Introduction

In the last decades, a common problem of coastal aquifers is the saltwater intrusion, which is the landward migration of freshwater/seawater boundary. The over-pumping of groundwater as well as geological heterogeneity of coastal aquifer, relative sea-level rise and coastal erosion (Ketabchi et al. 2016; Salaj et al. 2018; Meyer et al. 2019) are the main causes of this event. It can also be promoted by the presence of canals as well as other anthropogenic works, that have altered the landscape (Herbert et al. 2015), and natural subsidence related to soil oxidation (Doyle et al. 2007; Manda et al. 2014; White and Kaplan 2017).

The Mediterranean coastal zone, one of most densely populated areas in the world, is exposed to the groundwater over-exploitation (Scheidleger et al. 2004; Doll 2009; Leduc et al. 2017; Mastrocicco and Colombani 2021) and hence to the saltwater intrusion. This problem typifies several Italian coastal plains (e.g., Sardinia, Catania, Tiber, Versilia, Po) (Antonellini et al. 2008 and reference therein) where the lack of sufficient and good quality of water could also cause a reduction in fertile land and biodiversity (Zdruli 2012; Zahangeer et al. 2012; Herbert et al. 2015; Alfarrah and Walraevens 2018; Bhattachan et al. 2018).

The process of saltwater intrusion can be studied with direct and indirect methods. The first method includes the realisation of groundwater salinity profiles and groundwater sampling from monitoring wells (Kim et al. 2018 and reference therein; Jasechko et al. 2020 and reference therein). From these point measurements, the spatial complexity of saltwater dispersion can be captured by using multivariate statistical analyses tools such as factor analysis, principal component and cluster analysis (Kim et al. 2018; Busico et al. 2018; Corniello and Ducci 2014).

Anyway, the results obtained with traditional borehole survey technique reflect 'point' pollution situation. The spatial complexity of saltwater bodies could be achieved in a systematic, more rapid and relatively low-cost monitoring with indirect methods (geophysical approaches), such as the multi-electrode resistivity arrays and Airborne Electromagnetic (AEM) methods. These latter allow to map in detail the saltwater intrusion in both space and time domains (Viezzoli et al. 2010; Himi et al.; 2004, Sang-Ho et al. 2002). Among all, the 2D and 3D Electric Resistivity Tomography (hereafter ERT) gives the good results (Chitea et al 2011; Singh et al. 2013; Al-Sayed and El-Quady 2007; Gemail et al. 2004; Cianflone et al. 2018; Muzzillo et al. 2020; Ekwok et al. 2022).

This paper discussed the key role of 2D and 3D ERT to assess the salt/brackish water ingression in coastal plains. The novelty was the use of ERTs profiles and borehole stratigraphies as two mutually reinforcing types of data suitable to improve the knowledge of subsurface geology and the distribution of salty and/or brackish water in coastal areas. In this regard, geological data (boreholes and geological maps) made it possible to develop a preliminary scheme of the subsurface geology later improved by the spatial distribution and shape of geological bodies showed by the ERT profiles. In turn, the borehole stratigraphies, together with the down-hole and Multichannel Analysis Surface Wave, supported the identification of layers salt or brackish water saturated in ERT profiles. Furthermore, the potentiality of Geographic information System (GIS) was used to records, easily analyze, summarize data and to display results with thematic maps and graphics (Giménez-Forcada 2014; Tomaszkiewicz et al. 2014; Chabaane et al. 2018; Al-Halbouni et al. 2021).

