Overall seismotectonic modelling - a powerful tool in seismic hazard analyses: two applications in central Italy and California (U.S.A.)

A keystep in seismic hazard analyses is the assessment of the maximum potential earthquake which could occur in a studied area. For a world-wide application, use is made of empirical relationships between some characteristics at the surface of active faults and the surface-wave magnitude of the maximum related earthquakes. But the use of these relationships presents problems in terms of applicability and reliability, which are due mainly to the following factors: the difficulty of finding on the ground the key elements required as input to these relationships (active faults and/or seismic ruptures); this task is all but easy, especially when making only superficial geological surveys in tectonically complex areas, where brittle-faulted basements (and consequently most of the relevant seismic sources) are often hidden by disharmonic ductile sediment covers, the virtual exclusion in these relationships of the most significant parameters required by the Geomechanic for a correct location and evaluation of the maximum potential earthquake of a fault (respectively the deep geometry and the frictional strength of a fault plane). For these reasons, ai ENEL (National Electric Generating Board of Italy S.p.A.), in order to improve these relationships, a methodological approach has been developed, which is more suitable for tectonic settings such as the Italian one and which allows complete seismic hazard analyses (assessment of the earthquake sources and prediction of earthquake effects), where geological, mechanical and geophysical characteristics of the crust are taken into account ai the same time to formulate an overall seismotectonic model of a given area which explains all the available tectonic and seismological information. In this new methodological approach, evaluation of the seismic sources is done maintaining the basis of the existing relationships (which is to relate an active tectonic feature to the potential earquake); this basis has been made more effective by giving particular attention to the rheological conditions of the rocks involved in active strain, and by considering the resistant that, obstructing the motion of the brittle portion of an identified tectonically active structure, allows accumulation of elastic energy from which, if suddenly released, earthquakes can be generated. So, seven main categories of three-dimensional tectonic structures, to be filed on the basis of their seismotectonically relevant characteristics, have been proposed as reference for creating a seismotectonic data-bank of the most common active tectonic structures in the world. Thus, in the future, instead of referring only to one or more characteristics ai the ground surface of an active fault (as is done when applying the existing fault/magnitude relationships), a more complete comparison could be done with all active 31) tectonic structures, including their physical properties. li is hoped that such an approach, will be carefully evaluated by the existing Regulatory Commissions, and be as widely adopted as possible by the researchers so as to create seismotectonic data banks wherein world-wide seismically active structures can be filed on the basis of their overall seismotectonic characteristics. To give an example of the noteworthy possibilities of the method, two overall seismotectonic models are proposed: one for the central part of Italy and one for the western U.S.A. The guidelines of the overall seismotectonic models used to explain the seismic patterns in these two areas are: in the first case, the presence of a large zone with thin brittle crust softened by a geothermal positive anomaly, which is the residual effect of uprisen mantle material that ai present is acting against a slab of continental lithosphere on the NE, E and SE borders of the warm area; in the second case, the resistance given by a rooted cold body (probably correlated to the outcropping Sierra Nevada batholith) to the relative movement between the Pacific and North American Plates (in part taking place along the famous San Andreas fault), and the probable presence of soft crust in correspondence to areas with high heat flow densities in Oregon, Southern Idaho, Western Utah and Arizona. In this second case, the suggested model should only be considered very preliminary, due to the lack of sufficient information about the geometry and strength of the rocks belonging to active tectonic structures down to the maximum depth of the known seismicity; if studies aimed at completing the data set required by the present seismotectonic methodology were carried out, a remarkable improvement in the reliability of the western U.S.A. seismic hazard assessment could be obtained.