Active faults in the epicentral and mesoseismal Ml 6.0 24, 2016 Amatrice earthquake region, central Italy. Methodological and seismotectonic issues AN N ALS OF GEOPH YSICS, 59, Fast Track 5, 2016; DOI: 10.4401/ ag -7266 1 Active faults in the epi- central and mesoseismal Ml 6.0 24, 2016 Amatrice earthquake region, cen- tral Italy. Methodologi- cal and seismotectonic issues EMAN UELA FALCUCCI*, STEFAN O GORI*, FABRIZIO GALADIN I*, GIAN DOMEN ICO FUBELLI**, MARCO MORO*, MICH ELE SAROLI***, * *Istitu to N azion ale d i Geofisica e Vu lcan ologia em an u ela.falcu cci@in gv.it; Via d i Vign a Mu rata 605, Rom a **Un iversità d egli Stu d i d i Torin o ***Un iversità d egli Stu d i d i Cassin o e d el Lazio Merid ion ale Abstract The A ugust 24, 2016 A matrice earthquake (M l 6.0) struck a region of the central A pennines (Italy) where several active faults were known since decades, most of which are considered the surface expression of seismogenic sources potentially able to rupture during earthquakes with M of up to 6.5-7. The current de- bate on which structure/s activated during the mainshock and the possibility that conterminous faults may activate in a near future urged us gathering all the data on surface geological evidence of fault activi- ty we collected over the past 15-20 years in the area. W e then map the main tectonic structures of the 2016 earthquake epicentral and mesoseismal region. Our aim is to provide hints on their seismogenic potential, as possible contribution to the national Database of Individual Seismogenic Source (DISS) and to the Da- tabase of the active and capable fault ITaly HA zard from CA pable faults (ITHA CA ). I. IN TRODUCTION h e Ml 6.0 Au gu st 24, 2016 Am atrice earth qu ake occu rred on a N W-SE tren d - in g n or m al ru p tu re, m an ifestation of th e exten sion al tecton ic re gim e on goin g in th e cen tral Ap en n in e ch ain sin ce th e Late Pliocen e. Th e active N E-SW tren d in g stress gen erated sets of ch ain -p ar allel n orm al fau lt system s, arr an ged as tw o-to-th ree tecton ic rails, alon g w h ich m ajor h istorical an d in - stru m en tal seism icity con cen trates. Som e of T mailto:emanuela.falcucci@ingv.it ANNALS OF GEOPHYSICS, 59, Fast Track 5, 2016; DOI: 10.4401/ag-7266 2 th e ou ter m ost active n or m al fau lts of th e cen tral Ap en n in es are con sid ered as “silent”, th at is seism ic gap s (Galad in i an d Galli, 2000). Th e Am atrice earth qu ake occu rred in th e sector betw een tw o of th ese silen t fau lts, th e Laga Mts. an d Mt. Vettore n orm al fau lts. Th e form er affects th e base of th e w ester n slop e of th e Laga Mts. Its su rface exp ression is rep resen ted by a N W-SE ~26 km -lon g fau lt-scarp carved on to clayey -aren aceou s Miocen e flysch sequ en ces. Th e Mt. Vettore fau lt, N W-SE to N N W-SSE tren d in g, is m ad e of d ifferen t sp lays an d segm en ts w h ose scarp s are carved on to Meso-Cen ozoic lim e- ston e sequ en ces, exp osed alon g th e SW slop es of Mt. Vettore, Mt. Le Porch e an d Mt. Bove. Th ese com p ose an abou t ~27 km -lon g (at su rface) tecton ic stru ctu re. Seism ological, geod etic, rem ote sen sin g an d geological in vestigation s cu rren tly u n d er- w ay seem to in d icate th at p arts of th ese tw o stru ctu res p layed a p rim ar y role in th e seis- m ogen ic p rocess of th e Au gu st 24 m ain sh ock. Th e aftersh ock