To meet the increasing demand for metals to assist in a successful and rapid energy transition, it is crucial to discover more first-class mineral deposits. With most of the world's major deposits occurring near the surface, improved methods for detection at deeper levels are required. This paper summarizes the soil gas studies that have been published in English discussing the use of soil gas as a sample medium for mineral exploration. The potential and reliability of various methods and gas species (O2–CO2, sulfur gases, polymetallic studies, gaseous hydrocarbons, radiogenic daughters (He, Rn), hydrogen and other gases) are reviewed and the challenges for the broad-scale adoption of soil gas measurement as an exploration tool are discussed. Soil gas composition has promising potential for mineral exploration, but much remains to be understood about the origins and processes affecting it. There has been a great deal of variation among the studies in sampling and analytical techniques, targeted gas(es), targeted commodities and mineralization type, climatic conditions and environmental settings. Improvement is needed in technical consistency, systematic monitoring of the environmental factors shortly before and after sampling, and the impact of microbes on the composition of the gases. In addition, further study is needed into the impact of climate, the cover composition and structure as well as the biological impact of microbes and plant roots before soil gas composition is a reliable exploration method.

Supplementary material: Spatial data kmz file of gas study site locations and references presented in Figure 8 are available at https://doi.org/10.25919/5eww-8f16

Thematic collection: This article is part of the Reviews in exploration geochemistry collection available at: https://www.lyellcollection.org/topic/collections/reviews-in-exploration-geochemistry

Most of the current first-class mineral deposits occur close to the Earth's surface (<30 m; Haldar 2013), in part due to the limitations of current surface geochemical exploration techniques (Winterburn et al. 2019b). Future mineral exploration may lie in the development of novel approaches to ‘see through’ transported and residual material (referred to here as cover) to detect concealed orebodies. Current (bio)geochemical techniques encompass a range of multi-disciplinary approaches for understanding how the surface environment over mineral deposits is reflected in the plants, microbes, gases and groundwater. Geoscientists have expanded the types of media used for mineral exploration to include fine soil particles, plants, micro-organisms and groundwater (Anand et al. 2007; Leybourne and Cameron 2008; McFadden et al. 2012; Noble et al. 2013a, 2019; Gray et al. 2018; Simister et al. 2018), as well as a renewed interest in soil gas geochemistry (Noble et al. 2018; Plet et al. 2019, 2021; Lett et al. 2020a, b; Lett and Sacco 2021) as soil gases have the potential to travel through thick cover (Winterburn et al. 2019a).

The composition of soil gases is used as an exploration technique in oil and gas exploration (McCarthy and Reimer 1986; Klusman 1993), in environmental surveys to monitor for vapour intrusions (Wong and Agar 2009; Little and Pennell 2017), and in agriculture to monitor the production of greenhouse gases (Martins et al. 2016; McDonald et al. 2021). However, using soil gases that have passed through cover for mineral exploration remains largely understudied. The concept that gases are highly mobile and can migrate quickly and efficiently through and into soil is not novel. Soil gas has been used in the oil and gas industry since the 1930s, and for mineral exploration in the former Soviet Union since the 1960s (summarized in Lovell 1979; McCarthy and Reimer 1986; Klusman 1993; Hinkle and Lovell 2000) and elsewhere since the 1970s. The Barringer Research Laboratory (BRL) initiated the integration of airborne geochemistry and airborne geophysics to explore for mineral deposits during the early 1970s investigating gas, hydrocarbons and microparticulates sampled from low-flying aircraft (summarized in Bradshaw 2015). The BRL undertook some groundbreaking work in the field of surface geochemistry mineral exploration and has set solid foundation for future soil gas work. It is Lovell's PhD thesis (Lovell 1979) that builds on the BRL work and consolidated the use of soil gas applied to mineral exploration in the western world. Lovell developed protocols for collecting field samples as well as for technical and laboratory procedures. His thesis investigated O2-CO2, Hg, S and hydrocarbon gases and their potential to be indicators of several commodities (Au, Sb, Zn, Pb and polymetallic deposits). His work comprises the results of case studies completed worldwide, including Europe, Saudi Arabia and the USA. Since Lovell's thesis, there have been several studies investigating areas of known mineralization, where the geology of the overburden is defined. Three subsequent reviews have synthesized the main findings of numerous field studies and highlighted the limitations related to the use of soil gas in the exploration of mineral resources (McCarthy and Reimer 1986; Klusman 1993; Highsmith 2004).

A range of sampling techniques has been developed and implemented to improve the analyses of soil gas composition, including methods for conducting direct measurements, passively sampling gas within a container, collecting soil gas through an adsorbent, and thermal soil gas desorption from soils. Regardless of the analytical technique employed, many of these studies conclude that there is a biological impact by plants and/or microbes on the composition and abundance of gases measured at the soil surface (Butt and Gole 1985; Klusman and Jaacks 1987; Hinkle 1994). The studies also noted a difference in soil gas composition before and after rain (Lovell 1979; Klusman and Jaacks 1987; Rose et al. 1990; Johnston et al. 1993; Hinkle and Lovell 2000), confirming that weather also influences gases detected in the soil, a fact that many field studies have overlooked.

It is poorly understood how soil gases are affected by the weathering of ore, the impact microbes have on the formation and composition of these gases, and how these gases interact with their environment, all of which are needed to provide a baseline before reliably implementing improved techniques in this complex natural environment.

Three of the most common approaches for collecting and measuring soil gases are: active sampling by pumping the gases out of the soil; passive sampling by leaving the sampling material to accumulate gas over time ( around two months); and soil desorption by heating the soil sample under controlled laboratory conditions and measuring the released gases. Some studies use a combination of these techniques to study gases of interest. A summary of the soil gas field studies is presented in Tables 1 and 2, grouped according to the sampling technique and target gases.

CO2 and O2

The measurement of CO2 and O2 in mineral exploration for buried mineral deposits has been reported since the 1960s in the Soviet Union (Lovell 1979) and the late 1970s in the western world (Lovell 1979; Lovell et al. 1984). Analysis of CO2, and/or O2 is currently the most cost-effective soil gas technique and can be measured directly in the field (Lovell et al. 1984; McCarthy and Reimer 1986; Klusman 1993; Polito et al. 2002; Lett et al. 2020a, b; Lett et al. 2022) or collected and later analysed in the laboratory (Alpers et al. 1990).

