Structures in Shallow Marine Sediments Associated with Gas and Fluid Migration

: Geological structure changes, including deformations and ruptures, developed in shallow marine sediments are well recognized but were not systematically reviewed in previous studies. These structures, generally developed at a depth less than 1000 m below seaﬂoor, are considered to play a signiﬁcant role in the migration, accumulation, and emission of hydrocarbon gases and ﬂuids, and the formation of gas hydrates, and they are also taken as critical factors affecting carbon balance in the marine environment. In this review, these structures in shallow marine sediments are classiﬁed into overpressure-associated structures, diapir structures and sediment ruptures based on their geometric characteristics and formation mechanisms. Seepages, pockmarks and gas pipes are the structures associated with overpressure, which are generally induced by gas/ﬂuid pressure changes related to gas and/or ﬂuid accumulation, migration and emission. The mud diapir and salt diapir are diapir structures driven by gravity slides, gravity spread and differential compaction. Landslides, polygonal faults and tectonic faults are sediment ruptures, which are developed by gravity, compaction forces and tectonic forces, respectively. Their formation mechanisms can be attributed to sediment diagenesis, compaction and tectonic activities. The relationships between the different structures, between structures and gas hydrates and between structures and authigenic carbonate are also discussed.


Introduction
Shallow marine structures are sediment deformations and ruptures developed in a shallow environment less than 1000 m below the seafloor (mbsf) due to many kinds of driving forces such as overpressure, differential gravity, compaction forces, tectonic forces, etc. Berton and Vesely [1] identified comet pockmarks on acoustic profiles and suggested that they are formed by gas flows and modified by the bottom currents. Shakhova et al. [2] detected pockmarks using multi-beam sonar backscatter data and proposed that gas release causes the pockmarks. Chen et al. [3] identified gas chimneys on seismic profiles and indicated that they are related to thermogenic gas migration. Matsumoto et al. [4] identified gas pipes on seismic profiles and suggested that they are induced by thermogenic gas migration. Chen et al. [5] identified mud diapirs and mud volcanoes on seismic profiles and indicated that their formation mechanisms are related to an ample discharge of gasbearing fluids. Ho et al. [6] studied the morphology of pipe structures and polygonal faults and suggested that they are formed by overpressured gases and fluids. Previous studies show that shallow marine structures are generally considered as pathways for gas and fluid migration. In recent years, the structures related to gas hydrate accumulation have increasingly attracted researchers' attention.
Gas hydrates have been intensively studied as a potential clean energy source in the last decades [7][8][9]. They are stable under relatively low temperature and high pressure conditions, and are generally located in sediments several hundred meters deep or even deeper below the seafloor [10][11][12][13]. Massive gas hydrate dissociation can release vast amounts of methane into seawater and enhance ocean acidification, which is a significant process affecting the carbon balance in the marine environment [14][15][16]. The accumulation and distribution of gas hydrates are closely related to the gas and fluid migration which is primarily controlled by the structures in shallow marine sediments [17][18][19][20][21]. Wang et al. [22] studied three cold seeps in the Taixinan Basin and found massive gas hydrate and chimney structures associated with these cold seeps. Lu et al. [23] studied a mature pockmark in the Zhongjiannan Basin and indicated that the pockmark creation was induced by methane consumption. Natalia and Soledad [24] recognized gassy sediments, pockmarks and gas plumes in the Ria de Vigo in Spain and indicated that the underlying gassy sediments are the gas sources of the pockmarks. Berton and Vesely [1] identified mud volcanoes and underlying chimney structures on the continental margin of southeastern Brazil.They suggested that gas seeps occur along with the chimney structures. Taylor et al. [25] recognized salt diapirs and associated faults at the Blake Ridge Diapir, indicating that the gas and fluid might migrate upward through these faults. Lastras et al. [26] recognized shallow slides and pockmarks in the Eivissa Channel and suggested that the pockmarks occur near the slide headwall scars that reveal fluid escaping. Alrefaee et al. [27] studied polygonal faults and associated fluid migration in the Northern Carnarvon Basin and indicated that the polygonal fault systems can act as pathways for the fluid migration. Gasperini et al. [28] studied the tectonic fault in the Sea of Marmara and suggested that gas and fluid emissions increase along the tectonic fault system after an earthquake.
