Integrated Onshore-Offshore Investigation of a Mediterranean Layered Coastal Aquifer by Johanna Lofi1, Philippe Pezard2,Fred´ eric´ Bouchette2, Olivier Raynal2,3, Pierre Sabatier2,4, Nataliya Denchik2, Arnaud Levannier5, Laurent Dezileau2, and Raphael¨ Certain3

Abstract Most of the Mediterranean coastal porous aquifers are intensively exploited. Because of climatic and anthropogenic effects, understanding the physical and geological controls on groundwater distribution and flow dynamics in such aquifers is crucial. This study presents the results of a structural investigation of a system located along the coastline of the Gulf of Lions (NW Mediterranean). A key aspect of this study relies on an onshore-offshore integrated approach combining outcrops, seismic profiles, and borehole data analysis. This multidisciplinary approach provides constraints on pore-fluid salinity distribution and stratigraphic organization, which are crucial in assessing the modes of groundwater/seawater exchanges. Onshore, Lower Pliocene deposits dip gently seaward. They are unconformably overlain by Holocene clays in the lagoons. Offshore the Pliocene deposits either outcrop at the seabed or are buried below nonconsolidated sands infilling paleo-valleys. Beneath the lido, the groundwater salinity distribution consists of salty pore water, overlying fresher pore water. Active circulation of groundwater masses is inferred from the geophysical results. In particular, offshore outcrops and paleo-valleys may play an important role in salt water intrusion.

Introduction and to predict seawater intrusion (Bear et al. 1999; Bar- Coastal aquifer structure and dynamics have been low 2003). In most porous coastal aquifers, fresh water studied for many decades for groundwater resources, pri- flows seaward. However, close to the coast, due to fluid marily with the aim to assess fresh groundwater reserves density differences generating density flows, seawater with higher density penetrates into the aquifer, below the fresh water. This forms an inclined transition zone (TZ) 1Corresponding author: Geosciences´ —UMR where salt water/fresh water intermix. In natural condi- 5243—Bat. 22, Universite´ de Montpellier 2, Place E. Bataillon, tions, the geometry of this mixing zone (MZ) is related 34095 Montpellier Cedex 05, ; +33-4-67-14-93-09; fax: +33-4-67-14-32-44; johanna.lofi@gm.univ-montp2.fr to the hydrological properties of the aquifer, the physical 2Geosciences´ Montpellier—UMR 5243—Bat. 22, Universitede´ properties of both fluids (Henry 1964; Voss and Souza Montpellier 2, Place E. Bataillon, 34095 Montpellier Cedex 05, 1987; Croucher and O’Sullivan 1995), the geometry of France. the aquifer (Abarca et al. 2007), and/or the type of tides 3CEFREM, Universite´ de Perpignan , 52 Avenue Paul Alduy, 66860 Perpignan Cedex, France. (Brovelli et al. 2007). The geometry and the extent of the 4Universite´ de Savoie, Laboratoire Environnement Dynamiques MZ will also depend of the degree of heterogeneity of et Territoire de Montagne, CNRS, UMR 5204, Le Bourget du Lac, the aquifer (Dagan and Zeitoun 1998; Held et al. 2005). France. As a result, saline and brackish groundwater can be found 5 Schlumberger Water Services, P.O. Box 553, 2600AN Delft, far onshore in coastal plains, and, conversely, fresh and The Netherlands. Received November 2011, accepted October 2012. moderately brackish groundwater of meteoric origin can © 2012, The Author(s) extend offshore (Hathaway et al. 1979; Groen et al. 2000; Ground Water © 2012, National Ground Water Association. Kooi and Groen 2003). The anthropogenic influence on doi: 10.1111/j.1745-6584.2012.01011.x salinity patterns, such as inland aquifer pumping resulting

NGWA.org GROUND WATER 1 in seawater intrusion, has received by far the greatest Geological Context attention in hydro-geological studies (Bear et al. 1999). The study area is located along the coast of the Gulf However, salinity distribution relies strongly on the inter- of Lions’ passive margin, in the NW Mediterranean basin action between groundwater fluid dynamics and a geolog- (Figure 1). On this margin, the history of the coastal ical environment (commonly complex and anisotropic). porous aquifers is closely related to the Messinian Salinity The spatial distribution at several scales of porosity, Crisis event (MSC; Hsu¨ et al. 1973) which occurred hydraulic conductivity, and other hydraulic parameters approximately 5.5 Ma ago and exposed the Miocene in the subsurface are thus of major importance, and are margin to subaerial erosion (Lofi et al. 2005). This strongly related to geological context and depositional created deep valleys upstream (Clauzon 1973) and huge processes involved at geological time scales. Limited accommodation space for subsequent PQ sedimentation work has been done to understand the effect of geologic (Lofi et al. 2003; Duvail et al. 2005). heterogeneity on the exchange of groundwater and sea- The PQ sediments supplied from the Rhoneˆ and adja- water through coastal sediments, particularly offshore. In cent coastal rivers successively caused the progradation addition, beside modeling and geochemical studies per- of well-developed sedimentary wedges, building a new formed near submarine seeps or from borehole samples, continental shelf (Figure 1c). The current location of the there is little published about offshore aquifers. Geophys- onshore PQ depocenters is beneath the modern flat coastal ical investigations, generally carried out using seismic plains (see RCP, HCP, and RhCP in Figure 1a). They methods, allow the identification of subsurface geometries contain fresh groundwater resources enclosed in porous or active vent features (Kindinger et al. 1999; Swarzenski siliciclastic aquifers (Laurent 1993; Ambert et al. 1998; et al. 2001; Mulligan et al. 2007). Other techniques, such Dorfliger¨ 2003; Duvail et al. 2005; Aunay et al. 2006; as electrical tomography (Andersen et al. 2007) and air- Aunay 2007). These aquifers are associated with the borne or seafloor electromagnetics have proven valuable coarse-grained fraction of the PQ wedges. for evaluating offshore pore water salinities or distribu- In this context, our study focuses on the Maguelone tions (Hoefel and Evans 2001; Teatini et al. 2011). Studies area located on the northernmost part of the margin, using these techniques are rare and seldom combined with between the Herault´ and the Rhoneˆ coastal plains onshore studies (Evans and Lizarralde 2011; Teatini et al. (Figure 1b). The northern limit of the area is the Mesozoic 2011). Defining an accurate aquifer structure and het- limestone of the La Gardiole hill (∼200 m high; Figure 2a erogeneity both onshore and offshore therefore remains and 2b). As shown in Figure 2b, outcrops located at of primary importance for providing coherent conceptual the foot of the La Gardiole hill (gray) consist either of geological models which can lead to better predictions of Holocene (green) or Pliocene (Upl, yellow) formations. preferential flow pathways (Bowling et al. 2005; Cohen The latter also outcrop beneath the city of Villeneuve- et al. 2010). les-Maguelone` (VMc) and along the northern edge of Most of the Mediterranean coastal porous aquifers the Vic lagoon. Within the study area, the lagoons have are located in the Plio-Quaternary (PQ) deposits (Clau- been partially filled with Late-Holocene deposits with low zon 1973; Duvail et al. 2005; Boughriba et al. 2006; hydraulic conductivity (mainly clays and silts; Sabatier Aunay 2007; El Yaouti et al. 2009; Kouzana et al. 2010). et al. 2010) and have a water depth <2 m. They are These aquifers are “coastal detritic formations,” following limited southward by a barrier island (Maguelone lido) on the Custodio (2010) classification for European aquifers. which an experimental site has been located (Figure 2c, Understanding groundwater distribution and flows in see also “Data and Methods” Section). The Lez and such aquifers is of prime importance since in many Mosson rivers flow eastward of the study area (Figure 2a Mediterranean coastal areas, agriculture, drinking water and 2b). supply, tourism, and industry strongly depend on the available groundwater resources. Understanding physi- cal and geological controls on groundwater distribution Data and Methods and flow dynamics is thus critical and essential in order Studying the subsurface structure of the study area to anticipate salt water intrusion and to design ade- at several scales enables aquifer heterogeneity to be quate warning systems and water management tools. In assessed. Information on regional lithology and geometry such a context, this study focuses on a Mediterranean is provided by geological mapping (BRGM 1996). PQ porous system, set apart from the axis of the main Previous field works around VMc enabled interpretation coastal plains. We provide a careful integration, at several of sedimentary facies alongside dip measurements of scales, of various geological, hydrological, and geophys- outcropping strata (BRGM 1996; Raynal et al. 2010). In ical datasets to present a coherent image of subsurface this study, subsurface structure beneath the Maguelone aquifer structure, along an onshore-offshore cross section. lido has been characterized from nine boreholes (12 to Present-day hydrological processes are a significant con- 80 m deep) drilled in the framework of the ALIANCE, tribution to the observed distribution of groundwater MUSTANG, and ATIP PROGELAC projects (location in resources, and this emphasizes the importance of the off- Figure 2c). A series of downhole geophysical logs were shore structure of the geological skeleton in which water acquired. ASGR 512 (Spectral Natural Gamma probe, circulates. ANTARES) conducted in cased boreholes, provides a

