1 INTRODUCTION

Gas hydrates are solid, ice-like compounds consisting of hydrocarbon (mostly methane) or nonhydrocarbon gases trapped by water molecules (Kvenvolden, 1993; Sloan and Koh, 2007). Three types of gas hydrate of cubic structures Ⅰ and Ⅱ, and hexagonal structure Ⅲ were already found and reported (Stackelberg, 1949; Ripmeester et al., 1987). Gas hydrate is widely distributed in deep marine and sub-permafrost sediments (Milkov, 2004) in gas hydrate stability zone (GHSZ) under relatively high pressure and low temperature. Due to high concentration of methane in hydrates (~170-m³ methane for 1-m³ structure Ⅰ hydrate under standard pressure and temperature (SPT)) (Claypool and Kaplan, 1974) and large hydrogen to carbon ratio of methane (Marchetti, 1985), gas hydrates become a potential clean energy source and an important carbon reservoir. Moreover, hydrates are easily dissociated as temperature or pressure varies, which can cause methane release to induce cold seeps or trigger catastrophic submarine slides (Paull et al., 2002; Westbrook et al., 2009).

Formation of methane hydrates within GHSZ is controlled by a series of physical, bio- and thermochemical processes to degrade particulate organic carbon (POC) deposited on the seafloor (Fig. 1). Firstly, POC is mostly degraded by microorganisms in the bioturbated layer (0–10-cm sediment depth). Then other POC is further buried below 10 cm, which is partly degraded via anaerobic microbes in the sulfate reduction zone (Flögel et al., 2011), partly degraded in the methanogenic zone to generate biogenic methane (Fig. 1a), and partly buried below biosphere to decompose into gas, oil, and other hightemperature (>100 ℃) products under thermo-chemical processes (Fig. 1a). Biogenic and thermogenic methane are main sources for gas hydrates formed in GHSZ when methane concentration exceeds pressure, temperature, and salinity-dependent solubility limits of hydrate (Fig. 1b) (Burwicz et al., 2011). However, the geothermal gradient rises too fast with depth and gas hydrates occur only in upper sediment layer shallower than ~1 km (Fig. 1b) (Kvenvolden, 1993; Miles, 1995).

Gas hydrate was first discovered in high-pressure natural gas pipelines in the 1930s (Hammerschmidt, 1934), and was drilled on the Blake Outer Ridge in 1972 (Hollister et al., 1972). Thereafter, gas hydrates were widely found and several production tests were already conducted in Canadian Arctic, Alaska, and offshore Japan (Beaudoin et al., 2014). In the South China Sea region, multidisciplinary investigations also revealed occurrence of gas hydrates in offshore Taiwan Island (Liu et al., 2006; Lin et al., 2009), Zhujiang (Pearl) River Mouth Basin (Zhang et al., 2015; Qian et al., 2018; Li et al., 2020), and Qiongdongnan Basin (Jiang et al., 2021; Liao et al., 2021). Especially in the Qiongdongnan Basin, currently active "Haima" cold seeps were found by remotely operated vehicle in 2015 and 2016 (Feng et al., 2018; Zhao et al., 2020), showing abundant gas hydrate deposits. However, the spatial distribution of gas hydrates, the controlling factors, and the most favorable area for future exploration are still not well defined. Here we use the available highresolution seismic lines, well and heat flow data in the Qiongdongnan Basin to investigate the issues through calculating the thickness of gas hydrate stability zone (GHSZ) and estimating the inventory.

