The model selected for the Fucino basin source is a surface-breaking fault characterised by pure normal slip. The model agrees with direct observations of the coseismic fault scarps formed during the 13 January, 1915 earthquake, with the inversion of 1915 coseismic elevation changes, and with paleoseismological analyses of the main fault scarps observed in the Fucino Plain.
There are inconsistencies among different authors concerning the names used for the faults of the Fucino region. Following is a list of the different names used for each fault branch, along with the most recent references.
The three seismologcal focal mechanisms of the 13 January 1915 earthquake published so far used first motion polarities recorded at several stations wordlwide, but they are all different. Micci et al. (1975) propose a solution characterised by normal faulting along a NE-SW striking plane; Gasparini et al. (1985) propose left-lateral strike-slip on a NE-SW striking plane; Basili and Valensise (1991) propose left-lateral strike-slip faulting on a NW-SE striking plane. As it is pointed out by Basili and Valensise (1991) and by Galadini and Galli (1999), all the solutions are poorly constrained due to the low quality of the seismic dataset and are in conflict with all the available geological, geomorphological and paleoseismological evidence, which suggests that the Fucino basin is actively undergoing deformation related to NW- striking normal faults located at the eastern border of the Plain.
Ward and Valensise (1989) fit the coseismic elevation changes induced by the 13 January 1915 earthquake with pure normal slip on a NW-trending fault plane, constraining their results with field observations of coseismic fault scarps and with kinematic indicators on the bedrock scarp of the Serrone Fault.
These investigators claim that a subordinate lateral component of slip could not be resolved due to the poor quality of the dataset, but also that a substantial component of strike slip, particularly if left-lateral, is rather unlikely. Conversely Amoruso et al. (1998) use a joint inversion of seismic and geodetic data to obtain a solution characterised by a strong sinistral strike-slip component (rake=318û±8û) on a NW-striking, SW-dipping fault whose surface projection corresponds to the Serrone Fault.
Michetti et al. (1996) point out that, according to contemporary descriptions of the coseismic surface ruptures of the 1915 earthquake and to additional modern field surveys, nowhere along the fault scarps it is possible to detect significant coseismic or Holocene lateral displacement. Results of the paleoseismological trenching of the main Holocene fault branches found in the Fucino plain, summarised in Galadini and Galli (1999), confirm that motion is mainly dip-slip with no lateral component.
STRUCTURAL GEOLOGY: according to Piccardi et al. (1999), the most recent generation of slickensides seen on the bedrock fault scarps of the Serrone, Parasano and Ventrino normal faults, have a dextral component of slip and are superimposed on an older set that instead shows sinistral motion. According to their interpretation the right-lateral component of motion is due to counterclockwise rotation of the fault-bounded blocks induced by a Regional left-lateral shear zone, the surface expression of which is represented by the NNW-trending Ovindoli and Sangro-Giovenco faults.
In contrast Galadini and Galli (1999) contend that the kinematic indicators found along the bedrock scarp of the Serrone Fault, that is the southeastern continuation of the San Benedetto-Gioia Fault, are mainly dip-slip.
Piccardi (1995) reports that streams running across the scarp at the base of the Ventrino Fault are offset right-laterally; the slip vector he calculates using the morphological indicators is consistent with that obtained by means of structural analysis. No geomorphic offsets are reported along the San Benedetto-Gioia and Marsicana Highway faults. In fact, there is no consensus on the present activity of the Ventrino Fault, which based on previous fault compilations Galadini and Galli (1999) consider to be an Early Pleistocene fault.
An explanation for the contrasting indications arising from structural, geomorphological and paleoseismological observations, could be that the kinematic indicators found on bedrock scarps are not related to the presently active tectonic regime, and that the bedrock fault scarps have been exhumed by erosional processes. The formation of a white ribbon of fresh limestone fault plane reported formed along the southwestern slope of Mt. Serrone after the 1915 earthquake is suggestive of coseismic reactivation with prevalent dip-slip motion, although it is clearly difficult to fully discriminate between tectonic and gravitational movements.
