Df

JCa

b

c

d

a

ARRAA

KANHIB

1

is1bhRtemaacbJ

G

0h

Journal of Geodynamics 71 (2013) 43– 51

Contents lists available at ScienceDirect

Journal of Geodynamics

j ourna l h om epage: ht tp : / /www.e lsev ier .com/ locate / jog

eformation behaviour of the low-rate active Balanegra Fault Zonerom high-precision levelling (Betic Cordillera, SE Spain)

esus Galindo-Zaldívara,b,∗, Maria Jesus Borquec, Antonio Pedrerab,arlos Marín-Lechadod, Antonio José Gil c, Angel Carlos López-Garridob

Departamento de Geodinámica, Universidad de Granada, SpainInstituto Andaluz de Ciencias de la Tierra, CSIC-UGR, SpainDepartamento de Ingeniería Cartográfica, Geodesia y Fotogrametría, Universidad de Jaén, SpainInstituto Geológico y Minero de Espana, C/ Alcázar del Genil, 4, 18006 Granada, Spain

r t i c l e i n f o

rticle history:eceived 18 January 2013eceived in revised form 15 July 2013ccepted 26 July 2013vailable online 6 August 2013

eywords:ctive tectonicsormal fault

a b s t r a c t

The main segment of the active Balanegra Fault Zone (SE Spain) is monitored through high-precisionlevelling profiles in order to determine its short-term activity. It belongs to the NW–SE oriented normaland normal–oblique widespread fault set of the Betic Cordillera related to the Eurasian–African plateboundary, and constitutes the possible source of four moderate-size earthquakes (Mw 4.5–6.7) over thelast five centuries at a minimum ∼100-year recurrence interval. Successive measurements since 2006along two levelling profiles determined varying temporal behaviour. A general increase in vertical throwoccurred between May 2007 and September 2010, within a remarkable low seismic activity and deforma-tion build-up period, followed by a subordinate regional relaxation stage probably associated to nearby

igh precision levellingnterseismic periodetic Cordillera

low magnitude earthquakes. In the longest northern profile deformation of up to 1.88 mm and ratesreaching 0.78 mm/year with a mean rate of 0.14 mm/year are observed. The shortest southern profile,which is close to the area where the main earthquakes of the 1993–1994 series nucleated, records simi-lar interseismic deformation behaviour although with lower deformation than the northern profile. TheBalanegra Fault Zone illustrates complex behaviour in the period comprised between major earthquakes,exemplifying the temporal variability of deformation.

. Introduction

The behaviour of active faults and earthquake nucleation, a mainnterest of Earth Science, has far-reaching implications in hazardtudies (Sibson, 1977; Jones, 1984; Schwartz and Coppersmith,984; Wyss and Martyrosian, 1998; Saegusa, 1999). The vast num-er of factors involved in formulating fault-deformation modelsampers an overall understanding of the process (Faulkner andutter, 2003; Faulkner et al., 2006; Tullis, 2007). According tohe elastic rebound theory (Reid, 1910; Savage, 1983), a syntheticarthquake cycle comprises co-seismic and inter-seismic defor-ation. During the inter-seismic phase, stress in the rock mass

ccumulates before being released by the earthquake. Faults maylso partially relieve this strain accumulation through aseismic slip-

reep deformation, by generating small-magnitude earthquakes ory intermediate modes of slip behaviour (Tocher, 1960; Beroza andordan, 1990; Wdowinski, 2009; Marone and Richardson, 2010).

∗ Corresponding author at: Departamento de Geodinámica, Universidad deranada, 18071-Granada, Spain. Tel.: +34 958243349; fax: +34 958248527.

E-mail address: jgalindo@ugr.es (J. Galindo-Zaldívar).

264-3707/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.ttp://dx.doi.org/10.1016/j.jog.2013.07.003

© 2013 Elsevier Ltd. All rights reserved.

In recent years, permanent and non-permanent GPS networks,high-precision levelling profiles across specific tectonic structures,and InSar images allow for quantification of surface deformationthroughout the whole process (Bourne et al., 1998; Fialko, 2006;Biggs et al., 2007; Gourmelen et al., 2010; Sousa et al., 2010). Thesefindings, coupled with seismicity analysis and geological observa-tions of active-fault behaviour, help elucidate the earthquake cycle.Studies are especially relevant in slow deformation areas such asthe Betic Cordillera (Gràcia et al., 2012), where long observationperiods and highly accurate geodetic methodologies must be useddespite the seismic hazard.

The Betic Cordillera, located along the oblique convergentAfrican–Eurasian plate boundary, is deformed by moderately activeNW–SE to WNW–ESE normal and normal–oblique faults associatedwith seismicity (Sanz de Galdeano et al., 1995; Galindo-Zaldívaret al., 1999, 2003; Marín-Lechado et al., 2005; Alfaro et al., 2008;Pedrera et al., 2011, 2012a). In the northern Alboran Sea, the activ-ity of these faults bears relevance for natural hazards (Goy and

Zazo, 1986; Gràcia et al., 2012), and some of these active normalfaults are monitored onshore by non-permanent GPS networksand high-precision levelling profiles (Gil et al., 2002; Galindo-Zaldívar et al., 2003; Alfaro et al., 2006; Marín-Lechado et al., 2010).

44 J. Galindo-Zaldívar et al. / Journal of Geodynamics 71 (2013) 43– 51

Fig. 1. Geological setting of the study area. (A) Regional geological setting. (B) Digital elevation model with faults (black lines) recognised on field and lineaments offshoref ocks oo bles 1

Traa3betraC

ltfbafipyf

2

Wao2ea

rom slope gradient. (C) Seismicity within and near Balanegra Fault Zone. The aftershf 2006–2007 and the seismic period of 2010–2012 are differentiated. Details in Ta

he active faults presumably have a very low slip rate given theegional low strain magnitude (4–5 mm/year plate convergence in

NW–SE direction, Argus et al., 1989; DeMets et al., 1990, 1994)nd the width of the sector subjected to deformation (more than00 km). Moreover, while regional horizontal deformations haveeen determined by GPS measurements (Fadil et al., 2006; Vernantt al., 2010), high-precision levelling is used to accurately quantifyhe vertical local deformation caused by normal faults. Althoughegional levelling profiles are available (Giménez et al., 2009), sliplong local active structures remains to be demonstrated in thisordillera.