2 J. Lofi et al. GROUND WATER NGWA.org E3° E4° 010 km NW SE Holocene Plio-Quaternary FRANCE Pliocene PQ MARGIN PROGRADATION Messinian (syn-crisis) Miocene Miocene N43°30 RhCP Mosson Pre-Rift Pre-Rift Fig.2-A Lez coastline sedimentary wedges HCP Fig.1(b) 185m VMc Arn. L Time (sec TWTT) land continental shelf 0 N43°00 Fig.1(c) Vic L continental slope

14 Fig.6(b) 2 LA GARDIOLE extrapolated deep basin 234m data RCP 4 N42°30 Thau L. Fig.6(a) MES ? (c) ? -20m

FCP -200m (a) Sète -10m (b) 0 50 100 150km

Figure 1. (a) Relief map of the Gulf of Lions margin showing the location of the study area. Red line marks the limit of the PQ outcrops. FCP, Figueres coastal plain; RCP, Roussillon coastal plain; HCP, Herault´ coastal plain; RhCP, Rhoneˆ coastal plain. (b) Simplified geological map of the study area [modified from the BRGM geological maps of Sete` and Montpellier and from Raynal et al. (2009)]. Thau L., Thau lagoon; Vic L, Vic lagoon; Arn. L, Arnel lagoon; VMc, city of Villeneuve-les-` Maguelone. (c) Dip line drawing illustrating the structure of the Gulf of Lions margin (modified from Lofi and Berne,´ 2008). The section onshore is a schematic extrapolation. At the bottom, the pre-rift basement is structured in horsts and grabens. The overlying Miocene sequence is eroded at the top by the Messinian Margin Erosional Surface (MES). At the top, the thick PQ prograding sequence, made of siliciclastic wedges, extends from onshore to offshore. At the present-time, coastal plain aquifers are developed in the coarse-grained fraction of the top of these wedges. measurement of the total gamma ray emissions (TGR) Offshore, close to the coast and in the Vic lagoon, of the formation adjacent to the borehole, allowing the high and very high resolution seismic profiles acquired identification of the main sedimentary units (Serra 1984). in the framework of the BEACHMED-E project image Contents of individual elements that emit gamma rays the upper part of the PQ sequence (Raynal et al. 2009), (U, Th, and K) are also differentiated. DIL45 (Dual with a penetration ranging from 10 to 100 m below the Induction probe, ALT) conducted in PVC cased boreholes seafloor. The main seismic units and boundaries observed provides information about the electrical conductivity of on these profiles have been labeled according to the works the formation, which is essentially driven by lithology, by Raynal et al. (2009, 2010): “Upl” for the Pliocene formation porosity and permeability, saturation, and sequence; “RES” for the “Regional Erosion Surface” interstitial fluid salinity (Archie 1942; Biella et al. 1983). generated during the last glacial-maximum (LGM) (during Borehole MAG5, located in the experimental site which the sea level was approximately 120 m below (Figure 2c), has been equipped with a Westbay6 multi- present-day sea level) and subsequent transgression. level groundwater characterization and monitoring system LIDAR bathymetric/altimetric survey was acquired allowing the surveying of multiple zones in a single bore- along the coast in 2007 (BEACHMED-E program). The hole (Black et al. 1986). The deployed system contains mean spatial resolution of these surveys is about 3 to 4 eight measurement ports to enable pressure measurements points per meter square. A subset of these points was as well as fluid sampling at formation pressure. Three extracted and interpolated to create a digital elevation packers were inflated at 13, 14, and 17 m below sea model of the area offshore from the Maguelone lido. level (mbsl), which isolated a pumping zone located at 15 mbsl. Below, packers were not deployed as isolation was expected to result from squeeze of the soft sediment Results around the Westbay casing. Groundwater samples were collected in June 2011 for geochemical analysis (total dis- Subsurface Sedimentary Structure solved solid content, TDS). Analyses were performed at The structure of the subsurface deposits is discussed the CIRAD (Montpellier, France), together with a sample hereafter along an onshore-offshore transect. collected from borehole MAG1 (59 m deep, PVC cased with perforations below 55 mbsl) in February 2011. Villeneuve-les-Maguelone` City (VMc) Sedimentological analysis performed on cores taken The Pliocene unit (Upl) outcrops around VMc and from borehole MAG1 includes facies description, radio- forms a small high in the topography culminating at carbon dating, and grain size analysis using a Beckman 21 m above sea level (masl; Figure 2a and 2b). On Coulter© LS 13 320. Only the <1 mm fraction was ana- the geological map, Upl mainly consists of continental lyzed. In levels displaying a significant number of large deposits (floodplain, fluvial, or lacustrine with lateral clasts (e.g., horizons C and S; see “Maguelone Lido” facies changes), and locally of marine deposits (sands; Section), the grain size distribution is thus not representa- BRGM 1996). Several boreholes (20 to 50 m deep) tive of the whole material. Sedimentary facies from cores have been drilled around the city for domestic water have been correlated with downhole logs in order to accu- supplies. Drilling reports available from the BSS (Public rately reposition cores in depth. Subsurface Bank) of the BRGM provide information on subsurface lithology, which mainly consists of continental 6Mark of Schlumberger. deposits (yellow clays and red silts interbedded with few

NGWA.org J. Lofi et al. GROUND WATER 3 Mosson Lez RES. An angular unconformity is thus clearly observed between the Holocene and Upl (Figure 3). The latter ◦ 185m LA GARDIOLE VMc is gently dipping toward the SE at an angle < 1 , 196m Arnel consistent with dips measured in outcrops along the lagoon 21m Palavas northern edge of the lagoon.

