The seismic work in FUTUREVOLC relies on the existing permanent networks. With many of the most active volcanoes in Iceland located under ice caps, like Mýrdalsjökull and Vatnajökull (see Fig. 2), extending the networks to close proximity to the volcanoes has not been possible, except on the occational rock outcrop (nunatak) in the ice. Many of these volcanoes are therefore undermonitored. However to enter into the glacies with the monitoring networks will require significant strides to be made, because in addition to technological developments of the instruments themselves, several problems will need to be overcome to allow instrument operation in the harsh glacier environment. This will be the task of one of the FUTUREVOLC partners, Guralp Systems Ltd. who will develop a seismic instrument in the project suitable for deployment in the ice. Close monitoring of the subglacial volcanoes will increase their monitoring level and enable tracking magma movements through migration of microseismicity and through detailed analysis of earthquake source mechanisms. The emphasis will be on real-time processing of detected signals for early warning of volcanic eruption. Accuracy of the earthquake locations is dependent on the ability to properly represent the heterogeneous crustal structure at volcanoes. Therefore by including 3D velocity models in the location procedures, the resolution of the seismicity mapping could be significantly improved.
Seismic tremor is also a common sign of activity and unrest in volcanoes and generally accompanies eruptions. Understanding the physical processes generating seismic tremor and discriminating the different characteristis of each one can improve the monitoring potential of volcanoes and decrease the number of false alarms. Subglacial floods issue from regions in the Vatnajökull ice cap on average every other year and a few floods have come from Katla in Mýrdalsjökull (Fig. 2), making them ideal candidates for the project’s journey into the glaciers to study the sources of tremor. Two seismic arrays will be installed at the glaciers edge to locate and track the sources of tremor. Floods from Grímsvötn, Katla and Eyjafjallajökull volcanoes as well as floods from the Skaftárkatlar ice cauldrons are also known (Fig. 2) and recent volcanic eruptions at Hekla, Eyjafjallajökull and Grímsvötn have been generous sources of volcanic tremor.
Figure. 2 Left: Locations of the volcanoes Eyjafjallajökull, Hekla, and Katla (beneath Mýrdalsjökull) in southern Iceland. Right: Map of the Vatnajökull ice-cap in southeast Iceland. Existing and planned monitoring sites are shown on the maps, including the routing of floodwater from several known subglacial sources.
Volcanic volatiles dissolved in glacial rivers, sourced from the volcanes are presently inadequatley monitored and their potential for advance warning of increased volcanic activity will be examined in the project. The chemical analysis work of glacier waters will focus on Katla volcano in the Mýrdalsjökull glacier, taking advantage of, and adding to, the existing river monitoring networks at Mýrdalsjökull, currently monitoring temperature and electrical conductivity in addition to the river stage. Increased activity and repeated floods from Katla since 2011 make this volcano a good site for testing the new observation- and analysis systems of river water.
The infrasound work makes use of partner UNIFI’s (University of Florence, Italy) current 4-channel infrasound array near Hekla, Eyjafjallajökull and Katla volcanoes, with the addition of three similar arrays installed in other locations under WP7 by partner UNIFI. The borehole strain network is focused near Hekla volcano, with the most recent installation in 2010, 5 km from the volcano. An additional installation, through collaboration with CIW and the BGS (British Gelogical Survey) is planned near Katla in 2012. The strain analysis will therefore be focused on Hekla and possibly Katla volcanoes. Real-time data from the existing tilt meter operated on the caldera rim of Grímsvötn and an additional tilt meter to be installed at Hekla in 2012 will be utilized in processes developed for these two volcanoes. The GPS network monitors volcanoes, like Hekla with high resolution, while the ice covered volcanoes are less well monitored at present. Efforts will go towards installing GPS instruments on rock outcrops in the Vatnajökull glacier to improve resolution there also. Real-time data from an existing tilt meter operated on the caldera rim of Grímsvötn and an additional tilt meter to be installed at Hekla in 2012 will, together with the high-rate GPS data enable near-real time processing and analysis of geodetic data, to be incorporated with other monitoring systems utilized and developed.
Subglacial eruptions and events
Floods caused by geothermal and volcanic activity is the most frequent volcanic hazard in Iceland and large subglacial eruptions can cause catastrophic floods. Over 50% of all eruptions in Iceland occur within glaciers and start off as subglacial. However, determining the onset of subglacial eruptions presents very significant challenges. Onset detection is exclusively dependent on geophysical signals, which, currently, are not fully understood. Past Icelandic eruptions demonstrate that there are strong seismic signals associated with volcano driven subglacial processes, but at present it is problematic to unequivocally distinguish between volcanic signals at the magma/lava interface, and those associated with flowing melt water or boiling. Hence the challenge is to distinguish between magma/lava movement, boiling hydrothermal systems, water flow and moving ice.