This contribution presents the results of the first high-precisionevelling monitoring of the short-term slip of the main segment ofhe Balanegra Fault Zone (BFZ), one of the most active outcroppingaults in the Betic Cordillera. This fault caused a seismic sequenceetween 1993 and 1994 in which the maximum event registered

magnitude of 5.1 (Mw). The fault pattern was characterised fromeld as well as geomorphic data from a geodetic survey in theeriod 2006–2012. The new geodetic, field, and seismicity anal-ses are used to examine the deformation model of the active BFZault.

. The active Balanegra Fault Zone

The Campo de Dalías is widely deformed by NW–SE andNW–ESE trending normal and oblique-slip faults that have been

ctive since the Pleistocene. These faults are coeval with the devel-

pment of ENE–WSW open folds (Fig. 1) (Marín-Lechado et al.,005, 2007; Martínez-Díaz and Hernández-Enrile, 2004; Pedrerat al., 2012b). Two prominent fault zones, formed by discontinuousctive fault segments, can be recognised onshore (Fig. 1B). The Loma

f the 1993–1994 series, seismicity from 1994 to 1998, the cluster of microseismicity and 2 (data from IGN database, www.ign.es).

del Viento Fault deforms the central part of the Campo de Dalías,with dextral–normal kinematics. Moreover, the BFZ is formed bygenerally southwestward dipping fault segments trending NW–SE,which extends from the north of Berja towards the Alborán Sea,controlling the western end of the Sierra de Gador antiform andthe NW–SE oriented straight morphology of the coastline (Fig. 1B).The 1993–1994 seismic series occurred along the BFZ (Fig. 1C). Faultslip rate values estimated in the fault segment that limits the west-ern border of the Sierra of Gádor are comprised between 0.1 and0.3 mm/year (Martínez-Díaz and Hernández-Enrile, 2004), in thenorthern part of BFZ. Westward, Gràcia et al. (2012) identify theNW–SE trending Adra Fault that extends 20 km offshore, reachingthe seafloor surface (Fig. 1).

The NW–SE oriented, westward-dipping BFZ (Figs. 1 and 2)is an active fault zone, as can be inferred from its syn-tectonicQuaternary deposits, its evident topographic escarpment in thelandscape, and its associated seismic activity (Fig. 1). Outcroppingfault surfaces are only found near the village of Balanegra, whereN160◦E/75◦W fault segments crop out with slicken lines plung-ing 70◦S. Southwards, the BFZ gives rise to a 6-km-long N140◦Eoriented lineament parallel to the coastline, where the differencein topography associated with fault activity is of about 45 m, andPliocene sediments have cross-bedding laminations dipping up to30◦ towards the downthrown block. Offshore, high-resolution sea-floor imaging shows lineaments mainly parallel to the BFZ witha mean trending of N140–145◦E (Gràcia et al., 2012). Thus, theBFZ is formed by a set of overstepping segments extending sea-

wards, separating the eastward upthrown Campo de Dalías fromthe westward downthrown Alboran Sea. In this setting, the onshoreoutcrops allow only partial monitoring of the fault zone, yet theyprovide valuable data regarding its activity.

J. Galindo-Zaldívar et al. / Journal of Geodynamics 71 (2013) 43– 51 45

Fig. 2. Geological map of Balanegra Fault Zone. Levelling profiles (red lines). Legend: 1 – Alluvial deposits (Holocene), 2 – Dunes (Holocene), 3 – Beach deposits (Holocene),4 – Marine terrace (Upper Pleistocene), 5 – Marine terrace (Middle Pleistocene), 6 – Marine terrace (Middle Pleistocene), 7 – Marine terrace (Lower Pleistocene), 8 – Reddishs n), 11f retatiov

3

smaesLtoEm

ilts (Pleistocene), 9 – Sands and silts (Pliocene), 10 – calcarenites (Upper Tortoniarom a 10 m resolution Digital Terrain Model (Junta de Andalucía, 2005). (For interpersion of this article.)

. Seismicity

The southern Betic Cordilleras are characterised by continuoushallow seismic activity of low magnitude as well as less frequentoderate earthquakes, revealing the relatively high seismic haz-

rd of this region. Historical records describe at least 50 destructivearthquakes near the study area, with details of the main shocks,urface breaks, and related damage distribution (Sánchez, 1917;ópez-Marinas, 1977; Mezcua et al., 2004). The surroundings of

he BFZ underwent a long period of seismic activity during previ-us centuries, with earthquakes of considerable intensity on theMS-98 scale: VI–VII (1686), VIII–IX and Mw 6.4 macroseismicagnitude (August 25, 1804, near the village of Dalías), and VIII

– Limestones and dolomites (Alpujarride basement). Topographic data extractedn of the references to colour in this figure legend, the reader is referred to the web

and Mw 6.1–6.7 (June 16, 1910, by Adra, with a Mw 5.5 aftershock)(Table 1). The exact location of these earthquakes remains con-troversial. As example, while the 1910 earthquake occurred overthe northwestern part of the BFZ after Mezcua et al. (2004), theepicentre proposed by Stich et al. (2003) is located 10 km fromthe BFZ (Fig. 1). Therefore, the BFZ (Pedrera et al., 2012a), andthe nearby Adra Fault have been both proposed as the source ofthe 1910 earthquake (Gràcia et al., 2012). In fact, it is not possibleto unequivocally attribute all these events to the Balanegra Fault

Zone.