Two kinematic models have been proposed to explain the recent evolution of the Fucino region. Both of them involve left-lateral displacement on NW- to NNW- striking faults. The model proposed by Galadini (1999) is based on structural, geomorphological and stratigraphic data and predicts strike-slip or oblique-slip left-lateral movements on faults striking N130û to N160û, and pure normal movements on faults striking N125û to E-W. According to this worker the strike-slip deformation accommodates extension on faults whose geometry is inherited and that are not perpendicular to the current extension direction. Galadini and Galli (1999) calculate a NE-SW extension rate across the entire Fucino ranging between 0.6 and 1.0 mm/yr; they assume pure normal slip on the NW-trending fault branches. This calculation is in agreement with that made for the Ovindoli-Pezza Fault by Pantosti et al. (1996), both for direction and for slip rate. In contrast, the model presented by Piccardi et al. (1999) envisions the recent tectonic evolution of the area as governed by a deep-seated sinistral shear zone. According to these workers the N20û±10ûE directed extension, operating at a rate ranging between 1 and 3 mm/yr in the Fucino area, suggests a left-lateral component of motion between Adria and Tyrrhenia along the central Apennines. In this view the NNW-striking structures, such as the Ovindoli and Sangro-Giovenco faults, characterised by a more or less strong left-lateral oblique component of motion, are the main shallow expression of the deep shear zone, while the NW-striking faults terminating on them with a horsetail geometry are secondary features accommodating the left-lateral shearing.
The strike of the Marsicana Highway and the San Benedetto-Gioia normal faults is very similar, within few degrees, to that of the faults belonging to the Upper Sangro Fault Zone and of the Ovindoli Fault, that are considered to have left- lateral oblique and strike-slip kinematics. Detailed morphological and microtopographic analyses of the central part of the Ovindoli-Pezza Fault (D'Addezio et al., 1996) clearly show that, despite the abrupt change in strike of this fault with respect to the general Apennines trend (up to 30û), the kinematics remain prevalently normal, with a maximum horizontal movement not exceeding 30% of the vertical displacement. Trenching along the same branch (D'Addezio et al., 1996) demonstrates that the two most recent paleoearthquakes were characterised by an age of occurrence and extent of deformation similar to those observed for the two most recent surface faulting events identified along the Piano di Pezza segment, and hence that the two faults slipped together. The study of the Ovindoli-Pezza Fault stresses the fact that we should not expect significant changes in the direction of motion along faults characterised by slight changes in strike; and also that all the faults found in the Fucino region presumably accommodate NE-SW oriented extension, similarly to several other large seismogenic faults along the Appennines and in agreement with the indications from seismological data.
Michetti et al. (1996) identify two pre-1915 surface faulting events through trenching along the San Benedetto-Gioia fault. They hypothesise the existence of the youngest of the two (Event A of Low Middle Ages age) by interpreting the attitude of some sedimentary units as due to the presence of a scarp that could be either an erosional feature (i.e. a paleoshoreline of the old Fucino lake) or a coseismic fault scarp. They rule out the first hypothesis because of the absence of shoreline deposits commonly found elsewhere in the Fucino Plain. In contrast, Galadini and Galli (1999) point out that there is no evidence of a Low Middle Ages event in any of the trenches excavated across the same fault branch, and hence interpret the deposition of the sedimentary sequence as being influenced by an erosional scarp rather than by a fault scarp.
If the Low Middle Ages earthquake did occur, we would have seen three earthquakes in the past 1,500 years, with a mean recurrence time of about 500 years, much shorter than the average suggested by older events. A large earthquake (M=6.5-7.0) in the Fucino plain would also have produced in the city of Rome intensity effects similar to those generated by the 1915 event; in contrast current historical catalogues do not report any event during the interval between 800 A.D. and 1394 A.D., indicated by Michetti et al. (1996) as the most probable time span of occurrence of the postulated Low Middle Ages event. If that event really occurred, then the Fucino basin seismogenic source should be regarded as capable of generating large earthquakes with a recurrence interval much shorter than that expected for most other sources of the Italian peninsula.
1) What is the role played by the faults affecting the eastern Fucino margin (Marsicana Highway and San Benedetto-Gioia faults)? Galadini and Galli (1999) regard them as the main branches of a single NW-striking, SW-dipping seismogenic fault characterised by pure normal movement. According to them these faults slip always simultaneously with a quasi-periodic Holocene recurrence time and characteristic displacement per event; conversely the Trasacco and Luco dei Marsi fault branches would move only passively. Piccardi et al. (1999) assign to each of them a vertical slip rate, a recurrence time and a maximum expected magnitude, therefore considering each of them as a potential seismogenic source.
2) Are all the active faults identified in the Fucino Plain area individual seismogenic sources that can generate independent earthquakes? Or are they only the shallow expression of a single deep seismogenic source?