sequ en ce is d is- tribu ted in th e territory affected by th e Laga Mts. an d Mt. Vettore fau lts, an d beyon d th e area of m axim u m coseism ic d eform ation (Gru p p o d i Lavor o IN GV su l terrem oto d i Am atrice, 2016). Sin ce research es are still on goin g abou t th e cau sative fau lt(s) an d con sid erin g th e in h e r- en t v ariation of static stress in th e region s n earby, w e h ere rev iew d ata w e collected over th e p ast 15-20 year s of geological field su rveys in th e ep icen tral an d m esoseism al areas, to d escribe an d m ap th e m ajor active fau lts of th e region , basin g on critically se- lected geological criter ia for active fau ltin g d efin ition in ter m s of both ch ron ology an d su rface evid en ce. Fin ally, w e w ill p r op ose som e h yp oth eses abou t th e ch aracter istics of th e related seism ogen ic sou rces, as p ossible con tribu tion to th e n ation al Database of In - d ivid u al Seism ogen ic Sou rces (Basili et al., 2008) an d to th e Database of th e active an d cap able fau lt ITaly H Azard from CAp able fau lts (ITH ACA) (h ttp :/ / w w w .isp ram bien te.gov.it/ it/ p rogett i/ su olo-e-territorio-1/ ith aca-catalogo-d elle- faglie-cap aci). II. GEOLOGICAL “RATION ALE” FOR ACTIVE FAULT DEFIN ITION AN D MAPPIN G Th e vast geological liter atu re on th e Ap en - n in e active exten sion al tecton ics p r ovid es n u m er ou s evid en ce of active n orm al fau ltin g th at allow to d erive a con cep tu al m od el for d efin in g an active an d cap able fau lt as th e p rim ary exp ression at su r face of a seism o- gen ic sou rce. As for fau lt cap ability, i.e. a fau lt able to ru p - tu re th e su rface, w e ad op t th e 6.0±0.2 th resh old m agn itu d e for su rface fau lt in g, on th e ord er of th at p r op osed by Mich etti et al. (2000) for th e Ap en n in e fau lts. We th erefore con sid er th e fau lts w e m ap as th e su rface ex- p ression of seism ogen ic sou rces able to ru p - tu re w ith M≥6.0±0.2 earth qu akes. We also m ap oth er n orm al fau lts bu t w ith d ifferen t sym bology to m ark th at, based on several geological observ ation s, w e con sid er th em as n ot (or n o m ore) able to n u cleate seism ic even ts larger than M 6.0±0.2. As for th e tim e in terval to assess active an d cap able fau ltin g, w e ad op t th e criteria p r o- vid ed by Galad in i et al. (2012), for th e Italian exten sion al d om ain : a fau lt sh ou ld be con - sid ered as active an d cap able if it d isp lays evid en ce of activation in th e last 0.8 Myr, u n - less it is sealed by d ep osits or lan d form s n ot you n ger th an th e Last Glacial Maxim u m . Th is d efin ition h as been basically ad op ted by seism ic m icr ozon ation regu latory in force in AN N ALS OF GEOPH YSICS, 59, Fast Track 5, 2016; DOI: 10.4401/ ag -7266 3 Italy (Com m ission e tecn ica p er la m icrozo n - azion e sism ica, 2015). H en ce, to assess an d m ap an active an d ca- p able exten sion al fau lt / fau lt system as p ri- m ar y exp ression of a d eep seism ogen ic sou rce, w e d efin e con cep tu al an d factu al cr i- teria based on Qu atern ar y geological, geo- m orp h ological an d stru ctu ral field evid en ce. From th ese cr iteria w e d erive “requ isites” th at a fau lt m u st all h ave: 1) th e fau lt m u st sh ow evid en ce of d isp lace- m en t of d ep osits an d / or lan d form s (d erived from