Marja Uski

Earthquakes in northern Europe in 1375–1989

Abstract An updated version of the catalogue of earthquakes in northern Europe is presented. It comprises events inside a region window of about 55–80°N and 10°W–45°E, covering a 615-year period from 1375 to 1989. The data have been compiled using all the available historical publications, catalogues, studies and reports of the region, and for events from 1984 to 1989, a regional seismic parameter database maintained by the Institute of Seismology of the University of Helsinki. The number of events exceeds 5200, of which more than 40% have been registered since 1980. The huge increase of event reports is the result of the installation of new high-frequency networks for microearthquake studies and the upgrade of the seismic instrumentation. Because of the different degree of accuracy and homogeneity of the parameter data, the events were divided into historical (1375–1964) and instrumental (1965–1989) datasets. The source parameters are mainly macroseismic for historical events, and obtained by iterative location procedure for instrumental data. By assuming internal homogeneity between the different regional magnitude scales, we estimated the catalogue to be complete for events with magnitude ≥ 4.5 since the 1880s, magnitude 4.0–4.4 since the 1940s and magnitude 3.5–3.9 since the 1970s. Although the distribution of seismicity exhibits similar patterns in both datasets, the clustering of epicenters and offshore seismicity is more distinct in the instrumental period. The occurrence of the greatest events can be seen more clearly in the historical data. The earthquake epicenters are concentrated in the geologically younger tectonic provinces in the western part of the region and follow, with some exceptions, the tectonic fracture and fault zones. In the Fennoscandian area, 80% of the earthquakes occurred in the upper 20 km of the earths crust. The deepest epicenters are in the southwestern corner of the region where the crust is thinnest.

Focal mechanisms of three earthquakes in Finland and their relation to surface faults

Abstract Focal mechanisms for three recent earthquakes in Finland are determined using P-wave polarities together with SV/P and SH/P phase amplitude ratios. The events occurred on May 11, 2000 in Toivakka, Central Finland (ML=2.4), on September 15, 2000 in Kuusamo, northeastern Finland (ML=3.5), and on May 2, 2001 in Kolari, western Finnish Lapland (ML=2.9). In order to obtain reliable estimates of the source parameters, one-dimensional crust and upper mantle velocity models are derived for the epicenter areas from deep-seismic sounding results. The starting models are modified by one-dimensional ray tracing using the earthquake observations. The events are relocated by employing P- and S-phase arrival times from the nearest seismic stations and the final velocity models. Synthetic waveforms, calculated with the reflectivity method, are used to further constrain and verify the source and structural parameters. The Toivakka earthquake indicates thrust- or reverse-faulting mechanism at a depth of 5 km. After comparison with aeromagnetic and topographic data we suggest the eastward dipping nodal plane (358°/42°) was the fault plane. The best-fitting fault plane solution of the Kolari earthquake suggests pure thrust-faulting at a depth of 5 km. The nodal plane striking 035°/30° correlates well with surface observations of the postglacial, possibly listric fault systems in the source area. The Kuusamo earthquake (focal depth 14 km) has a normal-faulting mechanism with the nodal planes trending 133°/47° or 284°/47°. Preference is given to the SW-dipping nodal plane, as it seems to coincide with topographic and magnetic lineament directions that have been active after the last ice age. The three earthquakes have occurred in old Precambrian faults and shear zones, which have been reactivated. The reactivated faults are favourably oriented in the local stress field.

Regional spectral scaling relations of source parameters for earthquakes in the Baltic Shield

Abstract Spectral analysis of regionally recorded Lg waves is performed to determine source parameters such as seismic moment, source radius and stress drop of earthquakes in the Baltic Shield, and to derive regional spectral scaling relations. The data consist of about 350 Lg phases on short-period, vertical-component analog seismograms from earthquakes with magnitude ( M L ) ranging from 2 to 5.2. Source radii estimated from the corner frequency show only a slight increase with increasing seismic moment in the range 3 × 10 18 to 5 × 10 20 dyn cm, while the increase is more rapid for larger events. Baltic Shield earthquakes show increasing stress drops, ranging from about 0.1 to 10 bars, for increasing seismic moment. The relatively low stress drops could, in part, be explained by bias due to band-limited analog data. The slope of the curve relating seismic moment to corner frequency is steeper than −3, which suggests a departure from a constant stress drop scaling relation proposed for large earthquakes. The derived relationship between the seismic moment ( M o ) and magnitude ( M L ) is: log M o = 16.93 + 1.01 M L (for 2 ⩽ M L ⩽ 5.2 ).

