many areas. Data are obtained both directly and through association. Direct methods include continuously monitoring the geological, geophysical, and geochemical changes that occur on active volcanoes and geological mapping of ancient volcanoes whose internal structure has been exposed by eruption and/or erosion. Associations can be determined by analyzing the character and sequence of historical and prehistorical eruptive products from various types of volcanoes and then matching these and other data to conceptual models of how volcanoes work (Figure 5.18).
Present understanding of the dynamics of active volcanic systems is still in its infancy—particularly those volcanoes related to convergent plate boundaries, like the Cascade Range, the Andes, and the Aleutian Islands. In the past 40 years research on active volcanoes has expanded from field observation and description to include experimental and theoretical studies. The remaining challenge is the monumental task of assembling all the parts to create a better understanding of how volcanoes work. Two aspects of basic research that would lead to specific reductions of volcanic risk are discovery of the mechanisms that trigger volcanic eruptions and determination of the frequency of large explosive eruptions on a regional and global basis.
Knowing what triggers an eruption could lead to more successful searches for detectable precursors. Several precursors have already been recognized. These include a dramatic increase in earthquake activity beneath a potentially active volcano; a swelling of the ground surface near the volcano's summit (see Figure 5.5); and a continuous ground vibration, called volcanic tremor, which is detectable on sensitive seismometers. These signs indicate the injection of molten rock into the shallow roots of the volcano and definitely point to the potential for eruption.
Sometimes, when researchers monitoring a volcano have alerted the public to an imminent eruption, the shallow intrusion of molten rock will stabilize and cool without reaching the surface. When this occurs, the scientists may be accused of issuing a false alarm. A better term for this scenario would be an aborted eruption. More precise methods are needed to distinguish between shallow intrusions of molten rock that will erupt to the surface and those that will not. In addition, scientists need to learn how to communicate the uncertainties involved in forecasting eruptions to the governing officials and the public at risk near potentially active volcanoes.
Volcanoes in Hawaii, especially Kilauea, have been thoroughly monitored over the past several decades. Research indicates that molten rock rises