field observation s an d p aleoseism olo g- ical tren ch in g across fau lt traces) in d iver se sectors of its tr ace, for d ifferen t ages (Mid d le Pleistocen e-H olocen e in terval), w ith offsets th at m u st in crease w ith th e age of d isp laced featu res (Fig. 1). Th e latter issu e is to av oid to con sid er as p rim ary th ose secon d ary ru p - tu res su ch as extr ad os or ben d in g fractu res, n ot p rim arily con n ected w ith th e seism ogen - ic fau lt at d ep th , as th ey in h eren tly close at certain d ep th . Th is can be on ly u n raveled by con sid erin g th e d isp lacem en t of d ep osits an d / or lan d for m s over lon g tim e sp an s. Figure 1. Structural scheme of the different displace- ment features related to primary surface faulting and secondary (extrados) displacement. We p referen tially con sid er th e d isp lacem en t of th e base of a given stratigrap h ic u n it across th e con sid ered fau lt. In d eed , th e top su rface cou ld be affected by ero sion - al/ d ep osition al p r ocesses th at m ay alter its actu al sh ap e an d m orp h o-stratigrap h ic sig- n ifican ce. As best p ractice, geologi- cal/ geom orp h ological d ata m u st ascertain th at th e con sid ered d isp laced top su rface is p rim ary (or su b-p rim ary) an d / or ascribable to a d ep osition al/ erosion al sp ecific ord er. Th is asp ect is fu n d am en tal to av oid to err o- n eou sly correlate acr oss a fau lt su rfaces (er o- sion al/ d ep osition al) p ertain in g to d ifferen t ord ers or h avin g d ifferen t m o rp h o- stratigrap h ic sign ifican ce. 2) th e scarp of a m ajor n or m al fau lt h as to be con tin u ou sly d etectable for several kilom e- tres. Th is relates to th e observ ation s th at fau ltin g occu rs an d m ain tain s on th e d iscrete stru ctu r al elem en ts for lon g tim e p eriod s, an d stru ctu r al featu res m u st h en ce be geo- logically an d geom orp h ologically recogn isa- ble in th e field . Min or syn th etic or m in or an - tith etic str u ctu res, or secon d ary fau lts m ay p rovid e evid en ce for d etectin g p rim ary a c- tive fau ltin g in th e field . Bu t as th ese fau lts m ay n ot accom m od ate a large am ou n t of slip an d , th erefore, m ay n ot be rep resen tative of th e m ajor fau lt beh aviou r, w e h ave n ot r e- p orted th ese secon d ary fau lts. 3) a several km -lon g n or m al fau lt segm en t, sp lay or br an ch for w h ich evid en ce of d is- p lacem en t of d ep osits an d / or lan d form s are n ot sp ecifically av ailable h as to be con sid - ered as p rim ar ily con n ected to th e seism o- gen ic fau lt if it eith er d isp lays overlap w ith or is p ar allel to oth er stru ctu ral fea tu res for w h ich th ese evid en ce are available, u n less sp acin g exceed s ~4 km (for syn th etic sp lays) in m ap v iew (Wesn ou sky, 2006); if so, th e fau lts m ay in d eed h ave d ifferen t kin em atic h istory an d beh av iou r an d cou ld be related to d ifferen t seism ogen ic str u ctu res. ANNALS OF GEOPHYSICS, 59, Fast Track 5, 2016; DOI: 10.4401/ag-7266 4 At th e sam e tim e, m ajor cau tion s m u st be taken before d efin in g an active fau lt as p r i- m arily lin ked to a d eep seism ogen ic stru c- tu re: a) It m u st be con sid ered if n orm al fau ltin g h yp oth esised on th e lateral con tact of d iffer- en t sed im en tary bod ies or of sed im en tar y bod