Introduction

Volcanic environments associated with active hydrothermal systems display a variety of surface expressions and abundant features of hydrothermal alteration. Controlled by various parameters (temperature, pressure, rock type, permeability, fluid composition, and duration), hydrothermal alteration affects petrophysical and mechanical rock properties (Browne 1978; Robb 2004; Mormone et al. 2011; Pola et al. 2012; Heap et al. 2014a; Wyering et al. 2014; Heap et al. 2014b; Mayer et al. 2015; Mormone et al. 2015; Mayer et al. 2016; Heap et al. 2017; Mayer et al. 2017). Alteration and mineral precipitation may further change the degassing processes, for example, by the development of a low-permeability cap-rock causing the pressurization of the system (Edmonds et al. 2003; Christenson et al. 2007; Ball et al. 2015; Mayer et al. 2015; Vignaroli et al. 2015; Mayer et al. 2017). If alteration and precipitation cause a build-up of pore pressure, a wide variety of eruptions type can result from steam-driven (e.g., phreatic or hydrothermal eruptions; Barberi et al. 1992; Christenson et al. 2010; Mayer et al. 2015; Montanaro et al. 2016a; Montanaro et al. 2016b; Mayer et al. 2017) to phreatomagmatic (Bertagnini et al. 1991; Houghton and Nairn 1991). Upper regions of hydrothermal systems may also be characterized by steam-heated fumarolic alteration driven by acidic sulfate-rich fluids (Hedenquist and Browne 1989; Hedenquist et al. 2000; Rye 2005). Leaching caused by these fluids can increase porosity and permeability of rocks, and eventually create vuggy silica, which facilitates a faster gas escape in the shallow zone (Ellis and Mahon 1964; Hedenquist and Lowenstern 1994; Boyce et al. 2007; John et al. 2008; Robert et al. 2014; Piochi et al. 2015; Mayer et al. 2016).

In this very complex environment, the effects of alteration in the subsoil region (i.e., within first decimeter from the surface), and the correlation of altered subsoil with the surficial part of hydrothermal systems, have remained largely uninvestigated. Instead, soil gas surveys in volcanic areas have been extensively used to estimate the budget of volatiles in volcanic systems as well as to identify and characterize active degassing structures which allow the release of gases from deep sources, i.e., magmatic bodies and/or hydrothermal systems (e.g., Allard et al. 1991; Chiodini et al. 2001; Chiodini et al. 2005; Notsu et al. 2006; Inguaggiato et al. 2012). A lesser number of studies have considered the effect of the subsoil heterogeneity on the surficial degassing behavior (Tassi et al. 2013; Bagnato et al. 2014; Bloomberg et al. 2014). A few others have considered the effect of shallow system permeability on the spatial and temporal distribution of thermal emission (Aubert et al. 2009; Schöpa et al. 2011).

In this study, we investigated the structure and composition of the subsoil and the effects of hydrothermal alteration in the crater of Solfatara volcano, within the Campi Flegrei caldera, Italy (Fig. 1). Solfatara volcano represents an area where most of the volcanic-hydrothermal activity is presently concentrated at this active caldera (Chiodini et al. 2001; Caliro et al. 2007). Here, the local fault system appears to have a major control on degassing, which in turn leads to a strong alteration of the volcanic products (Byrdina et al. 2014; Isaia et al. 2015; Mayer et al. 2016). Additionally, the maar-diatreme nature of the crater, and its filling by more recent volcanic deposits (Isaia et al. 2015), has resulted in a complex, fractured, and multi-layered structure through which the hydrothermal fluids rise. Across the crater floor, hydrothermal alteration affects the rocks, including pyroclastic fallout ash beds, pyroclastic density current deposits, and breccias differently (Mayer et al., 2016). The changes induced to their original microstructure and their physical and mechanical properties further control the system’s degassing behavior at the surface.

Fig. 1
figure 1

a Satellite image (Google Earth™ image, 2016) of Solfatara crater showing the location of investigated sites for (1) subsoil characterization and measurements of petrophysical properties, as well as for sampling of soil material for laboratory analyses (white circles) and (2) characterization of cracks around crusted hummock (orange circles). The location of a new series of electrical resistivity tomography (ERT) profiles around a selected hummock (green circle) is also reported. In the inset, the shaded relief of the central sector of Campi Flegrei shows the location of Solfatara crater and Pisciarelli in respect to close-by eruptive centers. b Geological cross-section (×2 vertical exaggeration) of the study area showing stratigraphic units and shallow fluid flows (modified from Isaia et al. 2015)

Full size image

Most geochemical and geophysical studies in this area have been carried out to investigate the subsurface of the crater floor in its first few (uppermost) hundred meters (Chiodini et al. 2001; Bruno et al. 2007; Caliro et al. 2007; Petrosino et al. 2012; Moretti et al. 2013; Caliro et al. 2014; Chiodini et al. 2015; Serra et al. 2016; De Landro et al. 2017). Results to date highlight the presence of resistive gas bodies below the main fumaroles in Solfatara, overlain by conductive descending bodies of liquid condensates (Bruno et al. 2007; Byrdina et al. 2014; Isaia et al. 2015), with fluid flows driven by both rock matrix permeability and fracture systems. Very little effort has been focused on surficial alteration within the Solfatara crater (Glamoclija et al. 2004; Piochi et al. 2015). Overall, a detailed description of the subsoil structure, the effects of alteration on it, and the consequent influences on degassing is thus lacking.

Here, we report results from a field campaign conducted in July 2015 aimed at investigating the in situ physical (temperature, humidity, permeability) and mechanical (strength, stiffness) properties of the subsoil within the Solfatara crater. In order to define the lithologies affected by distinct alteration effects, and their distribution within the crater, a detailed mapping of surficial hydrothermal features was carried out. Laboratory measurements and geochemical characterization were also performed on selected soil samples.