2 GEOLOGICAL SETTING

The northern South China Sea (SCS) passive margin (Fig. 2a) is formed under multiple-episodic rifting in the late Cretaceous to Oligocene (Ru and Pigott, 1986; Sun et al., 2021) prior to seafloor spreading in the Oligocene to early Miocene (Taylor and Hayes, 1983; Briais et al., 1993; Li et al., 2014). Several large sedimentary basins such as the Qiongdongnan and Zhujiang River Mouth Basins are developed along the margin (Fig. 2a). In the postspreading period, the margin is featured by fast basement subsidence and high sedimentation (Zhao et al., 2015, 2018b). In the west Qiongdongnan Basin, the sedimentation rate exceeds ~1 000 m/Ma since the Middle Pleistocene (1.8–0 Ma) (Fig. 3b) (Zhao et al., 2015, 2018b).

The Qiongdongnan Basin covers an area of ca. 8.4×10⁴ km² (Fig. 2b). Water depths span from 100 to 2 500 m and increase toward southeast (Fig. 3a). The basin started rifting in the middle Eocene (~45 Ma) and consists of three major structural units: the northern uplift, central depression, and southern uplift (Fig. 2a). The central depression comprises the LDS, LSS, SNS, BDS, CCS, and BJS (Fig. 2b), in which, the Cenozoic sediment reaches ~10 km thick (Fig. 3c) (Zhao et al., 2015). The heat flow also increases from continental shelf to slope, which is ~50–70 mW/m² in the northern shelf and upper slope, ~70–85 mW/m² in the central depression, and ~85–105 mW/m² in the eastern sub-basins (Wang et al., 2014).

3 DATA AND METHOD 3.1 Data

GHSZ is principally controlled by temperature and pressure conditions within sediment. Calculating the thickness of GHSZ thus requires bathymetry, sediment thickness, bottom water temperature, heat flow, and geothermal gradient. The first two data sets are obtained from high resolution, dense seismic grid, and industrial wells (Fig. 2b) in courtesy of China National Offshore Oil Corporation (CNOOC) (Zhao et al., 2018b). The seismic data contains 500 seismic lines with a resolution of 1 km×1 km and 8 seismic patches, which were collected using a 7.5-km long streamer at a trace interval of ca. 12.5 m (Fig. 2b). The industrial wells, consisting of 38 logging data and 21 vertical seismic profile (VSP) data, are used to constrain the strata ages, lithology and time-depth conversion (Zhao et al., 2018b). Under constraints of the seismic and well data, the bathymetry (Fig. 3a) and sediment thickness since the Middle Pleistocene (1.8–0 Ma; Fig. 3b) and during the Cenozoic (45–0 Ma; Fig. 3c) are derived, respectively, on 10 666 grid nodal points.

Bottom water temperature, heat flow, and geothermal gradient are also key parameters controlling gas hydrate stability. The previously observed bottom water temperature versus water depth in the SCS region (Fig. 4a) (Yang et al., 2018) are used to fit the relationship between bottom water temperature and water depth, which is used to calculate the bottom water temperature in the Qiongdongnan Basin (Fig. 5a). The measured heat flow (Fig. 5b) and sediment thermal conductivity (Fig. 4b) derived from 44 industrial drills in the shallow water and 110 seafloor heat flow probe stations in the deep water area (Wang et al., 2014) are applied to obtain the sediment geothermal gradients (Fig. 5c) and temperature. The heat flow data has a resolution of ~2.6 drills per 1 000 km².

3.2 Calculating the thickness of gas hydrate stability zone (GHSZ)

Gas hydrates are stable at relatively low temperature and high pressure. A fourth order polynomial was provided to define the gas hydrate stability boundary (Miles, 1995):

where a=1.559 474×10^-1; b=4.827 5×10^-2; c=-2.780 83×10^-3; d=1.592 2×10^-4; P_b is the pressure (in MPa) and T is the temperature (in ℃).

Temperature in sediment is:

where T_z is sediment temperature (℃) at depth z (meter below seabed); T⁰ is bottom water temperature (℃) at seabed; and ΔT/Δz is geothermal gradient within sediment.

Conversion between pressure (P_s, MPa) and depth (D, meter below sea level) in sediment is:

where C₁=[5.92+5.25sin²L]×10^-3, L is the latitude (in degree); C₂=2.21×10^-6; D=z₀+z, z₀ is water depth below sea level (m).