Automatic classification of seismic events within a regional seismograph network

This paper presents a fully automatic method for seismic event classification within a sparse regional seismograph network. The method is based on a supervised pattern recognition technique called the Support Vector Machine (SVM). The classification relies on differences in signal energy distribution between natural and artificial seismic sources. We filtered seismic records via 20 narrow band-pass filters and divided them into four phase windows: P, P coda, S, and S coda. We then computed a short-term average (STA) value for each filter channel and phase window. The 80 discrimination parameters served as a training model for the SVM. We calculated station specific SVM models for 19 on-line seismic stations in Finland. The training data set included 918 positive (earthquake) and 3469 negative (non-earthquake) examples. An independent test period determined method and rules for integrating station-specific classification results into network results. Finally, we applied the network classification rules to independent evaluation data comprising 5435 fully automatic event determinations, 5404 of which had been manually identified as explosions or noise, and 31 as earthquakes. The SVM method correctly identified 94% of the non-earthquakes and all but one of the earthquakes.The result implies that the SVM tool can identify and filter out blasts and spurious events from fully automatic event solutions with a high level of accuracy. The tool helps to reduce the work-load and costs of manual seismic analysis by leaving only a small fraction of automatic event determinations, the probable earthquakes, for more detailed seismological analysis. The self-learning approach presented here is flexible and easily adjustable to the requirements of a denser or wider high-frequency network. Fully automatic method for classification of seismic events.The method is based on Support Vector Machine.Effective in filtering out blasts and spurious events from automatic event bulletins.The method is flexible and easily adjustable to denser or wider networks.

A new local magnitude scale for the Finnish seismic network

Abstract A total of 1259 synthesized Wood-Anderson recordings from 216 local and regional earthquakes were used to establish a new ML scale for Finland. The hypocentral distances of the events range from 25 to 1940 km and the reported magnitudes from 1 to 4. The attenuation function (−log A0 in Richters original definition of ML), magnitudes, and station corrections for the eleven stations used in this study were determined by a multiple linear regression analysis. The shape of the attenuation function was modelled by two parametric forms: the “bilinear” function, −log A0(f), was described in terms of a frequency-dependent anelastic attenuation, and a geometrical spreading with separate rates for Sg and Lg domains. The −log A0 function comprised a single geometrical spreading term and a frequency-independent anelastic attenuation term. The results are as follows: -logA0(f) = 1.27log R + 0.00041f0.36R + 0.36 + S for R ⩽ 350 km -logA0(f) = 0.83log R + 0.00041f0.36R + 1.48 + S for 350 km -logA0(f) = 1.42log R + 0.00008R + 0.15 + S for R ⩽ 1900 km where R is hypocentral distance in kilometers, f is frequency (Hz) and S is station correction. The absolute levels of the attenuation functions were adjusted so that at 60 km the −log A0 and −log A0(10 Hz) curves equal the revised −log A0 for southern California. A correction was determined to accommodate the use of vertical seismograms in the ML formula. The logarithm of the horizontal (H) to vertical (Z) shear wave amplitude ratio was found to be distance-dependent: log ( H Z ) = 0.00009R . The difference in magnitudes computed from the above distance correction functions was less than 0.1 units. The frequency-independent form is preferred as it is simple to apply in routine bulletin work. The new ML scale can be written as: ML(f) = log A + 0.00017R + 0.15 + S for R ⩽ 350 km where A is the synthesized Wood-Anderson amplitude (mm) measured from the vertical component seismogram.