Crater geology and degassing

Solfatara volcano is located within the Campi Flegrei area (Fig. 1a), west of Naples, which is characterized by an active resurgent caldera system (see Vitale and Isaia 2014 and references therein). Solfatara developed 4200 BP, at the end of a series of eruptive events, which resulted in the formation of a maar-diatreme structure (Isaia et al. 2015). The volcanic activity included both phreatic and phreatomagmatic events, as well as lava dome-forming eruptions in the area around the present crater rim (Isaia et al. 2009; Vitale and Isaia 2014; Isaia et al. 2015). Geophysical (Bruno et al. 2007; Petrosino et al. 2012; Byrdina et al. 2014; Di Giuseppe et al. 2015) and field data (Isaia et al. 2009; Isaia et al. 2015) allowed reconstruction of the following: (1) the stratigraphy of the shallow part of the crater down to a depth of 150 m and (2) the complex hydrothermal system located below the Solfatara crater which includes a mix of upwelling fluids, gases, and meteoric water (Fig. 1b). Isaia et al. (2015) suggested that the deeper part of the system (below 150 m) comprises fractured rocks of the pre-eruption basement, which are mostly composed of deposits from the Agnano-Monte Spina eruptive sequence (De Vita et al. 1999), and which form a stair-step-like structure that is cut by concentric, steep ring faults (Fig. 1b and 2a). These are covered by the Solfatara deposits, which are mainly represented by phreatic and phreato-magmatic ash, lapilli, as well as breccias and collapse breccias (Isaia et al. 2009; Isaia et al. 2015). The Solfatara unit is in turn overlain by the pyroclastic deposits from Astroni volcano (less than 2 km north of Solfatara; Fig. 1) that cover most of the central sector of the caldera, and by more recent deposits. This collapsed sediment-filled structure hosts a complex system of fluid circulation, which is controlled by both rock matrix permeability and fractures (Byrdina et al. 2014; Isaia et al. 2015). Moreover, hydrothermal aquifers are present in the lower, middle, and upper part of the system, the latter emerging from the mud pool at Fangaia (Fig. 2b; Bruno et al. 2007; Petrosino et al. 2012; Isaia et al. 2015).

Fig. 2
figure 2

a Map of temperature distribution at the surface (modified from Byrdina et al., 2014). Faults (buried or surfacing) are indicated by dark red color (as reported in the geological map of Isaia et al. 2015). b Two-dimensional resistivity vertical cross-section along the Bocca Grande to Fangaia profile with overlapped Google Earth™ image (2016) draped over a DEM of the area. The principal units are the vadose zone below the crater rim, the shallow aquifers, the gas dominated reservoir, and the mixture of CO2 and gaseous/liquid water (modified from Byrdina et al. 2014)

Full size image

The fluids circulating in the shallow part of the hydrothermal system derive from a deeper zone where magmatic gases mix with meteoric water and rise as a plume to shallow depth (Caliro et al., 2007). Above the mix zone, fluids flow essentially as gas phase from a high-temperature area (> 350 °C) to a shallow zone (< 300 m) where temperatures range from 190 to 230 °C (Caliro et al., 2007). Electric resistivity tomography (ERT) profiles of this shallow zone (100–300 m) revealed the existence of a two-phase system represented by liquid- and gas-dominated zones in the subsurface of Fangaia mud pool and the fumaroles of Bocca Grande (BG) and Bocca Nuova (BN), respectively (Figs. 1 and 2b; Byrdina et al., 2014). In this shallow steam-heated part of Solfatara crater, a predominant advanced argillic alteration occurs, where sulfuric acid is created at, or above, the water table by the oxidation of H2S (Rye et al. 2005). This type of solfataric alteration causes the formation of mainly amorphous silica and alunite (Piochi et al. 2015; Mayer et al. 2016).

The hydrothermal activity at the surface is characterized by an intense soil diffuse degassing, both inside and outside of the crater (see Cardellini et al. 2017 and reference therein). Acidic (pH ~ 1.7) high-temperature (≥ 160 °C) fumaroles are mainly located in the eastern part of the crater, whereas hot springs, steam-heated pools (45–95 °C), and fumarolic vents are concentrated in its center (Glamoclija et al. 2004; Valentino and Stanzione 2004; Chiodini et al. 2011). The absence of vegetation represents the area of diffuse soil degassing as well as the solfataric alteration of the rocks (Fig. 1). Previous studies have indicated that hydrothermal alteration produced supergenic sulfate in Solfatara, with native sulfur and alunite dominating the newly formed mineral phases (Valentino and Stanzione 2004; Piochi et al. 2015). Fumarole temperatures measured at the surface range up to 145 and 150 °C for BG and BN, respectively (Fig. 2a; Valentino and Stanzione, 2004; Chiodini et al., 2015). The main components of the fumarolic emissions are H2O and CO2, followed by H2S, H2, N2, and CH4 (Caliro et al. 2007). Due to the buffering of magmatic fluids by the large hydrothermal system, components like SO2, HCl, and HF are not detected at the surface (Chiodini et al. 2001; Moretti et al. 2013). The substantial release of CO2 by fumarolic and diffuse degassing within the Solfatara-Pisciarelli area is anomalously high (> 2000 ton/day), typical of volcanoes in an eruptive phase (Chiodini et al. 2010; Chiodini et al. 2011; Aiuppa et al. 2013).