The study area is the Volturno Coastal Plain (VCP); it is located in the northwestern part of the Campania Region (southern Italy). The increasing demand of freshwater for agriculture and buffalo breeding requires knowledge on the fresh water availability and possible contamination by sea water intrusion. As the other deltaic plains, it is the mosaic of sediments testifying past environments and natural events occurred in the whole catchment. During the Last Glacial Maximum (LGM, e.g., 24 to 21 ky BP), the lowering of sea level exposed the coastal plains to river erosion whose valleys were successively buried by post-LGM deposits (Boyd et al. 2006; Dalrymple et al. 2006; Tropeano et al. 2011). The subsurface geology of VCP is very peculiar since late Pleistocene its geomorphological evolution was strongly controlled by the emplacement of volcanic deposits erupted by the Campi Flegrei and Somma-Vesuvio volcanoes. This area, which remained natural until the early 1800s, was characterized by four main rivers (Volturno, Savone, Regia Agnena, Clanio) and marshy areas near the coastline (Alberico et al. 2018). In the last century, this territory has been deeply modified by human presence and anthropogenic activities (Alberico et al. 2018; Ruberti et al. 2022) and in particular between the 1950s and 1980s, urban expansion increased by about 80% (Alberico et al. 2017). The land use change occurred in the eighties, characterised mainly by the transformation of agricultural and natural areas into urbanised areas, has deeply impacted on surface runoff, aquifer recharge and groundwater flows. Only in the mid-1980s-early 1990s did the rapid decrease of new settlements (new urbanised area = 0.76%) lead to a certain stability of the coastal system and the achievement of a new balance that we observe today.

This study, which can be reproduced in all coastal plains for which geophysical and geological data are available, is fundamental as it provides the information needed to achieve a balance between availability and consumptions of freshwater within a framework of sustainable management of natural resources.

Geological and hydrogeological setting

The investigated area is the coastal plain close to the Volturno River mouth (northwestern part of Campania Region, Southern Italy). This river is 175 km long and drains an area of about 545 km2 encompassing the Molise and Campania Regions (Fig. 1). Several geological reconstructions (Ippolito et al. 1973; Aprile and Ortolani 1978; Ortolani and Pagliuca 1986; Cinque et al. 1987; Cinque et al. 2000; Bruno et al. 2000) describe the Campania Plain as a morpho-structural basin developing since Late Pliocene, along the palaeo-Tyrrhenian margin of the Apennines fold-thrust belt.

Fig. 1
figure 1

Location map of the study area at regional (a) and national scale (inset map). The shaded relief shows the landscape morphology, and the yellow rectangle includes the study area

Full size image

The Campania Plain is bounded by NW–SE, NE–SW and E–W trending faults. The main extensional event linked to the activity of NE–SW normal faults, active between 700 and 400 ka (Milia et al., 2003), produced  half-grabens structures  (from northwest to southeast: Volturno River, Patria Lake, Campi Flegrei-Acerra and the Naples Bay basins) (Fig. 1) filled by more than 5 km of Quaternary deposits (Mariani and Prato 1988; Milia et al. 2003, 2013).

Starting from the early Pleistocene, the Campania Plain has recorded the onset of rapid tectonic subsidence and marine sedimentation. From the mid-late Pleistocene, continental deposits and the products of intense explosive eruptions related to the Somma–Vesuvio district and Campi Flegrei volcanic fields  took place (De Vivo et al. 2001; Milia et al. 2003; Romano et al. 1994; Cinque et al. 1997). During the Last Glacial sea-level fall, a lowstand condition led to a deep erosional phase testified by a surface carved into mid-late Pleistocene substratum and successively dissected by multiple rivers and buried by the Campania Ignimbrite (CI; 39.85 ± 0.14 ka, Giaccio et al. 2017 and references therein). After this event of rapid vulcanoclastic aggradation, the incision of the plain restarts until reaching the minimum eustatic of 18 ky BP (Siddall et al. 2003; Rohling et al. 2014), which led to an almost complete removal of CI from the valley axis (Romano et al. 1994; Corniello et al. 2010; Amorosi et al. 2012).

The latest Pleistocene-early Holocene sea-level rise promoted the rapid flooding of lower Volturno Plain (Amorosi et al. 2012; Sacchi et al. 2014).

