basin (EGLB). The broad zone of small-displacement faults that marks the CLFS in the lower Paleozoic sedimentary section and the uppermost basement may not provide the structural environment for generation of earthquakes in western New York. If this interpretation is correct, most seismicity is generated within the Niagara basement megablock beneath or west of the CLFS. Consequently, we may have to look to the deeper tectonic regime of basement megablocks to understand the distribution of modern seismicity in the EGLB.