Quantitative models of trace element behavior in magmatic processes

It was the deep-seated conviction of Paul Gast that the trace element approach was at least as powerful as experimental petrology to solve the major petrological problems. In this paper, we review the present status of the utilization of trace elements in igneous processes, which acknowledges his opinion. After a study of the historical development of trace element geochemistry which includes a brief discussion of the analytical methods presently available, we review the various models which describe the trace element behavior. A few of them are added, which concern mixing processes, and a set of graphical notations is proposed, which can be extended to any model. Partition coefficients, the methods for their analytical determinations, and their effective dependence on temperature, pressure and composition are then considered. We discuss various methods for the recognition of the process responsible for the variation of trace element concentrations in a suite of rocks. Once this “process identification” is achieved, the approach of inverse problems allows the determination of the characteristic parameters of this process. We detail this approach, exemplified by the fractional crystallization process, as applied to a series of alkali basalts (Terceira Island). We conclude that the solution of such inverse problems and further determinations of partition coefficients based on ion probe data might be the future development of trace element geochemistry.