Methods—subsoil characterization

Petrophysical and mechanical soil properties were determined both in the field and in the laboratory and were correlated with the geochemical and mineral composition of the rocks. The map in Fig. 1a illustrates the locations of investigated sites for (1) subsoil characterization and measurements of petrophysical properties and (2) sampling of soil material for further laboratory analyses.

Field-based characterization of petrophysical properties

A PL-300 soil permeameter (Umwelt-Geräte-Technik) was used to measure permeability and moisture of unconsolidated material (Fig. 3). This instrument has proven to be a useful tool in previous studies on the permeability of volcanic soils (Heap et al. 2017; Mayer et al. 2017). Samples were taken by using stainless steel cylinders (l = 61 mm; d = 72 mm) and air permeability was obtained within a measurement range of 5.6 × 10−16 m2 to 6.5 × 10−13 m2. The device determines the gas volume flow through the connected pore network of a sample based on Darcy’s law. An internal vacuum pump produces the inflow of ambient air through the sample, which is defined over a throat of calibrated permeability within the apparatus. The pressure difference over the sample is recorded by a sensor, which provides the pressure gradient of the air flow. Comparison of the pressure gradient, with respect to a second gradient over the calibrated permeability of the internal throat, enables determination of the volumetric flow of air through the sample, which consequently allows calculation of its permeability. At least three individual samples were taken with the sampling cylinder and connected to the soil permeameter for measurement to account for sample heterogeneities. The obtained air permeability characterizes the water-free pore space of the sample and is therefore a function of the water content (Makó et al. 2009; Umwelt-Geräte-Technik 2012). Water content of the sample is indirectly obtained by a sensor consisting of two stainless steel rods (l = 10 cm), which measure the dielectric constant (ε). Comparing the obtained dielectric value with calibrated standards of pure water (ε = 80) and dry soil (ε = 2) yields the water content. The volume affected by the measurement is ~ 1000 cm3. Cylinders containing the samples were weighed to determine bulk density. Selected samples were immediately sealed and wrapped for transport. This technique allowed non-destructive sampling and preservation of sample texture and water content required for further analysis in the laboratory.

Fig. 3
figure 3

Field equipment used for this study. PL-300 air (soil) permeameter: (1) the connections for the sample chamber and (2) the connection for the soil moisture sensor. TinyPerm 3 Portable (rock) air permeameter: measurements are automatically stored and cataloged on an Android™ device that communicate wirelessly with the TinyPerm (© 2016 New England Research, Inc.). A thermometer was used for soil temperature measurements. A pocket vane tester and penetrometer were used for shear and unconfined compressive strength tests, respectively

Full size image

A portable air permeameter (New England Research, TinyPerm III) was used to measure the in situ permeability of cemented sandy layers and very stiff/hard fine-grained material, as well as the permeability of surficial patinas found covering the subsoil at various sites (Fig. 3). During the measurements, the selected sites were chosen to avoid the presence of fractures, which could greatly affect permeability measurements (Heap and Kennedy 2016). This handheld device calculates the Darcian permeability based on the monitored response function of the transient vacuum at the interface between the nozzle and the rock during the evacuation of air from the rock (Possemiers et al. 2012; Farquharson et al. 2015; Vignaroli et al. 2015; Heap et al. 2017). Air permeability could be obtained within a measurement range of 10−16–10−12 m2. The field characterization of the investigated unconsolidated material additionally involved two handheld devices usually applied in soil science (Fig. 3). A pocket penetrometer allowed measurements of the unconfined compressive strength (Zimbone et al. 1996) and the undrained shear strength was measured by a torvain (Farquhar 2001).

The subsoil at the investigated sites was excavated to a depth of 20 to 50 cm to measure temperature, humidity, and petrophysical properties such as permeability and strength of various layers. Several petrophysical properties exceeded the measurement range of the applied instruments (Table 1). We note that although the field methods used in this study provide valuable estimates for permeability and rock strength, their accuracy and precision are often not comparable with well-constrained laboratory methods (Heap et al. 2017).

Table 1 Measured permeability (k average), soil temperature (T), unconfined compressive strength (σ), and undrained shear strength (τ)
Full size table

Laboratory-based characterization of petrophysical properties

Soil samples were further characterized in the laboratory to determine their rock physical properties. Samples were dried in an oven at 65 °C for at least 24 h until fully dry. Weighing of samples prior and post drying allowed determination of their water content. Dry powder density of all samples was obtained by helium pycnometry (Ultrapyc 1200e®, Quantachrome). Knowing the volume of the sampling cylinder, as well as the water content and dry density, allowed determination of the porosity of the samples, which is a key parameter controlling their mechanical behavior (Pola et al. 2014).

Geochemical characterization

Wavelength dispersive X-ray fluorescence (WD-XRF) and X-ray powder diffraction (XRD) analyses were used to determine the bulk geochemical as well as the mineralogical composition of selected samples. Details about these characterizations as well as XRD and XRF data tables, are reported in the Supplementary Material.

Geophysical characterization

In January 2016, a series of electrical resistivity tomography (ERT) profiles were performed to characterize one of the investigated hummock located in the Solfatara crater (Fig. 1). We used the Wenner configuration (Loke, 2002) with inter-electrode spacing ranging from 0.4 to 3 m. Thus, our 3-D survey has a metric resolution compared to the study of Byrdina et al. (2014), where 2-D survey was performed with a spatial resolution of 10 m (Fig. 2a). Electrode coordinates were measured using a real-time kinematic global positioning system (GPS) with 1 cm accuracy. A mesh area was constructed with 5182 tetrahedral elements with a maximum depth fixed at 15 m below the surface. A total of 504 resistivity measurements were successfully inverted using E4D code (Johnson et al. 2016) in five iterations each completed when the RMS value reached one. Soil temperature measurements were carried out along one resistivity profile (SW-NE) at a depth of 30 cm following the protocol of Revil et al. (2008). In addition, an infrared thermal image of the Solfatara crater and its surroundings, captured using an airborne multispectral sensor Daedalus 1268 ATM enhanced (Borfecchia et al. 2013) on 19 December 2013 at 04:00 (UTC), was used to highlight the hummock thermal anomaly. We orthorectified and geo-referenced this 1-m-ground-resolution image with 70 benchmark points.