12m In terms of hydrological properties, the lagoonal silty- N clays are expected to have a relatively low hydraulic −7 Vic lagoon conductivity (< 10 m/s) and are likely to act as a Fig.8 confining unit, bounding the Upl unit. MagueloneMEDITERRANEAN Lido (a) 0 5 km SEA Maguelone Lido Mosson Lez Holocene Beneath the lido, core analysis from borehole MAG1 Pliocene marine/continental: allows the recognition of two main depositional units GARDIOLE lacustrine carbonate, LA marine sandy-clays (Figure 4) consisting of Holocene deposits discordantly continental: sands overlying Upl. (c) conglomerates, red silts and clays The Holocene unit extends from the surface down marine sands to approximately 9 mbsl and consists of approximately Miocene 5 m of gray shelly sands overlying approximately 4 m of Pre-Rift carbonate Fig.3 green lagoonal silty-clays. Radiocarbon dating confirms -10m Fig. drillings for water supply or geothermic purposes the age of this sequence (two samples giving ages 6(a)

14 (b) 0 5 km Fig.6(b) scientific drillings <6000 years cal. B.P., see Figure 4a and Raynal et al. 2009). The lagoonal clays are underlain by a coarse- VMc Arnel Lagoon grained level containing shell fragments and interpreted Prevost Palavas as the RES. This level is associated with a small shift in Lagoon Vic the TGR (Figure 4c). Such a shift is also observed in the Lagoon Maguelone Lido surrounding boreholes (Figure 5), thus allowing the base MAG3 EXPERIMENTAL SITE of the Holocene sequence to be traced in this area. Maguelone SAR Lagoon MAG7 The Upl unit extends from 8 mbsl to the bottom MAG4 N MAG2 of the borehole (59 mbsl). It consists predominantly of o MAG9 MAG1 Pierre Lid e relatively fine-grained sediments (mean clay content: 28% Blanche lon MAG6 MAG8 lagoon 0 1 km 0 10 m (c) Mague MAG5 and mean silt content: 57%; Figure 4b). These deposits are mainly of continental origin with evidence of overprinted Figure 2. (a) Relief map and (b) simplified geological map soil-forming processes ranging from light gray to reddish- of the study area (modified from BRGM, geological map of brownish oxidized horizons. Dark organic rich levels with Sete,` and Raynal et al. [2009]). To the north, the carbonate ∼ of the La Gardiole forms a relief on which the PQ deposits plant remains are observed in several places (e.g., 16, are pinching out. Pliocene formations outcrop beneath the 22.5, 32.5, 57.5 mbsl), often correlating with peaks on city of VMc. The Arnel and Vic lagoons are partly filled the U-curve (Figure 4c). Gray horizons with marine shell with Holocene deposits. Black dotted lines mark the location fragments are found locally (∼36, 56, and ∼57 mbsl). of the seismic lines illustrated in this study. Black stars are Between 9.2 and 12.2 m, a carbonate rich level with altitudinal points. (c) Detailed aerial map of the study area (IGN). Black dots and squares mark borehole location. algal cysts is interpreted as an equivalent facies to the lacustrine deposits that outcrop approximately 3.5 km landward (along the northern edge of the Vic lagoon, Figure 2b). meter thick sandy or gravelly horizons). Upl pinches out As shown in Figure 4a and 4b, the clayey fraction is northward on the La Gardiole hill and outcrops southward relatively high throughout the Upl. However, two remark- on the northern edge of the Vic lagoon. There, it consists able coarse-grained horizons are found between approxi- of white lacustrine limestones, dipping at low angle mately 37 and 40 mbsl (non-consolidated marine/estuarine toward the SE. sands “S”) and approximately 14 to 16 mbsl (mostly non- consolidated fluvial gravels and sands “G”). These hori- Vic Lagoon zons are characterized by low values on the TGR log As imaged on seismic profiles (Figure 3), Upl (Figure 4c). Comparing the shapes of the TGR curves extends beneath the Vic lagoon, appearing as a relatively allows horizon G to be laterally correlated across bore- transparent seismic unit containing sub-continuous low- holes, and indicates this level covers broad spatial extent frequency seismic reflections dipping SE-wards. Upl at the scale of the lido (Figure 5). The other horizons is bounded above by the RES, which truncates the forming Upl in MAG1 (in particular the U rich horizons) underlying Upl reflections (e.g., Figure 3a, eastern part can also be correlated laterally with very good confidence of the profile). Upl is buried below lagoonal Holocene to the surrounding boreholes (e.g., to MAG4, Figure 5). deposits (Sabatier et al. 2008), evidenced by continuous Such lateral correlation is less obvious with boreholes high-frequency sub-horizontal reflections, onlapping the located further away (e.g., MAG3).

4 J. Lofi et al. GROUND WATER NGWA.org 0 200 m Holocene lagoonal deposits of low hydraulic conductivity Line (d) Line (c) 0 W Line (b) E

2 RES 4 Depth (m) 6 Upl Truncated reflection (a)

N Line (a) Line (a) N S Line (a) S NW SE 0 0 0 (e) 01 km HoloceneRES Holocene 2 Holocene 2 2 Vic RES Lagoon 4 RES 4 4 Upl

Depth (m) A MAG3 6 D 6 6 C (b) Upl (c) Upl (d) Upl B

Figure 3. Seismic profiles CAL-IV-46 (a), CAL-IV-28 (b), CAL-IV-31 (c), and CAL-IV-34 (d) acquired in the northern part of the Vic lagoon. Upl is evidenced by discontinuous sub-parallel reflections dipping toward the SE and truncated at the top by the RES. Above, sub-horizontal reflections correspond to discordant Holocene lagoonal deposits. Vertical exaggeration: 23. (e) Location map of the seismic profiles.