Even if volcanoes are not covered by glaciers, bad weather (blizzards, dense clouds) can mean that, visually, weak eruptions may go undetected for a few hours emphasising the need for real time detection of changes in geophysical parameters. Even when such changes are detected the exact location of an eruption and length of eruptive vent/fissure may not be known. Onset of open-vent eruption can be detected by infrasound observations. The presence of infrasound arrays in Iceland, strengthens the possibility of instrumentally detecting eruption onset, when visual observations are prevented.
Determination and evolution of eruption source parameters
Recent eruptions in Iceland (Eyjafjallajökull 2010, Grímsvötn 2011; Figure 1.3) and in South America (Chaiten 2008, Puyehue 2011) demonstrated the large impact that explosive, ash-producing eruptions can have on aviation, even though none of these events can be classified as major. Eyjafjallajökull 2010 caused unprecedented disruption to global air traffic while the ash from Puyehue circumvented the globe disrupting aviation in Australia and New Zealand, after travelling across the South Atlantic and Indian Oceans (Smithsonian, 2011). To make forecasts, sophisticated atmospheric dispersion models such as the NAME model of the UK Met Office are applied. However, the accuracy of the dispersal predictions depends critically on the model input. The most important and critical input is the mass eruption rate (the source term). Determination of this term is highly uncertain, and an estimate is usually obtained from a simple nonlinear empirically-derived power law relating the plume height with eruption rate. Other meteorological factors that may influence the plume are usually not taken into account although theories exist describing the effects of wind on plume height. More accurate methods for determining the mass eruption rate in order to make further improvements in the prediction capability of VAACs possible are a priority.
Figure 3. The eruption plumes of Eyjafjallajökull in 2010 (left) and Grímsvötn 2011. The difference in magnitute is apparent from the much larger dimensions of the Grímsvötn eruption.
It is the aim of FUTUREVOLC to address the issue of improving the estimates of the mass eruption rate in explosive eruptions in a decisive way through a multi-parameter approach. To achieve this, a variety of sensors will be implemented and combine into a unique system that estimates mass flow rate from a volcano in near real time to real time. One of the two largest work packages of the project is dedicated to this problem (workpackage 7; WP7). Emphasis will be on real time sensors (an array of radars, lighning detection systems, infrasound, optical cameras, electrical field sensors, tephra samplers, gas analyser systems) and their pre-eruption calibration, since the eruption source parameters can be highly variable with time, with significant changes occurring at time scales of minutes to hours. It is expected real time estimates of mass eruption rate can be achieved in all cases of significant explosive eruptions, which would account for 80% of all eruptions in Iceland, and near real time estimates of fully subglacial or effusive eruptions. A mobile laboratory will be taken out into the field in the event of an eruption to analyse the chemistry of the erupting magma and characterize grain sizes. The system to be developed will be a major advancement in the science of explosive volcanism requiring a wide range of expertise. Efforts from 15 of the partners of the consortium are required, including three SMEs, the development of new equipment and the merging of the various data into a unique single system. This new system will lead to more accurate input into atmospheric dispersion models, benefitting both local populations and risk assessment for aviation on a regional scale. The work on the mass eruption rate in WP7 is used as key input for the dispersal studies in WP8.
Transport of volcanic emissions
The emissions from volcanoes include gases, aerosol and silicate particles. Airborne aerosol injected into the atmosphere pose hazards to aviation. The ocean productivity may increase due to iron supplied by ash fallout. The increase ocean productivity may lead to a reduction of atmospheric carbon dioxide. Once into the stratosphere the volcanic aerosol impacts atmospheric chemical cycles and the solar and terrestrial radiation budgets, and thus influences the climate.
Winds can transport the ash and gases from eruptions rapidly and in multiple directions depending on the wind speed. Within the jet stream, wind speeds may easily reach 100 ms-1 (360 km hr-1) so that transport over long distances in a few hours is possible. The long-range influence of volcanic clouds requires a global observational perspective that can only be achieved by space-based (satellite) measurements. The Meteosat Second Generation Spin Enhanced Visible and Infrared Imager (MSG-SEVIRI) allows ash to be detected and followed day and night at 15 min. temporal resolution. During the recent Eyjafjallajökull eruption the combination of satellite data and a Langrangian transport model by an inversion scheme allowed the determination of time- and height-resolved volcanic ash emissions. However, an urgent need for measurements of the time-varying vertical source strength has been identified. This may be achieved by near-field measurement of ash and gas concentrations utilizing multi-spectral IR cameras allowing the retrieval of ash particle size, mass and optical depth.