The most intensive seismic activity recorded recently in thisregion (IGN database, www.ign.es) occurred from December 1993to January 1994, including Mw 4.8 and Mw 5.1 main shocks

46 J. Galindo-Zaldívar et al. / Journal of Geodynamics 71 (2013) 43– 51

Table 1Significant historical and instrumental earthquakes in the study area.

Date latitude (N) longitude (W) intensity (EMS-98 scale) magnitude location Time span betweenmajor earthquakes(years)

–/–/1686 36.85 2.95 VI–VII Ms 4.5b Berja. Al –25/08/1804 36.59 2.83 VIII–IX Mw 6.4c Dalías. Al 11825/08/1804a 36.59 2.83 VII Ms 4.7b Dalías. Al –29/08/1804a 36.59 2.83 VII Ms 4.7b Dalías. Al –

16/06/191036.60 3.50 VIII Mw 6.1d Adra.

Al10636.84c 3.00c Mw 6.7c

16/06/1910a 36.60 3.50 VII Mw 5.5d Adra. Al –23/12/1993 36.78 2.94 VII mbLg 5.0 S Berja 8304/01/1994a 36.57 2.82 VII mbLg 4.9 N Alborán –

a Foreshock.b From maximum intensity, using the D’Amico et al. (1999) relationship.

(iwbrrfs

amRpo

qha

TI(

c Mezcua et al. (2004), using the procedure by Bakun and Wentworth (1997).d Stich et al. (2003).

maximum intensity VII, MSK) (Stich et al., 2001). Over the follow-ng three months, 350 events (Md ≥ 1.5) were recorded in the area,

ith most hypocentres at a depth of 0–12 km. The spatial distri-ution of seismicity defines an approximately NW–SE elongatedegion, limited to the north and south by the two main shocksunning parallel to the NW–SE BFZ; their focal mechanisms showault-plane solutions involving normal faulting with an obliquelip component (Stich et al., 2001) (Fig. 1).

The distribution of earthquakes during the 1993–1994 seriesnd the following seismicity until 1998 reveals well-defined linea-ents striking N120–130◦E parallel to the BFZ (Stich et al., 2001).

elative locations between several nearby clusters illustrate theositions of adjacent active fault segments and the fragmentationf the crust into small fault blocks (Stich et al., 2001).

During the time span of the levelling surveys, no large earth-uakes took place directly in the BFZ (Fig. 1, Table 2). Near the fault,owever, a total of 29 shallow earthquakes were registered onshorend offshore, with magnitudes below 2.4.

able 2nstrumental seismicity during levelling survey period near Balanegra Faultwww.ign.es).

Date Time Latitude Longitude Magnitude

13/09/2006 00:05:38 36.6805 −2.8599 1.109/10/2006 12:06:11 36.6590 −2.8239 1.513/11/2006 18:22:39 36.6504 −2.8952 1.804/12/2006 15:00:23 36.6672 −2.8247 1.517/12/2006 00:33:04 36.6584 −2.8880 2.022/12/2006 17:19:42 36.6755 −2.8373 1.513/01/2007 05:46:26 36.6753 −2.8341 1.005/02/2007 04:48:05 36.7048 −2.8560 0.806/03/2007 12:53:36 36.6517 −2.8952 2.128/03/2007 07:18:35 36.6714 −2.8863 1.506/04/2007 02:30:06 36.6820 −2.8792 1.023/04/2007 17:44:48 36.6775 −2.8419 1.125/07/2007 17:54:40 36.6818 −2.8610 1.702/08/2007 01:00:38 36.6534 −2.8969 1.629/12/2007 19:44:31 36.6756 −2.8545 1.515/05/2008 04:21:36 36.6603 −2.8569 1.715/05/2008 06:26:49 36.6627 −2.8681 1.920/10/2008 17:45:12 36.6444 −2.8209 1.129/11/2008 01:47:25 36.6708 −2.8851 1.622/02/2010 11:45:40 36.6683 −2.8713 2.419/06/2010 23:46:24 36.6808 −2.8522 1.412/07/2010 06:53:23 36.6457 −2.8949 1.728/10/2010 03:37:35 36.6317 −2.8715 1.307/11/2010 05:04:33 36.6715 −2.8197 1.428/11/2010 23:30:45 36.7217 −2.8450 1.121/06/2011 19:11:59 36.7789 −2.8272 2.111/10/2011 10:55:21 36.6474 −2.8429 1.715/01/2012 20:26:46 36.7117 −2.8624 1.317/06/2012 21:12:33 36.7699 −2.8510 1.2

4. High-precision levelling

The quantification of vertical movements through comparisonof high-precision levelling data over several years is a techniqueused to monitor active normal faults, including seismic and creepdeformations, as it determines vertical changes with high accuracy(Giménez et al., 2009; Marín-Lechado et al., 2010). The method isbased on the comparison of height changes measured several timesalong the same levelling lines, providing information on the amountof vertical deformation around the main fault segment.

Two levelling profiles were measured across the southern partof the BFZ, the sector closest to the 1993–1994 seismic swarm, in sixfield surveys performed in 2006, 2007, 2009, 2010, 2011 and 2012.These profiles crosscut roughly orthogonal to the main onshoresegment of the BFZ, from the upthrown block up to the coastline,although the fault zone extends offshore. The Greenhouses profile(Fig. 3) is nearly 900 m long and stretches from benchmarks 997 to996, whereas the Old Guard Fortress profile has a length of 400 mapproximately and runs from benchmarks 993 to 991. Two bench-marks are located at each profile end, embedded vertically in rocksto guarantee their stability. The lowest benchmarks of both profileswere installed close to the coastline because the water table thereis almost constant, controlled by a stable sea level with very lowtidal variations.

