Gravity core (SSD-045/Stn-9/GC-01) was recovered onboard R/V Sindhu Sadhana (cruise no: SSD-045) from the K-G offshore basin (Latitude: 16°18.6027'N; Longitude: 82°37.9809'E; water depth: 1,673 m; core length: 3.16 m). The core was sub-sampled at the every 2 cm interval. Light gray to deep black colored soft sediments in this core yielded a strong odor of hydrogen sulfide after opening the sediment core onboard. The vertical sediment distribution in this core showed the dominance of clay and silt sized fractions. Numerous layers of light gray colored authigenic carbonates were found throughout the core. Dead shells of Calyptogena Sp. were found at 27 cmbsf and 51 cmbsf in the studied sediment core. Fracture-filled type gas hydrates were recovered from the core catcher and bottommost sediment interval at this site. Sediment sub-sampling for magnetic measurements were carried out in presence of high purity nitrogen flushing to avoid atmospheric conditions which could oxidize hydrogen sulfide or iron mono-sulfide present in the sediments. For rock magnetic analysis, 158 sub-wet sediment samples were weighed, and packed in a 25 mm cylindrical plastic sample bottles. Measurements were carried out at the Paleomagnetic laboratory of CSIR-National Institute of Oceanography (NIO), Goa, India and Center for Advanced Marine Core Research (CMCR), Kochi University, Japan.

Low-field low frequency (0.47 kHz) magnetic susceptibility (χ) was measured using a MS2B Bartington Instruments magnetic susceptibility meter. An anhysteretic remanent magnetization (ARM) was applied in a direct current bias field (50 μT) in the presence of 100 mT peak alternating field and the remanent magnetism was measured using an AGICO JR-6A automatic spinner magnetometer. An isothermal remanent magnetization (IRM) was applied in an inducing field of +1 T in the forward direction and was demagnetized by DC backfields at −20, −30, −100, and −300 mT using a MMPM10 pulse magnetizer. The respective remanences were measured using AGICO JR-6A automatic spinner magnetometer. Mass-normalized IRM acquired at a peak field of 1T is considered to be the saturation IRM (SIRM). S-ratio is calculated as the ratio between the IRM at −300 mT and SIRM (IRM_−300mT/SIRM_1T; Bloemendal et al., 1992). This ratio indicates the relative proportion of high coercivity minerals compared to soft ferrimagnetic minerals. Thermomagnetic measurements were carried out at Paleo and Rock Magnetism Laboratory of CMCR. About 12 selected dried sediment samples were analyzed on a Natsuhara Giken (Model NMB-89) magnetic balance (Range: Room temperature to 700°C) with a heating rate of 10°C/min in a 0.3-T field. Hysteresis loops, first-order reversal curves (FORC) and back-field demagnetization curves were also measured for 10 selected dried sediment samples with a saturating field of 1 T (averaging time of 200 ms, slew rate limit of 1 T/, and field increment of 4 mT) at the CMCR. FORC distributions were generally processed using a smoothing factor (SF) of 4, but SF = 6 was used where possible (Roberts et al., 2000). FORC diagrams (Pike et al., 1999) were processed using the FORCinel software (Harrison and Feinberg, 2008).

Low-temperature magnetic measurements were made on six selected dried sediment samples with a Quantum Design Magnetic Properties Measurement System at CMCR. A RT-SIRM was imparted at room temperature (300 K) in 2.5 T. Samples were later cooled to 5 K and warmed back to 300 K in a zero magnetic field. A LT-SIRM was then applied at 5 K in 2.5 T. Samples were warmed up to 300 K in a zero magnetic field (termed “ZFC” for zero field-cooled). Samples were then cooled to 5 K in the presence of a 2.5 T magnetic field. A LT-SIRM was again imparted at 5 K, and samples were warmed to 300 K in a zero magnetic field (termed “FC” for field-cooled). Magnetic parameters were normalized by sample mass. Derived parameters (δ _FC and δ _ZFC) were calculated following Moskowitz et al. (1993).

Magnetic particles were extracted from the bulk sediment following Petersen et al. (1986). A scanning electron microscope (SEM; JEOL JSM-5800 LV) captured the images of magnetic particles in a secondary electron imaging mode at energy levels between 15 and 20 keV. An energy dispersive X-ray spectroscopy (EDS) probe attached to the microscope was used to determine the composition of magnetic particles. The magnetic mineralogy of the selected samples from different sediment magnetic zones was determined using a Rigaku X-Ray Diffractometer (Ultima IV). The samples were run from 15° to 70° of 2θ at 1°/min scan speed using Cu Kα radiation (λ = 1.5414 Å).

Authigenic carbonates and chemosynthetic shells (Calyptogena sp.) were hand-picked during sub-sampling of the sediment core. These materials were washed and cleaned using a distilled water, and later dried at room temperature and stored in the plastic container.