Results

Solfatara subsoil main domains

Detailed mapping and in situ analysis allowed us to distinguish four main domains based on the prevalent type of the subsoil and the apparent differences in alteration (Fig. 4): (1) the Fangaia domain, (2) the silica flat domain, (3) the new crust domain, and (4) the crusted hummocks domain. Field characterization of the four domains included gas permeability, soil level temperature, unconfined compressive strength, and undrained shear strength (Table 1). Figure 5 displays representative core samples with their main petrophysical properties collected from the first three domains as well as a core taken for comparison from the vegetated area inside the crater.

Fig. 4
figure 4

Satellite image (Google Earth™ image, 2016) of Solfatara crater showing the distribution of main subsoil domains and the locations of in situ investigation sites (each site is indicated with a color associated with the domain). Stratigraphic logs including permeability (m2) of soil levels, as well as temperature (°C) and strength (kN/m2) profiles, for individual measurement sites within the crater can be find in the Supplementary Material (Figs. 1 to 14). The dashed line represents the profile of the conceptual cross section shown in Fig. 12

Full size image
Fig. 5
figure 5

Photographs of the cored samples representative of the semi-consolidated material from three of the four investigated domains within Solfatara crater floor: silica flat (Loc. 1 of Fig. 4), Fangaia (Loc. 4 of Fig. 4) and new crust domain (Loc. 9 of Fig. 4). A core of semiconsolidated silt from the (unaltered) vegetated soil (Loc. 13 of Fig. 4) is given for comparison. Below are shown the petrophysical properties determined in the field and laboratory, where δ b = bulk in situ density; δ p = dry powder density; Φ = calculated porosity; k = in situ gas permeability; m = in situ moisture measured with PL-300; w = water content determined by oven drying

Full size image

The Fangaia domain, containing two bubbling mud pools, is located in the southwestern (and most depressed) part of the crater (Figs. 1 and 4). Adjacent to, and almost enclosing, the Fangaia on the western side is the silica flat domain, a white, hard-packed, and smooth-surfaced unit. The domain name derives from the fact that the subsoil has a flat surface and is rich in silica. Located in the very center of Solfatara crater is the new crust domain which is located between the depression of the Fangaia and the fourth of the mapped domains, the crusted hummocks domain. The name of this latter domain derives from the fact that the subsoil develops on hummocky terrains, building up a crust-like appearance. The new crust name instead describes the similarity to the crusted hummocks, but in an early stage of formation. These domains are surrounded by areas characterized by encrustations, reworked material, and vegetation (Fig. 4). The encrusted area is displayed in the map but has not been investigated in this study as it is outside of the flat-floored crater. These encrusted and heavily altered areas have developed above the Astroni, Solfatara, and Agnano-Monte Spina products (Isaia et al. 2015) located along the eastern and southern crater walls and have been investigated in detail by Pola et al. (2012).

A large amount of reworked material derived from anthropogenic activity is distributed at various locations inside the crater (Fig. 4). Their distribution has been mapped, but further investigation was beyond the scope of this study. Vegetated areas are mainly located in the western part of the crater, together with a small isolated area at the central part of the crater (Figs. 1 and 4). A weakly altered dark gray to brown sequence of stiff silty levels and loose sand-pebble levels characterizes the subsoil of these vegetated areas (Figs. 5 and 6). The permeability of the predominately loose layers is generally high (> 6.5 × 10−13 m2), whereas stiff silty interlayers are less permeable (~ 10−15 m2). Soil temperatures range between 25 and 35 °C in the first 5–10 cm and rise to 45 °C at a depth of 50 cm (Fig. 13 of Supplementary Material and Table 1). The vegetated subsoil is very dry, with moisture values of ~ 12%. Mineralogical analyses revealed a composition affected by hydrothermal alteration with amorphous silica being the dominant phase (62%) accompanied by alunite (14%), sanidine (12.3%), kaolinite/halloysite (6.5%), jarosite (4%), and minor amounts of quartz (1.7%; Table 1 of Supplementary Material). Taking into account the intensity of hydrothermal alteration, which affects all crater floor lithologies, the vegetated subsoil sample serves as a comparison for the other samples investigated in this study.

Fig. 6
figure 6

Example of subsoil in the vegetated area within the Solfatara crater (Loc. 13 of Fig. 4). The stratigraphic log gives the in situ permeability (k, m2) measurements

Full size image

Fangaia domain

The Fangaia domain is located in a topographic low and occupies most of the southwestern sector of the crater. Its surficial material varies in color from gray to brownish and has two main types of subsoils (Figs. 1 and 4). Around the bubbling mud pools, it mainly consists of fine-grained levels, with light-to-dark gray tones (Fig. 7a). The temperature within these levels increases from 49.5 to 83.4 °C in the first 7 cm and rises up to 91° at 25 cm. These lithologies have a low permeability (4.7 × 10−16 m2; Fig. 5) and are very stiff (2–4 KN/m2) to hard in consistency (> 4 KN/m2; see Fig. 4 of Supplementary Material). The moisture obtained in the first 10 cm ranges between 60 and 84% and decreases to 28–34% for a slightly coarser layer at a depth of approximately 30 cm. At greater distance from the mud pools, fine-grained levels are intercalated with highly permeable (> 6.5 × 10−13 m2) loose-sand levels. Soil saturation and temperatures generally decrease (Fig. 7b; see Figs. 3 and 5 of Supplementary Material). Fine-grained levels show higher humidity (59%) and lower temperature (max 50 °C), whereas coarser layers are drier and hotter (max 62 °C). In some of the excavated pits, gas emission could be observed deriving from the sandy levels in correspondence with detection of higher temperatures. Mineralogical analysis revealed that the fine-grained levels from this domain are mainly composed of amorphous silica (72%) and alunite (24%), with minor amount of quartz (1.6%) and sanidine (1.6%: Table 1 of Supplementary Material).