The role of climate and microbes

To understand how the oxidation of sulfides at depth impacts the redox state of the overburden and affects the composition of gaseous CO2–O2 in the soil (Fig. 1; Hamilton 2007; Hamilton et al. 2004a, b), the local O2-CO2 background variations must be understood (Lovell 2000). The production of CO2 in soil is largely controlled by microbial activity, which mineralizes organic carbon into CO2 (Gougoulias et al. 2014). The composition of soil microbial ecosystems is affected by weather- and climate-induced variations in moisture, a prerequisite for microbial life (Bell et al. 2009; Nielsen and Ball 2015), which in turn can influence the O2–CO2 composition of soil gas (Ball et al. 1983, 1990). Other factors, such as temperature, soil composition and pore space, can also impact soil O2–CO2 composition (Hinkle 1994). Oxidation of organic matter in the soil, which consumes O2 and produces CO2, is also a contributing factor (Hinkle 1994). Therefore, in humid climates, the O2–CO2 variations in the soil gas associated with buried mineral deposits may be masked by increased microbial activity. Consequently, soil O2–CO2 measurements may be best suited to areas with arid conditions.

For O2–CO2 soil gas to reflect ‘true anomalies related to sulfide oxidation at depth’, the rate of gas production must be equal to or exceed the rate at which the air in the soil is completely renewed (Baver et al. 1972; Lovell 2000). This process can occur rapidly; for instance, Baver et al. (1972) noted that the total renewal of the air in average soil at a depth of 20 cm could occur in an hour. However, the type of vegetation, the activity of plant roots and microbes, and the soil organic matter content, which are dependent on seasons and climate, also impact the rate of complete air renewal in soil (Rolston 2005; Howe and Smith 2021).

Surveys repeated in the same geographical area but in different seasons can reveal similar anomalous patterns but of varying intensity (e.g. Johnson Camp area, Arizona; Lovell 2000) or they may indicate anomalous values during one sampling campaign but not during another (Colorado Plateau study site; Lovell 2000). These observations highlight that to reliably interpret CO2–O2 measurements, it is important to include control measurements to record seasonal variability outside of the area of interest. These baseline measurements are crucial for monitoring and recording the impact of meteorological variations on soil O2–CO2 composition.

Rabbit ears anomalies

A common feature in soil gas surveys over buried mineralization is the presence of ‘rabbit ears’ anomalies: a double-peaked anomaly in the edges of the cover over a buried mineral deposit (e.g. Fig. 1; Govett 1976; Govett and Atherden 1987; Hamilton 2000). The redox reactions that occur in the cover lead to the production of CO2 and consumption of O2 (Fig. 1), which has been shown in soil gas surveys by Hamilton (2000) and Hamilton et al. (2004a, b). Hamilton (2000) also observed that the reducing conditions in the sedimentary column can enable reduced gases (e.g. sulfur gases, hydrocarbon gases) to propagate through the cover to the surface, suggesting that there could be a greater potential for a soil gas anomaly in semi-arid to arid environments due to the relatively rapid migration of gases through the thick vadose zone, a finding supported by mineral exploration studies in arid to semi-arid environments (Lovell 1979; Hinkle and Dilbert 1984; Lovell et al. 1984; McCarthy and McGuire 1998; Polito et al. 2002; Cameron et al. 2004).

The use of CO2–O2 levels in soil gas as a pathfinder for mineral exploration is promising, but Hamilton (2000) notes that caution must be applied. A CO2 rabbit ears anomaly can also be associated with features at depth that are not related to economic mineralization, such as barren sulfides (e.g. pyrite), contacts between units with a strong redox contrast, barren graphite horizons, methane pockets and gas hydrates. Combining soil CO2 and O2 results with stable isotope analyses (13C and 18O) could help determine the origin of the anomalous gas (i.e. oxidizing sulfide v. soil respiration; Alpers et al. 1990) and limit false positives. Polito (1999) and Polito et al. (2002) suggest that an additional measure to increase the certainty of the origin of a soil CO2–O2 anomaly could be to also monitor other gases (such as gaseous hydrocarbons).

Sulfur gases

Sulfur gases are an obvious object of interest when targeting sulfide mineralization at depth. Weathering and oxidation of sulfide minerals can produce many sulfur gas species (see the thermodynamic modelling presented by Taylor et al. 1982 and Plet et al. 2019, 2021) and have been the focus of several field studies (Lovell 1979; Hinkle and Dilbert 1984; Oakes and Hale 1987) and laboratory investigations (Taylor et al. 1982; Stedman et al. 1984; Kesler and Gardner 1986; Plet et al. 2019, 2021). A detailed summary of work completed before 2000 can be found in Hinkle and Lovell (2000).

In the 1960s and 1970s, the analytical systems were not sensitive enough to measure soil sulfur gases (Glebovskaya and Glebovski 1960). Geoscientists’ attempts to detect mineralization using dogs trained to identify the distinctive smell produced during sulfide oxidation, and studies in Finland and the former Soviet Union, yielded encouraging results (Kahma 1965; Nilsson 1971; Kahma et al. 1975); however, similar work in Canada had low success rates in areas with known mineralization and this approach was discontinued (Brocks 1972).

During the 1970s and 1980s, a variety of sampling techniques were tested in areas with known mineralization, including (i) molecular sieves as passive adsorbents (e.g. Hinkle and Kantor 1978); (ii) soil desorption analyses (e.g. Lovell 1979; Hinkle and Dilbert 1984; Oakes and Hale 1987); and (iii) pumping of soil gases (McCarthy and Reimer 1986; McCarthy et al. 1986). However, due to the high reactivity of sulfur gases in nature, their use as mineral exploration pathfinders is a challenge.

Much of our understanding of the formation of sulfur gases during sulfide weathering is based on thermodynamic modelling and laboratory experiments under controlled conditions (Taylor et al. 1982; Plet et al. 2021). These controlled studies expanded our understanding of the interaction of sulfur gas with its environment and provided the groundwork for the use of these gases for mineral exploration (Taylor et al. 1982; Stedman et al. 1984; Kesler and Gardner 1986; Hinkle et al. 1990; Sun 2000; Plet et al. 2019, 2021).

Additional investigations highlighted unexpected differences between equilibrium thermodynamic predictions and laboratory experiments (as well as field results): thermodynamic predictions suggest the production of multiple sulfur gases, the most abundant phases being H2S and carbonyl sulfide (COS) (Taylor et al. 1982; Plet et al. 2021). However, laboratory experiments revealed three sulfur gases: carbon disulfide (CS2), COS and sulfur dioxide (SO2), with CS2 as the dominant phase (Taylor et al. 1982; Plet et al. 2021). Another key finding of these authors was the relatively greater production of sulfur gas under non-sterile conditions, suggesting that microbes influence the weathering of sulfide minerals and the production of sulfur gases (Taylor et al. 1982; Hinkle and Lovell 2000). Plet et al. (2021) examined the role of micro-organisms in the production of sulfur gases, in particular, CS2, in sterile and non-sterile experimental settings and showed that microbes enhance the production of sulfur gases. In the natural environment, microbes are active in the soil, the cover, and also directly with ore at depths, all of which impact the composition of the sulfur gases; the understanding of these processes is key to reliably using sulfur gases as a pathfinder in mineral exploration.