By comparing several boreholes, the difference in Offshore depths of a given Upl horizon allows the general geometry Previous studies based on very high-resolution seis- of Upl beneath the lido to be determined with precision. mic profile and shallow cores describe the organization Along section CD (Figure 5), perpendicular to the coast- of the main depositional units offshore (Raynal et al. line, the depositional units are thus dipping toward the 2009, 2010). On a seismic profile acquired approximately sea at a very low angle. Along the MAG3-MAG1 section, 2 km offshore and running perpendicular to the lido, ◦ conglomerate C is dipping at an angle of less than 0.3 Upl consists of low-to-high frequency slightly divergent toward the East. Upl strata are thus gently dipping toward reflections that dip gently seaward at a mean angle of ◦ ◦ the SE at angles < 1 , which is consistent with the dips approximately 0.5 (Figure 6a). These dips are consistent calculated from the seismic profiles acquired in the Vic with the ones measured beneath the lido, in the Vic lagoon Lagoon (Figure 3). and in outcrops. On strike profile, Upl reflections are rela- From the hydrological properties of Upl, the fine- tively continuous, almost sub-horizontal with large wave- grained levels are inferred to have a low hydraulic length undulations (Figure 6b). Locally, seismic units with conductivity (<10−7 m/s). A much higher hydraulic disorganized internal reflections and erosional bases are conductivity (∼ 10−3 m/s) is expected in the coarse- observed at the top of Upl. Raynal et al. (2009) interpreted grained, non-consolidated horizons of Upl (e.g., G, S, these as patches of Pleistocene deposits. or S3, Figure 5), thus allowing groundwater circulation. Upl (and Pleistocene above) is eroded at the top by In the Roussillon basin, continental deposits of equiva- the RES, as is clear on the seismic from the underlying lent age have variable hydraulic conductivities, ranging truncated Upl reflections (Figure 6). On strike profiles, the from 5 x 10−3 to 10−7 m/s according to their litholo- RES displays topographic lows corresponding to paleo- gies (Aunay 2007). Thus, groundwater fluxes in the fine- fluvial valleys, now in-filled with transgressive Holocene grained levels, although low, cannot be neglected. sands (unit U2 in Raynal et al. 2009). In the interfluves, close to the lido, Holocene deposits did not accumulate above the RES (Figure 6b). Upl is thus outcropping on the Maguelone Cathedral sea floor. These outcrops correspond to the Maguelone The Maguelone cathedral (Figure 2c) has been built submarine plateau, consisting of continental, lacustrine, on a topographic high culminating at 12 masl and onto or marine Pliocene formations, locally overlain by conti- which the modern sandy lido is attached. The Upl age of nental conglomeratic Pleistocene series (Alabouvette et al. the deposits forming the hill is suggested by the recogni- 2003). This plateau is clearly visible on the Lidar bathy- tion of horizon G on the TGR log acquired in borehole metric map and is characterized by a rough surface MAG3 (from ∼7to∼9 mbsl), drilled in the center of (Figure 7). Another submarine Upl outcrop is visible to the hill (Figure 5). Upl deposits have also been described the southeast (Aresquiers plateau). The smooth zones from outcrops, on the edge of the hill (Ambert 2003). separating these plateaus correspond to the transgressive In addition, the presence of tephra layers (0.5 masl to sands that infill the LGM paleo-valleys (Figure 7). In term 2 mbsl) dated from 4.2 to 3.8 Ma by the same authors, of hydrological properties, these coarse-grained and non- allows the age of Upl over the study area to be refined to consolidated transgressive sands are expected to have a the Lower Pliocene. high hydraulic conductivity (Custodio 2010).

NGWA.org J. Lofi et al. GROUND WATER 5 .Lfie l RUDWTRNGWA.org WATER GROUND al. et Lofi J. 6 mbsl For 14.5 approximately variations. at conductivity lithological CILM, low are the with example, that and correlate variations (CILD to frequency high interpreted MAG1 some show borehole conductivity 4d) Figure in formation the acquired study, a logs this water of (pore In concentration conductivity). conductivity electrolyte and pore the properties electrical both rock of on because depends salinity strongly formation 1942) sedimentary the (Archie characterize water qualitatively to help Data TZ, Borehole mbsl. 34 below Lido water the pore Beneath fresh Units to Hydrogeological brackish MAG5 and mbsl, and 32 sedimentary MAG1 above mbsl). The fine System. water 55 S). of Westbay pore & consists below salty (G the Upl to horizons casing unit). using grained brackish zone. (Upl 2011 (screened coarse Transition with deposits two saturated June Pliocene 2010 with TDS, is Lower in sediments (h) February the column non-consolidated log; sampled discordantly conductivity in continental, overlying water electrical predominantly depth deposits sampled formation grained, Medium Holocene fluid from (g) through logs; went content borehole Uranium boreholes and solid from TGR distribution dissolved (f) content size MAG5: total borehole grain Vertical solid in (b) dataset Gravelly dissolved of (2009); G, Compilation al. surface; total et erosion Raynal regional TDS, after RES, (e) B.P.) log. cal. lithological simplified (years dating (a) (fraction radiocarbon MAG1: Stars, borehole sandstone; in S, dataset horizon; of Compilation 4. Figure onoeeetia odciiymaueet can measurements conductivity electrical Downhole < m;()TRadUaimlg;()Fraincnutvt os ID(ep n IM(eimdepth); (medium CILM and (deep) CILD logs: conductivity Formation (d) logs; Uranium and TGR (c) mm); 1 55 50 45 40 35 30 25 20 15 10 5 0 Depth (m bsl)

gravels/sandstones Lithology > 45600 5573 5148 (a) G S 0% distribution 0-3.9 Grain size

3.9-63 ( µ

63-125 m) 100 (b)

125-250 BOREHOLE MAG1 >250 0 0 Uranium Ray TGR Gamma Bq/kg cps Total 400 sands c e f g (h) (g) (f) (e) (c) 80 20 TZ conductivity Formation mmho CILM Salty/brackish water CILD 2000

(d) Fresh to brackish water silts 0.44

10 TDS nt htstrt h elgclseeo Kat tal. 4d): et (Figure (Krantz distinguish clearly skeleton we geological Here, the 2004). saturate hydrological distinct that two units of determination the allows 4d). also and 4c signal (Figure electrical lithology the the by that frequency driven suggesting is high thus curve, the with TGR G correlate m, clearly the horizon 32 conductivity in of Beneath variations like bottom 4d). cyclic the and at 4a mbsl, clays (Figure 16 black around with below, G. correlates observed reservoir the the peak of to conductivity top related the The at be accumulated could gas, of low presence this Alternatively, (observed horizon cores). cemented on thick 20-cm a with correlates 20 g/l 30 h nepeaino h omto odciiylogs conductivity formation the of interpretation The clays RES Upl (Lower Pliocene) Holocene Series 0 0 Ray TGR Uranium Gamma Bq/kg cps Total carbonate 250 BOREHOLE MAG5 60 20 conductivity TZ Formation electrical mmho Salty/brackish water CILM 2000 Fresh to brackish water black clays 28.84 34.08 1.41 0.72 packer packer packer 2.96 TDS

13.90 10 7.19 20 g/l 30 • • water. fresh for exploited fresh horizon evidencing sandy S3, conductivities and toward water. MAG3, formation strata pore underlying MAG1, low Upl boreholes the the show of The of of MAG4 part dip laterally. lower and a The correlated horizon suggests SE-wards. be this level can rich of and Uranium depth TGR in G in the difference Horizon low m). MAG4, 3 a every resistivity In forms point subsurface lido. measurement permanent formation (one a Maguelone observatory by deep the provided the of is and some on resistivity in TGR drilled acquired logs boreholes the (CILD) of conductivity electrical Comparison 5. Figure ntetniga es taklmtrsaebnahthe distribution beneath vertical scale the kilometer on interpretation a above at The least lido. fresher at notably a extending an overlying with unit unit, saturated hydrological is sug- salty environment upper thus subsurface data the for that conductivity used gest borehole above and The from S3) interpretation. observations our (horizon the supporting mbsl At plants, 80 watering 5). hori- around (Figure coarse-grained located a depth from zon at pumped is water water time, fresh present of MAG3 presence the borehole In is 5). ductivity (Figure MAG4 MAG5 (drilled and boreholes neighboring 4) the hydraulic in (Figure configura- low observed hydrological also a is above as tion black The acts of barrier. probably with consisting which cores conductivity clays, on interval gray correlates thick and 5), 1-m and approximately 4d a (Figures 32 mbsl approximately 34 from to extending and units hydrological GAogJ o ta.GON AE 7 WATER GROUND al. et Lofi J. NGWA.org 80 70 60 50 40 30 20 10 0

mtl 4ms otebto ftebrhl,charac- borehole, the conductivities of formation approx- bottom by the from terized to extending mbsl unit, 34 water imately brackish to fresh conduc- a formation by characterized roughly mbsl, tivities approximately 32 extending to unit, 1 water from salty to brackish a Depth (m bsl) 0mmho 20 0 ti ot oigta h Zsprtn h two the separating TZ the that noting worth is It TGR A3SRMAG2 SAR MAG3 CILD cps ∼ S3 maa rmMG) h omto con- formation the MAG1), from away km 1 2000 0 5 5 200 250 250 300 ?

