G. Selvaggi

The 1997 Umbria‐Marche, Italy, Earthquake Sequence: A first look at the main shocks and aftershocks

A long sequence of earthquakes, six with magnitudes between 5 and 6, struck Central Italy starting on September 26, 1997, causing severe damages and loss of human lives. The seismogenic structure consists of a NW-SE elongated fault zone extending for about 40 km. The focal mechanisms of the largest shocks reveal normal faulting with NE-SW extension perpendicular to the trend of the Apennines, consistently with the Quaternary tectonic setting of the internal sector of the belt and with previous earthquakes in adjacent regions. Preliminary data on the main shocks and aftershocks show that extension in this region of the Apennines is accomplished by normal faults dipping at low angle (∼40°) to the southwest, and confined in the upper ∼8 km of the crust. These normal faults might have reactivated thrust planes of the Pliocene compressional tectonics. The aftershock distribution and the damage patterns also suggest that the three main shocks ruptured distinct 5 to 15 km-long fault segments, adjacent and slightly offset from one another.

Complex Normal Faulting in the Apennines Thrust-and-Fold Belt: The 1997 Seismic sequence in Central Italy

A long sequence of moderate-magnitude earthquakes (5 M 6) struck central Italy in September and October 1997. At the end of the sequence a year later, the seismogenic area extends for about 60 km along the Apennines. The analysis of historical seismicity suggests that this seismic sequence filled a 700-year gap in this portion of the chain. Other historical sequences in the same area are characterized by prolonged seismic release on adjacent fault segments, probably due to the in- volvement of shallow and complex structures inherited by the compressive tectonics. The distribution of seismicity and the fault-plane solutions show that the extension in this region is accomplished by normal faults dipping at relatively low angles (40) to the southwest. The focal mechanisms of the largest shocks reveal normal faulting with extension perpendicular to the Apenninic chain (northeast-southwest), consistently with the Quaternary tectonics of the internal sector of the northern Apen- nine belt and with previous earthquakes in adjacent regions. Three mainshocks oc- curred on distinct 5- to 10-km-long fault segments, adjacent and slightly offset be- tween each other. High-quality aftershock locations show that seismicity is confined within the sedimentary Mesozoic cover in the upper 8 km of the crust and that most of the aftershocks are shallower than the largest shocks, which nucleated at 6-km depth. Faults evidenced by aftershock locations have a planar geometry and show increased complexity toward the surface. Most of the aftershock focal mechanisms are dominated by normal faulting. Several strike-slip events occurred at shallow depths, reactivating portions of pre-existing thrust planes that segment the normal fault system. The spatiotemporal evolution of seismicity shows a peculiar migration of hypocenters along the strike of the main faults with multiple ruptures and the activation of fault segments before the occurrence of the main rupture episodes.

Spatio-temporal distribution of seismic activity during the Umbria-Marche crisis, 1997

We present the spatio-temporal distribution of more than 2000 earthquakesthat occurred during the Umbria-Marche seismic crisis, between September 26and November 3, 1997. This distribution was obtained from recordings of atemporary network that was installed after the occurrence of the first two largest shocks (Mw =, 5.7, Mw = 6.0) of September 26. This network wascomposed of 27 digital 3-components stations densely distributed in theepicentral area. The aftershock distribution covers a region of about 40 km long and about2 km wide along the NW-SE central Apennines chain. The activity is shallow,mostly located at less than 9 km depth. We distinguished three main zonesof different seismic activity from NW to SE. The central zone, that containsthe hypocenter of four earthquakes of magnitude larger than 5, was the moreactive and the more complex one. Sections at depth identify 40–50°dipping structures that agree well with the moment tensor focalmechanisms results. The clustering and the migration of seismicity from NW to SE and the generalfeatures are imaged by aftershock distribution both horizontally and at depth.

