2Dmodellingoflarge-scaleplatformmargincollapses … · 2018. 11. 12. · E-mail address: [email protected](G.Rusciadelli). PALAEO314013-10-03 Palaeogeography,Palaeoclimatology,Palaeoecology200

2D modelling of large-scale platform margin collapsesalong an ancient carbonate platform edge(Maiella Mt., Central Apennines, Italy):

geological model and conceptual framework

Giovanni Rusciadelli �, Nicola Sciarra, Massimo MangifestaDepartment of Earth Sciences, University ‘G. D’Annunzio’ of Chieti, 66013 Chieti Scalo, Italy

Received 20 December 2001; accepted 21 March 2003

Abstract

Results of numerical simulations of large-scale platform margin collapses along the edge of the Cretaceousplatform of the Maiella (Central Apennines) are presented, adding a new contribution to the understanding of the roleand quantification of the various factors that control triggering of these phenomena and the production of brecciasand megabreccias. For this purpose, a sophisticated numerical code, widely used in recent contexts to simulate overallmass movements and deep gravity-driven deformation, is applied for the first time to an ancient system on a scale ofseveral kilometres. The modelling carried out is a simplified attempt to understand a very complex phenomenonproduced over a very long time-span. It tests whether the perturbing causes in the model, acting together orseparately, are able to activate such large-scale collapse, and helps to understand the effects of perturbing causes andtheir variations. In particular, modelling was carried out considering both static (variations in sea level) and dynamic(seismic events without and with fault formation) conditions and combinations of them. Variations in sea level alonedo not produce all the effects observed but may certainly be a predisposing element. Seismic shock simulations showthat fracturing and shear stresses are concentrated on the platform margin, representing the most sensitive andvulnerable area. Earthquake swarms associated with an active fault are the most likely means of triggering large-scaleplatform margin collapses, such as those occurring during the Early Cretaceous in the Maiella.1 2003 Elsevier B.V. All rights reserved.

Keywords: large-scale platform margin collapses; mechanical modelling; Maiella Mountain; mid-Cretaceous platform

1. Introduction

The statement by Anderson and Crerar (1993)

T‘models do not represent reality, but our idea ofreality’T expresses perfectly one of the main limitsin the attempt to reproduce faithfully a geologicalphenomenon through modelling. The quantitativeapproach, however, has contributed and still con-tributes considerably to the progress of ourknowledge about the complex relationships thatgovern the development of depositional systems.

0031-0182 / 03 / $ ^ see front matter 1 2003 Elsevier B.V. All rights reserved.doi:10.1016/S0031-0182(03)00453-X

* Corresponding author. Tel. : +39-871-355-6579;Fax: +39-871-355-6454.

E-mail address: [email protected] (G. Rusciadelli).

PALAEO 3140 13-10-03

Palaeogeography, Palaeoclimatology, Palaeoecology 200 (2003) 245^262

www.elsevier.com/locate/palaeo

mailto:[email protected]

http://www.elsevier.com/locate/palaeo

The application of numerical simulation modelsin stratigraphy and sedimentology has, for exam-ple, enabled the evaluation and better understand-ing of geometric stratal patterns (e.g. Kooi andCloetingh, 1989), the architecture of carbonateplatforms (e.g. Bosence and Waltham, 1990), theirproduction and growth rates and the architectureof depositional sequences (e.g. Pitman and Golov-chenko, 1991). It also gives insights on the role ofchanges in sea level in the organisation of succes-sions and the relations between the various fac-tors that modify the accommodation space, suchas tectonic subsidence, eustasy, sediment load andcompaction (e.g. Bowman and Vail, 1999), etc.Nevertheless, little attention has been paid, interms of numerical simulation, to large-scale mar-gin collapses occurring on the edges of carbonateplatforms, both recent and ancient, to their sedi-mentary products (e.g. breccias and megabreccias)and inherited morphologies. The understanding ofthese events and their triggering mechanisms haveimportant implications for sequence stratigraphyinterpretation (importance of onlap geometriesnear indented morphologies, the position of thebreccias and megabreccias within the systemstracts) and for hydrocarbon research (spatial^temporal predictability, areas and volumes of po-tential reservoirs).For the mechanisms that trigger large-scale

margin collapses along the platform margin, thereare largely conceptual or qualitative models in ex-istence, based on present-day and ancient plat-form systems. In Recent carbonate systems, themost established cause that triggers these phe-nomena is the in£uence of seismic shocks associ-ated with movements along fracture zones (Hineet al., 1992). Other mechanisms such as sedimentload, storm waves, tsunamis, lowering of relativesea level, bioerosion, fracturing, bottom currents,etc., can cause or contribute to these events (Free-man-Lynde and Ryan, 1985; Schlager and Cam-ber, 1986; Paull et al., 1990). In ancient carbonatesystems, these phenomena and the consequentproduction of breccias and megabreccias are as-sociated, directly or indirectly, with the activity ofsynsedimentary faults during tectonic phases (Ber-noulli, 1964; Castellarin et al., 1978; Payros et al.,1999) and to variations in relative sea level (Sarg,

1988; Jacquin et al., 1991; Spence and Tucker,1997; Bosellini, 1998; among others).In this work, some results from the numerical

simulation of large-scale margin collapses occur-ring along the edge of an ancient carbonate plat-form are presented, giving a new contribution tothe understanding of the role and quanti¢cationof the various factors that control the triggeringof these phenomena and the production of brec-cias and megabreccias. For this purpose, a sophis-ticated numerical code (FLAC, 2000) was used, tosimulate the overall behaviour of the rock mass,in both static and dynamic conditions. This nu-merical code, widely used in recent contexts tosimulate mass movements and deep gravity-drivendeformation, in order to solve geotechnical andmechanical engineering problems (Dawson andRoth, 1999; Cala and Flisiak, 2001; Sciarra andCalista, 2001), is applied to an ancient system ona scale of several kilometres. The geological exam-ple modelled is the Maiella platform (CentralApennines), where one of the most spectacularexamples of ancient carbonate platforms charac-terised by large-scale margin collapses on the seis-mic scale, is found (Morsilli and Rusciadelli,2000; Morsilli et al., 2000, and in press). Themodelling of this system was exceedingly prob-lematic and complex, especially during the param-eterisation phase. The numerical code requires theknowledge of some parameters which cannot bedirectly obtained from an ancient system, butmust necessarily be deduced from recent, analo-gous systems, which are most similar to that an-alysed.The main purpose of this work is to illustrate

and discuss the geological situation and theadopted numerical model, the conceptual frame-work of the numerical code and the limits of itsapplication, variables and about how the resultsof the modelling are expressed.