Fig. 7
figure 7

a Example of subsoil within the Fangaia area in the proximity of the mud pools (Loc. 4 of Fig. 4) mostly dominated by low permeable silty levels. b In the western-distal part, the subsoil is characterized by the presence of highly permeable sandy-to-pebbly levels, which are intercalated with the silty levels (Loc. 3 of Fig. 4). The stratigraphic logs give the in situ permeability (k, m2) measurements

Full size image

Silica flat domain

In the western part of the crater, the subsoil is characterized by the presence of a white hard (> 4 KN/m2) and fine-grained level (Figs. 1 and 8). This compacted level, locally covered by a centimeter-thick layer of loose sand, has a thickness that varies between 10 and 25 cm and has a very low permeability (1.1 × 10−15 m2; Fig. 5). It overlays an alternation of brownish fine-to-coarse levels with a thickness of several centimeters. The sequence resembles that of the adjacent Fangaia domain. Locally, the coarser interlayers are cemented by sulfur (see Fig. 2 of Supplementary Material); gas emission from these levels was observed during the excavation for stratigraphic investigation. The top level of this domain generally shows low temperatures ranging between ~ 35 and ~ 45 °C and relatively low moisture values of approximately ~ 20%. Towards a depth of 40 cm, temperature rises to 75–83 °C, especially within permeable coarser levels (see Figs. 1 and two of Supplementary Material). Mineralogical analyses indicate that the uppermost homogenous part of this domain consists almost entirely of amorphous silica (98%) with very minor amounts of quartz (1.5%) and alunite (0.5%; Table 1 of Supplementary Material).

Fig. 8
figure 8

Example of silica flat subsoil at the western side of Solfatara crater, showing the hard silt level, which is white in color, on top of a Fangaia-like brownish, coarse level (Loc. 1 of Fig. 4). The stratigraphic log gives the in situ permeability (k, m2) measurements

Full size image

New crust domain

This domain is located in the very central part of the crater and characterized by a white loose sand and the presence of a red-to-brown fragmentary crust (Fig. 4). The subsoil down to 50 cm shows a repeated alternation of (1) permeable (> 6.5 × 10−13 m2) coarse levels, variably cemented by sulfur and (2) low permeability (≤ 5.6 × 10−16 to 10−15 m2) fine-grained levels (Fig. 9). The proportion of fine material increases in the vicinity of the Fangaia domain. The surficial fragmentary crust shows a low permeability patina (~ 10−15 m2) which is very stiff (2–4 KN/m2) to hard in consistency (> 4 KN/m2; Fig. 5 and Figs. 8, 9, and 10 of Supplementary Material). Temperatures within the first 5–10 cm range between 50 and 68 °C and rise to 93–94 °C at depth of 50 cm (see Figs. 8, 9, and 10 of Supplementary Material). Moisture values obtained in this area are relatively low with values of approximately 20% in the upper 10 cm. The material below the surficial crust consists mostly of amorphous silica (96%) and minor amounts of sanidine (1.6%), quartz (1.3%), and anatase (0.8%; Table 1 of Supplementary Material).

Fig. 9
figure 9

Example of subsoil within the new crust domain (Loc. 8 of Fig. 4). Precipitation of native sulfur (S) is focused along coarse layers. As in this example, the surficial patina can be locally covered by a level of sand. The stratigraphic log gives the in situ permeability (k, m2) measurements

Full size image

Crusted hummocks domain

The crusted hummocks represent the predominant morphology within the crater, occupying most of the central-eastern part. To the north, east and south they are bordered by encrusted areas. Reworked material partially covers the crusted hummocks in the northern and southern parts of the crater (Fig. 4). At the surface, hummocks are characterized by (1) a surficial dark gray patina, (2) the presence of cracks, and (3) several small fumarolic vents (Fig. 10a). Native sulfur occurs typically as yellow encrustations at these fumarolic vents. Most of the cracks present on the hummocks are associated with “wrinkles” (or creases) in the hummock surface which are parallel to the cracks (Fig. 10a). These “wrinkles” are also covered by surficial patina. Their geometry indicates a spreading-like motion occurring from the central cracks. The inner parts of the crusted hummocks consist of permeable (≥ 10−12 m2) sulfur-cemented sandy levels, alternating with low permeability (≤ 5.6 × 10−16–10−15 m2) fine-grained levels (Figs. 10b; see also Figs. 11, 12 and 15 of Supplementary Material). The former typically show a bright yellow color and a certain degree of consistency due to the sulfur cementation (Figs. 10b; see also Fig. 15a of Supplementary Material). The surficial patina covering the crusted hummocks shows a quite low permeability (10−14–10−15 m2) and is very stiff (2–4 KN/m2) to hard in consistency (> 4 KN/m2). Microprobe and optical microscopy analyses revealed a composition of rounded aggregates, consisting of sulfur, quartz fragments, and amorphous silica (Figs. 10a; see also Fig. 15b-d of Supplementary Material). The surficial patina shows temperatures ranging between 30 and 40 °C. Those temperatures rise to 60–90 °C in the vicinity of the cracks. At greater depth within the patina-covered hummocks, temperature generally rises up to 70–90 °C, in particular in correspondence with permeable coarser layers (see Figs. 11 and 12 of Supplementary Material). In some of the investigated pits, gas was visibly being emitted through this coarser material.

