GRUPPO NAZIONALE PER LA DIFESA DAI TERREMOTI
FRAMEWORK PROGRAM 2000-2002
REVISION OF THE THEORETICAL AND OBSERVATIONAL GROUNDS OF THE SEISMIC HAZARD ESTIMATES AT A NATIONAL SCALE
Dipartimento di Fisica, Università di Bologna
Viale Berti Pichat, 8 - 40127 Bologna
ANNUAL SCIENTIFIC REPORT
The research activities in the ambit of the present Project, effectively started on November 23, 2000 at the time when the funds became available to the research units. After the prorogation of all of projects, accorded on July 3 2001 in consideration of the delays accumulated in the activation phase, the investigations are still in course and will be completed only within the end of current year 2001. Hence this report cannot be exhaustive in the description of the results of such activities and only represents a "state of the art" lacking of the conclusive issues. As well known, seeing the strong reduction of the allotted funds with respect to the original request, the objectives and the products of the project had been reformulated according to the comments of the Evaluation Commission in order to privilege the researches considered "new and innovative" by the Commission. At the same time we also tried to save somehow the original aims of the project that would be completely compromised following literally the indication of the Commission. In particular, as already shown by the coordinator of the project on occasion of the presentation meeting, held in Rome on October 5 2000, a complete elimination of Tasks 1,2,8,10 and 11, would make the project illogical and almost completely lacking of links among the residual Tasks. For this reason a significant part of the funds (40%) were devoted for uses, generically indicated as "Other", concerning activities in the ambit of the Tasks considered not interesting (or maybe not well formulated). Among them, only activities "propaedeutical" to the hazard estimates have been considered, being that in the original project the hazard evaluation was scheduled for the second and third years (not approved by the Commission). We anyhow started, in collaboration with a INGV research group not formerly included in the project, some activities devoted to the formulation of an innovative time-dependent hazard model, based on a recently published space-time occurrence model (Console e Murru, 2001), that proved to own a better predictive efficiency with respect to the traditional Poissonian one.
Report of research activities carried on in the ambit of Tasks considered "new and innovative" by the Evaluation Commission.
Task 3 SEISMOGENIC MODEL (Resp.: Valensise)
Due to the financial limitations imposed by the severe budget reduction, the activities of this Task proceeded essentially within other independent research projects. The first release (1.0) of the "Database of potential sources for earthquakes larger than magnitude 5.5 in Italy" (Valensise and Pantosti eds, 2000) was made public, and a few sample copies distributed through CDs, during the Workshop "La sismogenesi in Italia" held at INGV (20-21 July 2000). During the present project new data have been acquired and included in the Database, the second version of which is being published on Annali di Geofisica (Valensise and Pantosti, 2001). The management software has also been improved. In particular, the new version of the software governing the Database will incorporate the catalogue of earthquake focal mechanisms elaborated within Task 7. Each mechanisms will be represented as a properly georeferenced and scaled "beach-ball", and the whole catalogue will benefit from the flexible architecture of the Database and of MapInfo. These improvements will become of public domain with a third release of the Database scheduled to appear in 2002.
Task 4 DYNAMICS OF THE SEISMIC SOURCES (Resp.: Bonafede)
The studies of this Task (not funded by the project) continued as scheduled by the research projects that funds them.
Task 5 STATISTIC OF SEISMIC SOURCES AND CATALOG COMPLETENESS (Resp.: Gasperini)
We performed an analysis of the instrumental seismicity from 1960 to 1996 (Lolli e Gasperini 2001) as it results from the combination of the "Catalogo Strumentale dei terremoti Italiani dal 1981 al 1996" (CSTI) (Instrumental Catalog Working Group, 2001) with the "Catalogo del Progetto Finalizzato Geodinamica (PFG)" (Postpischl, 1985). For this joint catalog we verified that the event rate is strongly variable with time, because of an incorrect calibration of the magnitude in the period before 1980 (see Fig. 1)
Fig. 1 Cumulative number of events in Italy with M³4.0 from the PFG catalog (from 1960 to 1980) and from the CSTI Catalog (from 1981 to 1996). The quotes indicate the annual rate in the different periods (excluding the aftershocks of major events).