2. The geological model

2.1. The relict morphology of the platform margin

The geological model used for the simulationderives from a Cretaceous carbonate platform lo-

PALAEO 3140 13-10-03

G. Rusciadelli et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 200 (2003) 245^262246

cated along the northern edge of the ApulianPlatform and magni¢cently outcropping in Maiel-la, in the Central Apennines (Fig. 1). In this area,in spite of Mio^Pliocene tectonic activity, theoriginal relationships between Cretaceous slopeand platform deposits are perfectly preservedand observable along natural sections at the seis-mic scale. These are de¢ned by a 1000-m-deeppalaeoescarpment marked by the abrupt contactbetween Lower and Upper Cretaceous platformfacies and Upper Cretaceous slope deposits (Cres-centi et al., 1969; Accarie et al., 1986; Vecsei,1991; Eberli et al., 1993; Morsilli et al., 2002).The geometric reconstruction of this contact intwo and three dimensions (Morsilli and Ruscia-delli, 2000) has shown that the edge of this sectorof the Apulian Platform was characterised by arather complex physiography, de¢ned by the al-ternation between prominent stable tracts and in-dented tracts cut into the platform facies of theLower Cretaceous (Fig. 2). This physiographyseems to have originated in the wake of an impos-ing phase of collapses of the platform margin,

which occurred between the Albian and the EarlyCenomanian, as can be deduced from the age ofthe breccias and megabreccias intercalated in thebase-of-slope to basinal successions north of theplatform (Morsilli et al., 2000, 2002). The in-dented plan geometry and asymptotic pro¢le ofthe detachment surface and the type of depositsassociated with them (poorly sorted megabreccias,with sub-rounded clasts and abundant matrix)suggest that the collapses are rock-avalanche-type phenomena.

2.2. The pre-collapse depositional pro¢le

For the simulation of the gravity-driven collap-ses, it was necessary to de¢ne the type and thegeometry of the platform margin previous to thecollapse phase. The stable, prominent portions ofthe margin are the most useful in the reconstruc-tion of the pre-collapse depositional pro¢le, sincethe original features are better preserved com-pared with the indented areas (Fig. 3). In spiteof this, these sectors also show signs of erosion

Fig. 1. Geological map of the Central Apennines showing the studied area and the broad distribution of Mesozoic depositionalenvironments (platform, margin to proximal slope and distal slope to basin) of the Lazio-Abruzzi Platform and the main struc-tural elements.

PALAEO 3140 13-10-03

G. Rusciadelli et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 200 (2003) 245^262 247

connected with the formation of the palaeoescarp-ment. The abrupt interruption northwards ofLower Cretaceous margin facies suggests, infact, that the platform margin extended beyondthe present position of the palaeoescarpment.The presence of proximal slope clinostrati¢ed de-posits near the base of the Selva Romana Valleyoutcrops (Fig. 4), provides a key element for re-constructing the geometry of the depositional pro-¢le of the platform and to hypothesise the originalextent of the areas eroded during the collapses.The clinoforms observed represent the basinwardcontinuation of the collapsed margin portion,originally located along the southern £ank ofthe Selva Romana Valley. In the outcrops belowCima delle Murelle, a little further south of SelvaRomana Valley, internal platform and margin fa-cies are found. They represent the landwardequivalents of the massive clinoforms. The mate-rial removed, which constitutes the clinoforms,must have accumulated progressively along theplatform £ank, corresponding to a slope thatwas already inclined and inherited from the devel-opment and growth of depositional systems dur-

ing previous stages. The presence of a fairly in-clined slope, in front of a pre-existing basindepression, would have limited the progradationof the platform and favoured its aggradation andthe formation of clinoforms.Another relevant aspect for constructing the

pre-collapse geometric model is the height of theslope. This may be evaluated by considering therelationship between the platform £ank and theposition of the slope to basin transition with re-spect to the platform edge. The latter is pin-pointed by the point where the morphologicalgradient of the platform depositional pro¢le con-sistently increases (Vanney and Stanley, 1983),while the slope to basin transition is shown atthe point where the inclination of the slope con-sistently decreases. At Maiella, this is not directlyobservable since it is not outcropping; neverthe-less an estimate of its distance from the platformedge can be deduced from seismic surveys of theburied and non-deformed margin of the ApulianPlatform o¡shore, in the Adriatic. Several seismicsurveys recording platform to basin transition,not a¡ected by Cretaceous faults and by erosional

Fig. 2. Three-dimensional reconstruction of the palaeoescarpment during the Late Cretaceous. Note the complex morphology ofthe palaeoescarpment de¢ned by the alternation between indented and prominent areas (modi¢ed after Morsilli and Rusciadelli,2000).