Polymetallic gases

There are a number of studies that have focused on the detection of metallic nanoparticles carried to the surface by deep gas and/or vapour fluxes. Whereas the BRL sampled and analysed the detected particles within 100 m above the Earth's surface using low-flying aircraft (Bradshaw 2015 and references therein), Malmqvist and Kristiansson (1984) reported results of an ascending gas flux in boreholes in mines (massive sulfide and stratabound sulfide mineralizations), which they termed ‘geogas’. In later work, Kristiansson et al. (1990) highlighted the accumulation of Au and As in snow immediately above a concealed massive sulfide orebody. They found that snow, with its large surface area, acted as a particle collector and adsorbed relatively abundant particles onto its surface. In addition, the snow had remained on the ground for a few months, which permitted averaging of the signal, indicating that the ascending gas flow was an ongoing process (Kristiansson et al. 1990). However, the use of snow samples was deemed not to be realistically useful for mineral exploration due to the risk of contamination and climate limitations.

Sampling methods evolved to a passive sampling technique, which averages the signal over one to two months (Kristiansson et al. 1990; Malmqvist et al. 1999), thus limiting the impact of short-term weather variations. A funnel collector containing a membrane was used to trap any metal nanoparticles in the ascending gas. The membranes were then analysed using a particle-induced X-ray emission (PIXE) method (Kristiansson and Malmqvist 1987). Experiments were conducted in both cold and hot semi-arid to arid climates. In cold climates, the metal particles detected were believed to have been transported to the surface by an ascending flux of gases (Malmqvist and Kristiansson 1984; Kristiansson and Malmqvist 1987; Kristiansson et al. 1990; Malmqvist et al. 1999). These studies were repeated in the hot conditions of Australia, over multiple locations of known buried orebodies. Two major differences were identified: (i) elements in the collected samples did not reflect the major elements of the orebody (to a depth up to 400 m; Johnston et al. 1992) and (ii) in hot arid environments, evapotranspiration processes are believed to drive the particles’ migration (Johnston et al. 1993) rather than deep flux.

In the 1990s, several studies investigated the technique, using various names to describe the approach: ‘geogas’ by Kristiansson et al. (1990) and Malmqvist et al. (1999); ‘SIROGAS’ by Johnston et al. (1992, 1993); ‘NAnoscale Metals in Earth Gas (NAMEG)’ by Wang et al. (1997a, b) and Xie et al. (1999).

Case studies in Asia

The geogas procedure evolved to streamline analyses, progressing from PIXE observations to instrumental neutron activation analysis, laser single atom detection and/or inductively coupled plasma mass spectrometry (ICP-MS). In the 1990s, studies in Asia were undertaken above mineral deposits using both the original and the modified sampling approach to investigate soil gases and their associated particles. Most of these studies were published in Chinese, but an English summary can be found in Wang et al. (2008). In Central China and Uzbekistan, NAMEG studies focusing on Au were conducted on extended areas in China (160 000 km2 using a sample spacing of 20–40 km) and in Uzbekistan (>1000 km transect using a 5–10 km sample spacing). In China, the NAMEG study, whose sample area included major faults and known Au deposits, revealed some large-scale anomalies (extending over 1000 km2 in the vicinity of faults). These anomalies reflect the deep Tanlu and Liaokao faults in China as well as some of the major deposits and areas of interest (Wang et al. 1997a). In Uzbekistan, the survey targeting Au and pathfinder elements As, Hg and Sb highlighted a prospective area in the vicinity of known giant Au deposits (Wang et al. 1997a; Xie et al. 1999). These studies indicate that this approach can be used in areas with varying types of cover, such as residual, marine or transported sediments, and desert sand (Xie et al. 1999).

In a study by Wang et al. (2007), a revised sampling apparatus was used, replacing the adsorbent material with a liquid collector, which decreased the risk of contamination and improved negative controls. The authors reported the findings from numerous case studies conducted in China over areas of buried deposits, highlighting the potential to monitor for both major elements and pathfinder elements known to be suitable for the targeted commodity. Using this improved technique, the multi-element approach (Fig. 2, Table 1) was successful in revealing the known concealed deposits (Wang et al. 2007).

Cao et al. (2009) further improved the sampling design by replacing the adsorbent with a transmission electron microscopy (TEM) grid. This change replaced the PIXE analyses with TEM coupled to an energy dispersive spectroscope and improved the characterization of the nanoparticles carried by the ascending gas flow (Fig. 3).

This design can be applied to passive sampling and active sampling via pumping (Cao et al. 2010). In passive sampling, the technique allows for the collectors to be left in place for a longer time (>45 days, Cao et al. 2009). This newer technique was deployed over known deposits (Au; Cao et al. 2009) and mines (Cu; Cao et al. 2010) and revealed that the nanoparticles detected above an orebody were richer in metals than the background samples. Based on these results, multiple studies using a similar approach were implemented throughout the 2010s (Cao et al. 2010, 2015; Cao 2011; Jiang et al. 2019). These studies found that the metal-bearing particles in areas of buried mineralization and their absence over barren areas indicated a connection between these particles and the concealed orebodies (Cao et al. 2009; Jiang et al. 2019). However, the surface expression (i.e. dispersion halo) of these buried ore deposits was not discussed.

Case studies in Australia

In Australia, the 1990s saw the first work on soil gas using the SIROGAS design (Fig. 4). The technique termed SIROGAS (Johnston et al. 1993) tested over a range of known deposit types (Cu–Au in the Trough Tank deposit in Queensland, Pb–Zn in South Australia, W at the W–Pb–Zn–Cu O'Callaghan deposit in Western Australia, and Ag–Pb–Zn–Cu in New South Wales) as well as in non-mineralized areas with known magnetic anomalies (Sydney Basin, Western Australia and New South Wales) to establish a background signal for non-mineralized areas. These locations have differing climates, differing cover types (e.g. weathered bedrock, aeolian sand) and a range of cover thickness (Johnston et al. 1993).

