3D P-wave velocity model of Ireland's crust from controlled source tomography

<p>We present a 3D P-wave velocity model of the crust and uppermost mantle below Ireland. In the absence of local earthquakes, we used quarry and mining blasts recorded on permanent stations in the Irish National Seismic Network (INSN) and during various temporary deployments. We compiled a database of 1,100 events and around 20,000 P-wave arrivals, with each event associated with a known quarry. The source location uncertainty is therefore minimal. Both source and receiver locations are fixed in time and we used repeating events to estimate the travel time uncertainty for each source-receiver combination. We created a starting 1D velocity model from previously available data, and then used VELEST to calculate a preliminary minimum 1D velocity model. The 1D velocity model enabled us to remove outliers from the data set, and to calculate the final minimum 1D model used as the initial model in the 3D tomographic inversion. The resulting 3D P-wave velocity model will shed new light on the 3D crustal structure of Ireland.</p>

Improved earthquake location in the area of North Euboean Gulf after the implementation of a 3D non-linear location method in combination with a 3D velocity model.

Considerably improved hypocentral locations of the 274 earthquakes, with magnitudes between 1.5 to 4.1 Ml recorded during the period from 2009 to 2010 by the Hellenic Unified Seismographic Network (HUSN), have been obtained for the area of North Euboean Gulf after the implementation of a 3D non-linear location algorithm and a previously calculated local 3D velocity model for both P and S wavephases. To assess the effectiveness of the 3D locations we compared the results with the solutions obtained with alternative 1D velocity models such as the minimum 1D model calculated with the VELEST algorithm and the 1D model used by the National Observatory of Athens (NOA) for daily earthquake analysis. We were further ableto assess the location accuracy of each model by comparing the location results for a number of quarry blasts that occurred in the area in that period. The use of the local 3D velocity model provides considerably more accurate than the minimum 1D model which in turn provides more constrained locations from the 1D model of NOA. The epicentral locations calculated by each model are almost similar; however the depth distribution of the events varies, with depth differences of up to 12 km for some earthquakes. The results prove that accurate, local models are necessary in order to achieve more accurate locations for the events in a local area.