PALAEO 3140 13-10-03


to by-pass slopes, have been investigated for thispurpose. In this area, the slope to basin transitionis marked by a break of slope and by the presenceof sharper, more regular seismic re£ectors, indi-cating the change from slope to base-of-slope de-posits. This transition takes place at a distance

varying between 4 and 6 km from the platformedge. With an initial acclivity of around 30‡,which tends to decrease rapidly as it movesaway from the platform edge, the height of theslope, at the transition with basinal deposits,reaches values of about 1500 m and 2000 m, cor-responding to the value of the bathymetry, with-out correction for compaction (Fig. 5). Geomet-rical and depositional features similar to thosereconstructed for the Lower Cretaceous pre-col-lapse pro¢le of the Maiella platform can be rec-ognised in numerous examples of platform to ba-sin transition, documented along the buriedmargin of the Friuli Platform (cf. Cati et al.,1987; Casero et al., 1990), the Dalmatian Plat-form (Casero et al., 1990) and the northern sec-tors of the Apulian Platform (Gargano area)(Morsilli and Borsellini, 1997).Fig. 6 represents the reconstruction of the pos-

sible geometry of the pre-collapse depositionalpro¢le of the Maiella platform at the end of theEarly Cretaceous, based on the interpretation of¢eld and subsurface data. On the basis of thisreconstruction it can be hypothesised that theedge of the platform extended around a further

Fig. 3. Cross section along a prominent sector of the platform margin, north-west of Cima delle Murelle. The Lower Cretaceoussuccession, underneath the palaeoescarpment, shows facies associations dominated by rudstones and grainstones with orbitolines,rudists and corals, representative of Aptian^Albian margin environments (modi¢ed after Morsilli et al., 2002).

Fig. 4. Lower Cretaceous northward dipping (around 30‡)clinoforms at the base of the Selva Romana Valley outcrops.Clinoforms are composed of alternating bio-lithoclastic grain-stones and rudstones and bio-lithoclastic breccias, in mediumto thick layers.

PALAEO 3140 13-10-03


1 km north compared with the present position ofthe palaeoescarpment, and that the maximumthickness of the sediments eroded following mar-gin collapses was around 300 m.

3. The numerical code

The numerical code used (FLAC, 2000) is atwo-dimensional (2D) ¢nite-di¡erence method of

Fig. 5. Seismic survey and linedrawing (a,b) of the buried and undeformed Apulia platform margin in the Adriatic o¡shore.Pl-Ple: Plio^Pleistocene; M: Messinian; P: Paleogene; UK: Upper Cretaceous; LK: Lower Cretaceous; J: Jurassic. (c) Recon-struction of the depositional pro¢le and estimation of the palaeobathimetry in the sector north of the Murelle.

PALAEO 3140 13-10-03


numerical analysis for calculations of continuummechanics. The system is represented by elementsin a grid constructed by the user in the shape ofthe object to be modelled. Each element behavesin accordance with a prescribed stress-deforma-tion law, linear or otherwise, in response to theforces applied or to the constraints imposed.FLAC is based on a ‘Lagrangian’ numericalscheme that adapts well to modelling large-scaledeformation and collapse (Sciarra, 2000; Sciarraand Calista, 2001).The general analysis consists ¢rst in an overall

re-equilibrium of the system and then in the studyof the rupture conditions. If the system is initiallyunstable, the numerical code is inappropriate tocarry out any types of analyses. The re-equilibri-

um process is thus necessary to verify the initialstress-strain conditions of the model and the con-gruence with the geometric pro¢le. Hydraulic con-ditions, represented by £uid £ow vectors, and thecalculus of the pore pressures must be knownduring this phase. This phase proceeds indepen-dently of any mechanical e¡ects. At the end of theanalysis of the re-equilibrium process it is possibleto represent the displacement and the deformationvelocity vectors at each node of the meshes. Inthis way the behaviour of the system can be ob-served with reference to its geometrical, physicaland mechanical peculiarities. After the re-equilib-rium process, the stability analysis allows to ex-plore limit conditions of equilibrium by varyinginternal (physical and mechanical parameters of

Fig. 6. Reconstructed geometry of the pre-collapse depositional pro¢le of the Maiella platform at the end of the Early Creta-ceous. Following this reconstruction, the pre-collapse depositional pro¢le corresponds to a depositional and essentially aggradingor weakly prograding platform rimmed by an escarpment inclined about 25^30‡.

PALAEO 3140 13-10-03


the rock, pore pressures, etc.) and external char-acteristics (application of loads, seismic events,etc.). The programme is able to analyse localand general stability of the system for each varia-tion imposed.

4. The geomechanical model

A grid with triangular and quadrangularmeshes that de¢nes homogeneous portions of ma-terial represents the mechanical situation used(Fig. 7). The average size adopted for the gridmeshes (40 m) represents the optimum value toreduce both the e¡ects of scale due to the modeldimensions (2.5U4.5 km) and the calculationtimes which were necessarily long in this case.Adopting shorter dimensions for the single

meshes produced no appreciable di¡erences inthe results.The values for the physical and mechanical pa-

rameters relating to the two main depositionalsystems (inner platform and platform margin) ofthe model are ¢xed in the centre of gravity of themeshes. Di¡erent physical and mechanical param-eters were assigned to the two depositional envi-ronments. For both environments Druker-Pra-ger’s yielding criterion was used (Chen and Han,1988), which is very suitable for elasto-plastic be-haviour in complex situations involving large-scale deformations.The di¡erent choice of mechanical parameters

was based on criteria concerning the di¡erentgrain sizes, texture and physical distribution ofthe materials. In particular, materials with lowermechanical resistance were associated with slope

Fig. 7. Scheme of the model with the mechanical conditions of constraint imposed on the borders. Grid is made up of triangularand quadrangular meshes with an average size of 40 m.

Table 1Values of physical and mechanical properties used in the calculation

Parameters Internal platform facies Margin to slope facies

Q (unit volume weight) ^ (kN/m3) 23.5 18.5cP (cohesion) ^ (Pa) 5.0W104 4.0W104

PP (friction angle) ^ (‡) 45 35E (elastic modulus) ^ (Pa) E ¼ Eo

c0h

c a

� �Kc

0h ¼ K Wc 0

v Eo = 2.2W109 Eo = 2.0W109

Tensile strength ^ (Pa) c tmax ¼

c0

tgP5.0W104 3.6W104

Poisson’s ratio 0.33 0.30Bulk modulus ^ (Pa) 1.83W109 1.66W109

Shear modulus ^ (Pa) 4.58W108 4.16W108

n (porosity) ^ (%) 20 25

PALAEO 3140 13-10-03


environments. Table 1 shows the values used inthe numerical calculation. These were taken fromprevious authors (Ciancetti and Sciarra, 1987),who derived them from studies of materials oflocal formations (Calcare Massiccio) whose phys-ical and mechanical characteristics are similar tothose of the Maiella platform during the collapsephases.Discontinuities in£uence parameter values fa-