As introduced above, these structures are closely related to each other, and the structures are closely associated with gas and fluid migration. However, the characteristics, formation mechanisms and effect of the structures are not systematically reviewed. This article aims to provide a systematic summary of the current knowledge about the shallow structures developed in shallow sediments, to classify the shallow marine structures, identify their characteristics and explore the mechanisms of the structures, and discuss their relationships with gas and fluid migration.

The Classification of the Structures
The structures are generally driven by different geological forces, including overpressure, gravity, compaction forces and tectonic forces.
Overpressure is widely developed in many hydrocarbon basins, and its building, transfer, and release generally induce different structures. A seepage occurs during the overpressure release ( Figure 1A), a gas chimney develops during the overpressure building and transfer ( Figure 1C), and a pockmark is a seafloor mark of the overpressure release site ( Figure 1B). Therefore, seepages, pockmarks and gas chimneys are classified as overpressure-associated structures. Seepage may be referred to as a natural phenomenon rather than a geological structure. However, seepage is an active gas and fluid release process, and the gas chimney indicates the process of gas and fluid accumulation and migration; they make up an integrated overpressure system. Therefore, the seepage is classified as an overpressure-associated structure.
Diapir structures are soft-sediment deformations driven by differential gravity and compaction [29][30][31][32]. The soft plastic layers such as mud layers and salt layers easily extrude into overlying layers because of differential forces. Therefore, the mud diapir ( Figure 1D) and the salt diapir ( Figure 1E) are classified as diapir structures. The mud volcano and mud mound are also included in shallow marine structures in this study. The overpressure of the gas-and fluid-rich mud sediment is a critical driving force for their formation. In this study, we classify mud volcanoes and mud mounds as mud diapirs. The two structures are regarded as the later stage of the mud diapir. When the mud diapir reaches the seafloor, a conic-like structure formed is termed a mud volcano, and a dome-like structure is termed a mud mound. gether the landslide, the polygonal fault, and the tectonic fault. In addition, the "shallow marine structure" is not restricted to a depth of less than 1000 mbsf due to the following considerations. First, the shallow marine structures are generally distributed in shallow unconsolidated marine sediments. Their structural characteristics are various and are easily influenced by many unpredictable factors. Second, some structures are developed in shallow and deep environments, and some structures are developed in a deep environment but extend into a shallow environment.  [23]); (C) Gas chimney from the seismic image, Cascadia margin, off the coast of Vancouver Island, Canada (modified from Wood et al. [34]); (D) Mud diapir from the seismic image, Lower Fangliao Basin, offshore SW Taiwan (modified from Hsu et al. [35]); (E) Salt diapir from the seismic image, upper Atlantic margin, SE Brazil (modified from de Mahiques et al. [36]); (F) Landslide from bathymetry image, Pearl River Mouth Basin, the northern slope of South China Sea (modified from Chen et al. [3]); (G) Polygonal fault from the seismic image, offshore Uruguay (modified from Turrini et al. [37]); (H) Tectonic fault from the sea-bottom picture, Tekirdag Basin, Sea of Marmara (modified from Armijo et al. [38]).