ies w ith th e bed rock, cou ld be in stead d u e to m orp h o-sed im en tary p rocesses, th at cou ld be m islead in gly in terp reted . For in - stan ce, sed im en tary u n its em bed d in g, flu vial en tren ch m en t an d terracin g. At th is p u rp ose, d etailed geom orp h ic an d sed im en tological an alyses, ch ar acterisation of th e m ech an ism of tran sp ort an d sed im en tation , an d (p aleo)en v iron m en t assessm en t of d ep osi- tion are cru cial asp ects. b) It m u st be evalu ated t h e p ossible occu r- ren ce of an y oth er n on -tecton ic m orp h oge- n etic p rocess th at cou ld im itate su rface fau lt- in g, esp ecially w h en occu rrin g alon g fau lt - scarp s. Am on g th ese factors, local an d areal lan d slid in g (i.e. of sed im en ts/ r ocks at th e base of a fau lt-scarp ), lar ge scale grav itatio n - al m ass m ovem en ts (su p er p osin g on a fau lt- scarp ), local sed im en t com p action (resem - blin g fau lt d isp lacem en t), local erosion (e.g. of th e d ebris d ep osited at th e base of th e fau lt-scarp ), m orp h o-selection p rocesses an d d ifferen tial er osion (i.e. d u e to d ifferen tial erod ibility of d ep osits/ rocks across a fau lt- scarp ), local grou n d su bsid en ce or collap ses (e.g. sin kh oles or kar stic featu res), h u m an activity (qu arryin g, excav ation s, d ebris ac- cu m u lation s) m u st be th orou gh ly ackn ow l- ed ged . A n u m ber of stu d ies d iscu ssed abou t th is (Bu cci et al., 2007; Fu belli et al., 2009; Kastelic et al., 2015). Most of th ese stu d ies d efin ed th at active fau ltin g on ly based on th e recogn ition of su p p osed m orp h otecton ic fea- tu res, su ch as th e sole local exp osition of th e fau lt p lan e at few p laces at th e base of a giv- en fau lt-scarp , or th e recogn ition in th e field of geom orp h ic featu res (e.g. trian gu lar fac- ets) su p p osed to be fau lt-r elated , can n ot be con sid ered satisfactory w h en taken alon e. III. DISCUSSION Data collected th r ou gh th e p ast 15-20 year s in th e area affected by th e 2016 Am atrice seism ic sequ en ce an d su rrou n d in gs allow u s to trace series of n orm al fau lt system s, th at w e con sid er as th e p rim ary exp ression at su rface of seism ogen ic sou r ces. As for th e Laga Mts. fau lt, evid en ce of Late Pleistocen e-H olocen e activ ity w as rep orted (Galad in i an d Galli, 2003). Th e fau lt can be sp litted in to tw o segm en ts, th e ~18 km -lon g sou th ern segm en t, bou n d in g th e Cam p otosto plateau, an d th e ~8 km -lon g n orth er n seg- m en t, bou n d in g th e Am atrice basin . Th e Cam p otosto segm en t sh ow s a con tin u ou s su rface fau lt-scarp an d d isp lays geological an d geom orp h ological evid en ce of Late Pleistocen e-H olocen e m ovem en ts; su rface fau ltin g ep isod es w ere id en tified by p aleo- seism ological tren ch in g. Con versely, n o ev i- d en ce at su rface of late Qu atern ary fau lt a c- tivity w as fou n d alon g th e Am atr ice se g- m en t. Galad in i an d Galli (2003), th erefore, con sid ered th is segm en t as in active or th at th e fau lt w ou ld n ot be n o m ore able to gen - erate earth qu ake lar ge en ou gh to ru p tu re th e su rface. N otew orth y, th is is con sisten t w ith th e su rface observation s m ad e after th e A u - gu st 24 m ain sh ock, as n o evid en ce of su rface ru p tu re occu