As magma moves toward the surface, it interacts with anything in its path: hydrothermal systems, cooling magma bodies from previous eruptions, and (or) the surrounding “country rock.” Magma also undergoes significant changes in its physical properties as pressure and temperature conditions change along its path. These interactions and changes lead to a range of geophysical and geochemical phenomena. The goal of volcano monitoring is to detect and correctly interpret such phenomena in order to provide early and accurate warnings of impending eruptions. Given the well-documented hazards posed by volcanoes to both ground-based populations (for example, Blong, 1984; Scott, 1989) and aviation (for example, Neal and others, 1997; Miller and Casadevall, 2000), volcano monitoring is critical for public safety and hazard mitigation. Only with adequate monitoring systems in place can volcano observatories provide accurate and timely forecasts and alerts of possible eruptive activity.

At most U.S. volcanoes, observatories traditionally have employed a two-component approach to volcano monitoring: (1) install instrumentation sufficient to detect unrest at volcanic systems likely to erupt in the not-too-distant future; and (2) once unrest is detected, install any instrumentation needed for eruption prediction and monitoring. This reactive approach is problematic, however, for two reasons.

• 1. At many volcanoes, rapid installation of new ground-1. based instruments is difficult or impossible. Factors that complicate rapid response include (a) eruptions that are preceded by short (hours to days) precursory sequences of geophysical and (or) geochemical activity, as occurred at Mount Redoubt (Alaska) in 1989 (24 hours), Anatahan (Mariana Islands) in 2003 (6 hours), and Mount St. Helens (Washington) in 1980 and 2004 (7 and 8 days, respectively); (b) inclement weather conditions, which may prohibit installation of new equipment for days, weeks, or even months, particularly at midlatitude or high-latitude volcanoes; (c) safety factors during unrest, which can limit where new instrumentation can safely be installed (particularly at near-vent sites that can be critical for precursor detection and eruption forecasting); and (d) the remoteness of many U.S. volcanoes (particularly those in the Aleutians and the Marianas Islands), where access is difficult or impossible most of the year. Given these difficulties, it is reasonable to anticipate that ground-based monitoring of eruptions at U.S. volcanoes will likely be performed primarily with instruments installed before unrest begins. 
  

• 2. Given a growing awareness of previously undetected 2. phenomena that may occur before an eruption begins, at present the types and (or) density of instruments in use at most U.S. volcanoes is insufficient to provide reliable early warning of volcanic eruptions. As shown by the gap analysis of Ewert and others (2005), a number of U.S. volcanoes lack even rudimentary monitoring. At those volcanic systems with monitoring instrumentation in place, only a few types of phenomena can be tracked in near-real time, principally changes in seismicity, deformation, and large-scale changes in thermal flux (through satellite-based remote sensing). Furthermore, researchers employing technologically advanced instrumentation at volcanoes around the world starting in the 1990s have shown that subtle and previously undetectable phenomena can precede or accompany eruptions. Detection of such phenomena would greatly improve the ability of U.S. volcano observatories to provide accurate early warnings of impending eruptions, and is a critical capability particularly at the very high-threat volcanoes identified by Ewert and others (2005).

  
  

For these two reasons, change from a reactive to a proactive volcano-monitoring strategy is clearly needed at U.S. volcanoes. Monitoring capabilities need to be expanded at virtually every volcanic center, regardless of its current state of unrest, with particular emphases on real-time data transmission and increasing the diversity, quality, and quantity of instrumentation at U.S. volcano observatories. In this report, we present recommendations for the types and numbers of instruments that should be deployed to monitor U.S. volcanoes. These recommendations are the result of discussions among members of a panel of government- and university-based scientists about the status and future directions of volcano monitoring, with discussions framed by the U.S. Geological Survey (USGS)’s National Volcano Early Warning System (NVEWS) initiative (Ewert and others, 2005).