Final Report Summary - GLOBALSEIS (NEW GOALS AND DIRECTIONS FOR OBSERVATIONAL GLOBAL SEISMOLOGY)
The Earth's mantle has not cooled much since the planet formed. Unlike Mars, which seems to have stopped tectonic evolution, it surface is always in the process of renewing itself, re-creating topography and with that the conditions for life. It has not experienced thermal events that are violent and frequent enough to make its atmosphere too poisonous for life to subsist. How Earth has been able to do that is a matter of lively debate: the abundance of radioactive elements is too low to counter the effects of cooling. Some process inside the Earth must therefore enable the mantle to retain a significant part of its calories. One hypothesis is that only subduction zones and thermal plumes contribute to mass exchange between upper- and lower mantle. To test that hypothesis, significant advances are need in our ability to image the Earth's interior, especially thermal plumes. The goal of this project was to provide this significant advance, in particular by developing the capability to observe seismic waves in the oceans, since they cover 2/3 of the Earth's surface and host, with very few exceptions, the hotspots that indicate the presence of plumes.
Tomographic imaging uses the small deviations in the time it takes seismic waves to traverse the mantle, exploiting the fact that hot thermal anomalies slow waves down, while cold slabs transmit the waves faster to picture temperature variations. When a seismic wave returns to the surface in the oceanic domain, it generates an acoustic wave in the water column. We were able to develop the Mermaid, an autonomous robot that is able to recognize the arrival of this acoustic wave despite the presence of many other acoustic signals. We modified a conventional oceanographic Apex float, equipped it with a hydrophone and two-way satellite communication, and a wavelet-based algorithm that provided sufficient intelligence to identify a seismic arrival from a strong earthquake at the other side of the Earth. We are routinely able to see earthquakes at large distance with magnitudes larger than 6.5 occasionally even below 6. At close distance (e.g. on an oceanic ridge) the smallest quake recorded had a magnitude of about 2. In two days, we recorded so many (235) small earthquakes of an earthquake swarm in the Indian ocean that the memory capacity was exceeded, since we had never expected the robot to be so successful. We now have two ongoing projects for sea-bases tomographic observations: one in the Ligurian Sea and one around the Galapagos islands, location of a suspected mantle plume. A Proof-of-Concept addition has enabled us to work with OSEAN, a small engineering firm, to develop a new float with a much longer life expectancy. This ill in principle make it possible to cover all of the oceans with a new seismic network.
In addition to the Mermaid development, we continued our theoretical work on global tomography and assembled a giant data set that will eventually be used to invert directly for temperature variations. Important advances were made in parameterization of the Earth using wavelets, in the measurement of seismic delays using cross-correlation, in the reduction of very large data sets without loss of information (reducing computing memory requirements by an order of magnitude), in developing an objective method for the estimation of standard errors in observed data, and - using local tomography in the Andes - in linking tomographic images to compositional variations. This involved considerable software development, which is shared via our web page. We also showed, that temperature variations alone (rather than changes in composition) can explain the statistics of seismic delays in the lower mantle, if the influence of signal frequency content is taken into account.
In summary, the project has made a giant leap forward in advancing observational seismology in the oceans, and has made crucial steps towards inverting giant seismic data sets jointly for the Earth's temperature variations. We expect to reach this stage in 2016.
Tomographic imaging uses the small deviations in the time it takes seismic waves to traverse the mantle, exploiting the fact that hot thermal anomalies slow waves down, while cold slabs transmit the waves faster to picture temperature variations. When a seismic wave returns to the surface in the oceanic domain, it generates an acoustic wave in the water column. We were able to develop the Mermaid, an autonomous robot that is able to recognize the arrival of this acoustic wave despite the presence of many other acoustic signals. We modified a conventional oceanographic Apex float, equipped it with a hydrophone and two-way satellite communication, and a wavelet-based algorithm that provided sufficient intelligence to identify a seismic arrival from a strong earthquake at the other side of the Earth. We are routinely able to see earthquakes at large distance with magnitudes larger than 6.5 occasionally even below 6. At close distance (e.g. on an oceanic ridge) the smallest quake recorded had a magnitude of about 2. In two days, we recorded so many (235) small earthquakes of an earthquake swarm in the Indian ocean that the memory capacity was exceeded, since we had never expected the robot to be so successful. We now have two ongoing projects for sea-bases tomographic observations: one in the Ligurian Sea and one around the Galapagos islands, location of a suspected mantle plume. A Proof-of-Concept addition has enabled us to work with OSEAN, a small engineering firm, to develop a new float with a much longer life expectancy. This ill in principle make it possible to cover all of the oceans with a new seismic network.
In addition to the Mermaid development, we continued our theoretical work on global tomography and assembled a giant data set that will eventually be used to invert directly for temperature variations. Important advances were made in parameterization of the Earth using wavelets, in the measurement of seismic delays using cross-correlation, in the reduction of very large data sets without loss of information (reducing computing memory requirements by an order of magnitude), in developing an objective method for the estimation of standard errors in observed data, and - using local tomography in the Andes - in linking tomographic images to compositional variations. This involved considerable software development, which is shared via our web page. We also showed, that temperature variations alone (rather than changes in composition) can explain the statistics of seismic delays in the lower mantle, if the influence of signal frequency content is taken into account.
In summary, the project has made a giant leap forward in advancing observational seismology in the oceans, and has made crucial steps towards inverting giant seismic data sets jointly for the Earth's temperature variations. We expect to reach this stage in 2016.