rred alon g th e Am atrice fau lt segm en t (Em ergeo Workin g Grou p , 2016). In terestin gly, th e d eep p ortion of th e Cam p otosto segm en t activ ated in its low er- m ost p ortion d u rin g th e 2009 seism ic se- qu en ce (Bigi et al., 2013). As for th e Mt. Vettore fau lt, Galad in i an d Galli (2003) d efin ed th at th e 18 km lon g su r- AN N ALS OF GEOPH YSICS, 59, Fast Track 5, 2016; DOI: 10.4401/ ag -7266 5 face exp ression of th e seism ogen ic stru ctu re is m ad e of d ifferen t segm en ts an d sp lays. Evid en ce for th e Pleistocen e-H olocen e activ i- ty of th e stru ctu re w as obtain ed by p aleo- seism ological in vestigation s. Th e Mt. Vettor e fau lts th erefore sh ow s geological evid en ce at su rface of recen t activity, d ifferen tly from th e Am atrice segm en t. Th erefore, th e abou t 5 km lon g su rface ru p tu re alon g th e easter n m ost sp lay of th e Mt. Vettore fau lt follow in g th e Au gu st 24 m ain sh ock (Em ergeo Workin g Grou p , 2016) w ou ld cor robor ates su rface fau ltin g p oten tial of th e Mt. Vettore fau lt. Ou r an alyses revealed th at th e fau lt sh ou ld be len gth en n orth w ard for ~8 km . IV. CON CLUSION S Th e Ml 6.0 Au gu st 24, 2016 Am atrice eart h - qu ake stru ck a region of cen tral Italy w h ere several n orm al active n or m al fau lts w ere al- read y kn ow n . Debate on th e seism otecto n ics of th e 2016 even t, th e p resen ce of m an y oth er n earby active n orm al fau lts an d th e fact th at som e of th ese u n d erw en t in crease of static stress after th e 2009 L’Aqu ila earth qu ake (Falcu cci et al., 2011) an d likely after th e 2016 seism ic sequ en ce, u rged u s to gath er an d su m u p th e resu lts of field in vestigation s w e p erform ed in th is r egion an d to m ap in d etail th e active n orm al fau lts w e con sid er as th e p rim ary exp ression at su rface of seism ogen ic sou rces able to ru p tu re d u rin g eart h qu akes w ith m agn itu d e of u p 6.5-7. In th is p er sp ec- tive, as for th e m axim u m d ep th reach ed by th e con sid ered fau lts, th e an alysis of th e seism ological d ata related to th e 2009 L’Aqu ila earth qu ake (e.g. Valor oso et al., 2013) an d 2016 Am atrice seism ic sequ en ce (Gru p p o d i Lavor o IN GV su l terrem oto d i Am atrice, 2016) su ggest th at m ajor seism o- gen ic fau lts in th e cen tr al Ap en n in es can reach ~10 km d ep th . Figu re 2 su m s u p all th e active fau lts w e m ap p ed , an d th e related seism ogen ic sou rces, as p ossible con tribu tion to th e Italian Database of th e active an d ca- p able fau lts (ITaly H Azar d from CAp able fau lts, ITH ACA) an d to th e Italian Database of In d ivid u al Seism ogen ic Sou rces (Basili et al., 2008). Figure 2. M ap of the active faults of the region (col- oured lines). Faults: M t.V ettore-Bove fault, M V BF; N orcia fault, N F; Cascia fault, CF; A matrice fault segment, A Fs; Campotosto fault segment, CFs; Capit i- gnano fault, CaF; Upper A terno V alley-Paganica fault ststem, UA V -PF. Seismogenic sources, white boxes; Purple boxes indicate seismogenic sources not able to generate surface faulting earthquakes. ANNALS OF GEOPHYSICS, 59, Fast Track 5, 2016; DOI: 10.4401/ag-7266 6 REFEREN CES Basili, R., Valen sise, G., Van n oli, P., Bu rrato P., Fracassi, U., Marian o, S., Tiberti, M. M., Bosch i, E. (2008). Th e Database of In d ivid u al Seism