1. We link sulfate, calcium, magnesium, and iodide pore-water gradients to the occurrence of underlying gas hydrates.

The gas hydrate is an ice-like substance consisting of water and gas (most is CH₄) that forms under conditions of low temperature, high pressure, and adequate gas concentration (Sloan, 1990). An accepted estimation of gas hydrate resources is about 3,000 trillion cubic meters (TCM), which is enormous when compared with the conventional gas resources (∼404 TCM) and shale gas (204-456 TCM) (Chong et al., 2016). Due to its considerable reserves and clean combustion products, gas hydrate may become a promising energy resource (Collett, 2002; Hesse, 2003; Makogon et al., 2007; Boswell and Collett, 2011; Chong et al., 2016). The methane contained within gas hydrates is a kind of potential greenhouse gas, 20 times more effective as a greenhouse gas than carbon dioxide (CO₂) (Mitchell, 1989; Ruppel, 2011). Therefore, any release of methane from gas hydrate decomposing due to temperature or pressure changes can cause a positive feedback cycle for atmospheric warming, thus impacting climate change (Borowski et al., 1996; Haq, 2000; Wallmann et al., 2014; Brown et al., 2016; Reay et al., 2018). Accordingly, the abundance, nature, and distribution of gas hydrate are of great importance and have been arousing a lot of research interests of many researchers.

Geophysical methods are widely applied to the study of gas hydrate distribution and abundance. Bottom-simulating reflectors (BSRs) are often coincident with gas hydrate occurrences at depth because dissociating gas hydrate produces in situ gaseous methane that causes an acoustic impedance contrast that generates its strong seismic reflector (Miller et al., 1991; Hornbach et al., 2012). However, a BSR is not necessarily associated with, or completely consistent with, the presence of gas hydrate (Kvenvolden et al., 1993; Le et al., 2015; Dumke et al., 2016). Thus, we try to improve the setup of regional gas hydrate exploration by employing some geochemical indicators. Geochemical methods have already been tried to confirm potential gas hydrate-bearing areas through the geochemical studies of pore water in sediment (Borowski et al., 1996; Dickens, 2001; Bhatnagar et al., 2008; Wu et al., 2013; Pogodaeva et al., 2020). One of the most commonly used geochemical tools for hydrate recognition in drill cores is the presence of freshening chloride anomalies coupled with δ¹⁸O increase in pore water (Hesse, 2003; Luo et al., 2014). Nevertheless, the cores need to penetrate through the hydrate occurrence zone (often hundreds of meters below the seafloor) for the recognition of these anomalies, which is an economic disadvantage for large-scale exploration projects. Alternatively, the profiles of other pore water species (e.g., SO₄ ^2-, I^-) that record early diagenetic reactions in shallow sediments can serve as a powerful tool in gas hydrate exploration (Schulz, 2006). Some reactions associated with methane could be employed to indicate the existence of gas hydrate because a large amount of methane may occur in the gas hydrate occurrence zone. Within the sulfate reduction zone, normally, sulfate is consumed by organoclastic sulfate reduction (OSR) (Berner, 1980; Gieskes et al., 1981): 2 C H 2 O+S O 4 2 − → H 2 S+2HC O 3 − (1)

However, in the sulfate-methane transition zone (SMTZ, or SMI for sulfate–methane interface) where methane upwells from deep sediments, sulfate supplied from seawater will be exhausted by the anaerobic oxidation of methane (AOM) reaction (Barnes and Goldberg, 1976), S O 4 2 − + C H 4 → H S − + 2 HC O 3 − + H 2 O (2)

Substantial depletion of sulfate through AOM causes the SMTZ to become shallower (Blair and Aller, 1995; Borowski et al., 1996; Dickens, 2001). Some researchers accordingly proposed that the linearity of sulfate gradients and the shallow depth of SMTZ could be used to indicate the existence of gas hydrate-bearing areas (Borowski et al., 1999; Kim et al., 2020). What's more, the calcium concentration of pore water is also sensitive to the AOM process (Rodriguez et al., 2000). It declines with depth due to the enhanced authigenic carbonate precipitation. At the SMTZ where intense AOM reactions occur, the sharp increase in pore water alkalinity may cause the rapid decrease in calcium concentration (Raiswell, 1988; Boetius et al., 2000; Rodriguez et al., 2000). Besides, concentrated iodine anomaly is also commonly observed in gas hydrate-bearing areas (Fehn et al., 2003; Fehn et al., 2006; Muramatsu et al., 2007). The abnormally high concentration of iodide in pore water is related to the decomposition of massive sedimentary organic matter, which represents a large potential gas source. As a result, the concentrated iodide concentration of pore water could also reflect the possible occurrence of gas hydrate. Based on previous research we mentioned above, we try to explore the feasibility of using gradients of sulfate, iodide, and authigenic carbonate precipitation as indicators for gas hydrate in the regional exploration of gas hydrate formation.

