Do we still need volcano observatories?

The missions of volcano observatories have continuously and significantly evolved since the official opening of Vesuvius observatory and the official speech of the first Director, Macedonio Melloni, in 1845. In his original view, a volcano observatory plays three major roles in the 1) discovery, caracterization and intepretation of the influence of sismo-volcanic activity on monitored physico-chemical parameters, 2) understanding of the structure and the functioning of our planet and 3) support to civil defense for risk mitigation and crisis management.Societal pressure, national policies and technical evolution determine the relative proportions of the contributions observatory can effectively provide in these 3 fields. We here analyze a set of recent examples issued from the monitoring of volcanoes having a variable rate of activity and contrasting eruptive styles, to explore the multiple roles volcano observatories play and their fast evolution with respect to their original definition.Policies aiming at improving society resiliency to volcanic hazards need trustworthy sources of information.Communication societies produce fast evolving communication media and face an increasing crisis of confidence. In this context, volcano observatories represent the source of reliable, credible, accurate and unbiased information on monitored parameters, able to integrate and validate multiple data sources and to deliver an impartial analysis of alert levels.However, at national levels, multiple sources of trustworthy information on hazard and risks exist and they are currently disseminated between several monitoring centers and academic institutions, whose coordination is critical for the delivery of reliable, credible and unbiased information to authorities and population.We argue that improving coordination and message delivery requires not only the acknowledgement of the fundamental role of scientific debate and the education of population and stackeholders about data uncertainty, but most important, the integration in a unified approach of highly contrasting time scales inherent to hazard (long term) and monitoring (short term) of continuosly evolving natural systems like volcanoes.