We broadly classified the rock magnetic profile of the sediment core SSD-045/Stn-9/GC-01 into three distinct sediment magnetic zones: MZ-1 (uppermost 23.0 cmbsf), MZ-2 (25.0–265.0 cmbsf), and MZ-3 (267.0–315.0 cmbsf) based on the down core changes in the magnetic mineral concentration, composition and granulometry data. χ _lf, ARM, and SIRM (Figures 2A–C) are generally used as a diagnostic for magnetic mineral concentration. In MZ-1, a systematic decrease in these parameters from 1 to 5 cmbsf indicate down-core reduction in the magnetic mineral concentration (Figures 2A–C). MZ-1 showed higher values of χ_lf, ARM and SIRM suggesting high concentration of magnetic minerals while lower values are observed in MZ-2 and show uniform values except at certain intervals (Figures 2A–C). The beginning of MZ-2 is marked by an abrupt decrease in the magnetite concentration, and we refer this interval as the zone of reduced magnetic susceptibility (χ_lf). Within this zone (MZ-2), three distinct sediment intervals (G1, G2, G3) showed minor rise in χ_lf, ARM, SIRM, and SIRM/χ_lf indicating the dominance of fine-grained (SP-size) ferrimagnetic particles (Figures 2A–C,E; Tarduno, 1995; Rowan et al., 2009).

The SIRM/χ_lf ratio, an indicator of grain size of magnetic iron sulfides (Peters and Thompson, 1998; Peters and Dekkers, 2003) showed a linear decrease in MZ-1 and MZ-2 with a noticeable rise in G1, G2, G3 intervals and the bottommost sediment interval (303.0–315.0 cmbsf) in MZ-3 indicating the presence of fine-grained magnetic particles (Figure 2E). The variations in S-ratio can be linked to the changes in magnetic mineralogy, and is generally high and close to unity throughout the core (Figure 2F). A noticeable drop in S-ratio is observed between 229 cmbsf to 241 cmbsf suggesting the presence of either higher coercivity magnetic mineral (hematite, goethite) or ferrimagnetic magnetic iron sulfides (greigite, pyrrhotite) in this interval (Figure 2F) (Peters and Dekkers, 2003; Roberts et al., 2011).

Thermomagnetic experiments on the representative sediment samples from MZ-1, MZ-2, MZ-3, G1, G2, and G3 are shown in Figure 3. A major drop in magnetization between 563°C and 595°C except on G3 sample (Figure 3I) suggest that the magnetite is the dominant magnetic mineral present in the sediments (Özdemir and Dunlop, 1997; Figures 3A–J). Increase in χ between 332°C and 480 °C can be attributed to the conversion of paramagnetic minerals into magnetite during heating process (Hirt et al., 1993; Passier et al., 2001; Philips, 2018; Figures 3A–J).

FORC diagrams provided additional constraints on the domain states and mineralogy of the magnetic particles (Figure 4). In general, closed contours in all the plots suggest that the magnetic particles exhibit vortex state to multidomain (MD) type behavior (Roberts et al., 2000; Roberts et al., 2017). FORCs indicate that the magnetic mineralogy of samples from MZ-1, MZ-2, and MZ-3 could be dominated by the ferrimagnetic iron oxides (Vortex state-MD type magnetite) and sulfides (SP Greigite) (Figures 4A–G) (Muxworthy and Dunlop, 2002). In all the samples, FORC distributions are characterized by a peak at coercive field B_C ∼ 10 mT with closed contours suggesting the dominance of vortex state behavior in magnetite (Lascu et al., 2018; Muxworthy and Dunlop, 2002; Roberts et al., 2017; Figures 4A–I). Greigite can exhibit both ultrafine-grained superparamagnetic (SP) (Rowan and Roberts, 2006; Rowan et al., 2009) and stable SD-type behavior (Roberts, 1995). Diagenetically reduced sediments enhances the formation of SP-size greigite particles (Tarduno, 1995; Rowan et al., 2009). Increase in SIRM/χ_lf in G1, G2 and G3 intervals confirms the presence of SP-sized greigite particles (Figure 2E; Maher and Thompson, 1999; Nowaczyk et al., 2012; Snowball and Thompson, 1990).