vouring conditions of instability and increasingthe chances of triggering collapses. In line withthe scale of the studied model, the presence ofdiscontinuities is essentially associated with thebedding, fracturing, faults and with the e¡ects ofundercutting phenomena due to sea level changes.Long-term emersions of the platform, associatedwith the development of karstic forms, may havecontributed to the formation of discontinuities.Excluding palaeo-karst forms related to the peri-odic, brief emersions of the platform top, the onlykarstic event of a size that would favour the for-mation of discontinuities able to weaken the massof rock is the mid-Cretaceous emersion. At Maiel-la, as in the greater part of the peri-Adriatic Cre-taceous platforms, a long-term hiatus, varioustypes of karstic forms and the formation of baux-itic soils are associated with this emersion event.The many studies carried out on mid-Creta-

ceous peri-Mediterranean platforms in plate inte-rior settings (D’Argenio and Mindszenty, 1991)have revealed the presence of karstic forms suchas karren, sinkholes, cavities and dissolution brec-cias, generally ¢lled with bauxitic soils (D’Argenioand Mindszenty, 1995), and palaeo-reliefs withdi¡erences in level up to a maximum of 40 m(D’Argenio et al., 1987). These karstic forms,and especially bauxites, are found fairly distantfrom the edge of the platform, as is suggestedby facies associations below the discontinuity.Their formation, essentially linked to dissolutionphenomena through the e¡ect of interaction be-tween meteoric waters and emerged carbonate re-liefs, is substantially di¡erent from that of karsticforms that characterise the coastal zones. In areasnear the edge of the platform, the high porosity ofthe deposits favours water in¢ltration and circu-lation, generally inhibiting the development ofkarstic cavities, especially in autogenic systems

(Ford and Williams, 1996) such as isolated car-bonate platforms. Nevertheless, karstic phenom-ena in coastal areas are also linked to the presenceof the so-called mixed corrosion zone. This zonecorresponds to a transitional band between thefreshwater layer and the more dense seawaterlayer. Following oscillations of the sea level, thiscorrosion zone moves vertically favouring the de-velopment of a considerable zone of dissolution(Ford and Williams, 1996).To evaluate the in£uence of karstic phenomena

on the formation of discontinuities on the edge ofthe platform, the possible geometric distributionof the water table in the Maiella platform wasreconstructed during the mid-Cretaceous emer-sion. Considering the Ghyben-Herzberg law andthe sea level position respectively at 340 and 380m from the emerging top of the platform, it canbe observed that the maximum depth at whichkarstic discontinuities associated with the £uctua-tion of the transition zone between fresh and saltwater could develop was around 150 m. In theabsence of deeper karstic phenomena linked toprevious phases of emersion, the mid-Cretaceousone causes instability only in the uppermost por-tion of the model, through structures whose depthis completely insu⁄cient to destabilise a portionof margin equal to that collapsed. On the basis ofthese elements, karstic phenomena cannot be con-sidered a factor of deep destabilisation and do nottherefore represent, in our opinion, one of thecauses triggering large-scale collapses along theplatform margin. Karstic phenomena, however,may have played an important role in the forma-tion of minor structures that are at present ob-servable along the palaeoescarpment such asnotches, or may have enhanced surface decay pro-cesses of the rock mass in tension fractures. Inaddition, by reducing the volume of the massand therefore the unit weight of the material(quanti¢able by varying the average porosity ofthe materials inside the meshes), karstic phenom-ena act by reducing the load at the top of theslope, thus increasing stability rather than insta-bility.As far as fracturing is concerned we can state

that this is not relevant in the initial conditionspreceding the collapse phase, since the platform

PALAEO 3140 13-10-03


has not yet been subject to phenomena able toproduce the relevant and several hundred metresdeep fractures. For bedding, it was possible tobuild the model so that the mechanical disconti-nuities associated with them were represented bythe con¢guration of the meshes. What has beenstated above for karstic phenomena is also truefor the undercutting linked to variations in sealevel and to wave action, which could cause localinstability in the emerged part of the platform.Another element relating to the model de¢ni-

tion is the knowledge of the system of constraintsimposed on the model. In Fig. 7, the mechanicalconditions of constraints imposed on the bordersof the model are shown. They allow only verticalmovement of the sides and no movement of thelower edge.

5. The conceptual framework

The modelling carried out must be understoodas a simpli¢ed attempt to understand a very com-plex phenomenon produced over a very long time-span (from 3 Ma to 10 Ma). This simpli¢cation isnecessary since it is not possible to evaluate theevolution of physico-mechanical parameters andthe seasonal variations of the hydrological re-gimes of a system of over 100 Ma; besides, noarithmetical algorithm is able to simulate the pro-gressive e¡ects of an environmental degradation,which has been going on for millions of years.However, what is interesting is if the perturbingcauses of the model, acting together or separately,are able to activate such a major phenomenon ofcollapse independently of the time required. Inany case, the analysis was carried out above allto try to understand the e¡ects induced by eachperturbing cause and by variations of these. Inparticular, fundamental importance was given tothe variations in sea level and to the presence ofseismic events. A magnitude of 6‡ was allocatedto these seismic events and they were applied sev-eral times for each variation in sea level. They arepurely indicative and served to clarify only thee¡ects on the general behaviour of the model,since the number and the real intensity of earth-quakes that occurred are not known. Regarding

sea level, analyses were carried out starting from acompletely submerged system with subsequentlowering of the sea level up to 3100 m followedby an overall rising phase.In terms of the results expected from the model,

the progress of the deformation of the meshesshould ¢rst of all be analysed to verify if move-ments conform to the present situation of theslope. The deformation is obtained by superim-posing both the static and the dynamic e¡ectsfor each calculation step. Another important ele-ment is the progress of the increases in shearstrain in parts of the model. The size of the modelis such that, especially in seismic periods, ruptureplanes independent of each other may be pro-duced. To recognise these alignments the zonesof di¡erent shear strains are shown, which arealso zones of possible triggering of lines of frac-turing families. From each run of the programme,we are able to obtain elements that should becritically evaluated and justi¢ed. Although resultsoften conform to the input data, they may notnecessarily represent the phenomenon exhaus-tively. Congruence is veri¢ed through the inertialbalancing of the system of stresses internal andexternal to the model (gravity, hydraulic forces,restraining reactions, loads) and if it is notreached it would mean either an incorrect valueof input data or the rupture of the model.