Substituting Eq.2 into Eq.3 gives hydrostatic pressure-temperature relationship Eq.4 of the sediment.

Simultaneously solving and comparing P_s and P_b can constrain the base of the gas hydrate stability and obtain the thickness of GHSZ in the Qiongdongnan Basin (Fig. 6).

3.3 Calculating gas hydrate inventory (GHI)

A transport-reaction model was constructed to explore controlling factors on gas hydrate accumulation (Wallmann et al., 2006). On basis of a series of modelling, a transfer function (5) is given to estimate the GHI for completely compacting margin sediments (Wallmann et al., 2012):

where a=0.002 848±0.000 49; b=1.681±0.027; c= 24.42±7.2; d=0.994 4±0.10; e=-1.441±0.19; f=0.392 5±0.032; GHI is gas hydrate inventory within GHSZ (in kg methane-C/m²); L_GHSZ is GHSZ thickness (in m), POC₀ (in wt%) and w₀ (in cm/ka) are POC concentration and burial velocity (cm/ka) in surface sediments. The function is valid for GHI≥0 over a range of L_GHSZ=0–1 000 m; w₀=10–150 cm/ka; and POC₀=0.2–5 wt%. Here the POC₀ is given a value of 5% to estimate the maximum value of GHI in the Qiongdongnan Basin (Fig. 7).

4 RESULT AND DISCUSSION 4.1 Spatial distribution of GHSZ in the Qiongdongnan Basin

Natural gas hydrates are of highly interest to academia and industry due to implications for submarine geohazards, a future energy source (~10 times greater than conventional reserves), global climate change (methane is a greenhouse gas ~25 times more effective at raising temperature than CO₂) (Kennett and Stott, 1991; Dickens et al., 1995) and carbon cycle (occupying a third of the world's mobile organic carbon) (Beaudoin et al., 2014). Multidisciplinary hydrate investigations were conducted in the South China Sea, including the seismic survey, pressure and non-pressure coring, logging while drilling, in-situ temperature and permeability measurements, pore water sampling (Wang et al., 2011; Song et al., 2014; Zhang et al., 2015; Ye et al., 2019). These studies revealed widespread bottom simulating reflectors (BSRs), an important indicator of the base of GHSZ (John et al., 1991; Kvenvolden et al., 1993), in the offshore Taiwan Island (Liu et al., 2006; Lin et al., 2009), Zhujiang River Mouth Basin (Zhang et al., 2015; Qian et al., 2018; Li et al., 2020), and Qiongdongnan Basin (Figs. 6–8) (Deng et al., 2021; Jiang et al., 2021), cored the disseminated and massive gas hydrate deposits (Wang et al., 2011; Zhang et al., 2015; Qian et al., 2018), and successfully performed the first test production in 2017 in the Shenhu area in the Zhujiang River Mouth Basin (Wu and Wang, 2018).

In the Qiongdongnan Basin, hydrate related gas chimneys (Fig. 8) (Ye et al., 2019; Deng et al., 2021; Jiang et al., 2021) and active cold seeps due to methane leakage (Zhao et al., 2020) were also discovered, which indicate the occurrence of abundant gas hydrate deposit in the basin. Calculation of GHSZ thickness further reveals that the GHSZ thickness ranges mostly between ~200 and 400 m at water depths >500 m (Fig. 6). The total gas hydrate inventory is ~6.5×10⁹ t carbon over an area of ~6× 104 km² (Fig. 7). Furthermore, three areas including the lower uplift to the south of the LSS, the SNS and BDS, and the CCS are found to have a thick GHSZ of ~250–310 m, 250–330 m, and 350–400 m, respectively (Fig. 6), where the estimated gas hydrate inventory all exceeds ~150 kg/m² (Fig. 7).