In order to make homogeneous the two catalogs we had to apply an empirical correction to the magnitudes of the PFG catalog. In particular diminishing the magnitudes of 0.5 units in the period from 1976 to 1980 and of 0.3 units in the period from 1960 to 1975 we obtained a catalog whose annual rate is almost constant, excluding the aftershock of major earthquakes, over the entire period from 1960 to 1996. The resulting dataset has been used, in the cited work by Lolli e Gasperini (2001) to compute the parameters of the Reasenberg and Jones (1989) model of aftershocks occurrence in order to predict the behavior of future sequences generated by strong earthquakes.
Task 6A SEISMIC WAVE TOMOGRAPHY (Resp.: Morelli)
Our studies consisted of preliminary steps necessary for using crustal structure models to correct body wave travel times before seismic tomographic inversion. We have considered several models, available in the literature, which contain different information and which are characterized by different spatial resolution. Due to the limited resources available, it has not been possible to do the proposed compilation of published data, but it has been necessary to resort to existing syntheses. In the meantime, the existing tomographic model has been analyzed in detail, to identify to which extent the scarce knowledge of crustal structure may bias results. In our current approach, the effect of the crust is removed in two distinct phases: with the use of station static corrections, and by the use of a 0-km depth layer in the inversion, which absorbs the effects of otherwise unmodeled crustal delays. The database of seismic phases, used in the inversion (Fig. 2), has also been updated with data made available in the last years.
Fig. 2 Seismic phase tomography for layer at 50 km of depth, after Piromallo and Morelli (2001a).
Task 6B, SEISMIC ATTENUATION TOMOGRAPHY (Resp.: Mele)
The activity in this field, essentially preliminary to the real researches, regarded the verification of the available data.
Task 6C, SEISMIC INTENSITY TOMOGRAPHY AND SITE EFFECTS (Resp: Gasperini)
Subtask 6C.1 Tomography of seismic intensity A physically grounded bilinear attenuation model (Gasperini, 2001a) represents the central reference of a tomographic inversion on a grid with sides of 25 and 50 km.
Fig. 3 Variations of the coefficient of the first trait of the bilinear attenuation model (Gasperini, 2001a) with respect to the central value of 0.051 degree/km.
In Fig. 3 we can observe how the zones with higher attenuation are concentrated mainly on the Tyrrhenian slope of Apennines and close to active volcanic areas (Gulf of Naples and Mt. Etna), while a narrow and prolonged band with low attenuation well follows the outer North-Apenninic Arc. The attenuation is also high in the Veneto-Friuli area (north-east), while an insulated spot with a relatively high attenuation is present in Piedmont almost exactly centered to the Ivrea area. The attenuation maxima (close to active volcanoes) correspond to a decrease of the intensity of one degree per 10 km while the minima (in the Po valley, in the Trentino region and in the Eastern Sicily) to about one degree per 100-200 km. The spatial variation of the slope in the second part of the curve (not shown here) shows a pattern quite different with minima located in the Northern area and maxima in the Apennines. The checkerboard tests demonstrated that the resolution of the method is good even if a random error of the same order (s≈1 degree) of the one really observed in the data (Gasperini, 2001a) is added to simulated observations.
Subtask 6C.2 Site effects in terms of intensity. The locality empirical residuals resulting from the application of the tomographic model described above have been computed. An analysis of the geological and geophysical properties of the soils have been approached to establish the possible correlation with the observed residuals.
Moreover, in collaboration with the "Progetto Finalizato Beni Culturali" of the CNR a detailed seismic zonation of the urban area of Florence (Boccaletti et al., 2001), based on a computer-aided analysis of the damages induced by the earthquakes of May-June 1895 have been carried out.