6. Results of the analysis

The ¢rst results obtained refer to the hypothesisof variations in sea level for an initial homogene-ous, intact model without fractures, faults or seis-mic events (quasi-static conditions). These varia-tions involved the fall and progressive rise in sealevel, in steps of 50 m. Fig. 8 shows the graphsrelating to the horizontal displacements of thegrid with bathymetric values equal to 0, 3100 mand 350 m (rising phase from 3100 m) of the sealevel. Phenomena of localised breaking, a¡ectingto the ¢rst few tens of metres depth, are producedalong the slope because of the weakening subse-quent to the fall in sea level and the action of newregimes of pore pressure. Shear strains diagramshows that maximal values are concentrated on

PALAEO 3140 13-10-03


the platform margin, which is therefore subject toconsiderable weakening (Fig. 9). During the runof the programme, areas of progressively largershear strains are generated on the platform aswell. In spite of this, sea level changes do notproduce exhaustively the e¡ects attended from

¢eld reconstruction. However, they may certainlyrepresent a predisposing element to promote in-stabilities in the platform margin sector.For dynamic analyses, an ideal accelerogram

for horizontal-component earthquakes of magni-tude 6 (Fig. 10) was used, applied to the base of

0.00E+00

5.00E-01

1.00E+00

1.50E+00

2.00E+00

2.50E+00

3.00E+00

sea level 0.00 m

0.000 1000 2000 3000 4000 5000(m)

500

1500

2500

(m)

a

0.000 1000 2000 3000 4000 5000(m)

500

1500

2500

(m)

0.00E+00

5.00E-01

1.00E+00

1.50E+00

2.00E+00

2.50E+00

3.00E+00

3.50E+00

sea level -50.00 m

b

0.000 1000 2000 3000 4000 5000(m)

500

1500

2500

(m)

0.00E+00

5.00E-01

1.00E+00

1.50E+00

2.00E+00

2.50E+00

3.00E+00

3.50E+00

sea level -50.00 m

c

Fig. 8. Horizontal displacements of the model grid with bathymetric values equal to 0, 3100 m and 350 m (rising phase) of thesea level.

PALAEO 3140 13-10-03


the model, with duration of 16 s and maximumamplitude of 9 m/s2. The number and frequencyof earthquakes remain unknown and therefore theaim is to check compatibility of the congruence ofthe deformations from an earthquake and the ge-ometry of the rupture planes of the collapse phe-nomena observed. In this regard, diagrams show-ing the largest vertical and horizontal movementsand the maximum shear strains recorded in themodel subsequent to the application of the seismicevent are illustrated in Fig. 11. The model wasinitially considered completely submerged and

subsequently static e¡ects mentioned above werecombined with dynamic e¡ects. In all cases thelargest localised collapses were observed in theplatform edge and in more inner positions. Inaddition, along the slope lineaments occur wherethe shear strains are largest (Fig. 11); these linea-ments are perpendicular as regards the earthquakedirection and dip at an angle of around 75^80‡ tothe horizontal. It may be assumed that the areasthat are most seismically vulnerable will be con-centrated along these lineaments from a singleearthquake, because of the concentration of shearstresses. Further seismic events will produce noth-ing more than a greater ampli¢cation in the de-formation along these alignments. These planescan be seen also as preferential zones in whichthe fracturing of rock mass is concentrated. Neg-ligible di¡erences on the rock mass behaviour areobtained when seismic activity is coupled with sealevel changes, testifying to the small in£uence ofthe position of the sea level.Previous analyses have shown that a variety of

e¡ects (fractures and changes in the shear strain)have been produced on the modelled system bythe application of changes in the sea level (staticanalysis), earthquake swarms (dynamic analysis)and their interactions. Moreover, it has beenshown that these e¡ects concentrate in the areaof the platform margin, and they are able to pro-duce small-scale collapses and changes in theequilibrium conditions. However, e¡ects producedby modelling are far from those observed in the¢eld and considered triggering factors are thus

sea level -50.00 m

(m)0.000 1000 2000 3000 4000 5000

500

1500

2500

0.00E+00

1.00E-03

2.00E-03

3.00E-03

4.00E-03

5.00E-03

6.00E-03

7.00E-03

8.00E-03

Fig. 9. Maximum shear strains in the model shown in Fig. 8, with sea level at 350 m rising from 3100 m.

Fig. 10. Ideal accelerogram of an earthquake of magnitude 6used in the dynamic analysis.

PALAEO 3140 13-10-03


insu⁄cient to produce large-scale collapses suchas those reconstructed for outcrops from geomet-rical and depositional features.A further analysis was carried out to include

the e¡ects of the progressive dislocations along afault located along the slope. The fault down-throw is generated progressively by incrementsof a few metres during the fault evolution. Here

Fig. 11. Maximum shear strain increment of the numerical model during dynamic analysis at di¡erent sea levels. The zones ofmaximum shear strains are concentrated along planes dipping 75^80‡, located on the slope and on the inner platform.