Ground-based microwave radar systems can have a valuable role in volcanic ash cloud monitoring as evidenced by available radar imagery. These systems represent one of the best methods for real-time and areal monitoring of a volcano eruption, in terms of its intensity and dynamics. The possibility of monitoring 24 hours a day, in all weather conditions, at a fairly high spatial resolution (less than few hundreds of meters) and every few minutes after and during the eruption is the major advantage of using ground-based microwave radar systems. They can provide data for determining the ash volume, total mass and height of eruption clouds. There are still several open issues about microwave weather radar capabilities to detect and quantitatively retrieve ash cloud parameters. A major impairment in the exploitation of microwave weather radars for volcanic eruption monitoring is due to the exclusive use of operational weather radars for clouds and precipitation observation. Several unknowns may also condition the accuracy of radar products, most of them related to microphysical variability of ash clouds due to particle size distribution, shape and dielectric composition. These issues will be addressed in the project (WP7) to enhance the use of radars in ash cloud detection and characterization.
Satellite measurements play a key role in providing continuous measurements of ash mass loadings which in turn may be used to constrain dispersion model forecasts and assist aviation planners. Sophisticated ash retrieval algorithms have been developed but these lack vital validation data that can come from detailed ground-based measurements. In FUTUREVOLC the combination of ground-based measurements with satellite data and dispersion model forecasting will constitute the most powerful tool available for providing advanced warnings to aviation and health authorities about volcanic ash and gas transportation. Further improvement of satellite volcanic ash retrieval algorithms can be achieved by combining dispersion modeling with a state-of-the-art radiative transfer model.
Viewing of a volcano by several multi-spectral IR from several directions will allow the collection of 4-D (3 space and time) data. In combination with new data retrieval methodologies this will provide unique data on the time-varying vertical source strength. During the Eyjafjallajökull event and also in the past, atmospheric transport models have shown substantial skill in calculating ash dispersion. Still there are important open questions that need to be addressed in order to forecast ash dispersion, the resulting atmospheric ash concentrations and the associated uncertainties in the most reliable way. This includes improved knowledge about source emissions, and knowledge about the uncertainties in meteorological data. These issues will be addressed by FUTUREVOLC.
From a gas monitoring perspective, Iceland is one of the least explored volcanic realms on the planet, with very few data on high-temperature magmatic gas emissions having been available until the Eyjafjallajökull eruption in 2010. Even during the Eyjafjallajökull eruption, gas data have only sporadically been taken, and only after the eruption onset; therefore, pre-eruptive degassing features, which may help to constrain modes and rates of magma storage and ascent in the upper crust, are virtually still un-characterised. The recent advent and wide diffusion in the volcanological community of new fully automated in-situ and remote instruments for gas monitoring, which FUTUREVOLC is planned to permanently use in Iceland for the very first time, promise to contribute to a decisive progress in monitoring of magmatic gas compositional features and fluxes over the country.
Trans-border communications and networks
Worldwide the focus of trans-border communications has been focused on the issue of ash. National volcano observatories (eg Iceland Met Office) are required (by ICAO) to update the regional VAAC about the progress of any eruption and in particular the height of the ash plume. The height of the plume is used to assess empirically the eruption rate. In this project we develop this much further and intend the London VAAC to receive much more advanced information and data from the FUTUREVOLC community in close to real-time. For example we plan to supply detailed ash plume height assessments from multiple sources with quantified uncertainty (WP7), we aim to provide this information as a time series to enable the London VAAC to modify their model to account for fluctuations. We aim to provide an assessment of grainsize distribution as rapidly as possible as well as an assessment of likely magma eruption rate and gas content of the plume. The FUTUREVOLC team will rapidly integrate data and information across disciplines to provide the best and most appropriate information in a timely manner. It will not just be the VAAC that receives information, we will develop further the needs of trans-border governments who require close to real-time information on ash composition, leachates and gas flux in order to consider any environmental or health impacts further afield than Iceland. The engagement of scientists from across Europe in FUTUREVOLC and the communications within the team that we envisage also ensures that we promote the IAVCEI protocol (IAVCEI 1999) ‘single message’ about the volcanic hazard across Europe to the media and on websites. The check lists and best practice that we devise in FUTUREVOLC will in particular concentrate on enhancing the already strong science-Civil Protection links in Iceland and promoting Best Practice in this area across Europe and potentially elsewhere. For example, checklists and Best Practice guidelines may be applicable for future eruptions in Greece where there is currently no experience of managing a volcanic crisis. We will have links with projects working on reducing risk and increasing resilience to volcanic risk worldwide for sharing knowledge in this rapidly expanding field of international communication and cooperation.
Interferometric analysis of SAR data (InSAR) is one of the key tools for space-based monitoring of volcanoes. Identification of Iceland as a GEO supersite would be make possible the full integration of space observations with the in-situ measurements described above, and the advances in the state-of-the-art monitoring that the FUTUREVOLC plans to deliver