The calculated GHSZ thickness is compared with the seismically measured GHSZ thickness along line 1 in the lower uplift to the south of the LSS and line 2 in the lower uplift to the south of SNS (see locations in Fig. 6). The base of seismically interpreted GSHZ is constrained by BSRs (John et al., 1991; Kvenvolden et al., 1993). The thickness of GHSZ between seafloor and BSRs along lines 1 and 2 are derived on basis of the two-way travel time using a P velocity of ~2 500 km/s (Hyndman and Davis, 1992) (Fig. 8). Results show that the calculated GHSZ thickness is consistent with the seismically measured GHSZ thickness and the maximum deviation is less than ~25 m along lines 1 and 2 (Fig. 8).

4.2 Main factor controlling GHSZ thickness in the Qiongdongnan Basin

GHSZ thickness depends on local sediment temperature-pressure conditions. Low bottom-water temperature, heat flow, geothermal gradient, and large bathymetry are in favor of thick GHSZ formation. In the Qiongdongnan Basin, gas hydrates stay stable at water depths >500 m (Fig. 6). In these deep water area, bottom water temperature varies slightly and decreases with water depth from ~6–2 ℃ (Figs. 5a & 6a) (Yang et al., 2018), which has thus little effects on GHSZ thickness. However, the measured heat flow increases with water depth. It is ~50–70 mW/m² in water depth < 500 m, and ~70–100 mW/m² in water depth >500 m (Fig. 5b) (Wang et al., 2014), on basis of which, the calculated geothermal gradient also increases with water depth. It is ~30–40 ℃/km in shallow water and ~40–60 ℃/km in deep water (Fig. 5c). High heat flow and geothermal gradient tend to reduce GHSZ thickness, while the thickest GHSZ of ~350–400 m still occurs in the CCS where the heat flow is the highest of 80–100 mW/m² and the water depth is the largest of ~2 400–3 000 m (Fig. 6). It is obvious that water depth is a dominant factor in controlling GHSZ thickness in the Qiongdongnan Basin.

4.3 Favorable area for gas hydrate accumulation in the Qiongdongnan Basin

GHSZ provides potential space for gas hydrates but it cannot guarantee accumulation of gas hydrates (Burwicz et al., 2011). In addition to GHSZ, hydrate accumulation also relies on gas sources and transport systems, which are commonly grouped into two end members, the diffusion and venting systems (Su and Chen, 2006). In the diffusion system, two phases of water and hydrate reach thermodynamic equilibrium in GHSZ. Beneath the base of GHSZ, biogenic gas accumulates to form bottom simulating reflectors (BSR). However, the venting system is in thermodynamic unequilibrium and controlled by kinetics. It requires assistance of conduits, such as faults, diapirs and/or gas chimneys, to transport deep thermogenic gas to GHSZ, where three phases of water-hydrate-gas coexist. In the venting system, no BSR formed but mud volcanoes and pockmarks are usually observed at seafloor.