Low-temperature magnetic measurements for representative samples selected from the different sediment magnetic zones MZ-1 (Figure 5A), MZ-2 (Figure 5B), MZ-3 (Figure 5E), G1 (Figure 5C), and G3 (Figure 5D) are presented. A well-developed Verwey (T_v 117–120 K) transition due to the presence of detrital magnetite is evident in RT-SIRM, ZFC-FC, and in the first derivative of magnetization curves (Figures 5A,B,E). During warming of RT-SIRM, remanence did not recover totally in these samples (MZ-1, MZ-2, and MZ-3) (Verwey, 1939; Muxworthy and McClelland, 2000; Özdemir et al., 2002; Chang et al., 2016). Samples (G1 and G3) from the zone of reduced magnetic susceptibility (MZ-2) did not show any indication of a Verwey transition in RT-SIRM, ZFC-FC, and the first derivative of magnetization curves (Figures 5C,D). We noticed that nearly 65% of the LT-SIRM imparted at 5K is lost during warming between 5 and 30 K (Figures 5C,D). This reflects the dominant presence of superparamagnetic (SP) magnetic particles in these samples (Tarduno et al., 1995; Passier and Dekkers, 2002). Kars and Kodama (2015) noticed a similar pattern of rapid decrease in LT-SIRM from 5 to ∼30 K and ∼50–60% loss of the imparted remanence in the sediment samples from Magesplay fault Zone of the Nankai Trough, offshore Japan. They attributed that the loss of remanence was due to the presence of SP particles which get unblocked very rapidly. The SP magnetic particles in these samples (G1 and G3) are most likely iron sulfides like greigite or pyrrhotite. These can be explained by the fact that fine-grained magnetic particles in these samples would get readily dissolved in the sulfidic environment (as seen through lowest χ_lf in MZ-2). In contrast, the SP greigite nanoparticles are thermodynamically more stable compared to the SP magnetite nanoparticles and would rather remain preserved (Rowan et al., 2009; Roberts et al., 2018). Lack of LT transitions in the magnetization derivation curves in these samples (G1 and G3) is consistent and further confirms the presence of greigite (Figures 5C,D; Chang et al., 2009; Roberts et al., 2011).

In addition to rock magnetic parameters, SEM-EDS data confirm and support the interpretation on the magnetic mineralogy in core SSD-045/Stn-9/GC-01. Detrital origin ferrimagnetic iron oxides (titanomagnetite) of various sizes and shapes are found in all magnetic zones (Figures 6A–L). Magnetic zones (MZ-1, MZ-3) contain well-preserved titanomagnetite grains. EDS data for these grains indicate the presence of titanium, iron, and oxygen with minor amounts of calcium, silicon, aluminum, and manganese (Figures 6B,K,L).

Diagenetically formed iron sulfides are found to occur as sub-micron, spherical-shaped particles in MZ-1, MZ-2, G1, G2, and G3 (Figures 6A,C,J), and EDS spectra contain Fe and S peaks. Presence of such particles in the magnetic extracts from the zone of reduced χ_lf (G1, G2, G3) suggest that a ferrimagnetic iron sulfide i.e., greigite or pyrrhotite is the major constituent of the particles occurring as overgrowth on the paramagnetic host substrate such as pyrite (Rowan and Roberts, 2006). Sub-micron sized iron sulfide particles were also seen to occur as overgrowth on the large titanomagnetite grain in MZ-2 (Figure 6E). Several titanomagnetite grains of various sizes and shapes were found in the magnetically enhanced zone (MZ-3), and EDS spectra contain Ti, Fe, and O peaks (Figures 6K,L).

Magnetic mineralogy is dominated by titanomagnetite, greigite, and pyrite particles occuring in varying proportion (Figure 7). XRD data confirms the dominant presence of pyrite along with minor peaks of greigite and titanomagnetite in G1, G2, and G3 (Figures 7F,I,K). Magnetically enhanced zone (MZ-3) contains titanomagnetite and pyrite (Figures 7N,O). Pyrite is the major magnetic mineral in the zone of reduced χ_lf (MZ-2) (Figures 7C–M). XRD peaks corresponding to quartz are seen in all the samples (Figure 7).

Gray color authigenic carbonates of various sizes and morphologies are found predominantly throughout the core. Representative images of the carbonates found within the greigite-bearing (G1, G2, G3) sediment intervals are presented in Figures 8A–I. Two relict shells of Calyptogena (family: Vesicomydae) were found in the sediment depth intervals of 26–28 cmbsf and 50–52 cmbsf respectively (Figures 8J,K; Table 1).

Bulk sediment grain size in the core SSD-045/Stn-9/GC-01is dominated by the silt and clay sized fractions (Figure 9). Silt and clay content in this core varies from 14.56 to 52.29 vol% and 47.62–83.56 vol%, respectively. Down-core increase in sand content with distinct sediment intervals containing relatively higher sand concentrations at 83, 135, 197, 239, and 315 cmbsf are observed (Figure 9). Higher χ_lf were found in some of these sand layers in MZ-1 and MZ-3 suggesting the presence of coarse-grained magnetic particles (Figure 9).