Since ca. 6.5 ky BP, the progressive change from transgressive to regressive conditions marked the establishment of a coastal progradation trend (ca. 4 ky BP). It was testified by the seaward migration of lagoon, beach barrier, alluvial plain and of Volturno River delta (Barra et al. 1989; Cinque et al. 1997; Amorosi et al. 2012) that continues until 1900 (Alberico et al. 2018). Several seismostratigraphic analysis of the outer shelf highlighted the continuity of this structure with a prograding wedge of Falling stage System Tract (late Pleistocene) passing upward to a well-defined Transgressive System Tract and to the Highstand System Track deposits (Iorio et al. 2014; Ferraro et al. 2017). In the plain zone, Plio-Pleistocene pyroclastic, lacustrine, palustrine and marine deposits are overlaid by CI and Holocene alluvial and pyroclastic sediments (Fig. 2). The CI is a gray ashy rock associated with black scoriae and molten lava, with a variable degree of permeability strongly related to the thickness and the physical characteristics (lithification, granulometry, amount of scoria, etc.) that play a key role, as semi-confining or confining bed, in the hydrogeological setting of plain (Corniello and Ducci 2014). It outcrops along the border of Volturno Plain with a thickness of 40–50 m and becomes thinner or absent toward the Volturno River. The main aquifer of the VCP, located in the alluvial, pyroclastic and marine porous sediments underlying the Campanian Ignimbrite, is recharged by the carbonate and volcanic  aquifers (Corniello et al. 2010). It is: (i) confined in the areas close to the carbonate reliefs (M.te Massico, M.te Maggiore), where the tufaceous deposits of CI are continuous, (ii) semi-confined in the southern sector of plain and (iii) phreatic near the coast (Corniello and Ducci 2014).

Fig. 2
figure 2

Simplified geological map (from sheet 171, “Gaeta”, 1968-modified) and piezometric contour of main aquifer (from: a Corniello et al., (2010), b Corniello and Ducci, (2014) and c Ducci et al. 2020, modified) overlaid to shaded relief of Volturno Plain

Full size image

A secondary phreatic aquifer occurs in the sand, sand-clay and clay-peat deposits lying on the IC. It is recharged by rainfall infiltration and is then primarily affected by variations in rainfall and temperature. From 2000 to 2018, the rainiest month was November and the driest August. Mean yearly temperature changed from 15.6 °C (years 2005) to 16.8 °C (years 2015 and 2018) and showed an increasing trend of + 0.43° in the period 2000–2019. The hottest months are July and August, while the coldest are December, January and February (Lasagna et al. 2019). Rainfall and temperature together with anthropogenic factors (e.g. high groundwater abstraction), can affect water exchanges, which vary both spatial and temporal domains, between the aquifer and the Volturno River (Celico et al. 2010).

This water echanges can affect the groundwater quality in fact, Corniello et al. (2010) identified high chloride values in the areas close to Volturno River Mouth and along the Volturno River, which has its bed at about − 3.5 m below sea level, and hypotized the intrusion of sea water into the river as far as Cancello Arnone. Busico et al. (2018) evidenced higher value of Na+ and Cl− due to seawater intrusion in the area close to the Volturno mouth and only recently, Schiavo et al. (2023) hypothesized that the presence of inland salt water (more than 2 km from the coastline) resulted from trapped palaeowater, rather than actual seawater intrusion, in areas with significant peat layers framed in the subsurface.

In the area close to the Volturno River and in coastal zone, the discontinuity of CI deposits favors the hydraulic continuity between aquifers that become a unique body. The transmissivity ranges from 7.4 × 10−3 to 8.3 × 10−3 m2/s (Working Group of Water Protection Plan ex Basin Authority of North-Western Campania 2006).

Materials and methods

The analysis of saltwater intrusion at Volturno Plain was realized into two sites: 1) the Volturno coastal zone, dominated by dense urban zones and 2) the rural area located north-eastward of Castel Volturno center, characterized by agricultural lands, sparse farmhouses and few residential buildings (Alberico et al. 2017). For this purpose, 13 2D-ERT (survey date: 2013), 4 3D-ERT profiles (survey dates: May and October 2013 and 2014) were surveyed (Fig. 3).

Fig. 3
figure 3

Location of geophysical survey overlaid to Google Earth image. The red lines are the traces of 2D ERT, the red rectangle evidences the area of 3D ERT survey, the blue light rectangles and the green point identify the location of MASWs and DH, respectively. The white rectangles are the drill-holes. In the inset map, the yellow points represent the DHs realized for the Municipal Urban Plan of Castel Volturno (2008)

Full size image

Furthermore, 2 Multichannel Analysis Surface Wave (MASW), 1 Down-Hole (DH) and 4 new boreholes were acquired to corroborate the interpretation of the ERT profiles (Fig. 3). The shear wave velocity (Vs), compressional waves velocity (Vp) and bulk modulus of 7 DHs (Castel Volturno Municipal Urban Plan 2021), the drill-hole A49 (Caserta Provincial Territorial Plan 2012) and CV001 (Amorosi et al. 2012), closer to the present shoreline, were also considered (Fig. 3).