ogen ic Sou rces (DISS), version 3: Su m m arizin g 20 years of r esearch on Italy's earth qu ake geology. Tecton op h ysics, 453:20- 43. Bigi, S., Caser o, P., Ch iar abba, C., Di Bu cci, D. (2013). Con tr astin g su rface active fau lts an d d eep seism ogen ic sou rces u n veiled by th e 2009 L’Aqu ila earth qu ake sequ en ce (It a- ly). Terra N ov a, 25: 21-29. Bu cci, F., D’On ofrio, D., Tavarn elli, E., Pr o s- ser, G. (2007). Trian gu lar facets or Flatiron ? A n ote of cau tion fr om th e Lu can ian Ap en - n in es, Italy. Com m ission e tecn ica m icr ozon azion e sism i- ca (2015). Lin ee gu id a p er la gestion e d el te r- ritorio in aree in teressate d a Faglie Attive e Cap aci (FAC), ver sion e 1.0 Con feren za d elle Region i e d elle Pr ovin ce Au ton om e – Dip ar- tim en to d ella p rotezion e civile, Rom a . EMERGEO Workin g Gr ou p (2016). Coseis- m ic effects of th e 2016 Am atr ice seism ic se- qu en ce: first geological resu lts; DOI 10.4401/ ag-7195.. Falcu cci E., Gor i S., Mor o M., Pisan i A.R., Melin i D., Galad in i F., Fred i P. (2011). Th e 2009 L’Aqu ila earth qu ake (Italy): w h at n ext in th e region ? H in ts from stress d iffu sion an alysis an d n orm al fau lt activity. Earth Pla- n et. Sci. Lett., 305:350-358. Fu belli, G., Gori, S., Falcu cci, E., Galad in i, F., Messin a, P. (2009). Geom or p h ic sign atu res of recen t n orm al fau lt activity versu s geological evid en ce of in activity: Case stu d ies from th e cen tral Ap en n in es (Italy). Tecton op h ysics, 476:252–268. Galad in i F., Falcu cci E., Galli P., Giaccio B., Gori S., Messin a P., Mor o M., Sar oli M., Scard ia G., Sp osato A. (2012). Tim e in terv als to assess active an d cap able fau lts for en g i- n eerin g p ractices in Italy. En gin eerin g Geo- logy, 139/ 140:50-65. Galad in i, F., Galli, P. (2000). Active tecton ics in th e cen tral Ap en n in es (Italy)—in p u t d ata for seism ic h azard assessm en t. N at. H azard s 22: 225–270. Galad in i, F., Galli, P. (2003). Paleoseism ology of silen t fau lts in th e Cen tr al Ap en n in es (It a- ly): th e Mt. Vettore an d Laga Mts. fau lts . An n als of Geop h ysics, 46:815-836. Galli, P., Galad in i, F., Pan t osti, D. (2008). Tw en ty year s of p aleoseism ology in Italy. Earth Sci. Rev. 88: 89–117. Gru p p o d i Lavor o IN GV su l terrem oto d i Am atrice (2016). Secon d o r ap p orto d i sin tesi su l Terrem oto d i Am atrice Ml 6.0 d el 24 Agosto 2016 (Italia Cen trale), d oi: 10.5281/ zen od o.154400. Kastelic, V., Bu rrato, P., Carafa, M., Basili, R. (2015). Progressive Exp osu re of th e Cen tral Ap en n in e’s “N astrin i”: a Fallin g Tecton ics Par ad igm ? GN GTS 2015.. Mich etti, A.M., Ferreli, L., Esp osito, E., Po r- fid o, S., Blu m etti, A.M., Vittori, E., Serva, L., Roberts, G.P., (2000). Grou n d effects d u rin g th e Sep tem ber 9, 1998, Mw =5.6, Lau ria earth qu ake an d th e seism ic p oten tial of th e aseism ic Pollin o region in Sou th ern Italy. Seis. Res. Letts. 71:31–46. Valoroso, L., Ch iar alu ce, L., Piccin in i, D., Di Stefan o, R., Sch aff, D., Wald h au ser, F. (2013). Rad iogr ap h y of a n or m al fau lt system by 64,000 h igh -p recision earth qu ake location s: Th e 2009 L’Aqu ila (cen tral Italy) case stu d y. J. Geop h ys. Res. 118:1–21. Wesn ou sky, S.G., 2006. Pr ed ictin g th e en d - p oin ts of earth qu ake ru p tu r es. N atu re 444: 358–360.