Fig. 10
figure 10

a Top view of a large crusted hummock showing the cracks and associated “wrinkles” parallel to the cracks; “wrinkle” geometry suggests a spreading-like motion occurring around the central cracks. Inset is a detail of the surficial patina and of the coarse sulfur (S) cemented grains. b Stratigraphic log, with permeability (k, m2), ranges of the different levels, and example of vertical profile of a crusted hummock. These structures are characterized by a surficial and low-permeable patina, variably enriched in fine material, covering an alternation of permeable sulfur-cemented sandy (from fine to coarse in size) and silty levels

Full size image

Discussion

Relationship between different domains

The relationship between the different subsoil domains and their properties (permeability, temperature, and strength of the individual levels) is complex and possibly reflects those of the underneath hydrothermal system. The Fangaia domain is most likely related to (1) erosional processes around the Fangaia depression (Fig. 4) and (2) the dynamics of the boiling mud pools, which represent the emergent water table. Based on historical account and images of the Solfatara crater in the last five centuries (Sicardi L (1956) Mercati 1570; Kozák and Čermák 2010), we can assume that in time the Fangaia area has always been situated in this deeper part inside the crater, thus representing a small endorheic catchment where altered and eroded sediments from the surrounding areas can accumulate. Moreover, the mud pools have been characterized by pronounced variations in the water table mostly due to increase or decrease in steam condensation, and reduction of rocks permeability driven by hydrothermal alteration (Chiodini et al. 2001; Bruno et al. 2007). Thus, we can infer that, during the maximum extension of the mud pools, fine-grained material has been deposited over a large area covering sandy/pebbly material accumulated by erosional processes. Different cycles of water level change may have produced the alternation which we observe in the subsoil today. Close to the mud pools, where the deposition of fine-grained material has been dominant, higher surface temperatures can be measured due to the vicinity to the emitting fluids and bubbling mud body. Temperature rapidly decreases south and west-northwest from the Fangaia area where the water table rapidly deepens (Bruno et al. 2007). Here, temperatures between 70 and 90 °C can be reached only below the first meter or can be encountered within the coarse sandy/pebbly levels representing high permeability pathways which facilitate gas escape. The mineralogy detected in the Fangaia samples suggests complex acidic sulfate conditions of fluids that mostly promote alunite crystallization from very hot and acidic liquids, and sulfur precipitations from gaseous (H2S) emissions (de’ Gennaro et al. 1980; Rye et al. 1992; Piochi et al. 2015; Heap et al. 2017).

The adjacent silica flat domain (Fig. 4) has a subsoil which overlies the Fangaia sequence and is affected by decreased level of degassing. A drier and relatively colder environment may have facilitated the replacement of the initial components and the precipitation of amorphous silica. North of the Fangaia domain, and topographically part of the same depression, there is the new crust domain (Fig. 4). The new crust subsoil in the proximity of the Fangaia domain shows abundant alternating sequences of cemented coarse and loose fine-grained levels. These levels may be interpreted as material deposited by the mud pools of the Fangaia domain to the north during increased fluid supply and maximum extension of the pools. In its central and northern part, thicker levels of sand and pebbles, many of which are sulfur-cemented, became dominant. The abundance of coarser material increases towards the crusted hummocks suggesting that these mounds may have been long present in the crater so that their erosion supplied sediments to the nearby depression. The new crust domain has developed above an area where a marked fluid degassing has occurred in the last 30 years (Chiodini et al. 2016; Cardellini et al. 2017), and where a fault has been envisaged through geophysical studies (Bruno et al. 2007; Byrdina et al. 2014). Thus, the degassing of dry, acidic gases through the subsoil, which is rich in coarse-grained material, may explain the formation of a low permeability patina.

The crusted hummocks domain overlays a concentric ring fault system at the eastern and northern margins of the crater floor (Isaia et al. 2015). These heavily fractured portions of the crater are characterized by high degassing as shown in Fig. 2. Here, we suggest that both the intense degassing and precipitation of sulfur, which characterizes the inner part of the crusted hummocks, can be considered as the cause for their formation. Our hypothesis is based on the findings related to smaller and similar hummocks located in a mud pool of the nearby Pisciarelli area (Figs. 1 and 11a). At Pisciarelli, tiny hummocks show very much the same structure as the larger crusted hummocks having the following: (i) a surficial patina, yellowish-to-greenish in color, can be observed together with (ii) cracks and parallel wrinkles indicating a spreading from a central crack, and (iii) an alternation of sandy levels, cemented by sulfur, and fine-grained levels. Additionally, and in correspondence with the main surficial crack, a vertical fracture-like feature can be recognized (Fig. 11a, b). Thus, we propose the following model for the structure and formation of both the crusted and tiny hummocks: (1) a fracture-like feature is present and acts as a pathway for rising fluids, (2) on the way to the surface thermal expansion of the ascending fluid produces cracks, (3) native sulfur is precipitated at surface contributing to the formation of a low permeability surficial patina, and (4) sulfur is also precipitated amongst the grains of the coarser layers leading to cementation. Similar effects of rim formation and cementation due to sulfur precipitation, which can contribute to a decrease in permeability and create sealing levels within surficial lithological units, have been also observed and measured at Vulcano island, in Italy (Harris and Maciejewski 2000), and at Rotorua, in New Zealand (Cody 2003). In this model, here called the “crack-healing hummock” model (Fig. 11c), sulfur precipitation adds mass, and crystal growth pushes apart the ground materials producing swelling and hence the hummock. In this process, the fine-grained levels most likely accommodate the local deformation produced by the swelling. At the surface, the highest temperatures can be measured along the cracks where gas is emitted, whereas the surficial patina appears to work as a thermal insulator having lower temperatures.