Overpressure-Associated Structure
Overpressure is generally induced by hydrocarbon generation, differential compaction, hydrothermal pressurization, clay mineral dehydration, and tectonic compaction [39]. The combination of temperature changes, pressure drop from water unloading, rapid burial by salt deposition, pressure increases during the following water loading, and relative sea level-rise play a fundamental role in overpressure development and release in the underlying evaporites sediments [40][41][42][43][44]. During the overpressure building, transfer,  [23]); (C) Gas chimney from the seismic image, Cascadia margin, off the coast of Vancouver Island, Canada (modified from Wood et al. [34]); (D) Mud diapir from the seismic image, Lower Fangliao Basin, offshore SW Taiwan (modified from Hsu et al. [35]); (E) Salt diapir from the seismic image, upper Atlantic margin, SE Brazil (modified from de Mahiques et al. [36]); (F) Landslide from bathymetry image, Pearl River Mouth Basin, the northern slope of South China Sea (modified from Chen et al. [3]); (G) Polygonal fault from the seismic image, offshore Uruguay (modified from Turrini et al. [37]); (H) Tectonic fault from the sea-bottom picture, Tekirdag Basin, Sea of Marmara (modified from Armijo et al. [38]). Landslides ( Figure 1F), polygonal faults ( Figure 1G), and tectonic faults ( Figure  1H) are driven by gravity, compaction forces and tectonic forces, respectively. All these structures have one or multiple prominent rupture surfaces; therefore, they are classified as sediment ruptures. The structural characteristics are usually different because they are not driven by a single force. The rupture surface is a reasonable connection that strings together the landslide, the polygonal fault, and the tectonic fault. In addition, the "shallow marine structure" is not restricted to a depth of less than 1000 mbsf due to the following considerations. First, the shallow marine structures are generally distributed in shallow unconsolidated marine sediments. Their structural characteristics are various and are easily influenced by many unpredictable factors. Second, some structures are developed in shallow and deep environments, and some structures are developed in a deep environment but extend into a shallow environment.

Overpressure-Associated Structure
Overpressure is generally induced by hydrocarbon generation, differential compaction, hydrothermal pressurization, clay mineral dehydration, and tectonic compaction [39]. The combination of temperature changes, pressure drop from water unloading, rapid burial by salt deposition, pressure increases during the following water loading, and relative sea level-rise play a fundamental role in overpressure development and release in the underlying evaporites sediments [40][41][42][43][44]. During the overpressure building, transfer, and release, the gas and fluid accumulation, migration, and release generally occur, resulting in pore pressure changes and associated structures [45,46].

Seepage
Marine seepages, also called "methane seeps" and "cold seeps", are active structures on the seafloor caused by gas and fluid release (Figure 2A). These structures were first reported in 1958 [28]. A seepage system generally consists of a gas and fluid source, a pipelike pathway and a leaking structure [47]. Seepages are recognized as gas flares/flames on bathymetry image and as gas bubbles in sea-bottom photographs and they are widely found in the South China Sea, the Gulf of Mexico, off the Costa Rica shore, off Svalbard in Norway, the Hydrate Ridge of America, off the shores of Ireland and New Zealand, the eastern Black Sea, the eastern Mediterranean, and the Cascadia margin [13,22,42,[48][49][50][51][52][53][54]. A gas flare as high as 750 m was reported in offshore NW-Svalbard [55], and numerous gas bubbles were reported in the northern South China Sea [56]. Seepages are generally distributed over marine or lacustrine hydrocarbon sediment basins [57] and are possibly related to salt diapirs and tectonic faults [57][58][59]. In addition, seepage is considered as the previous stage of pockmarks since active seepages are widely discovered above pockmarks [60,61].  Geochemical analysis of the sediment samples, pore water, and gas bubbles collected in the seepage areas indicate that the gas hydrate disassociation probably provides the seepage gas source [63,64,84,85]. In the Taixinan basin in the South China Sea, newle formed seepages are discovered next to paleo-seepage sites. Gas hydrates formed in the paleo-seepage site seal the migration pathway of the gases and fluids. As a result, the gases and fluids migrate through a new pathway and leak from a new seepage site [22]. The cases indicate that seepages are considered significant indicators of gas hydrate reservoirs [19,51,52,64,74,75]. Geochemical analysis of the gas seepage can help estimate the gas source and determine whether the seepage is active or not [48,65,[85][86][87].
Authigenic carbonates are good markers of methane seepage and environmental conditions [13,48,49,85,88,89], which generally exist as thick tight crusts and scattered multihole reefs [13,90]. The nodular high-Mg carbonate forms under low methane content conditions, and the aragonite carbonate crust forms under high methane or high Ca content conditions [48,82]. Microbial mats and benthos are discovered beside the active seepage [38,60,74,82,91], related to anaerobic oxidation of methane and sulfate reduction [48,79,82,92], which make a significant contribution to the formation of the authigenic carbonates.
Seepages are mostly discovered above pockmarks, with gentle or vigorous gas leakage, leaving crater-like depressions on the seafloor [90]. Gas pipes are also found below the seepages, considering as the pathways of gas and fluid migration [33]. In some cases, polygonal faults also exist as pathways for the seepages [20]. The structures show a close relationship with each other and the gas and fluid migration.