A good correlation between χ_lf and SIRM is observed in all the samples, but exhibits differences in their slope suggesting the presence of ferri-and paramagnetic mineral assemblages in the samples (Figure 10A). A cross plot of SIRM/χ_lf vs. χ_lf has been successfully utilized as a proxy for characterizing the magnetic mineral assemblages in the gas hydrate bearing sediments (Larrasoaña et al., 2007; Kars and Kodama, 2015; Badesab et al., 2019). We broadly classify the samples into three sub-groups (Figure 10B). Samples exhibiting higher χ_lf and SIRM/χ_lf values suggest the presence of fine-grained magnetite. Higher values of SIRM/χ_lf and relatively low χ_lf in the majority of samples indicate the dominance of ferri-and paramagnetic iron sulfides. Samples which lies within the intermediate range of SIRM/χ_lf and χ_lf values represent the mixture of magnetic iron sulfides and magnetite (Figure 10B). A good covariation between χ_lf and mineralogy diagnostic S-ratio is seen (Figure 10C). Samples exhibiting low S-ratio and relatively narrow range in χ_lf can be attributed to the presence of either higher coercivity magnetic minerals or ferrimagnetic iron sulfides (Figure 10C) (Peters and Dekkers, 2003; Roberts et al., 2011).

Mineralogy diagnostic rock magnetic parameters indicated that the bulk magnetic mineralogy is dominated by the vortex state titanomagnetite. Moskowitz et al. (1993) demonstrated that the origin of magnetic mineral assemblages in the sediment samples can be characterized by δ _ZFC and δ _FC. In the analyzed samples, we noticed that δ _FC/δ _ZFC values are close to 1, which suggest that the titanomagnetites are of detrital origin (Figure 10D; Housen and Moskowitz, 2006). All data falls on a 1:1 line. Samples exhibiting low values δ _ZFC and δ _FC indicate the presence of diagenetically formed ferrimagnetic iron sulfides, while higher values of δ _ZFC and δ _FC indicates the presence of mixed-type (magnetite, pyrite) magnetic mineral assemblages (Figure 10D; Kars and Kodama, 2015; Badesab et al., 2019).

Variations in magnetic mineralogies (titanomagnetite, pyrite, greigite) and concentration governed the down-core changes in the magnetic properties. Primary magnetic minerals in the uppermost MZ-1 are dominated by titanomagnetite (vortex) with minor amount of pyrite (Figures 2F, 5A, 6B, 7A,B). Gradual decrease of χ_lf in MZ-1 can be linked with the progressive dissolution of iron oxide and subsequent precipitation of iron sulfides. A distinct drop in χ_lf, ARM, and SIRM at the shallowest sediment interval (3–5 cmbsf) accompanied by the presence of pyrite and magnetite reflects the top-most dissolution front related to intense pyritization fueled by AOM-coupled sulfate reduction, and probably represent the depth of present-day SMTZ in the core SSD-045/Stn-9/GC-01 (Figures 2A–C, 6A, 7A). A systematic down-core increase in χ_lf and SIRM (from 7 to 17 cmbsf) manifested by the presence of magnetic particles below the proposed present-day SMTZ (3–5 cmbsf) indicate the period of weak AOM activity and minimum dissolution of the magnetic minerals (Figures 2A,C, 6B, 7A). Rapid burial and preservation of the detrital magnetic particles due to increased sedimentation specifically during this period is another possibility (Badesab et al., 2019). A marked depletion in χ_lf between 25–29 cmbsf and 43–52 cmbsf reflects the periods of increased magnetite dissolution and pyrite formation linked with at least two intense methane seepage events at the studied site (Figures 2A, 6C,D, 7C,E). These intervals most likely reflect paleo-SMTZ front, where upward migrating methane front interact with downward migrating sulfate front to trigger strong AOM.