Fig. 4 Spatial behavior of EMS92 intensity in the urban area of Florence, after the earthquakes of May-June 1895. The intensity value is computed by a computer-aided technique on overlapping cells with sides of 500 m with centers shifted by 50 m.
In Fig. 4 the results of such analysis are shown overimposed to a cartographic base almost coeval (1897) to the seismic event. We can see how, for the most of the analyzed area, the estimated intensity ranges from VI to VII. It must be noted that VI-VII is the intensity estimated by the computer-aided technique for the whole city. As well it also coincides with the evaluation made by macroseismic experts (Ferrari and Guidoboni, 1995). The most severe effects are concentrated in the north-eastern part of the city, where in some places intensity reaches the degrees VIII and IX. This may appear surprising as such area is placed at a longer distance from the seismic source (located about 15 km south east of the city, close to the small town of Impruneta) with respect to the south-eastern one, where instead the intensity is VI.
Subtask 6C.3 Site effects in terms of ground motion parameters. According to the reformulation of the project no activities were scheduled for this Subtask.
Task 7, FOCAL MECHANISMS (Resp.: Morelli)
Subtask 7.1 CMT Focal Mechanisms. The complete description of earthquake focal mechanisms, consisting of seismic moment tensors, is conveniently computed with the Centroid Moment Tensor (CMT) algorithm, routinely run on all events, globally distributed, having magnitude above 5-5.5 at Harvard University. An extension of the method allows the use of the fundamental mode surface wave train, thus reducing the minimum magnitude of analyzable events. We have applied the Regional CMT (RCMT) technique to all Italian earthquakes, which could be analyzed, as a detail improvement of the Mediterranean Catalog which we maintain (Pondrelli et al., 2001a). Newly analyzed events (years 1997-2000) are now being published (Fig. 5). In the meantime, we considered all events, for which seismic moment tensors are available -- either on Harvard or on our new Mediterranean catalog -- located in Italy and surrounding regions, to calculate the geometry and total amount of seismic deformation. Results have been presented to international conferences. Limitation in resources however hindered the realization of detailed studies, and has limited the time window which could be analyzed.
Fig. 5 - Map of all RCMT (Regional Centroid Moment Tensor) solutions presently available for Italian region. All mechanisms belong to moderate magnitude events (4.0 < M < 5.5). In black the mechanisms belonging to the RCMT Euro-Meditearranean Catalog (Pondrelli et al., 2001a) which includes all solutions for moderate magnitude events occurred between 1997 and 2000. In red the mechanisms for events occurred between 1968 and 1996.
Subtask 7.2 First motion focal mechanisms. A MS-Access Database, including presently 4800 Fault plane solution, had been designed and built. Its most prominent feature is the standardization of the formats for successive computations. It also allows importing the data coming from the CMT, INGV and ETH
Fig. 6 MS-Access mask for the visualization of output data.
This tool make it possible to easily access the data by MS-Access masks (Fig. 6), to generate input files for the most popular graphical packages (GMT, qplot, Mapinfo, ecc.), as well as to compute the strain and stress fields by the current methodologies (Kostrov, 1974; Gephart and Forsyth, 1984).
Task 9A, GEODETIC MEASUREMENTS (Resp.: Riguzzi)
SubTask 9A.1 Materialization of the vertices of the Apenninic network. A compilation of the permanent, semi-permanent and temporary sites that operated in Italy has been completed. Because of the strong reduction of funds we decided to concentrate our efforts on a reduced area from the surrounding of Aquila to Pollino Massif (instead of from Umbria to Calabria). The weakest parts of the network (coast of Lazio, southern Tuscany and Abruzzo), where the strengthening of the network is most urgent, has been determined. Many sites were carefully recognized to find geologically stable areas, taking into account the local networks and permanent stations already existing.