2D—electrical resistivity tomography

The Wenner–Schlumberger array configuration was adopted for the ERT survey to capture lateral and vertical variations in resistivity of coastal floodplains deposits. Each electrode is both a source of current and a tool to measure electrical potential. This schema is comparable with the horizontal distribution of data points in the pseudo-section of the Wenner array, but it allows to extent of about 10% the investigation depth (Pazdirek and Blaha 1996).

Furthermore, it provides, maintaining the same distance between the electrodes, a better vertical resolution than the Schlumberger array. In the coastal zone, 11 2D ERT profiles were acquired (Fig. 3). Two systems made up of 48 electrodes spaced 4 m apart and of 96 electrodes spaced 5 m apart were used to survey the profiles from 1 to 8 and from 9 to 11, respectively (Table 1, Fig. 3).

Table 1 The number of electrodes, the distance between electrodes and the total investigated length by the single ERT profile are listed
Full size table

In the rural zone, the 2D ERT profiles from 12 to 14 (Fig. 3) were realized using a system with 96 electrodes. The equidistance between electrodes varied from 5 to 2.5 m (Table 1).

The processing of ERT sections followed a workflow consisting of three main steps: (i) removal of data with an apparent resistivity < 0.01 Ωm and with standard deviation > 5%, (ii) calculation of apparent resistivity pseudo-sections; (iii) data inversion (Loke and Barker 1996a).

3D—electrical resistivity tomography

The 3D ERT survey was carried out with the Wenner–Schlumberger array along a cross-diagonal survey to reduce the number of measurements and acquisition time (Loke et al. 2014). The potential measured, over two consecutive years, by electrodes placed on the x- and y-axes and on 45° lines (Fig. 4a) made it possible to monitor saltwater intrusion in 4D (space–time domain).

Fig. 4
figure 4

Geometric field configuration made up of 216 3D electrodes (a) and spatial distribution of sampled points (b)

Full size image

The 3D ERT survey was realized for an area of 4600 m2. It was divided in a rectangular grid made up of 9 profiles, each of one composed of 24 electrodes (total electrodes, 216) spaced 5 m apart in both x and y directions (Fig. 4b).

The same workflow adopted for the evaluation of 2D resistivity model was also applied to the 3D model. The inversion procedure was based on the smoothness-constrained least-squares routine (Loke and Barker 1996b) that determines with an iteratively method the 3D resistivity model of the subsoil.

Seismic survey (MASW and DH)

MASWs are particularly useful for measuring the Vs of all geological layers within the geophone scheme, while the DH test provides a one-point velocity (Vp and Vs) measurement (Di Fiore et al. 2015, 2020).

The MASW-01 and MASW-02 (Fig. 5) were acquired at sites 102 and 200 along the profile ERT 9 (Fig. 3). The MASW schema is made up of 72 geophones (4.5 Hz) spaced 2 m apart, a seismic source (striking Hammer) with an optimum offset of 9 m and a 24-bit seismograph to record the data.

Fig. 5
figure 5

The diagrams display the Vs, Vp and the bulk modulus values plotted versus depth; the light blue color indicates the saturated zone. In the box on the lower right, the Vs versus depth for MASW1 and MASW2 is reported

Full size image

The DH test measures the time necessary to the P and S waves to travel from the seismic source to the geophones located into the borehole. The schema adopted for the DH test are composed by 4 geophones placed in an horizontal plane at 45° one from the other and 1 located in the vertical plane (Di Fiore et al. 2020). The arrival times of P and S waves plotted against the depth made possible to outline the dromocrone and calculate the velocities of P and S waves for each slope changes. This test reached a depth of − 30 m (Fig. 5).

The recorded data were stacked to improve signal-to-noise ratio, and the SurfSeis software (Park et al. 1999) was used to process the 1-D velocity profile of Vs. The MASW reached a penetration depth of − 70 m and − 90 m (Fig. 5).