Fig. 11
figure 11

a Snapshot of Pisciarelli (located in Fig. 1) showing the main hydrothermal features around the area (left), included the “tiny” crusted hummocks within the dry mud pool. A zoom of one of the tiny hummocks (right) shows surficial cracks and parallel wrinkles. b Cross-section of same tiny hummock showing features very similar to the larger crusted hummocks (Fig. 8) with (i) a surficial patina, yellowish-to-greenish in color at the surface, and (ii) an alternation of sandy levels, cemented by sulfur, and fine-grained levels in its inner part. Additionally, a vertical fracture-like feature can be recognized. c Conceptual “crack-healing hummock” model for the formation the crusted hummocks in the Solfatara crater (more details in the text)

Full size image

A conceptual cross section of the crater, in which the relationship between the different domains is captured and summarized, is given in Fig. 12. Generally, our findings are in agreement with previous studies on the spatial distribution of soil temperature at the Solfatara crater (Chiodini et al. 2001; Tassi et al. 2013; Byrdina et al. 2014). However, at individual measurement points, the soil temperature distribution can be affected by local heterogeneities in the soil properties controlling heat, and possibly gas, transfer (Tassi et al. 2013; Bagnato et al. 2014). We suggest that the presence of coarser levels in all the investigated subsoils acts as pathways for horizontal fluid circulation. This concept is supported by the fact that along coarser levels we observed (1) gas escaping during excavation; (2) the presence of native sulfur, most likely precipitated by circulating fluids (e.g., in Fig. 9); and (3) spikes in temperature profiles (e.g. in Figs. 2 and 11 of Supplementary Material). In contrast, the presence of relatively low soil permeability due to hydrothermal alteration or presence of fine-grained levels may prevent gases emission at the surface even in sites characterized by high temperature at 10 cm depth. Therefore, across the different domains, the presence of stiff and low permeability levels such as (1) fine-rich levels and (2) surficial patinas represents barriers for the gas and heat emission, which escape to the surface only along fractures (e.g., cracks in the hummocks).

Fig. 12
figure 12

Conceptual cross section through the subsoil domains (more details in the text). Section profile is in Fig. 4

Full size image

Relationship with shallower structures

In the northern part of the new crust domain, hardened crusts (i.e., the surficial patina) are present on top of hot and dry sandy/pebbly levels. The very high temperatures of the soil may be associated with hot gases rising through a highly permeably pathway. A NW-SE shallow fault (or fracture system) has been located exactly below this area by Bruno et al. (2007). Following our “crack-healing hummock” model, the encrusted features above locally sulfur-cemented sandy/pebbly levels can be seen as crusted hummocks in their initial stage. Accordingly, the presence of an inferred permeable structure underneath this area may be supported.

For a better understanding of the crusted hummock at the very shallow depth (10 m), a series of ERT profiles were performed (Fig. 13). We investigated a well-defined hummock with an area of 28 m2, in the SE part of the Fangaia mud pool (Figs. 1 and 13a). The survey consisted of five radial resistivity profiles centered on this crusted area and a circle one carried out to delineate the external area (Fig. 13a). In Fig. 13b, c, the crusted hummock is shown as a resistive structure between 50 and 100 Ωm and extends 10 m beneath the surface. It is surrounded both laterally, and at depth, by more conductive layers, especially in the north-western part of the hummock. The global resistive structure matches with infrared thermal anomalies and high soil temperatures that reach 90 °C in the central area, and which are located mostly along the cracks described above (Fig. 13a). At the surface, mineral precipitation and amorphous silica-formation create a very thin low-permeability layer (< 10−15 m2) which acts as a barrier to the emission of gas and heat. More permeable and porous sandy materials are located beneath this low permeability patina; thus, we suggest that a partially gas-saturated area can be formed in this layer and this would explain the high resistivity values at depth in the hummock.

Fig. 13
figure 13

a Location of the ERT profiles performed on the hummock (represented in green; see Fig. 1 for the hummock location) with soil temperature measurements (red dots); thermal and elevation profiles are given at the bottom of (a). Resistivity structure of the investigated hummock showing a high-resolution 3D resistivity model of the hummock, draped with the infrared image (b) and a sensitivity map of the 3D resistivity model (diagonal JTJ matrix with J being the Jacobian matrix). The model has a high accuracy for values above 10−2. The maximum sensitivity of 100 is located in the center of the hummock structure (c)