PALAEO 3140 13-10-03


are reported three steps corresponding to threesubsequent moments during the running of theprogramme. The ¢rst step depicts the situationat 100 m of fault dislocation, the second step cor-responds to the time at which the fault dislocationproduces a cumulative downthrow of 300 m,whereas the third step (500 m) represents the sit-uation at the end of the fault evolution, whenmodelling has produced geometries comparablewith geological data. Both static (without earth-quakes) and dynamic (earthquakes) conditionswere coupled during this analysis.Fig. 12 illustrates model results in quasi-static

conditions, e.g. without earthquake swarms asso-ciated to the fault dislocation. Although an in-crease in the vertical displacement (Fig. 12a) andan increment of the shear strains are recordedalong the slope (Fig. 12b), the rock mass restsstable and collapse phenomena are hampered dur-ing the simulation.A di¡erent behaviour of the rock mass is ob-

served when earthquake swarms are applied to

the fault dislocation (dynamic conditions) (Fig.13). Vertical displacements (Fig. 13a) and shearstrain increments (Fig. 13b) show critical valuesaround the platform margin, where an intense de-formation is recorded as indicated by the plastic-ity indicators (Fig. 13c). These conditions pro-mote the loss in the cohesion of the rock massfavouring the collapse of discrete portions of themargin (Fig. 13a).When the modelling produces the collapse of a

discrete portion of the margin, the simulation isblocked, the part corresponding to the collapsedmaterial is removed and the slope pro¢le is man-ually re-shaped according to the slide plane geom-etry created (compares di¡erent slope pro¢les ofthe three steps in Figs. 12 and 13). This re-shapingof the slope allows to better evaluating the rule ofdi¡erent geometric conditions along the slopeduring successive modelling phases. Successivelythe modelling re-starts throughout the fault dis-placement, simulating ¢rst static and then dynam-ic conditions.

sea level 0.00 m

sea level 0.00 m

sea level 0.00 m

sea level 0.00 m

sea level 0.00 m

sea level 0.00 m

800

1200

1600

2000

2400

1200 1600 2000 2400 2800 3200 (m)

-3.00E-01

-2.50E-01

-2.00E-01

-1.50E-01

-1.00E-01

-5.00E-02

0.00E+00 (m)

Dislocationfault 100 m

a

800

1200

1600

2000

2400

1200 1600 2000 2400 2800 3200 (m)

(m)


a

-3.00E+00

-2.50E+00

-2.00E+00

-1.50E+00

-1.00E+00

-5.00E-01

0.00E+00

800

1200

1600

2000

2400

1200 1600 2000 2400 2800 3200(m)

(m)


a

-3.00E-01

-2.00E-01

-1.00E-01

0.00E+00

1.00E-01

2.00E-01

750

1250

1750

2250

1250 1750 2250 2750 3250 (m)

b


0.00E+00

2.50E-04

5.00E-04

7.50E-04

1.00E-03

1.25E-03

1.50E-03

1.75E-03

2.00E-03

0.00E+00

1.00E-02

2.00E-02

3.00E-02

4.00E-02

5.00E-02

6.00E-02

750

1250

1750

2250

1250 1750 2250 2750 3250 (m)

b


0.00E+00

5.00E-04

1.00E-03

1.50E-03

2.00E-03

2.50E-03

3.00E-03

3.50E-03

4.00E-03

750

1250

1750

2250

1250 1750 2250 2750 3250(m)

b


Fig. 12. Numerical modelled vertical displacements (a) and maximum shear strain increments (b) at di¡erent dislocations of thehanging wall of the fault without dynamic e¡ects due to earthquakes.

PALAEO 3140 13-10-03


Some important results carried out from thisanalysis :1 ^ the quasi-static simulation (without earth-

quakes) shows that the formation of a morpho-logical downthrow does not represent alone adeterminant condition to trigger large-scale col-lapses at the platform margin. This is probablydue to the presence of well-cemented materialon the platform margin;

2 ^ introduction of earthquake swarms to thefault dislocation generates the collapse of discreteplatform margin portions;3 ^ long time activity of the fault coupled with

earthquake swarms allows reproducing geometriescomparable with those observed in outcrops.Although the software does not allow simulatingtemporal evolution of faults activity, the totalamount of displacement (500 m) suggests a pro-

-2.50E+00

-2.00E+00

-1.50E+00

-1.00E+00

-5.00E-01

0.00E+00

750

1250

1750

2250

1250 1750 2250 2750 3250(m)

(m)


a

-6.00E+00

-5.00E+00

-4.00E+00

-3.00E+00

-2.00E+00

-1.00E+00

0.00E+00

750

1250

1750

2250

1250 1750 2250 2750 3250(m)

(m)


a

750

1250

1750

2250

1250 1750 2250 2750 3250(m)

(m)


a

-2.50E+01

-2.00E+01

-1.50E+01

-1.00E+01

-5.00E+00

0.00E+00

0.00E+00

5.00E-03

1.00E-02

1.50E-02

2.00E-02

2.50E-02

3.00E-02

750

1250

1750

2250

1250 1750 2250 2750 3250(m)


b

0.00E+00

1.00E-02

2.00E-02

3.00E-02

4.00E-02

5.00E-02

6.00E-02

7.00E-02

8.00E-02

750

1250

1750

2250

1250 1750 2250 2750 3250(m)


b

0.00E+00

5.00E-02

1.00E-01

1.50E-01

2.00E-01

2.50E-01

750

1250

1750

2250

1250 1750 2250 2750 3250(m)


b

250

750

1250

1750

2250

1000 1500 2000 2500 3000 3500 (m)

Plasticity

2

1

3


c

250

750

1250

1750

2250

1000 1500 2000 2500 3000 3500 (m)

Plasticity

2

1

3


c

250

750

1250

1750

2250

1000 1500 2000 2500 3000 3500 (m)

Plasticity

2

1

3


c

sea level 0.00 m

sea level 0.00 m

sea level 0.00 m

sea level 0.00 m

sea level 0.00 m

sea level 0.00 m

sea level 0.00 msea level 0.00 msea level 0.00 m

Fig. 13. Numerical modelled vertical displacements (a), maximum shear strain increments (b) and plasticity indicators (c) at dif-ferent dislocations of the hanging wall of the fault during dynamic routine.

PALAEO 3140 13-10-03


tracted activity of the fault, necessary to repro-duce similarity between the geometries obtainedby software simulation and the outcroppingones. This activity could be referred to a timeinterval spanning from Albian to Early Cenoma-nian, corresponding to the time interval of occur-rence of the megabreccia deposits ;4 ^ the geometry, which best approaches the

palaeoescarpment reconstructed from outcropdata, is obtained after several collapse events oc-curring during the fault activity; this series ofcollapses produces the progressive backsteppingof the platform edge and di¡erent megabrecciaevents.