The AOM increases the concentration of hydrogen sulfide in the pore-water and favor dissolution of primary magnetic mineral assemblages (Kasten et al., 1998; Jørgensen et al., 2004; Dewangan et al., 2013). Similar low χ_lf intervals due to paleo-methane expulsion events were reported from the K-G basin and South China Sea (Mazumdar et al., 2009; Peketi et al., 2012; Dewangan et al., 2013; Hu et al., 2019). The presence of chemosynthetic clams like Callyptogena sp. and methane derived authigenic carbonates provided further evidences of intense methane seepage events at these sediment intervals (Figures 2A, 6C,D, 7C–E, 8J,K; Table 1). Fossil chemosynthetic community and authigenic carbonates were reported from site MD161/Stn-8 and NGHP-01–10D in K-G basin (Collett et al., 2008; Mazumdar et al., 2009, Peketi et al., 2012), Haima Seep, South China Sea (Feng and Chen 2015; Liang et al., 2017), southwest Barents and Norwegian Sea (Crémière et al., 2016), Mediterranean Sea (Aloisi et al., 2000), and Gulf of Mexico (Bian et al., 2013). Downward decrease in ARM/SIRM (9–19 cmbsf in MZ-1) is followed by the consistent low χ_lf and ARM/SIRM values in MZ-2 (25–265 cmbsf; Figures 2A,D). As confirmed through the XRD (Figures 7C–M) and SEM-EDS (Figures 6F,H), the dominant magnetic mineral in MZ-2 is titanomagnetite (vortex) and pyrite. It is most likely that fine SD magnetite grains might have been preferentially removed by diagenetic dissolution (Figures 2A–D) owing to larger surface area (Karlin and Levi, 1983; Liu et al., 2004; Garming et al., 2005). The dominance of pyrite and titanomagnetite within 59–265 cmbsf interval of MZ-2 coupled with abundant occurrence of authigenic carbonates suggest that the strong SMTZ fronts sustained due to uniform supply of methane gas from the deep-seated reservoir (Figures 2A, 6E–H,J, 7F–M). It is interesting to note that a lower S-ratio interval (229–241 cmbsf) is observed in MZ-2. Horng, C.S. (2018) and Kars and Kodama (2015) reported that sedimentary pyrrhotite and greigite can have high coercivities which could result in low S-ratio. A dedicated rock-magnetic study conducted by Kars and Kodama (2015) on gas hydrate bearing sediments from Nankai trough reported that authigenically formed ferrimagnetic iron sulfides such as greigite and pyrrhotite can exhibit high coercivities. The lower S-ratio interval observed in MZ-2 could be attributed to the presence of such ferrimagnetic iron sulfides (greigite/pyrrhotite) which were no longer stable and got converted into pyrite as evident through low χ_lf within 209–231 cmbsf, (Figure 2A) and SEM-EDS data (Figures 6H,I).

Lowermost sediment magnetic zone (MZ-3) is characterized by elevated χ _lf, ARM, SIRM, and contains the mixture of abundant detrital titanomagnetites and minor amount of diagenetically formed pyrite (Figures 2A–C, 5E, 6K,L, 7N,O). High χ _lf accompanied by mixed trend of ARM/SIRM values could be attributed to the presence of both fine and coarse-grained detrital magnetic particle which probably survived diagenesis (Figures 2A,D). Fracture-filled gas hydrates deposits were recovered from MZ-3 (305–315 cmbsf, Figure 1B). Trapping of H₂S within the hydrate deposits hindered the pyritization processes (Housen and Musgrave, 1996), and lead to the preservation of detrital magnetite particles in MZ-3 (Riedinger et al., 2005) (Figures 2A, 4I, 5E, 6K,L, 7N,O). Rapid burial of magnetic particles is another important factor which can explain the existence of high χ_lf in gas hydrate bearing sediments (Badesab et al., 2019). We believe that both these processes hindered the diagenetic reactions and favored high χ_lf values in the gas hydrate bearing sediments at site Stn-9.

A dedicated magnetic study on pelagic sediments from western equatorial Pacific Ocean revealed that diagenetically reduced sediments could enhance the formation of SP-size greigite particles (Tarduno, 1995). Increase in SIRM/χ_lf in G1, G2, G3 intervals confirms the presence of SP-sized greigite particles (Figure 2E; Tarduno, 1995). FORC distributions from the same greigite bearing samples G1, G2 and G3 also showed closed contours indicative of vortex-type magnetic behavior (Figures 4A–I; Lascu et al., 2018; Muxworthy and Dunlop, 2002; Roberts et al., 2017). These observations suggests that magnetic particles in G1, G2, G3 and MZ-3 (313 cmbsf) in core SSD-045/Stn-9/GC-01 is dominated by SP greigite and vortex-state magnetite. Hence the minor rise in SIRM/χ_lf (Figure 2E) manifested by the presence of pyrite (Figure 7O) in MZ-3 (301–315 cmbsf) can be explained by the fact that ferrimagnetic greigite which was formed earlier might have got transformed into more stable pyrite (Larrasoaña et al., 2007; Kars and Kodama 2015).

Occurrence of active methane seep are often associated with shallow gas hydrates and presence of seep-carbonates and has been well documented worldwide; for example offshore southwest Africa, Congo deep sea fan (Pierre and Fouquet, 2007), eastern Mediterranean Sea (Aloisi et al., 2000), northern Apennines and Tertiary Piedmont Basin of Italy (Dela Pierre et al., 2010; Argentino et al., 2019), Nyegga complex pockmark, Norwegian Sea (Mazzini et al., 2006), South China Sea (Tong et al., 2013; Han et al., 2014), Blake Ridge Diapir (Naehr et al., 2000), and Vestnesa Ridge, NW Svalbard (Schneider et al., 2018). At active seeps, methane gets consumed and this process generates dissolved bicarbonate, thereby increasing porewater alkalinity and favors precipitation of authigenic carbonate close to the seafloor (Boetius et al., 2000). Thus, the authigenically formed carbonates are considered as potential archives of past fluid flow (Aharon et al., 1997; Bayon et al., 2013; Feng and Chen, 2015). Repeated occurences of authigenic carbonates (Figures 8A–K; Table 1) from 45 to 219 cmbsf in core SSD-045/Stn-9/GC-01 indicate the episodic events of intense AOM due to enhanced methane flux. This also show the variability in past methane fluxes could be due to episodic ventilation of deep-seated methane gas or dissociation of gas hydrates or opening/closing of fault/fracture system driven by neo-tectonic activity (Dewangan et al., 2020).