The accuracy of levelling operations depends upon the qualityof the instrument used and its adjustment (Ihde, 1987). A LeicaDNA03 digital level, two 3-m Invar bar-coded staffs, two sturdysteel-spiked base plates for a stable staff position on all types ofterrain between benchmarks, and a tripod with fixed-length legswere used in the high-precision levelling surveys. The manufac-turer guarantees a standard deviation of ±0.3 mm/

√km of double

levelling, but there are many factors affecting the exactitude ofreadings in the development of a high-precision levelling profile.Therefore, it is necessary to consider, on the one hand, the staff cal-ibration; and on the other hand, the method of operation with thelevel.

The measurements were taken automatically, free of the influ-ence of the Earth’s curvature due to an option activated on thedigital level. The collimation error was calculated by means of theFörstner Method. The level automatically corrects all the measure-ments as well.

Each levelling run always began and ended with the same rodin order to avoid zero-point differences between rods. Every pro-file was observed back and forth. The levelling method used was

BFFB (backsight–foresight–foresight–backsight). The time intervalbetween readings was kept as short as possible. Likewise, backsightand foresight distances were kept as close to one another as pos-sible, up to a few cm per set-up, meaning that the total difference

J. Galindo-Zaldívar et al. / Journal of Geodynamics 71 (2013) 43– 51 47

F ault ZT Andal

twtc

b

�

wdb

p

�

mc

5

vtoaf2adsppetGtfit2

ig. 3. Detailed location of the levelling profiles in main segment of the Balanegra Fopographic data extracted from a 10 m resolution Digital Terrain Model (Junta de

hroughout the line was <10 cm. A staff scale range of 0.5–2.5 mas used. The line of sight was at no point closer to the ground

han 50 cm to prevent atmospheric refraction errors. Temperaturehanges during the observation campaigns were minor.

The standard deviation ��H in mm of the height difference, �H,etween benchmarks was calculated by:

�H = ±�ISO−LEV

√k (1)

here �ISO-LEV is the experimental standard deviation of a 1-kmouble-run levelling, and k is the length in kilometres betweenenchmarks.

The standard deviation � when different campaigns are com-ared is computed by:

=√

(��H1)2 + (��H2

)2 = ±�ISO-LEV

√k√

2 (2)

Fig. 4 shows the height differences between consecutive bench-arks measured in the 2006, 2007, 2009, 2010, 2011 and 2012 field

ampaigns.

. Vertical deformation

The vertical deformation determined from levelling measuresaried over the monitoring period and corresponds with part ofhe interseismic signal associated to the BFZ, which is distributedver a broader area (Figs. 4 and 5). In the Greenhouses profile, after

quiet period between May 2006 and May 2007, most of the sur-ace deformation occurred within the period May 2007–September010 (1.88 mm), with maximum rates found between May 2007nd May 2009 (1.45 mm, 0.72 mm/year. To compare the verticaleformation of the two profiles, it has been considered that thehortest distance between the profile ends in the Old Guard Fortressrofile (260 m length) is 2.5 times shorter than in the Greenhousesrofile (650 m length). After normalise the length, taking as refer-nce the Greenhouses profile (Table 3, Fig. 5), it can be observedhat the main deformation period was also recorded in the Olduard Fortress profile yet with moderate deformation, close to

he error bars. After September 2010, the Old Guard Fortress pro-le progressively recovers most of the deformation, in contrasto the Greenhouses profile. Between September 2010 and May011 the deformation suddenly partially recovered, and after May

one (above) and topography of the measurement tracks along the profiles (below).ucía, 2005).

2011 it continued to increase. The mean slip-rate value for theGreenhouses profile is 0.14 mm/year which is consistent with the0.1–0.3 mm/year long-term slip rate estimated in the northern partof the BFZ (Martínez-Díaz and Hernández-Enrile, 2004) and withthe 0.07 ± 0.03 mm/year calculated in the nearby Loma del Vientofault (Pedrera et al., 2012b). As the BFZ consists of an array of over-stepping segments extending seawards, rather than a single-slipfault surface, the vertical slips determined were held to representonly part of the short-term slip of the BFZ. The interseismic defor-mation is certainly distributed over a broader area. Nevertheless,these results provide valuable information regarding the Balanegraslip distribution and propagation.

6. Discussion

The quantification of vertical movements through a compari-son of high-precision levelling data over several years is a usefulmeans of monitoring slow active faults such as the BFZ. This methodprovides results that contribute to an understanding of the earth-quake cycle, with variable deformation rates in opposition to thecontinuous plate motions.

Surface deformations in scarps may be traced to different fac-tors, including landslides, variable water table level, differentobservation conditions, measurement errors, rock compaction ortectonic processes. We found no features associated with land-slides or rock compaction in the sector analysed, either during thefield mapping previous to the first measurement survey or dur-ing the subsequent surveys. Anyway, to consider the influence insurface deformation of water table level variation within the upper-most unconfined aquifer in Pliocene rocks, we have used as anindirect proxy the monthly cumulated rainfall from the nearbyAdra meteorological station (http://www.juntadeandalucia.es/agriculturaypesca/ifapa/ria). Fig. 6 illustrates the conceptual modelexpected during water table fluctuation within the homogeneousdetrital unconfined coastal aquifer and their influence in the ver-tical displacement between benchmarks. The lowest benchmarksof both profiles were placed by the coast, where low tidal vari-

ability ensured limited water table fluctuation and are consideredunaffected by this influence. During wet periods (Fig. 6A) the watertable rises, producing an increase of pore pressure within the satu-rated rock volume, leading the uplift of the upper benchmark and

48 J. Galindo-Zaldívar et al. / Journal of Geodynamics 71 (2013) 43– 51

Greenhouses profile Old Guards Fortress profile

28.96343

28.96150

28.96200

28.96250

28.96300

28.96350

28.96400

0.84342

0.843760.84352

0.84348

0.84334

0.843560.84300

0.84350

0.84400

0.84450

0.84500

0.84550

2005 2006 2007 2008 2009 2010 2011 2012 2013

998-997 (Control benchmarks footwall)