3) What are the relationships among the Fucino seismogenic source and its northern (the Ovindoli-Pezza Fault) and southern (the Sangro Valley Fault zone) counterparts? Are these three distinct seismogenic sources always generating earthquakes separately? Paleoseismological studies show that the age of occurence of paleoearthquakes along the Fucino and Ovindoli-Pezza faults cannot be correlated; we lack seismological and paleoseismological information for the area south of the Fucino, even if slope deposits younger than 27,000 years are reported displaced by NW-trending faults near Pescasseroli (Galadini et al., 1999); in this view the lack of historical seismicity points to the presence of a potential seismic gap in the Upper Sangro River Valley.
4) Has the Fucino basin seismogenic source a characteristic behaviour? The data obtained from trenching at several sites across different fault branches are not conclusive. Only at one site on the Marsicana Highway Fault it was possible to observe a costant displacement per event and calculate a quasi-periodic recurrence time.
5) When did the penultimate displacement event occur? Galadini and Galli (1996) hypothesise that it may have occurred either in 508 A.D. or 618 A.D., while Michetti et al. (1996) suggest that it may have occurred in 801 A.D..
Based on a survey performed immediately after the occurrence of the 1915 earthquake, this investigator describes and maps in detail the formation of several fractures in the plain and of a "voragine" running along the south- western side of Monte Serrone and downdropping the south-western side; he interprets these features as due to differential compaction between the sediments of the old lake and the slope detritus and the bedrock.
Gasparini et al. (1985)
These workers provide a fault plane solution of the January 13, 1915 earthquake by analysing polarities of the first motions of P-wave arrivals recorded at 23 seismic stations all over the word. The best solution given by the seismic inversion is a N298û striking, 39û dipping plane, characterised by almost pure left lateral slip.
Basili and Valensise (1986)
These investigators relocate the 1915 earthquake epicentre on the basis of the arrival time of the first motion in several Italian and European seismic stations; they calculate the fault plane solution of the event, giving a left lateral faulting mechanism with a little dip component on a N138û striking, 68¡ dipping best fitting plane.
Based on the analysis of aerial photos, this investigator analyses several faults cutting the sediments of the old Fucino lake and the Holocene-Upper Pleistocene terraces surrounding it. In the middle of the plain the faults are highlighted by linear contacts of soils characterised by different lithologies and moisture contents; closer to the basin margin they correspond to fault scarps displacing the terraces. According to the author these faults are of Holocene age and may have moved in historical time, although only few of them correspond to the scarps generated by the 1915 earthquake.
Bonasia et al. (1986)
These investigators use the geodetic levelling dataset formed by a pre-event survey performed in 1862 and by a post-event survey, in order to analyse static surface displacement associated with the January 13, 1915 earthquake, and give the fault plane geometry with a direct modelling. The authors provide as best result of their modelling a N112ûE striking, 70ûSW dipping, 18 km- long plane, located about 7 km SW of the Serrone Fault, in contrast with geological evidence of surface faulting along this fault.
Serva et al. (1986) and Blumetti et al. (1987)
Through a reinterpretation of the work of Oddone, achieved by interviewing several eye-witness of the 1915 earthquake, and based on a geomorphological survey carried out in the north-eastern area of the Fucino plain, these workers reconstruct and describe in detail the geometry of the fault scarps related to that event. According to them the surface faulting consisted of two sub- parallel about 10 km long fractures formed between the villages of Sperone and S. Benedetto dei Marsi and between the south-western side of Mt.
Parasano and the village of Cerchio, downthrowing the south-western side.
The authors also describe the stratigraphic succession exposed in a gravel quarry present in the San Veneziano area (south-eastern sector of the Plain) that shows the evidence of at least three surface faulting events in the last 13,000 years, post the deposition of the gravel.
Through a geomorphological study of the Fucino Plain and of correlations between ages of rock collapses in caves and ages of morphological features, this paper supplies a chronology of the inception of several fault scarps, present in the eastern side of the plain and near the town of Avezzano, and of the earthquakes related to them. In fact on the basis of morphological considerations he considers each scarp related to a single earthquake; he recognises and dates four events, three of pre-historic age (between 18-20,000 and 13-14,000 years B.P.; between 5,500 and 5,000 years B.P.; nearly 3,100 years B.P.) and one of Late Roman or Middle Ages age (484 or 508 A.D.?), related to a fault scarp cutting a Roman age channel.