Many gas hydrate-bearing areas around the world have been identified and investigated including the Blake Ridge (Paull et al., 1996; Paull and Matsumoto, 2000), Hydrate Ridge (Milkov et al., 2003; Torres et al., 2004), Gulf of Mexico (Boswell et al., 2009), Nankai Trough (Kastner et al., 1993) and Ulleung Basin (Park et al., 2008). In China, several geophysical and geochemical explorations of gas hydrate have been carried out on the northern slope of the South China Sea (SCS) since the 1990s (Zhang et al., 2002; Wu et al., 2004; Jiang et al., 2008; Wu et al., 2011; Zhang et al., 2012; Zhang et al., 2015; Jin et al., 2020; Wang et al., 2021). Numerous data from piston cores and drill sites have been accumulated, which allows us to make a regional comparison of hydrate distribution. The existence of gas hydrate in the northern SCS has been confirmed by geological, geophysical, and geochemical evidence including the occurrence of well-developed BSRs, methane-derived carbonate and shallow SMTZ depth (Liu et al., 2006; Song et al., 2007; Wang et al., 2011; Wu et al., 2011; Li et al., 2012; Li et al., 2015; Zhang et al., 2015; Chen et al., 2016; Jin et al., 2020). During the Sino-German Cooperative Project in early 2004, methane seepages in the SCS were verified for the first time by the discovery of Jiulong Methane Reef (Han et al., 2008). Three years later, natural gas hydrate samples were first recovered in a gas hydrate drilling expedition (GMGS1) which was initiated by Guangzhou Marine Geological Survey in the Shenhu area, northern SCS (Zhang et al., 2007; Wu et al., 2011). In 2013, GMGS2 was conducted in the east of Pearl River Mouth Basin of SCS, and massive gas hydrates were successfully recovered (Zhang et al., 2015). Recently, several visible gas hydrates were recovered in the Shenhu area again during the expedition GMGS3, which further constrain the gas hydrate reserves of the SCS (Yang et al., 2015). In 2016, GMGS4 was conducted in Shenhu area, and the new gas hydrates were found (Yang et al., 2017). In 2018, GMGS5 was conducted in QiongDongNan area of SCS, and massive gas hydrates were sampled (Ye et al., 2019).

Based on previous gas hydrate research, we correlate comprehensive geochemical indicators, the gradients of sulfate, calcium plus magnesium, and iodide, to the occurrence of underlying gas hydrate in marine sediments. The approach is based on the influence of the occurrence of gas hydrate on the geochemical characteristics of shallow pore water and was tested by correlating the collective gradients to known areas of gas hydrate. This result is compared with the hydrate detection results of GMGS1 and GMGS3. Based on this geochemical criterion, we also predict another promising area for future gas hydrate exploration in the northern SCS through the geographic information system. We believe that this research will benefit the future exploration of gas hydrate and will arouse a lot of interest from other researchers.

The South China Sea, tectonically controlled by interactions of Eurasian Plate, Pacific Plate, and Indo-Australian Plate, is one of the largest marginal basins in the western Pacific Ocean (Li, et al., 2012; Wang et al., 2013). A series of sedimentary basins (e.g., the Taixinan Basin, the Pearl River Mouth Basin, and the Qiongdongnan Basin) in the northern SCS are filled with Mesozoic and Cenozoic sediments, whose largest thickness is over 10 km. The thick sediments provide ideal methane sources for gas hydrate formation (Lüdmann and Wong, 1999). Moreover, the northern slope of the SCS has a water depth of 200∼3400 m and a bottom water temperature of 2∼5 °C, these temperature and pressure conditions are suitable for gas hydrate formation (Yao, 2001).

The study region, the Shenhu area, is situated in the middle of the northern slope of SCS (between the Xisha Trough and the Dongsha Islands) (Figure 1). In this area, quantities of deepwater sedimentary fan have developed since Neogene, with average sedimentation rates of 110, 38, 27 m Ma^-1 for Pleistocene, Pliocene, Miocene respectively (Wu et al., 2011). A series of high-angle faults and vertical fracture systems cut the sedimentary section since Pliocene. A series of central diapiric areas such as gas chimneys are well developed in the Shenhu area, providing favorable geological environments and structures for gas hydrate formation (Lüdmann et al., 2001; Wu et al., 2004). High-resolution seismic investigations show that most of BSRs are 150-350 m below the seafloor (Wu et al., 2013). In order to detect the occurrences and determine the distribution of gas hydrates, some surveys have been carried out by Guangzhou Marine Geological Survey since the 1990s. Hundreds of piston cores were recovered by research vessels of the Guangzhou Marine Geological Survey in recent years.

