CAN ZIRCONS REALLY RECORD THE THERMAL HISTORY OF MAGMAS?

Magmatic zircons can record both the time (from U-Th or U-Pb systematics) and temperature (based on Ti contents) of their formation. Thus, a relatively common approach is to interpret the crystallization temperatures of zircons formed at different times in an individual magmatic system in terms of the thermal evolution of stored magma. However it is also important to consider how adequately this approach records the thermal history of the magma in which they occur. While zircons record the temperature at which they crystallize, they provide no explicit information about the temperatures experienced since crystallization occurred.

There are a number of factors to be considered. Firstly, the growth of magmatic zircon is itself temperature dependent, and occurs when compositional and temperature conditions promote zircon saturation. As a result, zircons are unlikely to provide a representative sampling of the thermal state of a magma body throughout its residence. Secondly, uncertainties in measured zircon U-Th or U-Pb ages are large relative to the time that it takes for the analyzed volume of zircon to grow. The volume of zircon analyzed in a typical ion probe analysis would form in a few hundred to thousand years at typical growth rates. Thus, a limited number of short-lived zircon growth events can produce a broad apparent range of ages, and can collectively appear to define a trend. Finally, late mixing of zircons derived from different portions of the magma reservoir during eruption may produce zircons with an apparent range of temperatures and ages that are not representative of the overall conditions within the reservoir.

We investigate these issues using probabilistic forward models of zircon crystallization. These illustrate the limitations in the ability of zircon to record thermal histories and suggest that interpretations of thermal evolution from zircon populations are often non-unique. For example, several zircon data sets currently interpreted in terms of progressive thermal evolution can be readily explained by a restricted number of short episodes of zircon growth, interspersed with long periods of limited or no growth. In these cases, the total duration of zircon growth is less than a few percent of the total magma residence, and thus much of the thermal history goes unrecorded.