Pockmarks
Pockmarks are seafloor depressions firstly reported in the Emerald Basin in Canada [93]. Pockmarks are generally recognized as V-and U-shaped structures on acoustic profiles ( Figure 3A  The gas sources of seepages are generally composed of deep thermogenic gas and shallow biogenic gas [28,[62][63][64][65][66]. Deep thermogenic gas migrates upward from a deep source layer along the interface between a salt diapir and the surrounding sediments and the fault plane of an active fault [28,67,68]. When gases and fluids migrate through the evaporite layer, salt dissolves in the fluid and is transported to the seafloor. The density of the salt-rich fluid is higher than that of seawater; as a result, the salt-rich fluid deposits around the leaking site, and then forms a brine pool [40,67,69,70]. Shallow biogenic gas migrates upward through gas pipes to form a seepage [64]. The driving force of the seepage is considered to be overpressure, and the overpressure is generally induced by the continuous accumulation of gases and fluids. In the Hydrate Ridge, the accumulated gas and fluid in shallow layers opens micro-fractures to migrate through [51,58,[71][72][73]. In the Cascadia subduction zone, methane seepage was induced by the collision between ocean crust plates [74].
In addition, gas hydrate and authigenic carbonate possibly exist as trap layers resulting in gas and fluid accumulation and overpressure development. The seepage intensity is supposed to be dominated by the pressure of the accumulated gas and fluid. High gas pressure induces strong gas ejection flares on the seafloor, up to 750 m high [55,75,76]. Therefore, trap sealing is supposed to cause overpressure development and eruption [77]. During overpressure eruption, the gas hydrate mass and authigenic carbonate mass are blown out to the seafloor and form the gas hydrate outcrops and carbonate stacks, respectively [68,70,[78][79][80][81]. Seepages are discovered along with the fractures in the carbonate reef, indicating that the fractures exist as the pathway for seepages [58,82]. In the north Gulf of Mexico, carbonate crusts gradually form in the sediment sealing the seepage simultaneously [83]. The authigenic carbonates are different from the autochthonous carbonates, which were deposited in shallow-water environments and transported to the outer shelf due to the gravitational collapse of the platform margin [45].
The overpressure is the primary driving force to form a seepage. A strong overpressure leads to a vigorous seepage, and a weak overpressure leads to a gentle seepage. Seepage is a part of the overpressure-associated system that occurs during the overpressure release. The system is also composed of a source and a pipe structure during the overpressure accumulation and transfer, respectively ( Figure 2B).
Geochemical analysis of the sediment samples, pore water, and gas bubbles collected in the seepage areas indicate that the gas hydrate disassociation probably provides the seepage gas source [63,64,84,85]. In the Taixinan basin in the South China Sea, newle formed seepages are discovered next to paleo-seepage sites. Gas hydrates formed in the paleo-seepage site seal the migration pathway of the gases and fluids. As a result, the gases and fluids migrate through a new pathway and leak from a new seepage site [22]. The cases indicate that seepages are considered significant indicators of gas hydrate reservoirs [19,51,52,64,74,75]. Geochemical analysis of the gas seepage can help estimate the gas source and determine whether the seepage is active or not [48,65,[85][86][87].
Authigenic carbonates are good markers of methane seepage and environmental conditions [13,48,49,85,88,89], which generally exist as thick tight crusts and scattered multi-hole reefs [13,90]. The nodular high-Mg carbonate forms under low methane content conditions, and the aragonite carbonate crust forms under high methane or high Ca content conditions [48,82]. Microbial mats and benthos are discovered beside the active seepage [38,60,74,82,91], related to anaerobic oxidation of methane and sulfate reduction [48,79,82,92], which make a significant contribution to the formation of the authigenic carbonates.
Seepages are mostly discovered above pockmarks, with gentle or vigorous gas leakage, leaving crater-like depressions on the seafloor [90]. Gas pipes are also found below the seepages, considering as the pathways of gas and fluid migration [33]. In some cases, polygonal faults also exist as pathways for the seepages [20]. The structures show a close relationship with each other and the gas and fluid migration.