Association of chemosynthetic communities with methane seepage has been reported from several cold seep locations (Sellanes and Krylova, 2005; Levin et al., 2016; Amon et al., 2017; Mazumdar et al., 2018). Evidence for the strong methane seepage events at 25–29 cmbsf and 43–52 cmbsf, and subsequently anaerobic oxidation of methane comes from the occurrence of authigenic carbonate nodules, chemosynthetic communities (Calyptogena sp.), and pyrite followed by distinct χ_lf drop at this intervals (Figures 2A, 8J,K). Similar observations were reported from site MD161/Stn-8 in the K-G basin (Mazumdar et al., 2009; Dewangan et al., 2013). Variations of methane fluxes at seeps determine the migration of the SMTZ in marine sediments (Borowski et al., 1996). Delineation of SMTZ and the factors controlling its migration in a seep-impacted sediments could help in tracking the evolution of past methane seepage events. Previous studies employed various proxies to recognize the SMTZ and detect the past methane seepage events including Ba front (Dickens, 2001; Snyder et al., 2007), methane-derived authigenic carbonates, mineralogy and its carbon isotope composition (Teichert and Luppold, 2013; Kocherla et al., 2015), Sulfur isotope and Mo anomaly (Peketi et al., 2012) and magnetic susceptibility (Badesab et al., 2019). Fluctuations in methane fluxes in a complex gas hydrate bearing marine sedimentary system can also be tracked successfully using χ_lf records, for examples in K-G basin (Badesab et al., 2019; Dewangan et al., 2013; Usapkar et al., 2014), the Argentine continental slope (Garming et al., 2005), Nankai trough, Japan (Kars and Kodama, 2015), and Cascadia margin (Housen and Musgrave, 1996; Larrasoaña et al., 2007). Rock magnetic record of core SSD-045/Stn-9/GC-01 provides evidence of rapid SMTZ migration controlled by changing methane fluxes. Mineralogical analyses on the core revealed the presence of abundant diagenetically formed pyrite and minor amount of titanomagnetite within the zone (MZ-2) of reduced χ_lf (Figures 7D–M). Such phenomena can be attributed to the reducing environment associated with upward methane flux which enhances diagenesis of the highly magnetic detrital minerals such as titano-magnetite into nearly non-magnetic pyrite (Housen and Musgrave, 1996). In zone of reduced susceptibility (MZ-2), multiple episodes of seepage activities controlled the diagenetic alteration of magnetic minerals and hence have a direct impact on χ_lf. However, tracking paleo-SMTZ fronts can be challenging at times due to the differences in the rate of diagenesis constrained by changing methane fluxes. In such a scenario, the occurrence of authigenic carbonates would be useful in delineating the paleo-methane seepage activities. The abundant occurrence of authigenic carbonates in a depth range from 45 to 219 cmbsf within MZ-2 suggests that the SMTZ was quite variable in the past (Figures 8A–I; Table 1). We demonstrate that combination of rock magnetic and mineralogical (SEM-EDS, XRD) data together with authigenic carbonates are useful to identify the present and past SMTZ fronts fueled by methane seepages.