The Trough Tank deposit in Queensland, where mineralization occurs at a relatively shallow depth of c. 35 to 50 m, consistently returned anomalous values for many of the elements of interest measured (e.g. Pb, Au, Hg, Ca, K, Fe, Zn; Fig. 5) over the mineralization (Johnston et al. 1993). However, employing the same method at the O'Callaghan deposit in Western Australia, which occurs considerably deeper (c. 370 m), the mineralization was not detected.

In their study, Johnston et al. (1993) also compared the signals during dry and wet seasons and determined the optimal seasons for sampling varied with the target elements and the cover in the area. They concluded that (i) detection of the heavier elements (Pb, Zn, Au, Hg) was more successful during the wet season, and (ii) detection of lighter elements (Si, P, S, Cl, K, Ca, Ti, Cr, Mn, Fe, Cu, Ge, Sr, Br) was dependent on both the geological setting and the season. They found that the detection of light elements over sand plains and dunes was higher during the dry season, whereas over outcropping areas, the wet season returned higher measurements.

Johnston et al. (1993) suggested that in sandy areas, such as sand plains and dunes, the overburden may act as a reservoir for soil gas and they hypothesize that the surficial water vapour flux related to evaporation and transpiration increases at the end of the dry season due to higher temperatures.

In contrast, in areas with outcrop, Johnston et al. suggested that the water table recedes throughout the dry season, creating a strong and constant water vapour flux near the surface, returning low signals for light elements due to the deeper and more limited water table. They also suggested that the variations of heavy elements are controlled by the supply of metal ions. During the wet season, the high SIROGAS signal could be attributed to a more elevated water table, whereas the dry season shows lower signals due to the evaporation and recession of the water table.

Johnston et al. (1993) challenged the hypothesis that the carrier gases for metals come from the deep. Instead, these authors suggest that, in Australia, water, either as vapour or aerosol droplets, is more likely to be the principal transporting medium rather than a carrier gas derived from deep in the crust. This means that the signal detected by SIROGAS is of shallow origin, derived from groundwaters in contact with mineralization.

Multiple field studies were carried out in Australia during the 2010s that investigated soil gas variations in areas with differing regolith (cover) profiles and other surface media, such as vegetation, soils and regolith material (Anand et al. 2013). The gas collectors, which were based on Ore Hound GOCC® collectors that are designed to collect both hydrocarbon- and metal-bearing gases, were deployed over the Bentley volcanic massive sulfide Cu–Zn deposit in Western Australia and the metals from the collected gases were investigated using ICP-MS. Most of the element anomalies (i.e. Ni, Mg, Ca, Sr, Rb) that were studied highlighted the concealed orebody (Noble et al. 2013b).

Another study, which combined soil gas results in varying media, was undertaken on the North Miitel Ni deposit in Western Australia (Noble et al. 2013a). There, the polymetallic soil gas study detected a weak Ni anomaly over the buried mineralization.

Most of these studies supported a link between metal-bearing gas detected at the surface and concealed mineralization, but the processes and timing of metal migration via soil gas remained poorly understood. A series of long-term (two year) pit experiments were undertaken (Noble et al. 2013b; Anand et al. 2014, 2016). The design of the experiment (Fig. 6) permitted investigation of the processes liberating and carrying the metals to the surface. In these experiments, samples of known ores (i.e. Cu–Zn–Pb–Ag, Au, salts of Pb, Ni, Cu and Zn nitrate) were buried in a 2 m deep pit and covered by sand to sand–clay material. Soil gases were sampled using Ore Hound GOCC® collectors. The sampling media were replaced every three months, and the samples collected were analysed for metal and hydrocarbon content.

The results indicated that at the start of each summer, a single pulse of Cu and Zn was released at each pit, the strength of which varied depending on the pit setting. These differences were tentatively attributed to the biological processes related to which root and microbes were present in each pit, which may have impacted the volatilization of the metals. However, the overall results were considered inconclusive (Noble et al. 2013b).

These results from the pit experiments by Noble et al. supported findings previously reported by Johnston et al. (1993) and demonstrated that, in Australia, evapotranspiration processes are responsible for metal migration in gases. These evapotranspiration processes are largely driven by capillarity (Fig. 7), namely that metals, in water-soluble form, can easily be transported to the surface soil by water vapour (Anand et al. 2016). In arid climates, this is likely the driving force for the metal content in soil gases over ore at shallow depths.

These surveys indicated that climate is likely a primary factor for the differences observed between soil gas tested in cold and hot climates. Although the migration processes of the soil gas may vary by climate, the results obtained by this technique were similar, providing encouraging results for application to exploration for concealed orebodies.

Other polymetallic studies

A passive sampling survey by Pauwels et al. (1999) conducted over the giant massive sulfide deposits in the SW Iberian Pyrite Belt, Spain investigated several soil gases (CO2, He, Rn) and gas-carried metals as well as soil composition. The authors reported that metals adsorbed onto activated carbon showed anomalies over the concealed orebodies; however, these metal anomalies were not correlated to the metals in soils. They also noted that the findings were reproducible during both wet and dry seasons, suggesting that the cover effect, hypothesized by Johnston et al. (1993), may not always have an impact on soil gas composition.

Additional field surveys were conducted over Cu, Fe–Cu–Au and Cu–Au–Mo prospects in Chile (Castillo et al. 2015). In these studies, passively collected samples were processed using statistical and geostatistical evaluations, which allowed comparison of the results among the sites. At each site, elements were grouped using a statistical treatment of the data. Of the three case studies, only one appeared to link some of the metals (i.e. Co, Sn and Cu) to the buried lithology (Castillo et al. 2015). The anomalous values of others metals were thought to (i) be related to a structural control, or (ii) have an indirect relation to superficial deposits due to a combination of morphological characteristics and groundwater flow directions.

Hydrocarbons are organic molecules consisting exclusively of C and H. Hydrocarbon gases have routinely been used in oil and gas exploration and thus detection techniques for these gases are well developed. In contrast, links between the source of gaseous hydrocarbons in the soil and the occurrence of mineral deposits at depth remain unclear. In these settings, the concentration of gaseous hydrocarbons associated with a deposit is likely to be low and their composition complex.

The potential use of these gases as pathfinders for mineral exploration has been considered since the late 1970s (e.g. Arias et al. 1979, 1982; Lovell 1979). Later studies (e.g. Disnar and Gauthier 1988, 1989; Disnar 1990) suggest that evaluating the maturity of soil gas hydrocarbons released from a host sediment can contribute to identifying sediment-hosted buried mineralization. These authors’ findings relied on the size of the hydrocarbon molecules to determine the maturity of the soil gas hydrocarbons: the larger the molecule, the less mature it is. The relative proportion of ethane (C2H6) to methane (CH4) was used as a proxy to determine the maturity of the sedimentary host rock, and this is believed to highlight changes in maturity related to the contact with relatively ‘hot’ mineralizing fluids during deposit formation. The hypothesis was that these fluids increased the maturity of the host sediment (through oxidation and temperature) and therefore decreased the proportion of relatively larger molecules (i.e. ethane). When samples of the host sediment were heated, they were found to release the lighter methane hydrocarbons in greater relative abundance than the larger ethane molecule.