Pockmarks are seafloor marks generally induced by gas and fluid releases. In shallow marine sediments, continuous gas and fluid accumulation lead typically to a pore pressure increase. When the overpressure exceeds the limit of the overlying layers, the gases and fluids erupt into seawater, leaving a pockmark on the seafloor. The eruption model of gases and fluids was firstly proposed by Hovland [121] and is widely accepted by many researchers [2,20,90,106,122].
Pockmarks are seafloor marks generally induced by gas and fluid releases. In shallow marine sediments, continuous gas and fluid accumulation lead typically to a pore pressure increase. When the overpressure exceeds the limit of the overlying layers, the gases and fluids erupt into seawater, leaving a pockmark on the seafloor. The eruption model of gases and fluids was firstly proposed by Hovland [121] and is widely accepted by many researchers [2,20,90,106,122].
Many factors can lead to overpressure releases. Periodic sea level changes, glaciation and deglaciation, gas hydrate formation and disassociation, tidal and storm wave activity, rapid sedimentation and tectonic activity are supposed to break the balance between pore pressure and overburden pressure, inducing the leakage of the pore fluid and gas, leaving pockmarks on the seafloor [59,63,98]. Tidal cycles are considered the driving force of the gas and fluid accumulation and overpressure, and the overpressure eruption is triggered by storm waves [98,105,117,121]. Periodic sea level variation leads to the episodic eruption of the pore fluid and gas [63,75,94,123,124]. Meanwhile, overpressure eruption can be induced by deglaciation and earthquake activity [59,100,122,125]. In addition, a high sediment rate generally leads to overpressure. The sediment burial depth increases rapidly, causing a large amount of pore fluid in the sediment pore space, and the pore pressure increases quickly. The gas and fluid are driven along with the lithological interface, fault plane, and salt surface to release on the seafloor. In that case, pockmarks are prominent indicators of overpressure [59,95,112].
Gas hydrate disassociation can be induced by sea level drops and deglaciation, and the process can lead to pockmark formations [63,100]. When hydrocarbon gases and fluids migrate upward to the gas hydrate stability zone (GHSZ), the gas hydrate forms to seal Many factors can lead to overpressure releases. Periodic sea level changes, glaciation and deglaciation, gas hydrate formation and disassociation, tidal and storm wave activity, rapid sedimentation and tectonic activity are supposed to break the balance between pore pressure and overburden pressure, inducing the leakage of the pore fluid and gas, leaving pockmarks on the seafloor [59,63,98]. Tidal cycles are considered the driving force of the gas and fluid accumulation and overpressure, and the overpressure eruption is triggered by storm waves [98,105,117,121]. Periodic sea level variation leads to the episodic eruption of the pore fluid and gas [63,75,94,123,124]. Meanwhile, overpressure eruption can be induced by deglaciation and earthquake activity [59,100,122,125]. In addition, a high sediment rate generally leads to overpressure. The sediment burial depth increases rapidly, causing a large amount of pore fluid in the sediment pore space, and the pore pressure increases quickly. The gas and fluid are driven along with the lithological interface, fault plane, and salt surface to release on the seafloor. In that case, pockmarks are prominent indicators of overpressure [59,95,112].
Gas hydrate disassociation can be induced by sea level drops and deglaciation, and the process can lead to pockmark formations [63,100]. When hydrocarbon gases and fluids migrate upward to the gas hydrate stability zone (GHSZ), the gas hydrate forms to seal the migration pathway, and the free gas accumulates below the GHSZ to build overpressure. The gas, fluid, sediment, and gas hydrate are blown out to the seafloor when an overpressure eruption occurs and a pockmark forms [78]. Strong gas flares are widely discovered above pockmarks, which indicate that overpressure eruptions exist above the pockmarks [13,33,54,55,64,68,76,105,126].
Carbonate ridges and stacks are found beside pockmarks, and many researchers have indicated that authigenic carbonates occur as trap layers, leading to overpressure eruptions [79,87]. The authigenic carbonates are considered as products related to methane oxidation and sulfate reduction [63] that are import indications of hydrocarbon gas and fluid migration. Geochemical analysis of the sediment samples, pore water, and carbonate from the pockmark area can help predict the gas and fluid sources [85,90,98,100].