Ward and Valensise (1989) and Valensise(1989)
These investigators model static surface displacement related to the 1915 earthquake using the geodetic dataset formed by the pre-event and the post-event surveys, and model the geometry of the related fault using uniform and variable slip planar dislocation models. These analyses reveal a two lobed slip pattern separated by a central zone of low moment release, in agreement with geomorphological observations of landform features and of the fault scarps related to that event, with a N135û striking, 63¡ dipping, 24 km long and 15 km wide best fitting fault plane.
According to the authors the present shape of the Fucino basin could represent the long term result of repeated events that occurred on the 1915 Fault and on the Tre Monti Fault.
Quattrocchi et al. (1990)
This paper describes and measures natural degassation phenomena (mostly methane) observed at various locations along the southern side of the Plain.
The phenomena are interpreted as due to the degassation of small natural reservoirs due to the especially dry season and to the subsequent increase in rock premeability. However, their locations align along the Trasacco Fault, providing indirect evidence for its activity.
Brunamonte et al. (1991)
These investigators describe three exploratory trenches opened across one of the two fault scarps related to the 1915 earthquake, near the village of San Benedetto dei Marsi, and on the basis of the analyses of the stratigraphic sequence and of radiocarbon dating of samples recognise two paleoearthquakes referred to VI-IX century a.D. and X-XIV century a.D.
respectively.According to them these events may have been of the same magnitude as the 1915 earthquake.
Galadini and Messina (1994)
This paper describes the geological and tectonic evolution of the Fucino Basin, by means of stratigraphic, geomorphological and structural analyses.
According to these authors the ENE-trending Tre Monti fault zone, that borders the northern margin of the Fucino plain, was responsible for the first phase of opening of the basin during the Pliocene; later in the Upper Pliocene and Quaternary times the evolution of the basin was mostly influenced by the NW-trending faults of the eastern margin of the plain (San Benedetto dei Marsi-Gioia dei Marsi and Marsicana Highway faults), whose activity continued up to the present.
She estimates a long term-slip rate for the Fucino fault of about 1.7 mm/y on the basis of the uplifted "Pescina terrace", located on the footwall of the fault. Modeling of this presumably 700 ka feature suggests that 1,200 m of fault slip are needed to uplift it and tilt it from horizontal to slighlty NE-dipping.
Galadini et al. (1995).
These investigators recognise in the stratigraphic succession exposed at three distinct sites several surface faulting events: two trenches digged for pipeline supplies to the south of the village of Cerchio, and cutting the 1915 earthquake fault scarp, showed the evidence of four palaeoearhquakes, one of which occurred before 19100 yr B.P., two of them after this date and the last after 2800 yr B.P.; hand boring performed across the 1915 scarp near the village of Venere gave the evidence of an Upper Middle Ages age event; at last in an open quarry near the village of Casali d'Aschi they recognize four events with age ranging between 2,800 yr B.P. to the Middle Ages.
Galadini and Galli (1996)
By means of trenching at the intersection between a Roman-age channel and the Trasacco Fault, these investigators supply evidence for the occurrence of a surface faulting earthquake that displaced the channel excavated for drainage purpose by the Romans during the I-II century a.D.; on the basis of geological, historical and archaeological data they demonstrate that this event occured between VI and IX century a.D., and that it can be likely related to the 508 a.D.
earthquake that was responsible for severe damages to the Colosseum in Rome or in second hand to the 618 a.D. earthquake. They also hypothyze that this event could be considered as a twin of the 1915 earthquake, as paleoseismological trenching analyses in the eastern side of the Fucino Plain near the village of San Benedetto dei Marsi showed the presence of a surface faulting event of high Medieval age characterized by displacements with similar offsets along the same fault scarps of the 1915 event, giving then a first estimation of recurrence interval of about 1300-1400 years, for this seismogenic fault.
Galadini et al. (1996)
On the basis of paleoseismological analyses conducted at four sites along the Trasacco Fault, these investigators describe seven surface faulting earthquakes and give the age of the last three occurred before the 1915 event (6000-5000 years B.P.; 3400-3200 years B.P.; while the third may be related to the 508 a.D.
earthquake); they propose a recurrence time of about 1,500-2,000 years for events similar to the 1915 earthquake, and calculate the slip rate, which varies between 0.36 mm/yr and 0.15 mm/yr for the past 7,000 years moving northward along the Trasacco Fault.