Building on these results, Mulshaw (1996) compared three methods for extracting hydrocarbon gases within a mineral exploration context: (i) the ‘traditional’ desorption method, which was developed by Carter et al. (1988), consisted of heating coarsely crushed samples at 210°C for 150 min in a sealed container followed by analysing the headspace gases using gas chromatography; (ii) the ‘dry grinding’ method, which consisted of grinding a sample in a sealed mill for 30 min followed by analysing the headspace gases by gas chromatography; and (iii) crushing only those samples with a high limestone content and adding an ethylenediaminetetraacetic acid solution, which was then sealed and agitated for 15 h, after which the samples were cooled and the headspace gas analysed by gas chromatography. These two latter techniques released gases with higher concentrations of light hydrocarbons with ‘no apparent decrease in precision’ (Mulshaw 1996, p. 275).

However, some of the trends observed using these three methods reported anomalies that were not seen using the ‘traditional’ thermal desorption technique. Unfortunately, these newly detected anomalies could not be linked to any deposit or mineral discovery during subsequent exploration and were instead found to be related to faults and fractures.

Although these novel techniques did not reveal buried mineralization, their higher precision and sensitivity highlighted the importance of understanding the structural geology of the cover before being able to reliably use soil gas hydrocarbon to reveal buried mineral deposits.

Commercial laboratories have created a number of packages for analysing soil gas for hydrocarbons: soil gas hydrocarbon (SGH™) by Activation Laboratories, soil desorption pyrolysis (SDP™) by SDP Pty. Ltd. (no longer commercially available), and amplified geochemical imaging (AGI™) by Amplified Geochemical Imaging LLC). Of these commercial techniques, two release gases from soil samples (SGH™ and SDP™) and the other (AGI™) uses a specially engineered resin that is deployed for two months in the field and then recovered and submitted to the laboratory for desorption and analysis. The existence of these commercial techniques as well as several press releases by exploration companies demonstrates that there is an interest by companies in the use of soil gas as a medium for mineral exploration.

To date, only a few scientific publications have reported investigations of hydrocarbons over mineral deposits (Polito 1999; Polito et al. 2002; Luca Palacios 2012; Noble et al. 2013a, 2018; Pizarro Pavez 2016). These studies investigated soil gas hydrocarbon concentrations in transects over known mineralization in the arid regions of Western Australia and Chile. The analyses of hydrocarbon gases were coupled with traditional exploration geochemistry (Luca Palacios 2012) or with other gas proxies, such as O2–CO2 gases (Polito et al. 2002), or with analyses of metals adsorbed onto passive samplers and released using a partial leach technique (Noble et al. 2013a, 2018).

Noble et al. (2013a) reported variations in the hydrocarbon content that was detected in the soil gases on a transect over the North Miitel Ni–sulfide deposit (Australia) using the SGH™ technique. This dataset revealed the gaseous hydrocarbon content increased in the area over the concealed mineralization in addition to a rabbit ears anomaly directly above the buried mineralization.

Though these field studies showed encouraging results, the use of volatile hydrocarbons as a pathfinder for mineral exploration should be employed with caution. The lack of details provided by the commercial laboratories and the variability in the hydrocarbon compounds above concealed mineralizations impact the reliability of the approach. It is crucial to understand where and how these hydrocarbons formed, how they migrate and how they are impacted during their migration toward the surface.

To further the understanding of how soil gas hydrocarbons can be used in exploration, passive sampling for gaseous hydrocarbons was also included in the pit experiments discussed above (Noble et al. 2013b), with the major objective of monitoring and understanding gaseous hydrocarbon production over weathering ore. However, the setup contained abundant plastic material (a known source of hydrocarbon contaminants) and the paucity of detail in the commercial laboratory report did not allow for a meaningful interpretation of hydrocarbon results (Noble et al. 2013b). Adding to the uncertainty, the hydrocarbon compounds used as pathfinders in SGH™ analyses remain confidential.

Radon and helium in soil gases have both been studied as potential pathfinders for mineral deposits (Arias et al. 1979), uranium ore in particular (Butt and Gole 1985). Despite field studies (Pogorski and Quirt 1978; Reimer et al. 1979; Butt and Gole 1985) and some review work (Dyck 1976; Rose et al. 1990), Rn and He in soil gas have not shown to be promising indicators for uranium ore exploration (Butt and Gole 1985; Butt et al. 1987). This finding was further supported by studies that monitored the natural variations of Rn and He in non-mineralized areas (Klusman and Jaacks 1987; Hinkle 1994). These studies revealed that the large variations of Rn and He gases in soils are related to the seasons and environmental variations in temperature, humidity and barometric pressure. Klusman and Jaacks (1987) reported that these environmental parameters account for 83% of the Rn variance and 33% of the He variance, making the use of these gases as indicators for buried U mineralization challenging. Hinkle (1994) supported the theory that large variations in soil gases could be related to meteorological conditions but also suggested that the local conditions, such as soil type and organic content of the soils, may also impact these gases.

When Ball et al. (1990) reported field analyses for 222Rn and 220Rn (also referred to as Tn, thoron) in parallel to other gases (CO2, O2), they suggested the approach may be applicable in exploration for sulfide mineralization in some tropical environments. However, this interpretation appears mainly reliant on the CO2–O2 variations rather than the radiogenic daughters.

Overall, published work indicates that a wide range of environmental and climatic factors can affect both Rn and He concentrations in soil gases. As such it seems unlikely that these gases would present sufficient contrast in the presence of a mineral deposit to be detected against highly variable background concentrations.

The ongoing shift from fossil fuels to green energy has seen an increasing interest in natural hydrogen. The reader is referred to a review by Zgonnik (2020), who presents detailed maps of worldwide H2 seep locations. The seeps included in the report contain over 10% H2 in free gases, gas inclusions and groundwater. Zgonnik discusses the context and provides the data supporting the association of H2 with various rock types, including orebodies, kimberlites, Precambrian and ultrabasic rocks amongst others.