2005 2006 2007 2008 2009 2010 2011 2012 2013

995-998 (footwall-hanging wall difference)

-1.29142 -1.29137 -1.29140 -1.29144

-1.29162

-1.29143

-1.29300

-1.29250

-1.29200

-1.29150

-1.29100

-1.29050996-995 (Control benchmarks hanging wall)

2005 2006 2007 2008 2009 2010 2011 2012 2013

0.53512 0.53508 0.53510 0.53515 0.53511 0.535125

0.53400

0.53450

0.53500

0.53550

0.53600

0.53650993 994- (Control benchmarks footwall)

26.20950

26.21000

26.21050

26.21100

26.21150

26.21200992-994 (footwall-hanging wall difference)

0.55981

0.55971

0.55976

0.55979

0.55983

0.559745

0.55850

0.55900

0.55950

0.56000

0.56050

0.56100991-992 (Control benchmarks hanging wall)

2005 2006 2007 2008 2009 2010 2011 2012 2013

2005 2006 2007 2008 2009 2010 2011 2012 2013

2005 2006 2007 2008 2009 2010 2011 2012 2013

28.9621828.96198

28.96386

28.96279

28.96304

26.21093 26.21088

26.21135

26.21136 26.21134 26.211045

Heigh

t diffe

renc

e (m)

Heigh

t diffe

renc

e (m)

Heigh

t diffe

renc

e (m)

Year

Year

Year Year

Year

Year

Fig. 4. Levelling-profile results. Height differences in metres vs time in years in the Greenhouse profile: 70 m long from 998 to 997 benchmarks, 700 m long from 995 to 998benchmarks, 70 m long from 996 to 995 benchmarks. Height differences in metres vs time in years in the Old Guard Fortress profile: 30 m long from 994 to 993 benchmarks,300 m long from 992 to 994 benchmarks, 30 m long from 991 to 992 benchmarks. Errors bars are equal to two standard deviations.

Fig. 5. Vertical displacements between fault blocks relative to the first survey and cumulative deviation from mean monthly rainfall (29.46 mm/month in 2006–2012 period)from the nearby Adra village, as a proxy to analyse possible water table fluctuations within the detrital unconfined coastal aquifer.

J. Galindo-Zaldívar et al. / Journal of Geodynamics 71 (2013) 43– 51 49

Table 3Vertical displacement in profiles relative to the first measurement (mm).

Year Vertical displacement relative to the first measurement (mm)

Greenhouses profile (650 mlength)

Old Guard Fortress profile(260 m length)

Old Guard Fortress profile(×2.5 length normalised)

2006 0 0 02007 −0.2 −0.05 −0.1252009 +1.25 +0.42 +1.0502010 +1.68 +0.43 +1.0752011 +0.61 +0.41 +1.0252012 +0.86 +0.115 +0.288

Coastline

Coastline

Water table rise

Pore pressure increase

Vertical distance increase

Water table fall

Pore pressure decrease

A

Unsaturated

B

Saturated Homogeneous detrital

Lower benchmark

unconfined aquifer

Upper benchmark

between benchmarks

Vertical distance decreasebetween benchmarks

and compaction

F uctuatD

ttttAdti(iwfiatpwcmelrpa

tntwpd2

ig. 6. Influence in the vertical displacement between benchmarks of water table flry periods.

he increase of vertical displacement. During dry periods (Fig. 6B),he water table fall producing a decrease of pore pressure withinhe saturated rock volume leading compaction and the drop ofhe upper benchmark and the decrease of vertical displacement.lthough it seems to be a correlation between measured verticalisplacement and the cumulated rainfall (Fig. 5), the tendency ofhe vertical displacement between benchmarks is opposite to whats expected in the conceptual model (Fig. 6). During the dry period2006–2009), vertical displacement between fault blocks roughlyncreases. In contrast, vertical displacement decreases during the

et period (2010–2011). Therefore, relative movement betweenault blocks is most likely related to tectonic stresses and could benterpreted in terms of tectonic deformation of the active BFZ. Inny case, the water table effects cannot be completely discarded inhe vertical motion, because although the cumulated rainfall is aroxy to determine water table variation, there are no records ofater table level from nearby boreholes. We therefore chose well-

emented calcarenites and conglomerates to install the bench-arks. Additionally, two benchmarks were situated at the end of

ach profile to control for possible instrumental errors related toocal alteration of the benchmark height. Although the obtainedesults contribute to understand the behaviour of the southern BFZ,rofiles should be extended northeastwards in the future to include

broader sector including possible secondary fault traces.Fig. 5 shows heterogeneously distributed vertical deforma-

ion along the fault segment and its relationship with the mainearby earthquakes that took place during the surveys. A clus-er of microseismicity in the period 2006 and 2007 coincides

ith a slight decrease in the vertical fault throw along the tworofiles, interpreted as the surface response to partially elasticeformation dissipation. Subsequently, within the period May007–September 2010, there is a remarkable increase in the

ion within a homogeneous detrital unconfined coastal aquifer. (A) Wet periods. (B)

vertical deformation between fault blocks in both profiles (with arate of up to 0.72 mm/year in the northern Greenhouses profile),coinciding with an absence of earthquakes in the BFZ. This episodeis interpreted as a period of regional stress build-up and elasticenergy accumulation. Elastic deformation in the two profiles wasreleased after a new seismic period in nearby areas (2010–2012)although some deformation, may remain in Greenhouses profile.