Full size image

Outlook and implications

Detailed mapping of the Solfatara subsoil may serve as a basis for further investigation of hydrothermal activity in this and other crater areas affected by degassing soils, and fumarolic emission of gas and heat. Our survey represents a snapshot of activity in an environment subjected to time-varying surface processes which affect the spatial and temporal distribution of gas and heat emission, similarly to other systems such as Vulcano (Harris and Maciejewski 2000; Schöpa et al. 2011) and Nisyros (Pantaleo and Walter 2014) islands. It also highlights the presence of very low permeable levels widely distributed within the crater which play an important role in controlling the spatial distribution of gas and heat emission, as well as surface morphology. In case of a strong unrest and/or the passage of a seismic wave, the total amount of diffused gas can strongly increase (Gresse et al. 2016) as can the temperature of the fumaroles (Chiodini et al. 1995; Harris and Maciejewski 2000; Tranfaglia et al. 2011). These effects can create overpressure conditions if the medium cannot accommodate the increase in gas flow (Marini et al. 1993; Heap et al. 2017; Mayer et al. 2017). In seismically active areas, a similar scenario is not unlikely and could lead to a sudden and localized explosion with no warning (Barberi et al. 1992; Chiodini et al. 1995), as for comparable events at Nysiros (Marini et al. 1993), Dieng (Le Guern et al. 1982), and Rotorua (Cody 2003). At Solfatara and Pisciarelli areas, the 21 August 2017 earthquake of Casamicciola (about 20 km from Solfatara), with a magnitude of four, did not produce measurable effects on the hydrothermal activity (Osservatorio Vesuviano, internal reports). In contrast, stronger seismic events such as the 1930 Irpinia earthquake, which had a magnitude of 6.7 and occurred about 100 km from Solfatara, was followed by a small explosion in the Solfatara crater area as reported by Signore (1930).

Conclusions

A field campaign and laboratory analyses were conducted to characterize the subsoil nature in the Solfatara crater at Campi Flegrei and to estimate the effect of its lithology on the ongoing degassing processes. The hydrothermal activity across the crater floor of Solfatara produced very different morphologies and types of soils. We find that their formation and evolution is most likely related to the presence of (1) the Fangaia mud pools and (2) a system of ring-faults bordering them. The identified subsoil type, their main features, and distributions can be summarized as follows:

  1. 1)

    Zones dominated by thin, laminated deposits of fine-grained material are found around the two bubbling Fangaia mud pools. With distance from the pools, intercalation of eroded sand and pebble material becomes abundant. This Fangaia domain is located in a topographic low in the southwestern sector of the crater where overflow of the pools deposited layers of sediment.

  2. 2)

    A firm fine-grained level, mostly composed of amorphous silica, has formed in the western altered part of the crater (this being the silica flat domain) where this deposit covers that of the Fangaia mud pools.

  3. 3)

    A subsoil made up of subordinate fine-grained and mainly sandy/pebbly levels characterizes the center of the crater. Localized sulfur cementation of the sandy/pebbly levels and the presence of a hard surficial patina suggest an initial stage of crusted hummock formation in relation to a shallow buried fault (this being the zone of the new crust domain).

  4. 4)

    Large mounds made of sulfur-cemented sandy/pebbly levels and minor intercalated fine-grained levels, covered by a hard, surficial patina, occupy the northern, eastern, and southern sides of the crater. Here, the crusted hummocks domain develops above an intensely fractured area associated with a concentric ring fault systems.

A vegetated and unaltered soil, made up of alternating stiff silt and loose sandy-to-pebble levels, occupies the western sector of the crater, and an isolated spot in its northern sector. Finally, piles of reworked material are distributed around the crater floor covering some of the domains.

Our study provides the first detailed subsoil stratigraphy and map of the Solfatara crater, supported by a robust dataset of petrophysical and mechanical soil properties. Results show how degassing occurs mainly along dominant crack systems or in correspondence with the mud pools. Within the subsoil, fluid circulation is favored by the presence of coarse and highly porous sulfur-hardened levels, whereas surface emission is hindered by compacted fine-grained, low permeability layers. Moreover, a correlation was found between the investigated subsoil domain, and the shallow geology was interpreted through the use of geophysical data.

The work presented here, together with recent study from Valley of Desolation in Dominica Island (Mayer et al. 2017 this issue), and from Whakaari (White Island) in New Zealand (Heap et al. 2017), highlights the importance of the subsoil role in the surficial degassing processes within hydrothermally active environments. This is consistent with other studies of the distribution of degassing and thermal emissions, and their controls, at other fumarolic systems such as Nysiros and Vulcano. At these sites, the permeability of the near-surface environment and its changes in space and time have been shown to control the spatial and temporal distribution of heat and gas fluxes (Marini et al. 1993; Chiodini et al. 1995; Harris and Maciejewski 2000; Schöpa et al. 2011; Pantaleo and Walter 2014). Whakaari, which has a highly acidic environment (pH between 0.7 and 2) similar to Solfatara, produces alunitic and advanced argillic alteration of crater-filling material, a process which has also been shown to occur at fumarolic systems in Iceland and Vulcano (Arnórsson 1995; Fulignati et al. 1996). However, the heterogeneous nature of the subsoils and the way they are altered produce a twofold pattern reflected by (i) areas where the increase in porosity and permeability favors strong degassing and (ii) areas where reduction in permeability changes the preferred pathways for hydrothermal fluid circulation, and/or produces pore pressure augmentation that may results in explosive events (Mayer et al. 2015). We note, for example, that phreatic explosion craters around the Valley of Desolation are characterized by a moderated acidic environment (pH ≤ 4) that, in combination with a strong weathering effect, produces supergene intermediate and advanced argillic alteration (Mayer et al. 2017 this issue). In this case, a decrease in both permeability and strength of the main surficial lithologies produces changes in the style and intensity of degassing through time with occasional small explosions within the valley (Mayer et al. 2017). Nonetheless, there are many cases where sealing has not resulted in explosion as at the well documented case of Vulcano (Chiodini et al. 1995; Harris and Maciejewski 2000). Therefore, in these very complex hydrothermally active environments, an understanding of the surficial degassing processes, and their relation to eventual unrest phases, depends on a detailed understanding of the alteration style and intensity, as well as the spatial distribution of lithology that form and exist within the crater. Such surveys need to be placed in the context of the eruptive history of the system, as well as other geophysical measurements, thermal surveys, and gas flux/chemistry time series (Gambino and Guglielmino 2008; Ricci et al. 2015).