In the Qiongdongnan Basin, large GHSZ thickness occurs in three areas including the lower uplift to the south of the LSS, the SNS and BDS, and the CCS where water depths are ~1 000–1 600 m, 1 000–2 000 m, and 2 400–3 000 m, respectively (Fig. 3a), and sediment thickness is respectively of ~500–1 500 m, 200–600 m, and 0–200 m since the middle Pleistocene (1.8–0 Ma) (Fig. 3b). The lower uplift to the south of the LSS has relatively shallow water and high sedimentation rate (~270–830 m/Ma in 1.8–0 Ma) (Zhao et al., 2015). The rapid deposition with sufficient organic matters can provide biogenic gas to form diffusive hydrate in the GHSZ and the extensive hydrate-related BSRs which tend to cluster toward the sediment center (Fig. 6) (Jiang et al., 2021). Due to rapid and persistent subsidence, the surrounding areas also have a thick Cenozoic sediment (~500–15 000 m, Fig. 3) (Zhao et al., 2015). The syn-rifting Eocene-Early Oligocene organic matters are thus deep buried, over-matured and gasproduced, which is a thermogenic gas source (Zhu et al., 2011), but needs a channel to vertically migrate into the GHSZ. Previous studies showed that the Neogene polygonal faults (Wu et al., 2009; Wang et al., 2010), mud diapirs (Zhu et al., 2011), and gas chimneys (Wang et al., 2018; Jiang et al., 2021; Liao et al., 2021) are widely developed in the post-rifting strata, which can transport deep thermogenic gas to the GHSZ and seafloor to form hydrate, pockmarks (Wang et al., 2018), and active "Haima" cold seeps (Feng et al., 2018; Zhao et al., 2020). The methane carbon isotope composition of "Haima" cold seep also suggested a mixed gas source (Zhao et al., 2020). Therefore, the lower uplift to the south of the LSS has rich biogenic and thermogenic gas supplies (Zhao et al., 2020) and excellent transport systems (faults, diapirs, and gas chimneys). The coarse sediment with relatively high permeability in the low uplift is thus favorable for hydrate accumulation and the lower uplift is a priority for future exploration, especially where pockmarks (Wang et al., 2018) and active cold seeps (Feng et al., 2018; Zhao et al., 2020) are discovered (Fig. 6). Moreover, the seismically identified landslides (Ma, 2014) are far away located from seafloor pockmarks and cold seeps (Fig. 6), which are probably unrelated to methane release in the hydrate province.

The SNS and BDS have the sedimentation rate of ~110–330 m/Ma in 1.8–0 Ma (Zhao et al., 2015) where the gas hydrate was also found to distribute within a vertical chimney structure (Figs. 6 & 8) (Ye et al., 2019; Deng et al., 2021), implying another potential reservoir for gas hydrate accumulation. However, in the CCS, the sedimentation is extremely low of 0–110 m/Ma in 1.8–0 Ma (Zhao et al., 2015), probably indicating insufficient organic matters available for microbial methane formation although the calculated GHSZ thickness is the largest of ~350–400 m (Fig. 6).

5 CONCLUSION

The thickness of gas hydrate stability zone (GHSZ) in the Qiongdongnan Basin ranges mostly between ~200 and 400 m at water depths >500 m. The gas hydrate inventory is ~6.5×10⁹ t carbon over an area of ~6×10⁴ km². Three areas including the lower uplift to the south of the Lingshui sub-basin, the Songnan and Baodao sub-basins, and the Changchang sub-basin have a thick GHSZ of ~250–310 m, 250–330 m, and 350–400 m, respectively, where water depths are ~1 000–1 600 m, 1 000–2 000 m, and 2 400–3 000 m, respectively. In these deep waters, the bottom water temperatures fall in a narrow range of ~4 and 2 ℃. However, the heat flow increases with water depth and reaches a peak of ~80–100 mW/m² in the deepest water area of Changchang sub-basin. High heat flow tends to reduce GHSZ thickness, however, the thickest GHSZ still occurs in the Changchang sub-basin, highlighting the role of water depth in controlling GHSZ. The lower uplift to the south of the Lingshui sub-basin has high deposition rate (~270–830 m/Ma in 1.8–0 Ma), thick Cenozoic sediment, rich biogenic and thermogenic gas supplies and excellent transport systems (faults, diapirs, and gas chimneys), enabling it a favorable area for hydrate accumulation where hydrate-related bottom simulating reflectors, gas chimneys, and active cold seeps were widely revealed.

6 DATA AVAILABILITY STATEMENT

All data generated and/or analyzed during this study are available from the corresponding author upon reasonable request.

7 ACKNOWLEDGMENT

Thank two anonymous reviewers for valuable and constructive comments with which this manuscript is improved. The seismic and well data are provided by the China National Offshore Oil Corporation.