Traditional conceptual models of fault behaviour predict theperiodic recurrence of ‘characteristic’ earthquakes with similarrupture extents (Shimazaki and Nakata, 1980). The historical andinstrumental earthquakes of moderate magnitude near the BFZshow a minimum average time span in between of ∼100 yearsover the past 5 centuries and a characteristic magnitude of 5–6.5(Table 1). It may be considered that the 1910 earthquake, tracedto the nearby offshore Adra Fault Zone (Gràcia et al., 2012), mayalso be related in a broad sense to the BFZ (Mezcua et al., 2004)although this attribution cannot be fully corroborated. In addition,the attribution of other historical earthquakes to the BFZ, onlysupport the maximum possible values of fault activity. Accordingto the magnitude–maximum displacement prediction equationsof Wells and Coppersmith (1994), if we consider an average earth-quake of magnitude 5.7, a maximum displacement of 4 cm wouldbe necessary, and a minimum ∼100-year recurrence period wouldimply a vertical slip rate of 0.40 mm/year. For a purely seismicfault, the interseismic period would entail no slip on the faultsurface, with accumulation of elastic energy, and displacementoccurring only suddenly during earthquakes. We observed a morecomplex interseismic behaviour than that predicted by the model.

The monitored fault segment shows a general increase in throwfollowed by subordinate decrease in a period showing microseimicactivity in nearby fault segments. This behaviour may correspondrather to the relaxation of elastic deformation, typical of seismic

5 nal of

fGcr(tGp1

ofbtlcmbmws

7

mImtmissahw0mFu

BgpBoerhrctwc

A

Dt201a2

0 J. Galindo-Zaldívar et al. / Jour

aults, than fault creep. The 0.14 mm/year mean slip-rate for thereenhouses profile is probably partially related to shallow faultreep. Although it is close to the lowest range of the long-term slipate (0.1–0.3 mm/year) estimated in the northern part of the BFZMartínez-Díaz and Hernández-Enrile, 2004), it is anyway lowerhan the 0.40 mm/year maximum possible long-term slip rate.reenhouses profile is more deformed than Old Guard Fortressrofile located nearby the area where the main earthquakes of the993–1994 series nucleated.

Although plate boundaries have constant velocities, the activityf a single fault is discontinuous over time and may exhibit dif-erent creep and elastic deformations for each segment, the lattereing related to seismic behaviour. For this reason it is essentialo continue measuring vertical displacement with high-precisionevelling methods in coming years to better determine the verti-al deformation-rate variability and its relationship with low andoderate seismicity. These data help us determine the precise

ehaviour of slow active faults, discerning creep and elastic defor-ation components, and identifying locked segments associatedith high seismic hazard. Field geodetic data complete the offshore

eismic hazard studies developed in this region (Gràcia et al., 2012).

. Conclusions

The Balanegra normal fault zone, trending NW–SE, is one of theost active structures in the hinterland of the Betic Cordillera.

ts vertical throw in the main onshore fault segment is deter-ined by repeated measurements between 2006 and 2012 along

wo high precision levelling profiles. Both profiles record a maxi-um displacement during a low seismic activity period that can be

nterpreted as built-up of elastic deformation (2007–2010). Sub-equently, microseismic activity in nearby faults helps release thetresses and elastic deformation, producing a decrease in the throwcross the main fault segment (2010–2012). The northern Green-ouses profile shows remarkable deformation of up to 1.88 mm,ith variable rates reaching 0.78 mm/year and a mean rate of

.14 mm/year which roughly agrees with the long-term slip esti-ations of other sectors of the fault zone. The southern Old Guard

ortress profile, near the source of the 1994 (Mw = 4.9) earthquake,ndergone a lower deformation than the Greenhouses profile.

In any case, the mean rate of deformation determined for theFZ is lower than the roughly 0.4 mm/year highest estimation sug-ested by historical seismicity. Two possible explanations may beroposed: that part of the historical earthquakes occurs out of theFZ, or that stress accumulation with low related deformationsccurs during interseismic periods. The 5–6.5 maximum magnitudexpected for this fault in a minimum time span of ∼100 years rep-esents a seismic risk that should be taken into account for futureazard-mitigation decision making. The present study provides slipates, underlining spatial and temporal variability in the periodomprised between major earthquakes. Anyway, a long observa-ion period, extending onshore the survey profiles and includingater table level evolution of nearby boreholes are needed to better

onstrain the behaviour of the low slip rate Balanegra Fault Zone.

cknowledgements

The constructive comments of two anonymous reviewers andr. S. Martos-Rosillo have largely improved the quality of this con-

ribution. This research was supported by the Consolider-Ingenio010 Programme (Topo-Iberia project CSD2006-0041), CGL2008-

5155-E/BTE, CGL-2008-03474-E/BTE, CGL2010-21048, AYA2010-5501 and CGL2008-03679-E/BTE projects from Ministry of Sciencend Innovation of Spain and the P09-RNM-5388, RNM-148, RNM-82, and RNM-376 research groups of the Junta de Andalucía.

Geodynamics 71 (2013) 43– 51

References

Alfaro, P., Estevez, A., Blázquez, E.B., Borque, M.J., Garrido, M.S., Gil, A.J., Lacy, M.C.,Ruiz, A.M., Giménez, J., Molina, S., Rodríguez-Caderot, G., Ruiz-Morales, M., Sanzde Galdeano, C., 2006. Geodetic control of the present tectonic deformation ofthe Betic Cordillera (Spain). In: Sansò, F., Gil, J. (Eds.), Geodetic DeformationMonitoring: from Geophysical to Engineering Roles, IAG Symposium Jaén, Spain,vol. 131, pp. 209–216.

Alfaro, P., Delgado, J., Sanz de Galdeano, C., Galindo-Zaldívar, J., García Tortosa,F.J., López Garrido, A.C., López Casado, C., Marín-Lechado, C., Gil, A.J., Borque,M.J., 2008. The Baza Fault: a major active extensional fault in the central BeticCordillera (South Spain). International Journal Earth Science 97, 1353–1365.

Argus, D.F., Gordon, R.G., DeMets, C., Stein, S., 1989. Closure of theAfrica–Eurasia–North America plate motion circuit and tectonics of theGloria fault. Journal of Geophysical Research 94, 5585–5602.