Fracture-controlled fluid migration has been well-reported at cold seeps, for example in Vestnesa Ridge (Yao et al., 2019), Harstad Basin, southwest Barents Sea (Vadakkepuliyambatta et al., 2013; Crémière et al., 2018), Hydrate Ridge (Torres et al., 2002; Weinberger and Brown, 2006), Blake Ridge (Egeberg and Dickens, 1999), and Central Chile (Geersen et al., 2016). Fractures/faults facilitate efficient transport of methane-rich fluids from greater depth to shallower depth, and sometimes methane escape from sediment-water interface as flares (Berndt et al., 2014; Sahling et al., 2014). In methanic sediments, interplay between availability of reactive iron supply and relative production of hydrogen sulfide (HS−) control the pyritization process (Wilkin and Barnes, 1997). In case of excess HS− and high availability of reactive iron, the pyritization process is complete and the metastable iron sulfides are eventually converted to pyrite (Berner, 1967). Under low HS− concentration greigite may preferentially be precipitated over pyrite and will lead to the preservation of greigite. Migration of methane-rich fluids is found to be associated with the authigenic ferrimagnetic iron sulfide minerals like greigite and/or pyrrhotite (Larrasoaña et al., 2007). Greigite formation is enhanced in the vicinity of gas hydrate deposits as observed in Hydrate Ridge (Larrasoaña et al., 2007) and Blake Ridge (Wellsbury et al., 2000). For greigite formation and preservation, either limited amount of hydrogen sulfide or increased supply of reactive iron is required to constrain the pyritization process in methanic sediments (Liu et al., 2004). For example, Kao et al., 2004 demonstrated that widespread greigite preservation is likely if abundant reactive iron is available to react with sulfide during pyritization reactions. Roberts et al., 2011 suggested that the preservation of greigite depends on the balance between sulfide production and reactive iron concentration. In the studied core, three distinct greigite (G1, G2, and G3) bearing sediment intervals (81–93 cmbsf, 145–159 cmbsf, and 217–229 cmbsf) has been identified within the zone of reduced χ_lf. Presence of greigite has been further confirmed through rock magnetic (χ_lf and SIRM/χ_lf), SEM-EDS and XRD (Figures 2A,E, 6F–H, 7I–K). A general trend of downcore increase in magnetic grain size diagnostic proxy (ARM/SIRM) in MZ-2 is observed (Figure 2D; Maher and Thompson, 1999). ARM/IRM first decreases in MZ-1 because of preferential dissolution of fine-grained magnetite (Maher, 1988) in the upper part of the reduced zone, and rises downcore in response to the formation of fine-grained greigite in MZ-2. A rockmagnetic study conducted on a sediment core from Niger deep-sea fan by Dillon and Bleil (2006) observed that the iron redox boundary is very shallow (∼3 mbsf) and below, a pervasive magnetite dissolution has taken place. Similar to our observations, they noticed minor increase in ARM/SIRM with depth after an initial drop. They attributed the steadily increasing ARM/SIRM ratio in the anoxic sulfidic zone with the small gradually growing SD fraction of magnetic sulfides, for example authigenic formation of greigite is most likely in such environment and similarly reported in Mediterranean sapropels (Roberts et al., 1999; Passier and Dekkers, 2002) and Korean Strait (Liu et al., 2004). Increase in SIRM/χ_lf in G1, G2 and G3 intervals also suggest the presence of authigenically formed SP-sized greigite particles in the studied core (Figure 2E). Within the zone of reduced χ_lf (MZ-2), elevated methane concentration might have allowed huge production of sulfide which enabled complete pyritization and restricted the greigite preservation as seen through lower and uniform values of χ_lf, SIRM and abundant presence of pyrite in MZ-2 (Figures 2A,C, 7C–M). In contrast, decline in methane flux at G1, G2, and G3 intervals probably generated less sulfide, and favored the formation and preservation of greigite. Similar observation was made by Dewangan et al., 2013, where the presence of a thick authigenic carbonate layer beneath the SMTZ at site MD161/Stn-08 in K-G basin (Mazumdar et al., 2009) hindered the downward diffusion of sulfide favoring early arrest of the pyritization process and preservation of greigite. At the present site, low methane flux might have supported more distributed-type (vein-filling) hydrate deposits instead of massive deposits formed due to enhanced methane flux. We, therefore, propose that these three greigite bearing sediment intervals (G1, G2, and G3) are associated with distributed-type gas hydrate intervals at site Stn-9. However, we do not completely rule out the possibility of greigite formation due to increase in iron content in G1, G2, and G3.

A linkage between occurrence of greigite and increased clay content is evident (Figure 9). We propose a following plausible mechanism to explain the linkages between methane migration and formation of distributed-type (vein-filling gas hydrates), grain size and greigite formation and preservation. High resolution X-ray investigation conducted on the gas hydrates recovered from K-G basin revealed that hydraulic fractures created by upward migrating fluids was the dominant mechanism for the formation of veined-hydrate deposits at site NGHP-01–10 (Rees et al., 2011). At site Stn-9, it is highly possible that rapid migration of diffusive methane flux through the clay rich intervals in G1, G2, and G3 probably created multiple hydro-fractures which provided conducive environment for the formation of distributed–type (vein-filling) hydrates in these intervals. Growth of vein-filling hydrate network trapped the available sulfide within it and created a closed system by forming an impermeable layer which further restricted the supply of the downward diffusing sulfate (sea water) through the hydrate-bearing interval. A sulfide deficient zone is created which caused arrestation of pyritization and favored formation and preservation of greigite in G1, G2, and G3. Geochemical study conducted by Kao et al. (2004) on sediment cores from southwestern Taiwan reported that detrital minerals will survive in sediments if the methane fluxes are too small. At the studied site, fracture-filled gas hydrates were recovered from MZ-3 (305–315 cmbsf). Survival of detrital titanomagnetite particles in MZ-3 could be explained based on the fact that the lower methane (free) concentration in these intervals due to the presence of massive gas hydrate beneath this interval might have restricted the further diagenetic transformation of iron oxides into iron sulfides.

