Welcome to Iceland, welcome to Hveragerdi and welcome to this Penrose conference Plume IV: Beyond the Plume Hypothesis. The objective of this meeting is to discuss and assess alternative models for areas of anomalous volcanism. These include intraplate areas and areas on plate boundaries. For many years now, there has been something approaching a monologue regarding explanations for such volcanism, and the popular model is one involving hot upwellings, rising because of thermal buoyancy.

A great deal of data has been collected over the last 30 years or so, subsequent to the original suggestion of the thermal model. Many of these data do not fit this model, or do not accord with its predictions. For example, so-called “hotspots” appear not to be fixed relative to one-another. Seismic tomography has not clearly shown the structures hoped for in the lower mantle. Evidence for anomalously high temperatures in the mantle beneath volcanic regions is lacking or sparse.

Two approaches can be taken to the problem of new data not fitting a preferred model. First, the model can be changed to fit the observations. This is not, in itself, unreasonable. For example, it was done in the case of the plate tectonic model when it was discovered that some large regions such as the Basin & Range Province deform in a diffuse manner. The plate tectonic model remained strong, however, because its first-order predictions were confirmed. For example, it has been confirmed with GPS that large areas of the globe move in a coherent manner. Also the predicted senses of motion along plate boundaries such as the San Andreas fault have been confirmed.

The second approach to the problem of conflicting data is to propose and test entirely new models. Such an approach is vital in the case where extensive experimentation persistently fails to confirm the primary, testable predictions of the preferred model. In the case of the thermally buoyant upwelling model, the questions have become, what do “persistent”, “fail”, “primary” and “testable” mean? Have we done enough seismology and temperature investigations to be sure that hot, seismically detectable structures of the predicted size don’t exist? Do we all, in fact, agree that this is what the data show? What, fundamentally, are the primary, testable characteristics that the model must have, and without which it cannot stand?

Alongside increasing scrutiny of the standard model is a growing body of opinion that some, if not many, if not all, areas of anomalous volcanism are explained equally well or better by other models. These are models that do not involve high temperatures, homogenous mantle composition and uniform lithospheric stress. Models are being suggested that are based on the reverse picture – normal temperatures, inhomogenous mantle composition and variable lithospheric stress. Instead of thermal upwellings dynamically penetrating the lithosphere, we have an active lithosphere that ruptures and permits magma to rise. Instead of the magma volume being governed by temperature, it is governed by the melting point of the upwelling mantle. These, and many other ideas have been suggested but have not yet been fully developed nor come under close scrutiny by the wider Earth science community. In other words, many Earth scientists and students are not even aware that they exist.

This is the backdrop to this conference. Our objective this week is to discuss and assess models for anomalous volcanic regions that provide alternatives to the deep, thermal model. These models must make predictions and be testable. Many scientists from diverse disciplines are gathered here with data bearing on virtually every facet of the problem, so we should be able to make substantial progress. In addition, I hope that everyone will meet new colleagues and fruitful collaborations will result. In particular, we have several post-docs and students amongst us, and I particularly encourage them to vigorously engage and get as much as possible out of this meeting.

Gillian R. Foulger
Hveragerdi, Iceland
25th August, 2003