Bakun, W.H., Wentworth, C.M., 1997. Estimating earthquake location and magnitudefrom seismic intensity data. Bulletin of the Seismological Society of America 87,1502–1521.

Beroza, G.C., Jordan, T.H., 1990. Searching for slow and silent earthquakes using freeoscillations. Journal of Geophysical Research 95, 2485–2510.

Biggs, J., Wright, T., Lu, Z., Parsons, B., 2007. Multi-interferogram method formeasuring interseismic deformation: Denali Fault, Alaska. Geophysical JournalInternational 170, 1165–1179.

Bourne, S.J., England, P.C., Parsons, B., 1998. The motion of crustal blocks drivenby flow of the lower lithosphere and implications for slip rates of continentalstrike-slip faults. Nature 39, 655–659.

D’Amico, V., Albarello, D., Mantovani, E., 1999. A distribution-free analysis ofmagnitude-intensity relationships: an application to the Mediterranean Region.Physics and Chemistry of the Earth 24, 517–521.

DeMets, C., Gordon, R.G., Argus, D.F., Stein, S., 1990. Current plate motions. Geophys-ical Journal International 101, 425–478.

DeMets, C., Gordon, R.G., Argus, D.F., Stein, S., 1994. Effect of recent revisions tothe geomagnetic reversal time scale on estimates of current plate motions.Geophysical Research Letters 21, 2191–2194.

Fadil, A., Vernant, P., McClusky, S., Reilinger, R., Gomez, F., Ben Sari, D., Mourabit, T.,Feigl, K., Barazangi, M., 2006. Active tectonics of the Western Mediterranean:GPS evidence for roll back of a delaminated subcontinental slab beneath the RifMountains, Morocco. Geology 34, 529–532.

Faulkner, D.R., Rutter, E.H., 2003. The effect of temperature, nature of the pore fluid,and sub-yield differential stress on the permeability of phyllosilicate-rich faultgouge. Journal of Geophysical Research 108, 2227.

Faulkner, D.R., Mitchell, T.M., Healy, D., Heap, M.J., 2006. Slip on ‘weak’ faults by therotation of regional stress in the fracture damage zone. Nature 444, 922–925.

Fialko, J., 2006. Interseismic strain accumulation and the earthquake potential onthe southern San Andreas fault system. Nature 441, 968–971.

Galindo-Zaldívar, J., Jabaloy, A., Serrano, I., Morales, J., González-Lodeiro, F., Torcal, F.,1999. Recent and present-day stresses in the Granada Basin (Betic Cordilleras):example of a late Miocene-present-day extensional basin in a convergent plateboundary. Tectonics 18, 686–702.

Galindo-Zaldívar, J., Gil, A.J., Borque, M.J., González-Lodeiro, F., Jabaloy, A., Marín-Lechado, C., Ruano, P., Sanz de Galdeano, C., 2003. Active faulting in the internalzones of the central Betic Cordilleras (SE, Spain). Journal of Geodynamics 36,239–250.

Gil, A.J., Rodríguez-Caderot, G., Lacy, C., Ruiz, A., Sanz, C., Alfaro, P., 2002. Establish-ment of a non-permanent GPS network to monitor the deformation in GranadaBasin (Betic Cordillera, Southern Spain). Studia Geophysica and Geodetica 46,395–410.

Giménez, J., Borque, M.J., Gil, A.J., Alfaro, P., Estévez, A., Surinach, E., 2009. Compari-son of long-term and short-term uplift rates along an active blind reverse faultzone (Bajo Segura, SE Spain). Studia Geophysica and Geodaetica 53, 81–98.

Gourmelen, N., Amelung, F., Lanari, R., 2010. Interferometric synthetic apertureradar–GPS integration: interseismic strain accumulation across the HunterMountain fault in the eastern California shear zone. Journal of GeophysicalResearch 115, B09408.

Goy, J.L., Zazo, C., 1986. Synthesis of the Quaternary in the Almería littoralneotectonic activity and its morphologic features, western Betics, Spain.Tectonophysics 130, 259–270.

Gràcia, E., Bartolome, R., Lo Iacono, C., Moreno, X., Stich, D., Martínez-Díaz, J.J., Boz-zano, G., Martínez-Loriente, S., Perea, H., Diez, S., Masana, E., Danobeitia, J.J.,Tello, O., Sanz, J.L., Carreno, E., EVENT-SHELF Team, 2012. Acoustic and seismicimaging of the Adra Fault (NE Alboran Sea): in search of the source of the 1910Adra earthquake. Natural Hazards and Earth System Science 12, 3255–3267.

Ihde, J., 1987. Improved precision levelling for deriving significant vertical move-ments of the earth’s crust. Journal of Geodynamics 8, 313–320.

Jones, L.M., 1984. Foreshocks (1966–1980) in the San Andreas system, California.Bulletin of the Seismological Society of America 74, 1361–1380.

Junta de Andalucía, 2005. Modelo Digital del Terreno de Andalucía. Relieve y Oro-grafía. DVD edited by Junta de Andalucía, Sevilla, Spain.

López-Marinas, J.M., 1977. Estudio del terremoto de 22 Septiembre 1522 en Almería.Publicación Hidroeléctrica Espanola, 16 pp.

Marín-Lechado, C., Galindo-Zaldívar, J., Rodríguez-Frernández, L.R., Serrano, I., Pedr-

era, A., 2005. Active faults, seismicity and stresses in an internal boundary of atectonic arc (Campo de Dalías and Níjar, southeastern Betic Cordilleras, Spain).Tectonophysics 396, 81–96.