< Fresh to brackish water 0 Section AB Section CD Section Section AB 0mh eo 3ms,as suggesting also mbsl, 33 below mmho 50 TGR cps > 0 mo(i.e., mmho 300 0 0 A TGR cps

Arnel lagoon

B

RES G 1 km 0mmho 30 T.Z. Form. conductivity 0cps Mediterr. > sea TGR MAG4 < C 0 mS/m). 300 D 0 mmho. 300 3000 Fresh to brackish water 0mh 2000 mmho 20 T.Z. 0 MAG1 TGR CILD cps S

Fresh to brackish water 400 Salty water Upl (Lower Pliocene) Hol. Series otemaue ewtrcnetaini h area the in close concentration is seawater which g/L). measured g/L, (35.56 34.08 the the G, to horizon to up coarse-grained reaches the concentration In 4h). (Figure above lal niaetepeec flwslnt water salinity low (TDS of mbsl 35 presence than in the taken deeper samples indicate concentrations TDS water of System. analysis clearly Westbay pore the concentration geochemical using seven 2011 MAG5, June TDS on In performed a 2011 4e). was with (Figure February g/L in water, 0.44 sampled fresh water Geochemical indicates level). borehole ground to on above artesian down m analysis is 0.80 borehole PVC at The table below. with (water PVC cased screened with and currently m, MAG1 55 is from sampled MAG1 water MAG5. pore and/or borehole on Salinities and Conductivities Water below). Pore section (see samples anal- water geochemical on by performed confirmed ysis is salinities water pore of Discussion svr ls osaae ocnrto,sgetn a of types suggesting Two present-time. concentration, the at seawater sea the to with close connection g/L) water. (34.08 very salty G deposits the horizon is of in Holocene concentration origin water the the pore to and Indeed, relation Upl in considered both are of organization lithologies geometrical The and lido. be groundwater the to of expected salty landward part is located and unit this hydrogeologic fresh present-time, upper between the in the penetrating At 8) unit (Figure unconfined. MZ hydrogeologic being the area the thus study in the sands of result Holocene to clays to due non-consolidated these unit likely by Offshore, confining replaced conductivity. a hydraulic are as hydraulic low act low northward probably clays their a extending lagoons, Holocene clays, as the the Holocene in of acts These in base 4). this mbsl the (Figure 32 to that approximately barrier) infer conductivity at (we located from extends MAG1 interval that clay unit hydrogeologic the confined upper an Unit Hydrogeologic Salty Upper hydro- distinct two containing units. for system geologic layered account a inter- to as distribution, preted order groundwater hydrological present-day in fresher Present-day observed considered overlying the 5). be water and should pore 4 processes (Figures salty con- water to lido pore brackish vertical the The of beneath Upl. sists incises distribution clays, salinity lagoonal groundwater variable or con- sands of that of deposits unit), sist Holocene discordant (Upl RES, by The covered strata seaward. locally gently of Pliocene dip composed that Lower conductivity, is hydraulic of It sub- simple. set the relatively a area, study is the structure Over 8). surface (Figure transect established onshore-offshore be to an along cross-section ological eiso eceia nlsshsbe performed been has analysis geochemical of series A esgetta eet h io oeoe penetrated boreholes lido, the beneath that suggest We hydroge- synthetic a allow observations above The < /)adslywater salty and g/L) 3 (a)

(b)

Figure 6. (a) Dip high-resolution seismic profile (IX-Survey-07) acquired offshore of the lido. The Upl reflections are gently ◦ dipping toward the sea at an angle of approximately 0.5 . (b) Strike high-resolution seismic profile (IX-Survey-39) imaging in details Upl, which consist of low frequency, sub-parallel continuous reflections. Upl is eroded at the top by the RES, and deeply incised in the axis of the paleo-valleys. The Holocene permeable sands discordantly overlie the RES. Upl outcrops locally at the sea floor (see also Figure 7). Modified from Raynal et al. (2009). See Figures 2 and 7 for profiles location.

Maguelone MAG1 3.92° 3.83° Upl 3.86° cathedral 3.89° Holocene Fig.3(a) Fig.3(c) Fig.3(b) Fig.3(d) 43.50°

Vic lagoon Maguelone plateau Maguelone lido Holocene sands 43.48° (permeable) coastline Maguelone plateau (Upl submarine outcrop) Upl Fig.6(a) Holocene sands 43.46° (permeable) Fig.6(b) bathymetry altimetry 14 12 10 8 6 4 2 0 1 2 Aresquiers plateau (Upl submarine outcrop) metres

Figure 7. Bathymetric and altitude map from lidar data. Offshore, Upl strata are outcropping on the sea floor and form the Aresquiers and Maguelone plateaus. These plateaus can allow a direct hydraulic connection between Upl reservoir and the sea (Figure 8, Pathway 3). In between, the smooth areas in the bathymetry correspond to the Holocene sands accumulated in the paleo-valleys incised in Upl during the last-glacial phase. These sands are permeable and non-consolidated, possibly providing and indirect hydraulic connection to Upl (Figure 8, Pathways 4 and 5). The plateau and infilled paleo-valleys are also clearly visible in Figure 6b. In yellow, Upl outcrops onshore. connections are considered (Figure 8). The first one relies MZ (Figure 8) between seawater and fresh groundwa- on the direct penetration of seawater into the upper aquifer ter is expected to be located seaward of the lido, at an (Figure 8, Pathway 3) via the Aresquiers and Maguelone unknown distance from the coastline. Two recharge path- plateaus where Upl crops out at the sea floor (Figures 6b ways are likely to occur onshore to account for the origin and 7). In the second possibility, the connection is indirect, of the low salinity pore water recovered below 34 mbsl occurring via the clean sands overlying the RES (Figure 8, in MAG1 and MAG5 (Figure 4). The first one invokes pathways 4 and 5). a direct recharge by meteoritic water infiltrating onland through Upl outcrops. Because Upl strata are dipping SE- Lower Fresh Confined Hydrogeologic Unit ward, the area of VMc, is likely to constitute the main We suggest that boreholes drilled on the lido pen- recharge area (Figure 8, Pathway 1). A second possible etrated a lower confined hydrogeologic unit containing pathway involves an indirect recharge thanks to a lat- fresh water flowing seaward. At the present-time, the eral connection with karstic aquifers developed in the

8 J. Lofi et al. GROUND WATER NGWA.org 0 ~5 km NW SE La Gardiole Vic Lagoon Upl submarine outcrops hill (Aresquiers and Maguelone Maguelone plateaus) 50 Maguelone lido Infilled paleo-valley

MAG1 4 3 sea-level 0 1 5 ? MZ? sea-bottom ?