7. Conclusions

This work presents some results obtained fromthe 2D modelling of large-scale margin collapsesthat characterised the edge of the Maiella plat-form during the mid-Cretaceous. For the model-ling, a ¢nite-di¡erence numerical code was used,and this enabled the behaviour of the system indi¡erent static and dynamic conditions to be an-alysed, including both a homogeneous, intactmodel, and also one disturbed by a tectonic dis-continuity.The intact, homogeneous model allowed us to

check the importance of the dynamic simulationin the formation of sub-vertical fracture planesalong the slope; the variations in sea level seemto contribute little in the formation of fracturessurfaces. Even if the application of a single earth-quake to the intact model does not produce col-lapses equal in size to those observed, the forma-tion of preferential fracture planes represents animportant predisposing factor in the weakening ofthe platform margin.The best results were obtained coupling earth-

quake swarms with dislocations along a fault lo-cated along the slope and more or less parallel tothe direction of the pre-collapse platform margin.Families of sub-vertical fractures, asymptotic slideplanes and collapses of discrete parts of the mar-gin were produced by modelling in these condi-tions. A long-lasting seismic activity linked to thefault displacement is required to produce subse-

quent margin collapses, backstepping of the plat-form edge and rupture surfaces with similar ge-ometries to those observed along the EarlyCretaceous platform underlying the Murelle area.The results obtained strengthen the hypothesis

that the palaeoescarpment was caused by large-scale margin collapses linked to the tectonic dis-location of the system and associated earthquakeswarms. Moreover, the principles developed andthe preliminary conclusions reached may be appli-cable in general to carbonate platforms with fea-tures similar to those observed at the Maiella plat-form during the mid-Cretaceous.

Acknowledgements

This work was supported by research GrantsCOFIN 2000 to Prof. U. Crescenti and FondiRicerca Scienti¢ca ex 60% 2001 to G.R. The au-thors thank Prof. Uberto Crescenti for encourage-ment of this research, Dr Sergio Rusi for helpfuldiscussions on hydrogeology and Dr Monia Ca-lista for the support in the modelling. The authorsacknowledge A. Bosellini, J.B. Murray, P.W.Skelton and M. Morsilli for improving sugges-tions and constructive reviews.

References

Accarie, H., Beaudoin, B., Cussey, R., Joseph, P., Triboulet,S., 1986. Dynamique se¤dimentaire au passage plate-forme/bassin. Les facies carbonate¤s cre¤tace¤s du massif de la Maiel-la (Abruzzes, Italie). Mem. Soc. Geol. It. 36, 217^231.

Anderson, G.M., Crerar, D.A., 1993. Thermodynamics inGeochemistry; the Equilibrium Model. Oxford UniversityPress, New York.

Bernoulli, D., 1964. Zur Geologie des Monte Generoso. EinBeitrag zur Kenntnis der sudalpinen Sedimente. Beitr. Geol.Karte Schweiz, N.F. 118, 134 pp.

Bosellini, A., 1998. Scalloped vs Faulted carbonate platformmargins and the origin of basinal megabreccias. Mem. Soc.Geol. It. 53, 63^74.

Bosence, D., Waltham, D.A., 1990. Computer modeling theinternal architecture of carbonate platforms. Geology 18,26^30.

Bowman, S.A., Vail, P.R., 1999. Interpreting the stratigraphyof the Baltimora Canyon section, o¡shore New Jersey withPhil, a stratigraphic simulator. In: Harbaugh, J.W., Watney,W.L., Rankey, E.C., Slingerland, R., Goldstein, R.H.,

PALAEO 3140 13-10-03


Franseen, E.K. (Eds.), Numerical Experiments in Stratigra-phy: Recent Advances in Stratigraphy and SedimentologicComputer Simulations. SEPM Spec. Publ. No. 62, 117^138.

Cala, M., Flisiak, J., 2001. Slope stability analysis with FLACand limit equilibrium method. In: Billaux, D., Rachets, X.,Detournay, C., Hart, R. (Eds.), FLAC and Numerical Mod-eling in Geomechanics. Balkema, Rotterdam, pp. 111^114.

Casero, P., Rigamonti, A., Iocca, M., 1990. Paleogeographicrelationships during Cretaceous between the northern Adri-atic and the eastern southern Alps. Mem. Soc. Geol. It. 45,807^814.

Castellarin, A., Colacicchi, R., Praturlon, A., 1978. Fasi dis-tensive, trascorrenze e sovrascorrimenti lungo la ‘Ancona-Anzio’, dal Lias al Pliocene. Geol. Rom. 17, 161^189.

Cati, A., Sartorio, D., Venturini, S., 1987. Carbonate plat-forms in the subsurface of the northern Apennines. Mem.Soc. Geol. It. 40, 295^308.

Chen, W.F., Han, D.J., 1988. Plasticity for Structural Engi-neers. Sringer-Verlag, New York.

Ciancetti, G., Sciarra, N., 1987. Contributo alle conoscenza deimateriali lapidei nella Regione Marche. Quarry Constr. 18,89^96.

Crescenti, U., Crostella, A., Donzelli, G., Ra⁄, G., 1969.Stratigra¢a della serie calcarea dal Lias al Miocene nellaregione Marchigiano-Abruzzese (Parte II ^ Litostratigra¢a,biostratigra¢a, paleogeogra¢a). Mem. Soc. Geol. It. 9, 343^420.

D’Argenio, B., Mindszenty, A., Bardossy, G., Juhasz, E.,Boni, M., 1987. Bauxites of Southern Italy revisited. Rend.Soc. Geol. It. 9, 263^268.

D’Argenio, B., Mindszenty, A., 1991. Karst bauxites at region-al unconformities and geotectonic correlation in the Creta-ceous of the Mediterranean. Boll. Soc. Geol. It. 110, 85^92.

D’Argenio, B., Mindszenty, A., 1995. Bauxites and relatedpaleokarst: Tectonic and climatic event markers at regionalunconformities. Ecl. Geol. Helv. 88/3, 453^499.