Michetti et al. (1996)
This paper describes the results of trenching performed near the village of San Benedetto dei Marsi on one of the coseismic scarps related to the 1915 earthquake. They recognise at least two surface faulting events besides the 1915 one, almost characterized by similar coseismic displacement, and constrain their age with archaeogical materials and by radiocarbon dating of soil samples; according to them event B occurred between 550 and 885 a.D., and very likely it can be related to the 801 a.D. earthquake; they suggest the best estimate age range of event A is between the end of X century a.D. and 1349 a.D., then relating it to an earthquake of medieval age not recorded in the historical catalog. On the basis of the coseismic displacement registered on the fault scarp, they also hypothize that these two events were of the same magnitude of the 1915 earthquake.
Amoruso et al. (1998)
These investigators present a nonlinear inversion to determine the source parameters of the 13 Jan. 1915 earthquake, utilizing as input data both vertical displacement, deriving from the geodetic dataset formed by the pre-event (1862) and the post-event (1917) surveys, and seismic data (P wave first motions). The best fitting model geometry obtained with the assumption of uniform slip on a planar fault in a homogeneous elastic half-space, is a N143û striking, 55û dipping, 35 km-long fault plane, whose surface projection match with the serrone Fault. The authors point out that the right lateral component of slip used in the modeling is in agreement with structural, geological and geomorphological observations made on active faults found in the Fucino region.
This paper uses a vast integrated dataset to describe the Quaternary kinematic evolution of the Fucino plain and of other tectonic basins of the central Apennines. He shows that starting from the beginning of the Middle Pleistocene some NW-trending normal faults interrupted their activity, while others NW- to NNW-trending faults began to be characterised by left-lateral to oblique sinistral slip (N160û-striking portion of the Ovindoli- Pezza Fault and Upper Sangro Fault Zone); in the same time pure normal movements are observed along N125 to EW-trending faults (Velino, Marsicana Highway and San Benedetto-Gioia faults). According to his interpretation the change in fault kinematics may have been caused by the evolution of the northern Apennines arc characterized by progressive eastward migration of the compressive front followed by the extension, while the southeastern margin remained fixed.
Galadini and Galli (1999)
This paper present a summary of all the palaeoseismological data available for the Fucino Plain. The Fucino seismogenic source is intepreted to be formed by a primary fault system, consisting in the Marsicana Highway Fault and in the San benedetto-Gioia Fault, bordering the eastern side of the basin, with associated secondary faults, Trasacco Fault and Luco dei Marsi Fault, found in the middle of the plain and close to the western side of the margin.
Following this interpretation large magnitude earthquakes generated by the primary fault system would induce passive slip on the secondary faults.
Previous workers by means of paleosesimological analysis in different sites of the Fucino plain identified ten surface faulting events in the last 33,000 years, seven of which occured during the Holocene. Dating of geological and archeological features allowed a chronology and a calculation of the average recurrence interval for surface faulting earthquakes, that resulted ranging between 1,400 and 2,600 years. On the basis of the observed offset vertical slip rate were computed for individual fault branches, ranging between 0.24 and 0.5 mm/yr for the primary fault system, and between 0.1 and 0.27 mm/yr for the secondary faults. Galadini and Galli on the basis of the slip rate data collected calculate a minimum value for the extension rate across the Fucino basin ranging between 0.6 and 1 mm/yr in a NE-SW direction.
Piccardi et al. (1999)
These investigators carry out a detailed geomorphic and structural study of the faults found around the Fucino basin and give an estimation of their vertical slip rates.The authors consider the fault sytem responsible for present day active deformation in the Fucino plain to be formed by four NW- trending parallel faults (Serrone, Parasano, Muricci and Ventrino faults) found along the northeastern margin of the basin, and splying with a right- stepping horsetail geometry from the NNW-trending oblique left-lateral Sangro-Giovenco Fault found more to the east; to the north these faults are truncated by the ENE-trending oblique right-lateral Tre Monti Fault.
Following geomorphic analysis of the heights of the fault scarps, and assuming that the measured offset postdate the end of periglacial abrasion (14±4 kyr BP), the authors provide estimate of vertical slip rates, maximum expected magnitudes and average recurrence times for each individual fault branch; the ranges of the long term vertical slip rates calculated for the Serrone, Parasano and Ventrino faults result to be between 0.5 and 1.4, 0.5 and 1.0, and 0.3 and 1.1 mm/yr respectively. On the basis of measurements of the most recent slip indicators found on the bedrock scarplets and of geomorphological considerations the authors show that the NW-trending faults are characterized by a more or less component of right lateral oblique slip, and hypothesize that they accomodate extension in a N20°E direction; the calculated extension rate across the Fucino area, considering average fault geometry and throw rates, would be between 1 and 3 mm/yr.