All the geochemical data were measured at the state key laboratory for mineral deposits research, Nanjing University. The anions and cations of pore water except iodide were measured by an ion chromatography (Metrohm 790-1, Metrosep A Supp 4-250/Metrosep C 2-150). Anions were eluted by 1.8 mM Na₂CO₃ + 1.7 mM NaHCO₃. For cation system, ions were eluted by 4 mM dihydroxysuccinic acid + 0.75 mM pyridinedicarboxylic acid. In both systems, the flow rate of eluent was set at 1.0 mL/min, and the relative standard deviation of the measurement results was less than 3%. Iodide was measured using an inductive couple plasma mass spectrometry (ICP-MS) (Element II, ThermoFisher), and the analytical precisions were estimated to be <2%. Samples for iodide measurement was prepared by diluting in 1‰ aqua ammonia with 10 ppb of Rh as an internal standard.

The gradients of indicators are calculated by least-squares fitting, generating a correlation coefficient (r²) used to access the linearity of geochemical profiles. Detailed data of core sites are shown in Table 1. The r² value ranges from 0-1, which represents a strong correlation at value 1, a decoupling correlation at value 0. Both the gradient and corresponding r² values are mapped over the study area using Surfer® software.

We analyzed 246 pore water samples from 23 piston cores in Shenhu area (Figure 1). Due to the space limitation and since the gradient redox is what we mainly discuss, in this paper we only show the gradient and r² calculated from sulfate, iodide and calcium plus magnesium concentration profiles (Table 1).

The downward concentration gradients of sulfate, calcium plus magnesium ions and iodide vary from -0.2 to -3.72, from 0.94 to -3.15 and from 0.13 to 11.3, respectively.

HS428PC has the steepest decreasing gradient of sulfate and the steepest increasing gradient of iodide. HS251PC has the lowest sulfate decreasing gradient and HS60PC has the lowest iodide increasing gradient. For calcium plus magnesium, the downward gradient values exhibit both negative and positive. The highest downward increasing gradient of calcium plus magnesium occurs in HS08-6PC and the highest downward decreasing gradient is in HS296PC. Almost all the sites with relatively high absolute gradient values have the r² values close to 1.

The absolute values of sulfate and iodide downward gradient have a positive relationship, which means that when sulfate decreasing trend is steeper, the iodide has a steeper increasing trend. The calcium plus magnesium concentration gradient is more complicated, but in general, the absolute values exhibit a positive relationship with sulfate and iodide.

The sites with highest absolute gradient values of sulfate, iodide or authigenic carbonate intensity [i.e., the depletion of (Ca²⁺) + (Mg²⁺)] are invariably distributed in the northeast region of the study area (in red circles from Figure 2A, Figure 3A and Figure 4A) where the r² values are also high. The absolute gradient values gradually decrease from the northeast corner to the rest of the study area, with most of study area showing low absolute values except the northeast region. In addition, it has to be noticed that high r² values also exist in other regions except for the northeast area.

We investigated the concentration profiles of sulfate, calcium, magnesium and iodide in pore water from worldwide main gas hydrate-bearing sites. Each kind of ion has a similar trend in different sites, but shows different downward gradients.

The gradients of sulfate concentration, calcium plus magnesium concentration and authigenic carbonate precipitation intensity can indicate the rate of AOM, which is mainly influenced by the flux of methane. Iodide concentration is strongly affected by the decomposition of organic matter from gas source, and can indicate the methane flux from below or other sources. Thus, the higher the values of these indicators are, the more likely gas hydrate exists in the deep. Also, in some cases where irrigation, bioturbation or submarine slide displacement occurs, the depth of SMTZ is not accurate to indicate the flux of methane, and the sulfate gradient can be a more accurate one.

According to the response of pore water composition in shallow sediment to the high flux of methane in gas hydrate-bearing areas, we proposed that the gradients of sulfate, authigenic carbonate precipitation, and iodide together could be comprehensively employed as the indicators in the exploration of gas hydrate. These indicators were applied to the Shenhu area, SCS, and the results were presented in the form of contour maps. Satisfactory, the most promising area for hydrate occurrence predicted in the three contour maps is consistent with the discovery of GMGS1 and GMGS3 expeditions, which indicates the availability of this criterion. According to the criterion, a new promising area for the existence of gas hydrate was identified, though it needs to be tested by future exploration. We believe that this new geochemical criterion will benefit the exploration of gas hydrate in the future.