Dawson, E.M., Roth, W.H., 1999. Slope stability analysis withFLAC. In: Detournay, C., Hart, R. (Eds.), FLAC and Nu-merical Modeling in Geomechanics. Balkema, Rotterdam,pp. 3^9.

Eberli, G., Bernoulli, D., Sanders, D., Vecsei, A., 1993. Fromaggradation to progradation: the Maiella platform, Abruzzi,Italy. In: Simo, J.A.T., Scott, R.W., Masse, J.P. (Eds.),Cretaceous Carbonate Platforms. Mem. AAPG 56, 213^232.

FLAC, 2000. Fast Lagrangian Analysis of Continua. Version4.0, Itasca Consulting Group, Minneapolis, MN.

Ford, D., Williams, P., 1996. Karst Geomorphology and Hy-drogeology. Chapman and Hall University Press, London.

Freeman-Lynde, R.P., Ryan, W.B.F., 1985. Erosional modi¢-cation of Bahama escarpment. Geol. Soc. Am. Bull. 96, 481^494.

Hine, A.C., Locker, S.D., Tedesco, L.P., Mullins, H.T., Hal-lock, P., Belknap, D.F., Gonzales, J.L., Neumann, A.C.,Snyder, S.W., 1992. Megabreccia shedding from modern,low relief carbonate platforms, Nicaraguan rise. Geol. Soc.Am. Bull. 104, 928^943.

Kooi, H., Cloetingh, S.A., 1989. Intraplate stresses and thetectono-stratigraphic evolution of the Central North Sea.In: Tankard, A.J., Balkwill, H.R. (Eds.), Extensional Tec-tonics and Stratigraphy of the North Atlantic Margins.AAPG Mem. 46, 541^558.

Jacquin, T., Arneaud-Vanneau, A., Ravenne, C., Vail, P.R.,1991. Systems tracts and depositional sequences in a carbon-ate setting: study of continuous outcrops from platform tobasin at the scale of seismic lines. Mar Petrol. Geol. London8, 121^139.

Morsilli, M., Borsellini, A., 1997. Carbonate facies zonation ofthe Upper Jurassic^Lower Cretaceous Abulia Platform Mar-gin. Riv. It. Pal. Strat. 103-2, 193^206.

Morsilli, M., Rusciadelli, G., 2000. 2D and 3D reconstructionof an erosional to by-pass margin and related slope depositsof a Cretaceous carbonate platform (Montagna della Maiel-la, Central Apennines, Italy). Global Sedimentary GeologyProgram, Working Group 4 (Cretaceous Resources Eventsand Rhytms). Vieste-Gargano 25^28, Settembre 2000.

Morsilli, M., Rusciadelli, G., Vichi, G., Bosellini, A., 2000.Gravity-Driven Processes: Control on Margin Architectureand Slope Evolution in a Cretaceous Carbonate Platform(Montagna della Maiella, Italy). AAPG International Con-ference and Exhibition, October 15^18, 2000 (Bali).

Morsilli, M., Rusciadelli, G., Bosellini, A., 2002. Large-scalegravity-driven structures: control on margin architectureand related deposits of a Cretaceous Carbonate Platform(Montagna della Maiella, Central Apennines, Italy). Boll.Soc. Geol. It. Vol. Sp. 1, 619^628.

Paull, C.K., Freeman-Lynde, R., Bralower, T.J., Gardemal,J.M., Neumann, A.C., D’Argenio, B., Marsella, E., 1990.Geology of the strata exposed on the Florida escarpment.Mar. Geol. 91, 177^194.

Payros, A., Pujalte, V., Orue-Etxebarria, X., 1999. The southPyrenean Eocene carbonate megabreccias revisited: new in-terpretation based on evidence from Pamplona Basin. Sed.Geol. 125, 164^194.

Pitman III, W.C., Golovchenko, X., 1991. Modeling sedimen-tary sequences. In: Muller, J.A., McKenzie, J.A., Weissert,H. (Eds.), Special Symposium on Controversies in ModernGeology. Academic Press, London, pp. 279^309.

Sarg, J.F., 1988. Carbonate sequence stratigraphy. In: Wilgus,C.K., Hastings, S.B., Kendall, C.St.G., Posamentier, H.W.,Ross, C.A., Van Wagoner, J.G. (Eds.), Sea-level Changes ^An Intergrated Approach. SEPM Spec. Publ. 42, 155^181.

Schlager, W., Camber, O., 1986. Submarine slope angles,drowning unconformities, and self-erosion of limestone es-carpment. Geology 14/9, 762^765.

Sciarra, N., 2000. Rigid formation over plastic substratum:modelling of a lateral spread in central Italy. Proceedingsof the 8th ISSMGE and BGS International Symposium onLandslides, Cardi¡, Wales, 3, 1339^1346.

Sciarra, N., Calista, M., 2001. Modellazione del comportamen-to di formazioni rigide su di un substrato deformabile: ilcaso di Caramanico Terme. Mem. Soc. Geol. It. Roma 56,139^149.

Spence, G.H., Tucker, M.E., 1997. Genesis of limestone mega-

PALAEO 3140 13-10-03


breccias and their signi¢cance in carbonate sequence strati-graphic models: a review. Sed. Geol. 112, 163^193.

Vanney, J.R., Stanley, D.J., 1983. Shelfbreak Physiography:An overview. In: Stanley, D.J., Moore, G.T. (Eds.), TheShelfbreak: Critical Interface on Continental Margins.SEPM Spec. Publ. No. 33, 1^24.

Vecsei, A., 1991. Aggradation und progradation eines karbo-natplattform-randes: kreide mis mittleres tertiar der mon-tagna della Maiella, Abruzzen. Ph.D. Thesis, Mitteilungendes Geologischen Institutes der Eidgenossischen Techni-schen Hochschule und der Universitat, Zurich, 294, 171 pp.

PALAEO 3140 13-10-03


Documents

2Dmodellingoflarge-scaleplatformmargincollapses … · 2018. 11. 12. · E-mail address: [email protected](G.Rusciadelli). PALAEO314013-10-03 Palaeogeography,Palaeoclimatology,Palaeoecology200