The Geysers - Cobb Mountain Magma System, California (Part 1): U-Pb zircon ages of volcanic rocks, conditions of zircon crystallization and magma residence times

Abstract

Combined U-Pb zircon and ⁴⁰Ar/³⁹Ar sanidine data from volcanic rocks within or adjacent to the Geysers geothermal reservoir constrain the timing of episodic eruption events and the pre-eruptive magma history. Zircon U-Pb concordia intercept model ages (corrected for initial ²³⁰Th disequilibrium) decrease as predicted from stratigraphic and regional geological relationships (1σ analytical error): 2.47 ± 0.04 Ma (rhyolite of Pine Mountain), 1.38 ± 0.01 Ma (rhyolite of Alder Creek), 1.33 ± 0.04 Ma (rhyodacite of Cobb Mountain), 1.27 ± 0.03 Ma (dacite of Cobb Valley), and 0.94 ± 0.01 Ma (dacite of Tyler Valley). A significant (∼0.2–0.3 Ma) difference between these ages and sanidine ⁴⁰Ar/³⁹Ar ages measured for the same samples demonstrates that zircon crystallized well before eruption. Zircons U-Pb ages from the underlying main-phase Geysers Plutonic Complex (GPC) are indistinguishable from those of the Cobb Mountain volcanics. While this is in line with compositional evidence that the GPC fed the Cobb Mountain eruptions, the volcanic units conspicuously lack older (∼1.8 Ma) zircons from the shallowest part of the GPC. Discontinuous zircon age populations and compositional relationships in the volcanic and plutonic samples are incompatible with zircon residing in a single long-lived upper crustal magma chamber. Instead we favor a model in which zircons were recycled by remelting of just-solidified rocks during episodic injection of more mafic magmas. This is consistent with thermochronologic evidence that the GPC cooled below 350° C at the time the Cobb Mountain volcanics were erupted.

Introduction

The issue of whether silicic magma chambers can remain viable in the upper levels of the continental crust for long durations is a matter of ongoing debate (e.g., Huppert and Sparks 1988, Halliday et al 1989. Recent ion microprobe study of zircon from Quaternary volcanic rocks (e.g., Reid et al 1997, Brown and Fletcher 1999, Reid and Coath 2000, Bindeman et al 2001, Vazquez and Reid 2002 has highlighted the potential of this technique for assessing the longevity of large silicic magma systems. The sluggish diffusivity of U, Th, and Pb in zircon at magmatic temperatures Watson et al 1997, Cherniak and Watson 2001 coupled with the propensity of zircon to dissolve rapidly in hydrous silicic melts at temperatures above zircon saturation Harrison and Watson 1983, Baker et al 2002 give rise to the expectation that zircon U-Pb ages can record the onset of zircon crystallization in the magma (Reid et al., 1997). In contrast, ⁴⁰Ar/³⁹Ar ages of sanidine are conventionally expected to date the time of eruption (McDougall and Harrison, 1999). Measured discrepancies between zircon crystallization and eruption ages (e.g., Reid et al 1997, Brown and Fletcher 1999, Reid and Coath 2000, Bindeman et al 2001, Vazquez and Reid 2002 may thus have the potential to provide an estimate for zircon residence time in magmas (e.g., Reid et al., 1997) or, alternatively, detect crystal recycling from earlier intrusions (e.g., Bindeman et al., 2001). These contrasting conclusions bear directly on the nature and time scales of magma chamber processes.

The Geysers geothermal area in Northern California (Fig. 1) is an outstanding, economically important example of a long-lived, silicic magma-driven, geothermal system. Volcanic rocks erupted from Cobb Mountain directly overly the eastern flank of a subsurface plutonic body, the Geysers Plutonic Complex (GPC) that underlies the area of highest heat flow within the Clear Lake region (Schriener and Suemnicht, 1981). Despite the evidence for plutonic rocks in the GPC being as old as ∼1.8 Ma (see the Schmitt et al. companion paper in this issue), hydrothermal activity in the Geysers persisted in some form to the present day (e.g., Donnelly-Nolan et al 1981, Dalrymple 1993, Hulen et al 1997, Dalrymple et al 1999. Models explaining the current high heat flow of the Geysers - Clear Lake region (Fig. 1; Walters and Combs, 1992) invoke either a very large, continuously viable, mid to deep crustal magma body Isherwood 1981, McLaughlin 1981 or transient heating produced by episodic near-surface intrusions McLaughlin et al 1983, Williams et al 1993, Kennedy and Truesdell 1996, Stanley et al 1998, Stimac et al 2001.

By studying volcanic rocks from Cobb Mountain and two additional centers positioned at either end of the northwest-southeast trending GPC (Pine Mountain and Tyler Valley; see Fig. 2) we have (1) documented the physical conditions attending zircon growth through the analysis of major and trace elemental compositional data from zircon crystals, melt inclusions, and whole rock samples; (2) measured zircon U-Pb ages throughout the volcanic section that significantly predate newly obtained eruption ages for the same samples; and (3) assessed the zircon crystallization and residence history and the implications for the nature and longevity of the Geysers – Cobb Mountain magma system. While the crystallization interval for the zircons from Cobb Mountain is in excellent agreement with the main (∼1.2–1.4 Ma) intrusive phases within the GPC, the volcanic rocks contain no crystal record of the earliest ∼1.8 Ma intrusion (Schmitt et al., this issue). Our results suggest that both the GPC and spatially related eruptives from the Pine Mountain, Cobb Mountain, and Tyler Valley centers were fed by discrete batches of magma. Comparison with previously determined thermal histories for the GPC (Dalrymple et al., 1999) leads us to conclude that the older zircon U-Pb crystallization ages (relative to ⁴⁰Ar/³⁹Ar eruption ages) in the Cobb Mountain volcanics do not result from protracted magma storage in a single, long-lived (>0.2 Ma) shallow magma chamber. We instead propose that these differences are best explained by remelting portions of the just-emplaced, and rapidly cooled, GPC in response to episodic magma influx.