Seeps of natural H2 have been detected in a wide range of geological settings, and many studies have focused on characterizing the seeps and understanding the processes by which the H2 formed (Gole and Butt 1985; Prinzhofer et al. 2019; Myagkiy et al. 2020a; Boreham et al. 2021a, b). Other studies have focused on identifying and quantifying the processes that affect the concentration of H2 at the surface (Myagkiy et al. 2020a, b).

The relevance of natural H2 for mineral exploration, and thus to this review, may appear far-fetched; however, one hypothesis suggests that this natural H2 could be derived from water–rock interactions and redox processes in the crust, including Fe oxidation, and is believed to be the probable source of H2 in onshore emanations (Geymond et al. 2022; Milkov 2022). Additionally, in continental environments, one of the geological contexts within which H2 seeps have been identified is ‘stable intracratonic basins above Archean to Proterozoic basement’ (Moretti et al. 2021, p. 2), a setting in which mineral deposits also occur. Expanding on these premises, a recent study by Boreham et al. (2021b) investigated H2, CH4 and other hydrocarbons associated with a Au deposit in the Yilgarn Craton (Western Australia). The work not only focused on the gas concentrations but also included analyses of the gases for stable isotopes 2H and 13C. The findings of Moretti et al. (2021) are consistent with the premise that H2 forms through two processes, depending on the lithological setting: (i) in basic and felsic rocks, radiolytic reactions are suggested as the controlling process for the formation of H2, whereas (ii) in ultramafic lithologies, redox reactions are believed to be responsible (Boreham et al. 2021b). Extrapolating from these preliminary results, the natural H2 composition of soil gases could hold potential for mineral exploration; however, this very early work remains to be tested in the field.

Mercury

The Hg content in soil gas has shown interesting results, with anomalies reported over concealed deposits in diurnal and longer-term variations (McCarthy et al. 1969; McCarthy 1972; Klusman and Jaacks 1987). The short-term variations in Hg concentrations in soil gas have been shown to be related to air temperature and barometric pressure (McCarthy et al. 1969); however, there is a time lag observed between a change in temperature and a change in gaseous Hg anomalies, which is due to a buffering by the soil matrix (McCarthy 1972; Rukhlov et al. 2021).

Mercury becomes volatile through biomethylation and thus plants and soil microbial activity play a pivotal role in the formation of Hg anomalies in soil gas (Siegel et al. 1980). However, in a cold temperate environment, the Hg anomalies vary over the long term (e.g. seasonal) and are highest during the summer months. Cannon and Dudas (1983) suggest this is due to a decrease in microbial activity during the winter months and therefore less volatilization and increased adsorption of Hg onto soil particulates.

Carr et al. (1986) investigated Hg in soils and soil gases across Australia in 29 case studies. Although measuring Hg in soil gas has shown some success and is appropriate in some areas – particularly in areas with sandy cover – Carr et al.'s study showed that monitoring Hg levels in soils is largely preferable to monitoring Hg levels in soil gases. Fursov (1990), however, reported that Hg anomalies in soil gas have smaller dispersion haloes and as such better pinpoint buried deposits. Fursov (1990) published a review of Hg in soil gas measured at over 70 mineral deposit locations in the former Soviet Union. Anomalies of Hg in soil air were identified over blind deposits with a range of compositions, including Hg, Sb, base metal, Pb–Ag, chalcopyrite, Cu, Sn and Au. This work revealed that Hg highlighted the presence of both shallow (1 m) and deeply buried (>600 m) deposits and has led to discoveries. Other studies have had varying degrees of success in the detection of buried deposits using Hg content in soil gas and not all authors agree about the reliability of this approach. In a study by Fedikow and Amor (1990), the Hg content in soil gas was uneven and inconsistent, despite Hg enrichment in 10 base and precious metal deposits in Canada. This was generally attributed to the unsuitable technique used for the collection and measurement of Hg in the soil gases (Fedikow and Amor 1990). In addition, the authors questioned the sampling and analytical techniques employed and noted most analytical laboratories have Hg contamination (Fedikow and Amor 1990; Klusman 1993).

The lack of recently published work since the early 1990s suggests that using Hg in soil gas for mineral exploration has been abandoned. Challenges associated with diurnal and seasonal variations, and the role of biological activity in the formation of Hg anomalies in soil gases are likely to blame. However, new analytical techniques that permit rapid and economic analyses to measure Hg in soil gas in the field have been developed and are now available (Watson et al. 2015), which could trigger a renewed interest in this mineral exploration technique.

Organometal(loid) species

Organometal(loid)s can be formed as a result of biomethylation, both under aerobic and anaerobic conditions. Hirner et al. (1998) list three prerequisites for the formation of organometal(loid) species, the presence of (i) metal(oid)s, (ii) methyl donors from soil biomass, and (iii) suitable micro-organisms in or near the orebodies. Hirner et al. (1998) investigated two deposit types (Cu and Hg) in Germany. The protocols for sampling the soil gas included the use of a vacuum pump, the removal of water using CaCl2-filled drying tubes, and the use of liquid N2 to trap the gases. Multiple analytical steps, including low-temperature gas chromatography and ICP-MS were used. The results of this study are encouraging, particularly the finding that the abundant metals detected in the soil can be correlated to the presence of organic species (R2 >0.8). This preliminary study is singular, and even though some work on organometallic species has been published (Dubiella-Jackowska et al. 2007; Greenwood et al. 2013), their potential for mineral exploration remains unknown.

Halogens

Preliminary study of halogens detected over the Au–Ag–Cu prospect in Vancouver Island shows potential for these elements to become pathfinders for mineral exploration. In their work on soil gas Heberlein et al. (2017) used two types of adsorbents for passive sampling: activated charcoal and alkaline ion exchange resin; in addition they sampled snow above the deposits. The study combines and compares the soil gas results to vegetation geochemistry results for cross-validation of the prospect location. Overall, the activated charcoal returned more reproducible results than the resin. However, the water and organic content of the soil can impact the detection of halogens. Of the halogen elements Br showed rabbit ears anomalies for both activated charcoal and resin, whereas I only showed the anomaly with activated charcoal (Heberlein et al. 2017).