Boreholes

In the coastal plains, boreholes are the only direct investigation tool of the subsurface geology that is useful for the geological interpretation of the resistivity layers shown by 2D-ERT and 3D-ERT profiles. At this aim, we used the geological data retrieved from the boreholes S1, S2, S3, S4, realized for this work (Table 2) and from the A49 (Caserta Provincial Territorial Plan, 2005) and CV001 (Amorosi et al. 2012) closer to the coastal zone (Fig. 2).

Table 2 Thickness and lithology of boreholes drilled in the present study
Full size table

Furthermore, these boreholes together with those available from the scientific literature supported the reconstruction of palaeomorphology after the deposition of CI. This latter plays a key role in the hydrogeology; in fact, its presence can determine, as a relative impermeability, the separation between the main and shallow aquifers. The top of CI was generated with two steps: (i) the depth of the upper portion of CI, retrieved in 40 drill-holes (Pagliuca; 2009), was corrected according to the global mean sea level characterizing the MIS3.1 (Waelbroek et al. 2002) and the subsidence rate of 0.3 cm/yr defined for the last 50 ka by Ferranti et al. (2006); (ii) these values were interpolated with the inverse distance weighted method (Wu et al. 2021; Pando and Flor-Blanco 2022).

Data interpretation

The integration of 2D-ERT and 3D-ERT, MASW, DH, boreholes, information from Geological Map (sheet 171 “Gaeta”, 1968) allowed to improve the knowledge on the saltwater intrusion at VCP.

Stratigraphic schema of Volturno Coastal Plain

The VCP is characterized by three informal stratigraphic units (VU1, VU2, VU3) focused on the lithological features of deposits (Fig. 6).

Fig. 6
figure 6

A Geological sketch map of VCP. B Lithostratigraphic scheme of VCP reconstructed from boreholes. The tick marks show the location of the boreholes. C Stratigraphy of six boreholes representative of the subsurface geology. The single star indicates the boreholes made for the present work, while the double star signs the boreholes recover from scientific literature. The colors filling the boreholes indicate a connection with the layers (A, B, C, D, D') and sectors (sector1, sector 2, sector 3) drawn in the lithostratigraphic scheme (B)

Full size image

Moving landward from the present coastline the VU3 marks the transition from the dune system (layer A, sector 1) to the alluvial plain (layer A, sector 2, 3) and from the beach barrier (layer B, sector 1) to the lagoon-swamp zones (layer B, sector 2) and to alluvial plain (layer B, sector 3) (Fig. 6a, b).

The VU2 is made up of CI deposits (layer C, sectors 1 and 3) that is replaced by coarse alluvial deposits in the area of the Volturno River course (layer D, sectors 3) and is overlaid by prodelta clayey deposits close to the coast (layer D, sectors 1).

The CI is a relative impermeable placed at about − 20 m depth (measure referred to the present mean sea level) in the zones close to shoreline and of − 15 m at about 10 km landward (Fig. 7). These deposits are laterally discontinuous and have a thickness varying from 5 m, in the valley axes, to 40 m at the foot of slopes that border the plain, and it is often lacking in the area close to the Volturno River (Fig. 7).

Fig. 7
figure 7

Thickness (orange dots) and depth (lines from light cyan to brown) of Campanian Ignimbrite are displayed; the depth is referred to the present mean sea level. Data are overlaid to a gray tones shaded relief of VCP. The isobaths were derived from the Nautic map “Da Capo Circeo a Ischia e Isole Pontine” of Hydrographic Institute of Marina Militare Italiana

Full size image

The VU1 is made up of Late Pleistocene marine-alluvial deposits (layer D’, sectors 1–3). This layer is overlaid by the Campanian Ignimbrite (C) and by Holocene deposits (D) where CI is lacking (sector 2, Fig. 6a).

2D-ERT analysis

The resistivity (ρ) made it possible to identify areas of the VCP saturated with brackish and salt water.

