Although often spoken of in the same breath, volcanic eruptions and earthquakes are completely different. Volcanic hazards are far more diverse than those of earthquakes, ranging from sluggish lava flows engulfing houses to terrifying encounters between passenger aircraft and ash clouds high above the ground. And, of course, most eruptions are themselves accompanied by swarms of damaging earthquakes.

Seismologists are mostly concerned with trying to predict when an earthquake will happen, which is rather like trying to predict when a steadily inflated balloon will burst. Most large eruptions are preceded by obvious precursor events, which are hard to ignore on a well-monitored volcano. Thus, for volcanologists, it is as important to predict what will happen as when.

Systematic monitoring of volcanoes began in 1847 when Ferdinand II, monarch of the Kingdom of Two Sicilies commissioned the Osservatorio Vesuviano. Over the following 150 years, many more volcano observatories have sprung up around the world, usually in the wake of some major volcanic disaster, such as the destruction of St Pierre in Martinique in 1902, and of Armero in Colombia in 1985. Since 1847, the most useful instruments for volcano monitoring have been seismometers which detect the movements of magma beneath a volcano. A single seismometer will detect the precursor tremors that announce that a volcano is coming to life; a network of seismometers enables volcanologists to track the rising magma in the vent.

Unfortunately, though, many of the most serious eruptions in history have taken place on volcanoes not thought to be "active" and therefore not routinely monitored with even a single seismometer. Few volcanologists had ever heard of El Chich"n in Mexico prior to its lethal eruption in 1982. It will never be possible to install the expensive panoply of instruments on every volcano which just might possibly erupt. Fortunately, developments in technology have provided many new tools for volcano monitoring, enabling volcanologists to supplement conventional ground-based techniques with data from remote sensing satellites, enabling them to monitor any volcano on Earth at the drop of a hat. These, and the tried and tested ground methods are all reviewed in this timely book.

ÁñÁ«ÊÓƵ

Modern infra-red sensors are so incredibly sensitive that they can detect the radiance from a patch of glowing lava less than a square metre in area from an orbital height of 700 km. This provides a powerful means of detecting thermal changes on a volcano. In addition to their normal duties, weather satellites provide a cheap and efficient way of tracking ash clouds around the globe, an increasingly important task now that high-flying aircraft so commonly overfly volcanic regions. Even ground-based techniques are becoming space-oriented. Along with seismic activity, deformation of the ground surface around a volcano has long been recognised as a valuable precursor to eruptions. Hitherto, this was laboriously measured by staff and theodolite levelling. Now, the same task can be done by networks of Ground Positioning System (GPS) receivers, tuned to radio signals from a constellation of satellites. An even more sophisticated technique employs Differential Synthetic Aperture Radar (Dinsar) to measure differences in phase of reflected radar signals so as to measure changes in surface topography between successive overflights. Satellites are also used to relay telemetered data from remote volcanoes to laboratories on the other side of the world where specialists can interpret them.

These, and many other techniques, are summarised in this generally excellent book. As the only available text of its kind, it is bound to prove invaluable to scientists working in the rapidly growing field of geohazards. One failing is that the various contributors have responded very differently to their briefs: some have provided quite basic "how to" chapters, crammed with information useful to non-specialists, whereas others have focused on the theoretical backgrounds to their subjects. Many readers will find that the chapter on forecasting behaviour of lava flows is especially opaque, and a long way from the book's stated aim of showing how monitoring techniques have actually been successfully used on volcanoes.

ÁñÁ«ÊÓƵ

For the average reader, the most useful chapter will be one by Bob Tilling of the United States Geological Survey. Tilling discusses succinctly the practical issues of monitoring volcanoes, and presents some sobering case studies of successes and failures.

The good news is that volcanologists are getting much, much better at forecasting eruptions. The bad news is that with increasing technological complexity and swelling populations around the world, it is becoming ever more important that they should get it right.

Peter Francis is the author of Volcanoes: A Planetary Perspective.