Passive Seismology and Deep Structure in Central Italy

In the last decade temporary teleseismic transects have become a powerful tool for investigating the crustal and upper mantle structure. In order to gain a clearer picture of the lithosphere-asthenosphere structure in peninsular Italy, between 1994 and 1996, we have deployed three teleseismic transects in northern, central, and southern Apennines, in the framework of the project Geo ModAp (European Community contract EV5V-CT94–0464). Some hundreds of teleseisms were recorded at each deployment which lasted between 3 and 4 months. Although many analyses are still in progress, the availability of this high quality data allowed us to refine tomographic images of the lithosphere-asthenosphere structure with an improved resolution in the northern and central Apennines, and to study the deformation of the upper mantle looking at seismic anisotropy through shear-wave splitting analysis. Also, a study of the depth and geometry of the Moho through the receiver function technique is in progress. Tomographic results from the northernmost 1994 and the central 1995 teleseismic experiments confirm that a high-velocity anomaly (HVA) does exist in the upper 200–250 km and is confined to the northern Apenninic arc. This HVA, already interpreted as a fragment of subducted lithosphere is better defined by the new temporary data, compared to previous works, based only on data from permanent stations. No clear high-velocity anomalies are detected in the upper 250 km below the central Apennines, suggesting either a slab window due to a detachment below southern peninsular Italy, or a thinner, perhaps continental slab of Adriatic lithosphere not detectable by standard tomography. We found clear evidence of seismic anisotropy in the uppermost mantle, related to the main tectonic processes which affected the studied regions, either NE-SW compressional deformation of the lithosphere beneath the mountain belt, or arc-parallel asthenospheric flow (both giving NW-SE fast polarization direction), and successive extensional deformation (~E-W trending) in the back-arc basin of northern Tyrrhenian and Tuscany. Preliminary results of receiver function studies in the northern Apennines show that the Moho depth is well defined in the Tyrrhenian and Adriatic regions while its geometry underneath the mountain belt is not yet well constrained, due to the observed high complexity.

The April 1996 Irpinia seismic sequence: Evidence for fault interaction

The analysis of the Irpinia earthquake of 3 April 1996 (ML = 4.9), based on strong motion and short period local data, shows that it was a normal faulting event located within the epicentral area of the MS 6.9, 1980, earthquake. It was located at 40.67° N and 15.42° E at a depth of 8 km. The local magnitude (4.9) has been computed from the VBB stations of the MedNet network. The moment magnitude is Mw = 5.1 and the seismic moment estimated from the ground acceleration spectra is 5.0 1023 dyne cm. Spectral analysis of the strong motion recordings yields a Brune stress drop of 111 bars and a corner frequency of 1 Hz. The source radius associated to these values of seismic moment and stress drop is 1.3 km. The focal mechanism has two nodal planes having strike 297°, dip 74°, rake 290° and strike 64°, dip 25° and rake 220°, respectively. A fault plane solution with strike 295° ± 5°, dip 70° ± 5°, and rake 280° ± 10° is consistent with the S-wave polarization computed from the strong motion data recorded at Rionero in Vulture. We discuss the geometry and the dimensions of the fault which ruptured during the 1996 mainshock, its location and the aftershock distribution with respect to the rupture history of the 1980 Irpinia earthquake. The distribution of seismicity and the fault geometry of the 1996 earthquake suggest that the region between the two faults that ruptured during the first subevents of the 1980 event cannot be considered as a strong barrier (high strength zone), as it might be thought looking at the source model and at the sequence of historical earthquakes revealed by paleoseismological investigations.

Recent Seismic Activity and Earthquake Occurrence Along the Apennines

In this paper we analyze the seismicity along the Apennines in order to relate the seismicity pattern to the seismogenic environment and to the active tectonic processes. The analysis of crustal and subcrustal seismicity show that the northern and southern Apennines are characterized by two distinct patterns of seismicity. These two domains are separated by two important lithological discontinuities (the Ancona — Anzio and the Ortona — Roccamonfina lines). The seismicity along the Apennines is mainly concentrated in a narrow belt running along the chain, with an evident geometrical offset which corresponds to the Ancona-Anzio line. We discuss these observations considering both the seismological evidence and the present tectonic regime in which the earthquakes occur. We focus on the southern Apennines, where the largest earthquakes (6 ≤ M ≤ 7.0) have occurred. We analyze the main features of some recent seismic sequences in order to discuss the seismicity patterns in terms of the hypothesis of segmentation of the southern Apennine seismogenic belt. This work aims to be a preliminary contribution to the important goals of understanding the spatial and temporal patterns of seismicity and of identifying active faults.

THE KOZANI-GREVENA (GREECE) EARTHQUAKE OF MAY 13, 1995, A SEISMOLOGICAL STUDY

Abstract We present a detailed seismological study of the Kozani earthquake. We relocate the mainshock with regional data at depth of 14.2km beneath the Vourinos massif. We compute the focal mechanism by body waveform modeling at teleseismic distance and find a normal fault striking N240 ° and dipping 40 ° toward the NW with a centroid depth of 11 km. We installed a dense network of portable seismographs around the epicentral region and located several hundreds of aftershocks. The main cluster of aftershock seismicity defines a plane dipping north at an angle of about 35 °, consistent with the main-shock mechanism, while some seismic activity is also seen on an antithetic fault. Our results suggest the active fault plane to be the Deskati fault which dips at a constant angle and therefore branches on the Paleohori fault where surface breaks were observed. We also compute 181 focal mechanisms which mostly show normal faulting.