Three new sites were established (Fig. 7) to be occupied semi-permanently (Madonna di Cristo e Le Serre) or permanently (Roseto degli Abruzzi) by GPS receivers. Two new permanent stations are now working in the area under investigation (Potenza and Castel del Monte), also thanks to research funds coming from other projects. The site descriptions of the selected sites containing a short geological sketch, the access to the vertices and the WGS84 approximate coordinates were compiled. GPS observation analysis, coming from campaigns performed in the framework of other projects, showed with high significance the horizontal relative velocities of the southernmost part of the network; its estimation is about 4 mm/year and the deformation field has the main component of extensional style, normal to the Apennine chain.
Fig. 7 Map of permanent and semi-permanent stations in the investigated area
Subtask 9A.2 Execution of two measurement campaigns at least. No activities scheduled.
SubTask 9A.3 Data-Bank. To link powerfully the geodetic, the seismological and the tectonic databases, we started to develop a GIS by MapInfo 6.0 over Windows 98 platforms. The GIS allows to analyze and to manage in a unique dynamic and shared database the numerical information coming from the GPS surveys and the alpha-numerical information of the site descriptions.
SubTask 9A.4 Processing. No processing scheduled in this phase.
Task 9B, DEFORMATION FIELD (Resp.: Baldi)
Even in this case the activities are put off to next years.
Task 2A, INSTRUMENTAL SEISMIC CATALOG (Resp.:Monachesi)
Some investigators, participating to the present project, had made available to the GNDT researchers the database of the "Catalogo sismico strumentale dal 1981 al 1996" (Instrumental Catalog Working Group, 2001) resulting from the activities in the framework of the Subproject 5.1.3 of the former GNDT Executive Project 1998 (Resp: P. Gasperini and G. Monachesi). It consists of a CDROM containing both the data and the descriptive reports (in HTML format), which version 1.0 has been distributed in the GNDT ambit and is now available also over the Web at the address:
Since, on the basis of some test performed after the publication of the database, some error and inaccuracies (concerning some tenth of events) were found, the same group is now working to prepare a new corrected version of the database that will be made available (only on the Web) probably after the end of the present project.
Task 2B, MAGNITUDE (Resp.: Gasperini)
A work on the revision of the local magnitude, presently in press (Gasperini, 2001b), has been completed. This allowed the re-calibration of the duration and of the amplitude magnitudes of the events included in the database mentioned in Task 2A, by a comparison with the data of the two Wood-Anderson (WA) seismometers that operated in Italy in the seventies and eighties and with the synthetic WA traces deduced from Broad-Band stations. This research is still in course providing for the acquisition of new synthetic WA traces and the re-reading of real WA amplitudes from original seismograms (currently in course of recording by the Sismos Project of INGV). The aims of this research are the estimation of regionalized attenuation functions for the computation of magnitude and the increase of the database of WA magnitudes.
Task 1A, HISTORICAL SEISMIC CATALOG (Resp.: Guidoboni)
On the basis of the results of the revision of instrumental data (Task 2A) an update of the "Catalogo Parametrico dei Terremoti Italiani" (CPTI) has been started. This includes the recording and elaboration of the data coming from the Macroseismic Bulletin of the INGV from 1988 to 1996 and the upgrade of the database concerning the macroseismic works not yet utilized in the first version of the CPTI, like that the book on the Umbria-Marche Earthquakes (Boschi et al., 1998) and the version 3 of the Catalogue of Strong Italian Earthquakes from 461 B.C. to 1997 (Boschi et al., 2000).
Anzidei M., P. Baldi, G. Casula, A. Galvani, E. Mantovani, A. Pesci, F. Riguzzi, and E. Serpelloni (2001). Insights on present-day crustal motion in the central Mediterranean area from GPS surveys, Geophys. J. Int., 146, 98-110.
Boccaletti M., Corti G., Gasperini P., Piccardi L., Vannucci G. and Clemente S. (2001) Active tectonics and seismic zonation of the urban area of Florence, Italy. Pageoph, 158, 2313-2332.
Carletti F. and Gasperini P. (2001) Lateral variations of macroseismic intensity attenuation in Italy, (in preparation).
Gasperini P., (2001a). The attenuation of seismic intensity in Italy: a bilinear shape might indicates the dominance of deep phases at epicentral distances longer than 45 km, Bull. Seism. Soc. Am., 91, 826-841.