ERTs 1–8—The interpretation of these profiles was supported by the stratigraphy of borehole S1(Fig. 6), only the most representative profile was described. It has ρ ranging between 0.4 and 15 Ωm and show three main resistivity layers (Fig. 8). The layer A has ρ ranging from 8 to 15 Ωm up to 2.5 m depth; it may correspond with sandy clay and sandy silt of alluvial and lagoon-swamp environment, respectively (Fig. 6a, b, c). The layer B, located between 4 and 18 m depth, has a low ρ raging between 0.01 and 2 Ωm (Fig. 8); it well match with sandy silt deposits of lagoon-swamp environment seawater saturated (Fig. 6a, b, c). The layer D (depth: > 18 m) has ρ varying from 3 to 8 Ωm (Fig. 8); it could be characterized by less porous deposits of layer B seawater saturated.

Fig. 8
figure 8

Electrical Resistivity Tomography sections realized in the VCP (the location is reported in Fig. 6). The dashed lines identify the boundaries between layers characterized by the same range of resistivity values, while the capital letters refer to the geological interpretation reported in detail along the text.

Full size image

ERT 9—The analysis of this profile was supported by the borehole CV001 (Fig. 6a, b, c); it shows three layers characterized by ρ ranging from 0.07 up to 35 Ωm (Fig. 8).

The layer A encompasses the first 10 m depth, it shows ρ > 10 Ωm and a wedge shape geometry slightly dipping seaward, it extends from the shoreline up to 100 m landward. The lateral continuity of this layer is interrupted by a small zone, with lower ρ, placed from 100 to 275 m landward (Fig. 8). The wedge shaped of layer A and the small zones with lower resistivity, probably coincides with sand of present day beach deposits, while the inner sector, characterized by lateral continuity, may characterized by clayey silt and fine to medium sands of dune and interdune environment (Fig. 6a, b, c). Downward, a zone with lower ρ (< 2 Ωm), extended from − 10 m to − 30 m (layer B), contains several lenticular thick bodies with ρ > 10 Ωm. This layer is characterized a sequence of sand (beach barrier environment) and silt and clayey silt deposits (prodelta environment) seawater saturated confined at depth of about 30 m by the layer C (Fig. 6) with ρ > 30 Ωm (CI) (Fig. 8). The presence of these deposits is also corroborated by the changes of share wave velocity (Vs) recorded by the MASW_01 end MASW_02 (figs. 5 and 8).

ERT 10—This profile was interpreted with the support of borehole A49 (Fig. 6a, b, c). It has a layer A about 10 m thick. The continuity of this layer, showing ρ between 8 and 30 Ωm, is interrupted by two zones, located from 15 to 125 m and from 225 to 300 m, characterized by ρ ranging from 6 to 10 Ωm (Fig. 8). It could consist of pedogenized silt and clayey silt of alluvial environment and fine to medium sands deposits of beach barrier environment (Fig. 6a, b, c). The underlying layer B, with a bottom located between − 25 m and − 40 m depth, has ρ < 2 Ωm (Fig. 8). This features could correspond with a sequence of silt and clayey silt deposits of prodelta environment seawater saturated containing lenticular structures with higher resistivity (Fig. 8). The layer D reaches about -90 m; it is characterized by lenses-shaped bodies with high ρ (15–35 Ωm). At bottom of this layer, it is possible to observe a zone (layer D’) with ρ comparable to that typify the layer B. It could correspond with a saltwater lens connected with the layer B (Fig. 8).

The ERT 11—This profile was interpreted with the support of borehole A49 (Fig. 6a, b, c). It consists of a top layer A with ρ between 6 and 15 Ωm along the first 100 m; it reaches − 10 m depth and pinches out landward reaching about − 5 m. This layer may consist of pedogenized sandy clay of alluvial plain environment. The layer B shows a sub-parallel geometry with ρ < 3 Ωm; it may consist of sand deposits (beach barrier environment) silt and clayey silts deposits of prodelta environment saltwater saturated. The lowest layer D shows a slightly increases of ρ (3–6 Ωm) (Fig. 8). This layer could correspond to the Pleistocene sandy silt with peat deposits saltwater saturated (Fig. 6a, b, c).