Grain size also provides important control on the gas hydrate formation in marine sedimentary system (Clennell et al., 1999; Torres et al., 2008). In Blake Ridge, high methane hydrate concentrations were found in sediment layers containing high silt content (Ginsburg et al., 2000; Lorenson, 2000). Overall, clay and silt dominate the bulk sediment grain size in the studied core (Figure 9). Marked peaks of decrease in silt content and concurrent rise in clay content more specifically within gas hydrate bearing intervals (G1, G2, G3 and bottom most: 305–315 cmbsf) suggest that distributed methane hydrate accumulation in veins/fracture preferentially occurs in the sediment intervals containing high clay content (Figures 2A, 9).

Regulation of gas plumbing system due to opening/closing of the fractures/faults provides important constraints on the methane flux (Dewangan et al., 2011; Sriram et al., 2013; Mazumdar et al., 2019). Widespread occurrence (45–219 cmbsf) of authigenic carbonates in the studied core suggests that the multiple AOM driven by variability in paleo-methane fluxes is controlled by the opening and closing dynamics of the fracture/fault system due to prevailing shale tectonism (Dewangan et al., 2010; Gullapalli et al., 2019; Mazumdar et al., 2019). Sahling et al. (2014) also suggested that micro and macro fractures are often found associated with cold seep systems, wherein methane from deep-seated reservoir moves upward to sediment-water interface such as Hydrate Ridge (Torres et al., 2002) and Blake Ridge (Egeberg and Dickens 1999). At site Stn-9, fracture-controlled fluid transport supported the formation of gas hydrates (distributed and massive-type) (Mazumdar et al., 2019). We believe that hydrate crystallizes within the fault/fractures zone as methane gas migrate through GHSZ (You et al., 2019). It is interesting to note that several layers of hard authigenic carbonates are found above and within the proposed distributed-type gas hydrate bearing intervals (G1, G2, G2) (Table 1). It may also be possible that formation of authigenic carbonate layers coupled with clay deposits restricted the upward migrating methane and led to the formation of distributed-type hydrate deposits. The distributed gas hydrate deposits build a closed system, which prevented the completion of pyritization process and thus, favored the formation and preservation of SP size greigite.

We propose a plausible mechanism to explain the various features of an active cold seep system associated with shallow gas hydrates in the K-G basin (Figure 11). At active seep site, early diagenesis of magnetic minerals occurs very close to the sediment-water interface resulting in dissolution of primary detrital iron oxides and subsequent conversion to iron sulfides (pyrite). Distinct drop in concentration-dependent magnetic parameters manifested by the presence of diagenetically formed pyrite between 3 - 5 cmbsf represent the top-most dissolution front (recent) and can be attributed to the intense pyritization fueled by AOM-coupled sulfate reduction and probably represent the depth of present-day SMTZ in core SSD-045/Stn-9/GC-01. A couple of sediment intervals (25–29 cmbsf and 43–52 cmbsf) also show marked depletion in magnetic susceptibility and abundant presence of pyrite grains suggesting at least two events of huge H₂S build up by intense methane seepage events. Presence of chemosynthetic shells (Calyptogena Sp.), authigenic carbonates at these intervals confirm the paleo-methane seepage at site Stn-9. The existence of low and uniform magnetic susceptibility in MZ-2 (except G1, G2, and G3) reflects intense reductive diagenesis fueled by the elevated methane supply, which allowed the production of large concentration of sulfide, and led to transformation of primary iron-bearing magnetic minerals into paramagnetic pyrite. Repeated occurences of authigenic carbonates from 45 to 219 cmbsf in core SSD-045/Stn-9/GC-01 support multiple episodes of AOM intensification in the past. The opening/closing of fault/fractures due to neo-tectonic activities allowed migration of methane from deep-seated gas reservoir (Dewangan et al., 2011; Sriram et al., 2013; Dewangan et al., 2020; Mazumdar et al., 2019). Upon entering the shallow depths and encountering conducive P-T conditions, gas hydrates started crystallizing within the faults and fractures where methane concentration exceeded the solubility limit (You et al., 2019). We propose that the distributed-type gas hydrate formed in veins during periods of reduced methane flux at clay rich intervals (G1, G2, G3) arrested the pyritization process and favored the preservation of the intermediate iron sulfides minerals (greigite/pyrrhotite) (Larrasoaña et al., 2007).

High magnetic susceptibility in the bottom most sediment interval containing massive gas hydrate deposit (MZ-3) can be attributed to the survival of detrital magnetic grains due to both hinderance of diagenetic reactions due to trapping of H₂S within hydrate (Housen and Musgrave, 1996; Riedinger et al., 2005) and rapid burial and preservation of the detrital magnetic grains due to higher sedimentation events (Badesab et al., 2019). Such gas hydrates requires efficient methane migration from deep-seated reservoir through the fracture system as evident from the existence of seismic chimneys beneath this site (Dewangan et al., 2020). Some active faults may even transport methane to seafloor, which is observed as methane gas flares (Mazumdar et al., 2019).