Challenges associated with the use of soil gas for mineral exploration

For decades soil gases were presented as a promising new medium for mineral exploration. Yet this has not been realized and much remains to be understood before soil gases can be used as reliable pathfinders for mineral exploration. However, an article in the Western Mineralogist by Brocks (1972) titled ‘The use of dogs as an aid to exploration for sulfides’ and a 2012 feature article in Mining Technology titled ‘A nose for new mineral deposits: the ore-sniffing dogs transforming mine location’ (https://www.mining-technology.com/features/featureore-sniffing-dogs-for-mine-location/) plus the number of soil gas sampling and measurement techniques being offered by commercial laboratories all indicate that the mineral exploration sector has indeed been using soil gases in the field, though it does appear that these are not commonly used even after exploration companies have tested them. The success (or absence thereof) of the soil gas methods for finding new mineral deposits has not been widely disclosed. In fact, in Klusman's (1993) review, the author suggested another essential explanation, saying that ‘One of the most critical [reasons] has been due to the proprietary nature of much of the exploration research in the mineral industry’ and he goes on to state that ‘the mining industry is also not considered a leader in developing and applying new technology’. Most of the recent publications of soil gas studies have been by geological surveys or are related to PhD studies and have mostly focused on field experiments, reporting only empirical data.

The published studies discussed in this review generally lack the consistency needed to draw reliable comparisons, models or proxies. This lack of consistency in analytical techniques, types of deposits studied, and types of cover among individual studies and study sites, for example, has limited the development of soil gases as a robust medium for mineral exploration. Many studies investigating soil gases in barren areas have noted that gases showed variations related to seasons, i.e. related to precipitation and humidity (Haas et al. 1983; Solomon and Cerling 1987), but these factors are seldom consistently monitored. Published works have largely avoided drawing parallels between gases of interest and deposit type, cover type and structure and/or climate. This has resulted in limited knowledge of the fundamentals associated with soil gases and their potential for mineral exploration.

Understanding a complex medium requires greater control

Our current understanding of the role soil gases can play as pathfinders for mineral deposits is largely based on empirical field studies in a highly complex natural environment with many uncontrolled variables. To date, few results have been published of experiments attempting to constrain the processes by which soil gases are formed and altered. Most of these studies have been focused on measuring sulfur gases near massive sulfide deposits because sulfur gases are formed by oxidation of the sulfide ore (Taylor et al. 1982; Stedman et al. 1984; Kesler and Gardner 1986; Hinkle et al. 1990; Plet et al. 2019, 2021). However, the highly reactive nature of these gases limits their widespread use in mineral exploration.

A study by Luca et al. (2008) investigated the formation of gaseous hydrocarbons associated with bioleaching of well characterized ore under controlled conditions. In this study, the authors were able to identify some gaseous hydrocarbons of interest. Yet, in the field, where studies of hydrocarbon content in the soil gas successfully highlighted concealed mineralization (Disnar 1990; Noble et al. 2018; Simister et al. 2018), the gaseous hydrocarbon compounds were found to be distinct from those reported by the bioleaching study (Luca et al. 2008; Luca Palacios 2012).

Little is known about the underlying processes of soil gas formation and migration and crucial questions will need to be addressed. How and where do the gases form that can be detected in the soil? Do they form at depth, at the interface between the ore deposit and the cover, in the cover or the soil? Are the soil gases the result of gases migrating through the crust, do they form by chemical reaction or are they the result of biological interaction with the environment? Can all ore types be revealed using similar types of gases or would different ore types (or even different commodities) be best highlighted by different gases? These are but a few of the many questions in need of further investigation to eventually mimic natural environments. To reliably identify the processes influencing soil gas composition, more laboratory studies are needed in a range of conditions to allow comparison of closed v. open systems, sterile v. non-sterile settings and aerobic v. anaerobic environments. Thus, investigations of successively complex controlled systems are needed to provide insights into the source of soil gases and how cover type, climate and biology (e.g. microbial communities, insects, plants) play a role in the formation of surficial anomalies detected during soil gas surveys.

Sampling, target gases and analytical techniques

Collecting gas samples can be challenging due to their high mobility. Once a sample is collected, it is impossible to know if the sampling container has leaked, or if the sampling procedure was successful. To mitigate this, analyses can be made in the field; desorbing gases from the soil or rock samples or passing the sample through an absorbent are techniques that are commonly employed (Hinkle and Dilbert 1984; Oakes and Hale 1987; Kristiansson et al. 1990; Mulshaw 1996; Pauwels et al. 1999; Polito et al. 2002; Wang et al. 2007; Noble et al. 2013a). However, there is no consensus on which techniques work best, when to use each of the techniques, or even which gases to study. For example, Disnar and Gauthier (1988) targeted gaseous hydrocarbons from crushed outcrop rock using a desorption temperature of 210°C, whereas Hinkle and Dilbert (1984) desorbed gases from soils for 3 days at 47°C to study sulfur gases. Another sampling variation is the use of active v. passive sampling. Active sampling (i.e. the pumping of soil gases in the field) will provide a ‘snapshot’-type result, which only provides information at the moment of sampling/analysis takes place and does not capture the flux–pulse nature of many soil gases, which can result in more ‘false negative’ results. In contrast, passive sampling, which is similar to the desorption technique, provides an averaged signal over a longer period (typically months for passive sampling and longer for soil samples). Averaging the results provides a larger picture that not only limits pulse effects but also overcomes problems related to humidity and precipitation. References for papers presenting various sampling techniques for soil gases of interest are presented in Table 2.

The question of how to draw comparisons between studies remains. Are results from desorbed sulfur gases in soil comparable to those measured in crushed core rocks? Are actively sampled and field-analysed CO2–O2 gases (McCarthy and McGuire 1998; Muntean and Taufen 2011) comparable to desorbed CO2–O2 gases (Hinkle and Dilbert 1984)?

Some discrepancy in analytical procedures is to be expected due to the improvement of analytical facilities and techniques over the 60 years that soil gases have been studied; however, such a wide range of techniques and target gases means analogies have yet to be established and therefore prevents a critical assessment of the soil gas technique.

Climate

Field studies using various gas species to target mineralization have been undertaken all over the world under a wide range of climatic settings (Fig. 8, Plet and Noble 2022) and though the subsequent reports include a discussion of whether the surveys were considered successful in targeting mineralization, they often do not include whether the results varied with changing climatic conditions, a factor known to affect soil gas results. Reports usually only list the specific precipitation and temperature conditions at the time the soil gas samples were collected (Ball et al. 1983; Klusman and Jaacks 1987; Hinkle 1994). Whether the gases are measured directly in the field (e.g. CO2, O2, CH4) or require more complex analytical workup (e.g. SIROGAS, geogas, sulfur gases), the results from the soil gas samples will vary by seasonal and even diurnal fluctuations.