Altitude (m NGF) ? ? -50

MZ Karstic aquifer ? 2 -100 Holocene clays and silts MES: margin Messinian (low hydraulic conductivity) saturated with fresh water Erosion Surface Lower Pliocene MZ mixing zone Holocene sands sequence (Upl) Miocene sequence (high hydraulic conductivity) saturated with salty water horizons possibly acting as confining RES: Ravinement Erosion Surface unit (low hydraulic conductivity) Mesozoic limestones

Figure 8. Schematic 3D organization of the deposits over the study area. The continental Pliocene unit (Upl) can be assimilated to a stratified aquitard dipping toward the sea and containing small size aquifers developed in coarse-grained levels. This aquitard shows an atypical hydrological configuration beneath the lido, with salty to brackish water overlying fresher water. Black/gray clay horizon within Upl at the fresh/salt water transition is interpreted as acting an impermeable barrier. Beside lithology and structure of Upl, several present-day and past processes are possibly involved to explain the origin of the groundwater distribution (numbering): onland recharge by fresh water from outcrops (1) or deep karsts (2). Salt water contamination resulting from a connection with the sea, directly in places where Upl outcrops offshore (3) or indirectly via the permeable Holocene sands overlying Upl (4) or infilling the paleo-valleys inherited from the LGM (5).

La Gardiole limestones (Figure 8, Pathway 2). A bore- Possible Role of Paleo-Valleys in Subsurface Flows hole drilled at the foot of the hill, northward of VMc, The relatively simple subsurface structure of the passed through a series of caves located at 88 mbsl (Pub- study area allows assessment of the possible impact of lic Subsurface Bank of the BRGM). The Upl strata are the offshore LGM paleo-valleys in groundwater/seawater geometrically pinching out on the limestones, and thus the exchanges (Figure 8, Pathway 5). Previous studies pointed lateral contribution of karstic flows in recharging the Upl out that paleo-incisions infilled with permeable sediment can act as preferred pathways for both vertical and hori- aquifer cannot be excluded. Whatever the pathway consid- zontal groundwater flows, and should thus be considered ered, we suggest that the fresh pore water below the lido as sites of increased vulnerability to salt water intrusion is gravity driven, and that seaward flows occurs at least (Foyle et al. 2002; Barlow 2003; Mulligan et al. 2007; in the levels of Upl showing the highest hydraulic con- Green 2009). In particular, simulation modeling indicates ductivities (e.g., S in Figure 4 and S3 in Figure 5). This that when such structures breach a confining unit, they interpretation is confirmed by drilling operations from the enhance fluid exchanges, with fluid outflow from the observation of flowing artesian conditions in MAG1 at aquifer to the sea occurring primarily along the channel a depth of approximately 40 mbsl that roughly correlate flanks, and inflow from the sea occurring along the chan- with the depth at which horizon S is observed (Figure 4). nel axis (Mulligan et al. 2007). In this study, beneath the The degree of confinement of the lower hydro- lido, the transition between the upper and lower hydrogeo- geologic unit can be discussed as follows. As TDS logic units fits with the presence of a clayey layer possibly decreases continually with depth (Figure 4), the lower separating the two aquifers (Figures 4 and 8). Some of the aquifer could be interpreted as a weakly confined aquifer confinement may also come from the finer grains levels where upwelling flows through the formation could gen- observed below unit G. If such a confining unit is breached offshore by the RES, it may result in a connection between erate a thick MZ. However, the clear step observed the lower fresh water compartment and the sea, enhancing around 33 mbsl (TZ) in the formation conductivity logs fluid exchange across the seafloor (Mulligan et al. 2007). (Figure 4), the hydraulic head at 0.80 masl and the Beneath the lido, this barrier currently lies at a depth of artesian conditions observed during the drilling phase approximately 32 mbsl. Assuming that this clayey layer rather suggest a well-confined system. Leakage across the has a broad extension and is dipping seaward with an ◦ confining beds however is not excluded. In the future, angle of approximately 0.5 , as measured from the seis- pumping tests and time series of salinity and pressure mic profiles (Figure 6a), this confining unit would be lying chronics should help refining the hydrological character- at approximately 48 mbsl offshore, at approximately 2 km istics of this lower hydrogeologic unit. from the coastline. This is deeper than the depth of the

NGWA.org J. Lofi et al. GROUND WATER 9 paleo-valley thalwegs (40 mbsl) observed in this area on information to stress the spatial distribution of ground- the seismic profile shown in Figure 6a. This suggests that water resources. In particular, marine geophysical data this barrier has not been breached by river incision during allow knowledge of the onshore geologic characteristics the LGM, and that the lower fresh water hydrogeologic to be extended offshore and can thus play a key role in unit may be well separated now from the upper salty understanding the effects of offshore subsurface geologic hydrogeologic unit. Vertical diffusion in this layer is not heterogeneity on groundwater/seawater exchanges. excluded and will be strongly dependent on its hydraulic This study shows that beneath the lido, the subsur- conductivity. face can be simplified as a two-layered aquifer, with an The same calculation is applied to the permeable upper salty hydrogeologic unit overlying a fresher unit. horizon G (top depth ∼14 mbsl beneath the lido). Several parameters and processes may interact at varying ◦ Assuming a dip of 0.5 seaward, this horizon should spatial and temporal scales to account for these observa- be recovered approximately 2 km offshore at a depth tions. We suggest that as a result of an onshore recharge, of approximately 30 mbsl. As evident from Figure 6b, it gravity driven fresh groundwater extends at depth beneath can conceivably have been incised in the paleo-valleys. the lido. We infer that it is overlain by salty ground- Horizon G forms a high hydraulic conductivity layer water resulting from seawater intrusion in coarse-grained compared to the sediments just above and or below. Thus, strata. We link this intrusion to the presence offshore of if breached, it may favor salt water intrusion, providing a submarine outcrops and paleo-valleys infilled with non- good hydraulic connection to the sea. This could explain consolidated sands. These paleo-valleys locally incise Upl the origin of the high salinities measured in horizon G, over a few tens of meters (e.g., ∼40 mbsl at 2 km from the with values close to seawater concentrations (Figure 4h). coastline). These could thus provide an indirect hydraulic The biggest difficulty in aquifer modeling is in connection with the sea at great depth within the sedimen- knowing what structure to put into the model, since the tary column, particularly if subsurface confining units are aquifer is essentially invisible from the surface. This is breached. The presence of paleo-valleys offshore is widely even more true in the offshore domain where subsurface known, but previous research relating to their impact on data are not easily accessible. In this study, the onshore- groundwater/seawater exchanges is limited. The proposed offshore approach appears critical, with the geological conceptual model can be used to test other Mediterranean heterogeneity very evident, reinforcing the inaccuracy porous coastal aquifers that are set in Pliocene series, apart of simply extrapolating onshore observations to the from the main coastal plains, and in which the subsurface offshore domain. Especially, determining the distribution structure is expected to be relatively similar. and depth of paleo-channels infilled with transgressive Further studies should focus on testing the proposed deposits seems critical to the assessment of possible conceptual model and on its inclusion in a groundwater groundwater/seawater exchanges in deep confined coastal model in order to examine the specific flow processes porous aquifers. that are active in this environment. In particular, time Moreover, integration of various datasets allowed series of salinity and pressure in the various layers a coherent conceptual model of subsurface aquifer in and pumping tests will allow characterization of the cross section to be presented in a way that could not hydrological parameters of the system. be obtained from individual datasets. The multimethod approach also allows assessment of scale problem and aquifer heterogeneity. This approach illustrates significant Acknowledgments new perspectives for the study and the management Part of this work was performed in the framework of coastal porous aquifers that are under increasing of the European projects ALIANCE, MUSTANG, and anthropogenic and climatic pressures, especially in the PANACEA, as well as ARPE regional funding from Mediterranean basin. Languedoc-Roussillon. The site has been equipped with a Westbay System in the framework of a collaboration with Schlumberger Water Services. Coastal seismic pro- files were acquired during the BEACHMED-E Project. Conclusion One core was funded by the project ATIP PROGELAC. As summarized by Kooi and Groen (2001), the geom- Regional platform GLADYS provided granulometry facil- etry of the interface/TZ between fresh and salty ground- ities, seismic devices (laboratoire IMAGES, UPVD) and water in coastal areas varies according to the subsurface coring facilities. Littoral offices at DREAL Languedoc- conditions. The Maguelone area is characterized by a rel- Roussillon are acknowledged for their support. G. Henry, atively simple geological context that consists of Pliocene D. Neyens, S. Barry, J. Paris, and M. Geeraert are thanked sub-parallel strata dipping seawards at low angle, over- for their help with field data acquisition. B. Aunay is grate- lain discordantly by Holocene deposits. This area has fully acknowledged for the constructive comments on a been spared from multiple cut and fill events related first version of this article. Jenny Inwood is thanked for to successive Pleistocene eustatic cycles. For the above language and grammar review. We also thank the editor, F. reasons, this site provides a natural laboratory to study Schwartz, for advice, and the three anonymous reviewers porous coastal reservoirs. The combined use of seismic who improved the manuscript through constructive com- profiles, outcrops and borehole data provides significant ments.

