Volcanic activity affects populations around volcanoes and the productive activity of the neighborhoods. This is why civil authorities and the public need to be aware of the volcanic activity. This is a non-trivial task as all volcanoes have particular behaviors. Therefore, confident forecasts are a prime importance for the human populations and governments of concerned South American countries. Moreover, erroneous forecasts may be prejudicial and may imply the loss of observatories credibility. This project aims to help the dedicated institutes (like OVDAS in Chile) to generate faster and more efficient risk reporting for the government entities in charge of the evacuation plans.

To observe volcanic activities, observatories use different data acquisition techniques. Remote monitoring is the most widely used technique for the observation of seismic activities. They deploy high precision seismometers to measure ground vibrations. Having enough of the right instruments located in strategic places is especially important for detecting volcanic events. However the cost of this solution does not permit to observatories to multiply the number of observation stations.

The objectives of this project are to combine efficient data acquisition techniques based on Wireless Sensor Networks (WSNs) and efficient seismic events identification systems using several source of information provided by the wireless sensor network. This is motivated by the fact that:

• At this time, identification system is only based on seismic signals and there is no correlation analysis with other variables (gases emanation, ground deformation, temperature, etc). Consider other variables could lead to improve the seismic identification system and generate more accurate forecasts.
• WSN are well-adapted for large-scale environmental monitoring. A WSN consists in densely deployed sensor devices (autonomous nodes connected wirelessly) to monitor physical or environmental conditions, collect and fuse information and communicate with the global data processing unit (figure 3). In contrast with previous data acquisition techniques, they can provide multi-sensors measurements at high temporal and spatial resolution which are needed to understand and relate the biological response from environmental observations.