However, the gas and fluid flow in shallow marine sediments is still unclear because the process is hard to monitor [109]. Active pockmarks with an overlying active seepage and dormant pockmarks with an underlying gas pipe are widely discovered indicators of shallow gas accumulation [107,118]. Pockmarks are generally distributed above the hydrocarbon reservoir and fault system, indicating the location of the reservoir and migration pathway [20,68,104,111,122,[127][128][129].

Gas Chimney
Gas chimneys are also reported as "gas pipes" induced by gas and fluid percolation and migration in shallow marine sediments. The structures are generally considered as gas and fluid migration pathways. On seismic profiles, "gas plumes" or "gas clouds" are recognized as the indicators of gas pipes, which appear in a columnar geometry or a conic geometry ( Figure 4A) [130]. According to the comprehensive literature investigation, the gas chimney geometries are divided into pipe-like, conic-like, corona-like, and cloud-like shapes ( Figure 4B). Gas chimneys are supposed to be driven by adequate gas accumulation, composed of deep thermogenic gas and shallow biogenic gas [3]. Gas chimneys are recognized as acoustic blank zones, turbulent zones, ambiguous zones and pull-down reflections. The acoustic blank zone generally indicates gas-rich sediments [131], and the pull-down reflection possibly indicates free gas occurrence [132]. However, in other cases, researchers have recognized high amplitude pull-down reflections in the Lower Congo Basin. They interpreted them as stacked paleo-pockmarks [133], and the pull-up reflection can be induced by gas hydrate or carbonate occurrence [20].
Carbonate ridges and stacks are found beside pockmarks, and many research indicated that authigenic carbonates occur as trap layers, leading to overpressu tions [79,87]. The authigenic carbonates are considered as products related to oxidation and sulfate reduction [63] that are import indications of hydrocarbon fluid migration. Geochemical analysis of the sediment samples, pore water, and c from the pockmark area can help predict the gas and fluid sources [85,90,98,100] However, the gas and fluid flow in shallow marine sediments is still unclear the process is hard to monitor [109]. Active pockmarks with an overlying active and dormant pockmarks with an underlying gas pipe are widely discovered indi shallow gas accumulation [107,118]. Pockmarks are generally distributed above drocarbon reservoir and fault system, indicating the location of the reservoir an tion pathway [20,68,104,111,122,[127][128][129].

Gas Chimney
Gas chimneys are also reported as "gas pipes" induced by gas and fluid pe and migration in shallow marine sediments. The structures are generally consi gas and fluid migration pathways. On seismic profiles, "gas plumes" or "gas clo recognized as the indicators of gas pipes, which appear in a columnar geometry o geometry ( Figure 4A) [130]. According to the comprehensive literature investiga gas chimney geometries are divided into pipe-like, conic-like, corona-like, and cl shapes ( Figure 4B). Gas chimneys are supposed to be driven by adequate gas ac tion, composed of deep thermogenic gas and shallow biogenic gas [3]. Gas chim recognized as acoustic blank zones, turbulent zones, ambiguous zones and pull-d flections. The acoustic blank zone generally indicates gas-rich sediments [131], pull-down reflection possibly indicates free gas occurrence [132]. However, in oth researchers have recognized high amplitude pull-down reflections in the Lowe Basin. They interpreted them as stacked paleo-pockmarks [133], and the pull-up r can be induced by gas hydrate or carbonate occurrence [20].  Previous studies have proposed that free gas is supposed to be sealed by the capillary force of the sediments. When the overpressure exceeds the sealing force, free gas starts to migrate upward along the gas pipe. During the migration, the overlying pore fluid in the pipe is expelled out, and the underlying pore fluid is pumped into the pipe. When the gas releases into seawater, the underlying free gas is sealed again by the capillary force produced by the pore fluid in the gas pipe [103]. The process of gas migration and emission is episodic.