10 J. Lofi et al. GROUND WATER NGWA.org References Croucher, A.E., and M.J. O’Sullivan. 1995. The Henry problem Water Resources Research Abarca, E., J. Carrera, X. Sanchez-Vila, and M. Dentz. 2007. for saltwater intrusion. 31, no. 7: 1809–1814. Anisotropic dispersive Henry problem. Advances in Water Custodio, E. 2010. Coastal aquifers of Europe: An overview. Resources 30, no. 4: 913–926. Hydrogeology Journal 18, no. 1: 269–280. Alabouvette, B., M. Demange, J. Guerang´ es-Lozes,´ and P. Dagan, G., and D.G. Zeitoun. 1998. Seawater-freshwater inter- Ambert. 2003. Notice explicative, carte geologique´ de la face in a stratified aquifer of random permeability dis- France (1/250 000), feuille Montpellier (38). Orleans.´ tribution. Journal of Contaminant Hydrology 29, no. 3: Ambert, P. 2003. Early Pliocene volcanism in Maguelone (Mont- 185–203. pellier area, Herault,´ France). Comptes Rendus Geosciences Dorfliger,¨ N. 2003. The state of the French Mediterranean coastal 335, no. 15: 1051–1058. aquifers. In Tiac’03,ed.J.A.Lopez-Geta,´ J. de Dios Gomez,´ Ambert, P., J.-P. Aguilar, and J. Michaux. 1998. Evolution J.A. de la Orden, G. Ramos, L. Rodríguez, 187–206. geodynamique´ messino-pliocene` en Languedoc Central: Le Alicante: IGME. paleo-r´ eseau´ hydrographique de l’Orb et de l’Herault´ (sud Duvail, C., C. Gorini, J. Lofi, P. Le Strat, G. Clauzon, and de la France). Geodinamica Acta 11, no. 2–3: 139–146. A.T. Dos Reis. 2005. Correlation between onshore and Andersen, M.S.G., L. Baron, J. Gudbjerg, J. Gregersen, D. offshore Pliocene-Quaternary systems tracts below the Chapellier, R. Jakobsen, and D. Postma. 2007. Discharge Roussillon basin (eastern , France). Marine and of nitrate-containing groundwater into a coastal marine Petroleum Geology 22, no. 6–7: 747–756. environment. Journal of Hydrology 336, no. 1–2: 98–114. El Yaouti, F., A. El Mandour, D. Khattach, J. Benavente, and Archie, G.E. 1942. The electrical resistivity log as an aid O. Kaufmann. 2009. Salinization processes in the uncon- in determining some reservoir characteristics. Journal of fined aquifer of Bou-Areg (NE Morocco): A geostatistical, Petroleum Technology 25, no. 5: 1–8. geochemical, and tomographic study. Applied Geochemistry Aunay, B. 2007. Apport de la stratigraphie sequentielle´ ala` 24, no. 1: 16–31. gestion et alamod` elisation´ des ressources en eau des Evans, R.L., and D. Lizarralde. 2011. The competing impacts aquiferes` cotiers.ˆ PhD Thesis, University of Montpellier II. of geology and groundwater on electrical resistivity around Aunay, B., N. Dorfliger, C. Duvail, F. Grelot, P. Le Strat, Wrightsville Beach, NC. Continental Shelf Research 31, no. M. Montginoul, and J. Rinaudo. 2006. Hydro-socio- 7–8: 841–848. economic implications for water management strategies: Foyle, A.M., V.J. Henry, and C.R. Alexander. 2002. Mapping The case of Roussillon coastal aquifer. In Proceedings of the the threat of seawater intrusion in a regional coastal Colloque International sur la Gestion des Grands Aquif`eres, aquifer–aquitard system in the Southeastern United States. 150`eme Anniversaire de la loi de Darcy, 50`eme Anniversaire Environmental Geology 43, no. 1–2: 151–159. DOI: de l’AIH. Dijon, France. 10.1007/s00254-002-0636-6 Barlow, P.M. 2003. Ground Water in freshwater–saltwater Green, A.N. 2009. Palaeo-drainage, incised valley fills and environments of the Atlantic Coast. US Geological Survey. transgressive systems tract sedimentation of the Northern Bear, J., A.H.D. Cheng, S. Sorek, and D.H.I. Ouazar. 1999. Sea- Kwazulu-Natal continental shelf, South Africa, SW Indian water Intrusion in Coastal Aquifers—Concepts, Methods, Ocean. Marine Geology 263, no. 1–4: 46–63. and Practices. Dordrecht/Boston/London: Kluwer Aca- Groen, J., H. Kooi, V. Post, and J.J. De Vries. 2000. Fresh and demic Publishers. moderately brackish groundwaters in coastal plains and Biella, G., A. Lozej, and I. Tabacco. 1983. Experimental study continental shelves: Past and ongoing natural processes. In of some hydrogeophysical properties of unconsolidated Proceedings of the 16th Salt Water Intrusion Meeting, Wolin porous media. Ground Water 21, no. 6: 741–751. Island, Poland. Black, W.H., H.R. Smith, and F.D. Patton. 1986. Multiple- Hathaway, J.C., C.W. Poag, P.C. Valentine, R.E. Miller, D.M. level groundwater monitoring with the MP System. In Schultz, F.T. Mannheim, F.A. Kohout, M.H. Bothner, and Proceedings of the Surface and Borehole Geophysical D.A. Sangrey. 1979. US Geological Survey core drilling on Methods and Ground Water Instrumentation Conference the Atlantic Shelf. Science 206: 515–527. Held, R., S. Attinger, and W. Kinzelbach. 2005. Homogeniza- and Exposition. Denver, Colorado. tion and effective parameters for the Henry problem in Boughriba, M., A. Melloul, Y. Zarhloule, and A. Ouardi. 2006. heterogeneous formations. Water Resources Research 41: Extension spatiale de la salinisation des ressources en eau W11420.1–W11420.14. DOI: 10.1029/2004WR003674 et modele` conceptuel des sources salees´ dans la plaine des Henry, H.R. 1964. Effects of dispersion on salt encroachment Triffa (Maroc nord-oriental). Comptes Rendus Geosciences in coastal aquifers. U.S. Geological Survey Water-Supply 338, no. 11: 768–774. Paper 1613-C: C71–C84. Bowling, J.C., A.B. Rodriguez, D.L. Harry, and C. Zheng. 2005. Hoefel, F.G., and R.L. Evans. 2001. Impact of low salinity Delineating alluvial aquifer heterogeneity using resistivity pore water on seafloor electromagnetic data: A means and GPR data. Ground Water 43, no. 6: 890–903. of detecting submarine groundwater discharge? Estuarine, BRGM. 1996. Carte geologique´ de Sete` 1016. BRGM Editions. Coastal and Shelf Science 52, no. 2: 179–189. Brovelli, A., X. Mao, and D.A. Barry. 2007. Numerical mod- Hsu,¨ K.J., M.B. Cita, and W.B.F. Ryan. 1973. The origin of eling of tidal influence on density-dependent contami- the Mediterranean evaporites. In Initial Reports of the nant transport. Water Resources Research 43, no. 10: Deep Sea Drilling Project 13, Part 2, ed. W.B.F Ryan, W10426.1–W10426.15. K.J. Hsu,¨ and M.B. Cita, 1203–1231. Washington, D.C.: Clauzon, G. 1973. The eustatic hypothesis and the pre-Pliocene U.S. Government Printing Office. cutting of the Rhoneˆ valley. In Initial Reports of the Deep Kindinger, J.L., J.B. Davis, and J.G. Flocks. 1999. Geology and Sea Drilling Project, 13, ed. W.B.F. Ryan and K.J. Hsu,¨ evolution of lakes in North-Central Florida. Environmental 1251–1256. Washington, D.C.: U.S. Government Printing Geology 38: 301–321. Office. Kooi, H., and J. Groen. 2003. Geological processes and Cohen, D., M. Person, P. Wang, C.W. Gable, D. Hutchinson, management of groundwater resources in coastal areas. A. Marksamer, B. Dugan, H. Kooi, K. Groen, D. Lizarralde, Netherlands Journal of Geosciences/Geologie en Mijnbouw R.L. Evans, F.D. Day-Lewis, and J.W. Lane Jr. 2010. 82, no. 1: 31–40. Origin and extent of fresh paleowaters on the Atlantic Kooi, H., and J. Groen. 2001. Offshore continuation of coastal Continental Shelf, USA. Ground Water 48, no. 1: 143–158. groundwater systems; predictions using sharp-interface