Gasperini, P. (2001b) Local magnitude revaluation for recent Italian earthquakes (1981-1996). Journal of Seismology, (in press).
Instrumental Catalog Working Group (2001), Catalogo Strumentale dei terremoti italiani dal 1981 al 1996, Version 1.0, Clueb, Bologna, CD-ROM.
Lolli B. and Gasperini P. (2001) Aftershocks prediction in Italy Part I: Estimation of time-magnitude distribution model parameters and computation of probabilities of occurrence, Journal of Seismologi (submitted).
Piromallo C., and A. Morelli (2001a) P-wave tomography of the top 1000 km under the Alpine-Mediterranean area J. Geophys. Res (submitted).
Piromallo C., and A. Morelli (2001b) Improving seismic location: an alternative to three-dimensional structural models, Pure and Applied Geophysics, 158, 319-347.
Piromallo C., A. Vincent, D. Yuen, and A. Morelli, (2001) Dynamics of the transition zone under Europe inferred from wavelet cross-spectra of seismic tomography, Phys. Earth Planet. Int., 125, 125-139.
Pondrelli S., A. Morelli, G. Ekstrom, S. Mazza, E. Boschi, and A. M. Dziewonski, (2001a) European-Mediterranean Regional Centroid Moment Tensors catalog: 1997-2000 Phys. Earth Planet. Int. (submitted).
Pondrelli, S., Ekstrom, E., and Morelli, A., (2001b) Seismotectonic re-evaluation of the 1976 Friuli, Italy, seismic sequence, J. of Seism., 5, 73-83.
Riguzzi F., G. Pietrantonio, M. Crespi and M. Anzidei (2001). True or false GPS-derived deformations ?, Annali di Geofisica, 44, 3, 593-604.
Serpelloni E., M. Anzidei, P. Baldi, G. Casula, A. Galvani, A. Pesci and F. Riguzzi, (2001). Geodetic deformations in the Central-Southern Appennines (Italy) from repeated GPS surveys, Annali di Geofisica, 44, 3, 627-647.
Valensise, G., and D. Pantosti, (Edited by) (2000) Database of Potential Sources for Earthquakes Larger than M 5.5 in Italy, Public Release No. 1.0.
Valensise, G., and D. Pantosti (edited by) (2001). Database of Potential Sources for Earthquakes Larger than M 5.5 in Italy P. R. 2.0. Annali di Geofisica, Vol. 44, Suppl. 1, with database on CD-ROM (in press).
Boschi E., Guidoboni E., Ferrari G. and Valensise G. (1998) I terremoti dell'Appennino Umbro-Marchigiano, Compositori, Bologna, pp.267.
Boschi E., Guidoboni E., Ferrari G., Mariotti D., Valensise G. and Gasperini P. (2000), Catalogue of Strong Italian Earthquakes from 461 B.C. to 1997, Annali di Geofisica, 43, n. 4, e CDROM allegato.
Console R. and Murru M. (2001) A simple testable model for earthquake clustering, J. Geophys. Res., 106, 8699-8711.
Kostrov V.V. (1974) Seismic moment and energy of earthquakes and seismic flow of rocks. Izv. Earth Phys, 1, 23-40.
Gephart J.W. and Forsyth W.D. (1984) An improved method for determining the regional stress tensor using earthquake focal mechanism data: application to the San Fernando earthquake sequence. J. Geophys. Res., 89, 9305-9320.
Guidoboni E: and Ferrari G (1995) Historical cities and earthquakes: Florence during the last nine centuries and evaluation of seismic hazard, Annali di Geofisica, 38, 617-648.
Postpischl, D. (Edited by) (1985) Catalogo dei terremoti italiani dall'anno 1000 al 1980. Quaderni della Ricerca Scientifica 114 2B, pp. 239.
Reasenberg, P.A. and Jones, L.. (1989) Earthquake hazard after a mainshock in California. Science, 243, 1173-1176.