To be able to compare soil gas measurements between sites, the sample spacing must either be small and the measurements either need to have been taken under similar climatic conditions or the survey area must not have received precipitation for the duration of the passive sampling. These requirements may be achievable in arid climates, but it becomes more challenging in temperate environments, suggesting that soil gas techniques may have greater potential for mineral exploration in dry climates. However, Anand and Butt (2010) report little success in using soil gases in the hot arid conditions of the Yilgarn Craton (Australia), and additional studies by Noble et al. (2013a, 2018) have reported varying degrees of success. For soil gas surveys to be a more reliable tool for mineral exploration, climatic conditions should be monitored before and during soil gas measurements and sampling. Plet and Noble (2022) provide the reader with the dataset that was used herein, and can be interacted with, in order to evaluate how broadly a type of gas has been investigated in a given climate.

Cover composition and structure

It is well known that cover affects the surface geochemistry and the thicker the cover, the more diluted the geochemical signal of a concealed ore deposit becomes. As previously discussed, cover that is only a few metres thick can often mask a mineral deposit at depth. The cover also affects surface geochemistry in other ways. For instance, the mineralogy of the cover, the porosity and presence of cemented horizons, and the permeability of the sediments are all factors that impact the vertical dispersion of gases.

The structure of the cover plays a role in the way geochemical signals, and gases in particular, migrate from a concealed deposit to the surface. Gases migrate via the path of least resistance, such as along faults and fractures caused by geogenic, biological or anthropogenic sources. For instance, a cover containing abundant cemented horizons, such as duricrusts, is not very permeable. Consequently, a paucity of surface gaseous geochemical anomalies may be a reflection of the structural geology of the cover and not an indication that the rocks are barren.

In addition, the type and mineralogy of the non-indurated material in the cover can impact the soil gas geochemistry. For example, it is well known that hydrocarbon gases are selectively adsorbed onto clay minerals and organic matter (Cheng and Huang 2004) and therefore, the total organic content and the clay composition of the cover concealing a mineral deposit also affect the gaseous hydrocarbon anomalies detected at the surface. Organic matter and iron oxides are also believed to impact the ability to detect Hg gases at the surface, as Hg may have adsorbed within the cover and never reach the surface (Carr et al. 1986). If soil gas studies are to become more reliable, it is important to determine the nature of the cover material in the study area as well as its structure.

Biology

Living organisms, such as microbes, insects and plants, are known to consume complex hydrocarbons and produce simple gases such as O2, CO2, N2 or CH4 as well as more complex hydrocarbon gases. Microbes and micro-organisms (e.g. bacteria and fungi) are the most likely to influence the soil gas signature because they are by far the most ubiquitous biomass in soil and the deep subsurface (Bar-On et al. 2018) and they are known to vigorously produce and consume gases. Biologists continue to expand their knowledge of the extremely complex microbiome present in soils by monitoring a wide range of gases, from abundant gases such as CO2 and O2 to trace gases such as CH4, CO, COS, H2 (Conrad 1996; Roscioli et al. 2021).

Experiments have shown that microbes can affect the production of sulfur gas during sulfide oxidation and preliminary work indicates that there is a correlation between hydrocarbons and the composition of microbial communities in the soils (Taylor et al. 1982; Plet et al. 2021). In studies conducted over kimberlite deposits in Canada, soil gases and RNA/DNA were analysed at locations proximal and distal from known occurrences (Simister et al. 2018, 2019; Winterburn et al. 2019a; Iulianella Phillips et al. 2021). The authors consistently reported concurrent anomalies in trace element geochemistry and light hydrocarbons (butane) over the buried kimberlites. In addition, the diversity of bacteria in surface soils over kimberlite decreased with increased content of particular microbial species. Though the genomics investigation of soils overlying mineral deposits continues (Reith et al. 2012, 2015; Simister et al. 2018; Bohu et al. 2019), identification of microbial populations and their role in the distribution of hydrocarbon anomalies that are spatially related to buried mineralization requires further study.

Vegetation can also affect soil gases. Isotopic fractionation and plant root respiration can directly influence their composition. In addition, the roots of plants create a particular environment (rhizosphere) in which compounds exuded by the roots feed the microbes, which in turn impacts the chemistry of the immediate microenvironment. The biological composition of soil is highly complex, as soil may not only host general microbial communities – it may also host a large number of specialized communities within the niche microecosystems, all of which can affect the gases migrating to the surface.

Hamilton (2000) suggest that O2 and CO2 anomalies in soil over buried metal deposits may largely be derived from abiotic redox processes. Biology and biotic processes are efficient catalysers of abiotic reactions, increasing the reaction rates. Thus, the composition of trace gases detected by soil gas studies is most likely a reflection of a combination of processes, both abiotic and biochemical (Conrad 1996). Therefore, soil gases are intrinsically linked to the biological activity in the soil, cover and deposit.

Conclusions and Future Directions

Through reviewing past and current literature on the use of soil gases in mineral exploration, we have found a general lack of consistency among the studies. Though it is crucial to continually develop new techniques and improve technology, no attempts have been made to develop standardized techniques that would allow the development of a comprehensive database. The inconsistencies in the published data prevent any robust comparison among deposits because of the range of gases reported, the analytical techniques used, the types of deposits investigated, the climate at the survey sites, and a lack of detailed information about the cover, to mention only a few.

Nevertheless, this review also highlights directions in which research could develop to greatly improve the understanding of soil gases for mineral exploration, such as laboratory studies to improve background knowledge that can then be linked to field studies. In addition, some of the crucial aspects that should be taken into consideration when developing field surveys include the following:

  • a systematic monitoring of the weather (i.e. precipitation, humidity, temperature) in the days prior to and during sampling;

  • characterization of the cover, such as the presence of indurated horizon, the organic matter content and mineralogy of the different sedimentary horizons, and the structure of the cover;

  • a set of standard gases should be used for each study, in contrast to solely reporting gases that show a strong local anomaly;

  • geomicrobiological analyses should be integrated with all soil gas studies, as there is increasing evidence that soil microbes play a pivotal role in regulating and producing soil gases.

Until a deeper knowledge of the interrelations between soil gases, geology and biology is established, the use of soil gases as a pathfinder to mineralization will remain an untapped technique within the exploration industry.

The authors acknowledge CSIRO Mineral Resources. CP acknowledges the Research Plus CERC fellowship funded by CSIRO. Anicia Henne is thanked for her feedback on previous versions of the manuscript and Elizabeth Ambrose for her editing work. Both have improved the quality of the manuscript. Ray Lett is thanked for his review of the manuscript.

CP: data curation (lead), investigation (lead), writing – original draft (lead), writing – review & editing (lead); RRPN: funding acquisition (lead), supervision (supporting), writing – review & editing (supporting)

This work was funded by the Commonwealth Scientific and Industrial Research Organisation.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.

This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 License (http://creativecommons.org/licenses/by/4.0/)