NGWA.org J. Lofi et al. GROUND WATER 11 approximations and variable-density flow modelling. Jour- lagoonal environment controlled by inherited coastal mor- nal of Hydrology 246: 19–35. phology (Northern Gulf of Lions, France). Bulletin de la Kouzana, L., R. Benassi, A. Ben mammou, and M. Sfar felfoul. Soci´et´eG´eologique de France 181, no. 2: 211–224. 2010. Geophysical and hydrochemical study of the seawater Raynal, O., F. Bouchette, R. Certain, M. Seranne,´ L. Dezileau, intrusion in Mediterranean semi arid zones. Case of the P. Sabatier, J. Lofi, A. Bui Xuan Hy, L. Briqueu, P. Pezard, Korba coastal aquifer (Cap-Bon, Tunisia). Journal of and B. Tessier. 2009. Control of alongshore-oriented sand African Earth Sciences 58, no. 2: 242–254. spits on the dynamics of a wave-dominated coastal sys- Krantz, D.E., F.T. Manheim, J.F. Bratton, and D.J. Phelan. tem (Holocene deposits, Northern Gulf of Lions, France). 2004. Hydrogeologic setting and ground water flow beneath Marine Geology 264, no. 3–4: 242–257. a section of Indian River Bay, Delaware. Ground Water 42, Sabatier, P., L. Dezileau, M. Barbier, O. Raynal, J. Lofi, L. no. 7: 1035–1051. Briqueu, M. Condomines, F. Bouchette, R. Certain, U. Van Laurent, A. 1993. La gestion en bien commun des eaux Grafenstein, C. Jorda, and P. Blanchemanche. 2010. Evo- souterraines: La nappe des sables astiens de Valras lution tardi-Holocene` d’une lagune dans le Golfe du Lion (Herault),´ Une operation´ pilote en Languedoc-Roussillon. (Sud de la France). Bulletin de la Soci´ete G´eologique de PhD Thesis, Universite´ de Montpellier II. France 181, no. 1: 27–36. Lofi, J., and S. Berne.´ 2008. Evidences for Pre-Messinian Sabatier, P., L. Dezileau, M. Condomines, L. Briqueu, C. Colin, submarine canyons incisions on the Gulf of Lions Miocene F. Bouchette, M. Le Duff, and P. Blanchemanche. 2008. slope (Western Mediterranean). Marine and Petroleum Reconstruction of paleostorm events in a coastal lagoon Geology 25, no. 8: 804–817. (Herault,´ South of France). Marine Geology 251, no. 3–4: Lofi, J., C. Gorini, S. Berne,´ G. Clauzon, A. Tadeu Dos Reis, 224–232. W.B.F. Ryan, and M.S. Steckler. 2005. Erosional processes Serra, O. 1984. Fundamentals of Well-Log Interpretation. Ams- and paleo-environmental changes in the Western Gulf of terdam, The Netherlands: Elsevier. Lions (SW France) during the Messinian salinity crisis. Swarzenski, P.W., C.D. Reich, R.M. Spechler, J.L. Kindinger, Marine Geology 217, no. 1–2: 1–30. and W.S. Moore. 2001. Using multiple geochemical tracers Lofi, J., M. Rabineau, C. Gorini, S. Berne, G. Clauzon, P. De to characterize the hydrogeology of the submarine spring Clarens, A. Tadeu Dos Reis, G.S. Mountain, W.B.F. Ryan, off Crescent Beach, Florida. Chemical Geology 179, M.S. Steckler, and C. Fouchet. 2003. Plio-Quaternary pro- no. 1–4: 187–202. grading clinoform wedges of the Western Gulf of Lion Con- Teatini, P., L. Tosi, A. Viezzoli, L. Baradello, M. Zecchin, and tinental margin (NW Mediterranean) after the Messinian S. Silvestri. 2011. Understanding the hydrogeology of the salinity crisis. Marine Geology 198, no. 3–4: 289–317. Venice lagoon subsurface with airborne electromagnetics. Mulligan, A.E., R.L. Evans, and D. Lizarralde. 2007. The role of Journal of Hydrology 411, no. 3–4: 342–354. paleochannels in groundwater/seawater exchange. Journal Voss, C.I., and W.R. Souza. 1987. Variable density flow and of Hydrology 335: 313–329. solute transport simulation of regional aquifers contain- Raynal, O., F. Bouchette, R. Certain, P. Sabatier, J. Lofi, ing a narrow freshwater–saltwater transition zone. Water M. Seranne,´ L. Dezileau, L. Briqueu, P. Ferrer, and T. Resources Research 23, no. 10: 1851–1866. Courp. 2010. Holocene evolution of a languedocian

12 J. Lofi et al. GROUND WATER NGWA.org