ERTs 12 and 13—The interpretation is supported by the borehole S2 (Fig. 6a, b, c). These profiles point out a top layer A with ρ ranging from 10 to 25 Ωm extended from field plain to about − 15 m depth. It could be made up of sandy clay of alluvial environment (Fig. 6a, b, c). The layer B, typified by ρ varying from 1.5 to 4 Ωm, is probably made up of clayey silt and silty clay with fine to silty sand deposits of alluvial plain environment (Fig. 6a, b, c) salty-brackish water saturated and contains lens-shaped bodies with lower resistivity. The lowest layer C, located at a depth of about 40 m, shows ρ > 15 Ωm and may correspond to the CI deposits. The latter, in profile 13, is locally interrupted by a body with lower ρ (2–4 Ωm) located along the profile between 160 and 290 m (Fig. 8).

The ERT 14—The interpretation is supported by the borehole S3 (Fig. 6a, b, c). This profile consists of a top layer A characterized by ρ higher than 10 Ωm (Fig. 8). It may correspond with the sandy clay with peats layer of alluvial plain environment (Fig. 6a, b, c). At a depth of − 10 m, the intermediate layer B, with ρ ranging between 4 and 10 Ωm (Fig. 8), shows a sub-parallel geometry which continues downward for about 20 m with a sequence of gray sandy silt and clayey sand of lagoon-swamp environment salty-brackish water saturated. The lowest layer C shows an increases of ρ ranging from 10 and 15 Ωm (CI deposit) (Fig. 8).

3D-ERT models

The 3D resistivity survey made possible to monitor the saltwater intrusion up to − 20 m in both spatial and temporal domains. The horizontal slices of four resistivity seasonal models point out small variations of ρ: (a) 4.36–31.2 Ωm; (b) 4.66–30.36 Ωm; (c) 7.45–32.7 Ωm; (d) 2.97–35 Ωm. In the spring period, the slices show ρ ranging from 10 to 15 Ωm in the first 6 m depth, particularly the slices of May 2014 pointed out a resistivity of 15 Ωm at the 2 m depth, while in 2015 this value is recorded at about 6 m depth. The ρ increase to 25–30 Ωm is at a depth of 18 m to return successively at lower values (5–15 Ωm) at depth of about 20 m (Fig. 9). In the autumn, all slices show higher resistivity than in the springer period and particularly, ρ ranging between 10 and 15 Ωm and between 15 and 30 Ωm at 2 and 6 m depth, respectively (Fig. 9).

Fig. 9
figure 9

3D resistivity model of May (a) and October (b) 2013 and in May (c) and October (d) 2014

Full size image

The bottom layer (at a depth of about 20 m) is characterized by discontinuous bodies with resistivity values comparable with those of layer B passing laterally to bodies with lower resistivity. The topmost layer shows low resistivity for largest zones with (for the first 4 m depth) that may correspond with more relative permeable lithology saturated by a surficial aquifer. On May, the topmost layer preserves the shape of geological bodies but appears less resistive; it is probably due to the rise of groundwater after the recharge of the springer period. The 3D-ERT better discriminated the resistivity variations in the first 6 m depth; they do not show the presence of saltwater in this zone.

Multichannel analysis of surface waves and down-hole tests

The analysis of MASW-01 points out a Vs < 200 m/s for the first 20 m depth identified as layer A in the ERT 9. At about 40–50 m depth, the Vs close to 400 m/s supports the interpretation of layer C as the upper part of CI deposits (Guadagno et al. 1995). Similarly, the MASW-02 also highlights a velocity changes at about 70 m depth evidencing a slightly deepening of CI deposits (Fig. 5).

The DH data analysis evidenced two important results: (i) the presence of a saturated layer for a depth greater than − 7 m as evidenced by the distribution of compressional wave velocity (Vp > 1200 m/s—Zelt et al. 2006); (ii) the presence of thin denser layers, located at about − 10 m and − 20 m depth, testified by the increase of shear wave velocity (Vs) and of the bulk density (Fig. 5). The DHs aided to corroborate the soil saturation evidenced by the ERTs even though these measurements were not made at same times. This was possible for the slow response of groundwater systems to weather variability or climate change (Ducci and Polemio 2018). A campaign conducted in 2015 showed for the main aquifer almost the same levels as in 2003, favored by stable rainfall (Ducci and Polemio 2018).