Marín-Lechado, C., Galindo-Zaldívar, J., Rodríguez-Fernández, L.R., Pedrera, A., 2007.Mountain front development by folding and crustal thickening in the internal

nal of

M

M

M

M

P

P

P

R

S

S

S

of the western Mediterranean: implications for the kinematics and dynam-

J. Galindo-Zaldívar et al. / Jour

zone of the Betic Cordillera–Alboran Sea boundary. Pure and Applied Geophysics166, 1–21.

arín-Lechado, C., Galindo-Zaldívar, J., Gil, A.J., Borque, M.J., de Lacy, C., Pedrera, A.,López-Garrido, A.C., Alfaro, P., García-Tortosa, F., Ramos, I., Rodríguez-Caderot,G., Rodríguez-Fernández, J., Ruiz-Constán, A., Sanz de Galdeano-Equiza, C., 2010.Levelling profiles and a GPS network to monitor the active folding and faul-ting deformation in the Campo de Dalias (Betic Cordillera, Southeastern Spain).Sensors 10, 3504–3518.

arone, C., Richardson, E., 2010. Learning to read fault-slip behavior from fault-zonestructure. Geology 38, 767–768.

artínez-Díaz, J.J., Hernández-Enrile, J.L., 2004. Neotectonics and morphotectonicsof the southern Almería region (Betic Cordillera-Spain) kinematic implications.International Journal of Earth Sciences 93, 189–206.

ezcua, J., Rueda, J., García-Blanco, R.M., 2004. Reevaluation of historic earthquakesin Spain. Seismological Research Letters 75, 75–81.

edrera, A., Ruiz-Constán, A., Galindo-Zaldívar, J., Chalouan, A., Sanz de Galdeano,C., Marín-Lechado, C., Ruano, P., Benmakhlouf, M., Akil, M., López-Garrido,A.C., Chabli, A., Ahmamou, M., González-Castillo, L., 2011. Is there an activesubduction beneath the Gibraltar orogenic arc? Constraints from Pliocene topresent-day stress field. Journal of Geodynamics 52, 83–96.

edrera, A., Galindo-Zaldívar, J., Marín-Lechado, C., García-Tortosa, F.J., Ruano,P., López-Garrido, A.C., Azanón, J.M., Peláez, J.A., Giaconia, F., 2012a. Recentand active faults and folds in the central–eastern internal zones of the BeticCordillera. Journal of Iberian Geology 38, 191–208.

edrera, A., Marín-Lechado, C., Stich, D., Ruiz-Constán, A., Galindo-Zaldívar, J., Rey-Moral, C., Mancilla, F., 2012b. Nucleation, linkage and active propagation of asegmented Quaternary normal-dextral fault: the Loma del Viento fault (Campode Dalías, Eastern Betic Cordillera, SE Spain). Tectonophysics 522–523, 208–217.

eid, H.F., 1910. The California earthquake of April 18, 1906; The mechanics of theearthquake, 2, 192. Carnegie Inst. Wash. Pub. 87.

aegusa, A., 1999. Japan tries to understand quakes, not predict them. Nature 397,284.

ánchez, M., 1917. Datos referentes a la sismicidad de la porción oriental de la Penín-

sula Ibérica. Memorias de la Real Academia de Ciencias y Artes de Barcelona 13(8), 179–188.

anz de Galdeano, C., López Casado, C., Delgado, J., Peinado, M.A., 1995. Shallowseismicity and active faults in the Betic Cordillera. A preliminary approach toseismic sources associated with specific faults. Tectonophysics 248, 293–302.

Geodynamics 71 (2013) 43– 51 51

Savage, J.C., 1983. A dislocation model of strain accumulation and release at a sub-duction zone. Journal of Geophysical Research 88, 4984–4996.

Schwartz, D.P., Coppersmith, K.J., 1984. Fault behavior and characteristic earth-quakes: examples from the Wasatch and San Andreas fault systems. Journalof Geophysical Research 89, 5681–5696.

Shimazaki, K., Nakata, T., 1980. Time-predictable recurrence model for large earth-quakes. Geophysical Research Letters 7, 279–282.

Sibson, R., 1977. Fault rocks and fault mechanisms. Journal of the Geological Society133, 191–213.

Sousa, J.J., Ruiz, A.M., Hanssen, R.F., Bastos, L., Gil, A.J., Galindo-Zaldívar, J., Sanz deGaldeano, C., 2010. PS-InSAR processing methodologies in the detection of fieldsurface deformation – study of the Granada basin (Central Betic Cordilleras,Southern Spain). Journal of Geodynamics 49, 181–189.

Stich, A.G., Alguacil, G., Morales, J., 2001. The relative locations of multiplets inthe vicinity of the Western Almería (southern Spain) earthquake series of1993–1994. Geophysical Journal International 146, 801–812.

Stich, D., Batlló, J., Morales, J., Macià, R., Dineva, S., 2003. Source parameters of the1910 Mw = 6.1 Adra earthquake (Southern Spain). Geophysical Journal Interna-tional 155, 539–546.

Tocher, D., 1960. Creep on the San Andreas fault: creep rate and related measure-ments at Vineyard, California. Bulletin of the Seismological Society of America50, 396–404.

Tullis, T.E., 2007. Friction of rock at earthquake slip rates. In: Schubert, G. (Ed.),Treatise on Geophysics, vol. 4. In: Kanamori, H. (Ed.), Earthquake Seismology,vol. 5. Elsevier Ltd., Oxford, pp. 131–152.

Wdowinski, S., 2009. Deep creep as a cause for the excess seismicity along the SanJacinto fault. Nature 2, 882–885.

Wells, D.L., Coppersmith, K.J., 1994. New empirical relationships among magnitude,rupture length, rupture width, and surface displacement. Bulleting Seismologi-cal Society of America 84, 974–1002.

Vernant, P., Fadil, A., Mourabit, T., Ouazard, D., Koulalid, A., Martin-Dávila, J., Garate,J., McClusky, S., Reilinger, R., 2010. Geodetic constraints on active tectonics

ics of the Nubia–Eurasia plate boundary zone. Journal of Geodynamics 49,123–129.

Wyss, M., Martyrosian, A.H., 1998. Seismic quiescence before the M7, 1988, Spitakearthquake, Armenia. Geophysical Journal International 134, 329–340.