The density of this layers is consistent with a different diagenesis (i.e., variation in degrees of consolidation, in porosity, in textural and mineralogical characteristics) observed in the late Pleistocene-Holocene clayey units in several Mediterranean coastal plains. The higher density could be related to an over-consolidation process of sediment caused by the vertical fluctuations of the water table (Corazza et al. 1999; Bozzano et al. 2000). Such vertical fluctuations seem to be induced, in the lower parts of the river systems and close to the coastline, by the quasi-still stand of the sea level at the time of the Maximum Food Surface (see Posamentier and Allen 1999). The consolidation processes could be also due to the original depositional environments and to the consequent climatic changes (Rohrlich et al. 1995; Bonardi et al. 2004; Mancini et al. 2013; Pepe et al. 2015).

Discussion and conclusive remarks

The present research highlighted the key role of ERT surveys to delineate the spatial extent of saltwater intrusion in coastal plains. The integration into a Geographic Information System of MASWs, DHs, borehole stratigraphies and geological maps made it possible to draw a stratigraphic schema of Volturno Coastal Plain useful for the interpretation of ERT sections. This schema together with the difference in resistivity, that characterizes fresh (10–100 Ω) and saltwater (0.2 Ω) (Moulds et al. 2023 and reference therein), allowed us to recognize the distance reached by the saline wedge from the coastal belt.

The subsurface was found to consist of five resistivity layers labelled A, B, C, D and D’, the layer B is saltwater saturated and the layer D is brackish water saturated. The lateral discontinuous of all layer favors the vertical migration of groundwater that is avoided only in the zones characterized by a continuity of layer C (Campanian Ignimbrite).

More important is the presence of thin denser layers at -10 and -20 m depth highlighted by the comparison of Vs, Vp and bulk modulus of DHs tests (Fig. 5). They represent a type of relative impermeable that makes more difficult the vertical groundwater seepage between the shallow local aquifer, characterizing the layer A, and the main aquifer of layer B.

Furthermore, the discontinuity of the CI makes this deposit a non-perfect aquiclude and exposes the deep aquifer to possible contamination by surface aquifers.

The 2D ERTs evidenced two zones water saturated, also pointed out by the DHs (Fig. 5), with different salinity.

The first zone, extended from the coast up to 1.7 km inland (Fig. 10), is characterized by a very low resistivity (≤ 2 Ωm) that testifies the presence of seawater from − 4 m up to about − 30 m depth. The second zone, extended from 1.7 to 3 km inland, is characterized by a resistivity ranging from 2 to 5 Ωm that indicates the presence of salty-brackish water within a depth range from − 10 m to − 30 m (Fig. 10). The latter may be due to the mixing of salt and fresh water (salt wedge transition zone) and/or brackish water of the Volturno River that feed the aquifer (Corniello et al. 2010).

Fig. 10
figure 10

The spatial distribution of saltwater and brackish water is shown

Full size image

The 3D-ERT better discriminated the resistivity variations in the first 6 m depth and gave 2D shape of geological bodies. The resistivity values characterizing the first meters of layer A exclude the saltwater contamination in this zone as also confirmed by the groundwater samples from elm xylem and soil profiles which showed negative δ18O values, indicating the continental or meteoric nature of the groundwater (Esposito 2017, PhD Thesis).

This work shows the full potential of 2D ERTs and 3d ERT to study saltwater intrusion despite the absence of hydro-physical parameters measurements in wells suitable to validate the ERT interpretation (Folorunso et al. 2021). Anyway, the availability of DHs, to corroborate the saturation of ERT layers, together with borehole stratigraphies allowed us to recognize the presence of salt and brackish water layers.

Future analyses will be focused on some aspects that still need more investigation such as the freshwater contamination both at distances greater than 3 km inland and at greater distances from the Volturno River and others channels. Furthermore, the presence of paleo-channels, which may represent a priority passages for saltwater intrusion, as also the presence of lenses of fossil saltwater, which can contaminate surficial groundwater, will be investigated in the next future.

In the context of integrated coastal zone management, this information can help stakeholders gain insight into saltwater intrusion and provide some of the information needed to take the right actions (e.g., optimizing pumping rates, identifying the site for installing water wells) to preserve water quality status.