In offshore SW Taiwan, extensive gassy sediments exist [134], which indicate the gas migration in shallow marine sediments. Gas percolation generally induces gas accumulation and overpressure, which results in dome structures on the lithological interface. Continuous gas accumulation leads to overpressure increase and small fracture formation [50]. The gas migrates upward along the small faults and forms a gas chimney. The geometries of gas pipes are possibly related to the gas pressure and sediment lithology. The pipe-like gas chimney is inclined to form in high gas pressure conditions and muddy sediments, and the cloud-like gas chimney is inclined to form in low gas pressure conditions and sandy sediments ( Figure 4B).
In many cases, the gas chimneys are located above the bottom simulating reflectors (BSRs), below the BSRs, and through the BSRs, which indicate a good relation with gas hydrates [22,33,34,55,132,133,135]. Gas pipes are also found below pockmarks and seepages, indicating vertical pathways for pockmarks and seepages [127]. The gas migrates upward along salt tectonic faults and forms gas pipes at the upper end of the faults. The gas and fluid then migrate to the seafloor through the gas pipes [20]. When the gas migrates along faults and accumulates below the GHSZ, small fractures form due to overpressure, and periodic earthquakes drive the gas release along with the fractures [125]. Therefore, gas chimneys are supposed to record the multi-active seepage controlled by earthquake activities [122]. Deep thermogenic gas migrates along the deep faults and forms gas pipes that extend to the seafloor in the central North Sea [64]. In the Lower Congo Basin, researchers have recognized gas pipes located above the triple junction of the polygonal faults, indicating a close relationship between them [65].
When a gas pipe extends through the GHSZ, gas hydrates then form in the pipes [20]. When the gas hydrates disassociate to free gas, the free gas then migrates upward to form a gas pipe [13,136], therefore, gas pipes can help locate free gas reservoirs and the gas hydrate layers.

Diapir Structures
Mud and salt diapirs have been extensively studied previously. Mud and salt layers are soft plastic layers and are easily deformed by differential gravity and compaction forces. In many gas hydrate studies, the two structures are recognized and facilitate gas and fluid migration.

Mud Diapirs
Mud diapirs are also reported as "mud volcanos" in the Olimpi diapiric field in the Mediterranean Ridge [137][138][139] and as "mud mounds" in the mud volcano area in the Gulf of Cadiz [140] ( Figure 5A). The structures are focused and widely studied regarding gas and fluid migration [141][142][143]. The mud layers ( Figure 5(B1)) saturated with gas and fluid move upward and form diapirs along with fractures or faults ( Figure 5(B2)). When they erupt on the seafloor, the gas and fluid escape into the seawater, leaving the muddy sediment deposited.The conic-like mud structures are termed mud volcanoes ( Figure 5(B3)) [81,144,145]. The mud diapirs are considered as the previous stage of mud volcanoes, indicating that the overpressure exists in muddy sediments before the mud volcano eruption [134]. Mud diapirs are mainly recognized as dome-like structure on and beneath the seafloor [12,18,146,147]. A previous study proposed that the deep hydrocarbon fluid migrated upward with liquified muddy sediments, solid blocks and shallow biogenic gas, and then erupted on the seafloor [70]. The slope angle of the mud volcano is possibly related to the viscosity of the muddy sediment [5]. The mud volcano is recognized as the bull-eye shape in map views [83]. On the seismic profile, the up-warp diapir geometry is recognized as a mud diapir [35]. Mud volcano outcrops are observed in Azerbaijan, with a methane emission volume of as much as 800 m 3 per year [148,149]. Similar outcrops are observed in Diglipur Island, India [150], northern Italy [151], and Indonesia [152]. and then erupted on the seafloor [70]. The slope angle of the mud volcano is possibly related to the viscosity of the muddy sediment [5]. The mud volcano is recognized as the bull-eye shape in map views [83]. On the seismic profile, the up-warp diapir geometry is recognized as a mud diapir [35]. Mud volcano outcrops are observed in Azerbaijan, with a methane emission volume of as much as 800 m 3 per year [148,149]. Similar outcrops are observed in Diglipur Island, India [150], northern Italy [151], and Indonesia [152]. Previous researchers have proposed the following mud volcano formation mechanism. Firstly, the deep gas percolates upward to the upper plastic clay layer, and the pore spaces are filled with gas, resulting in gravity instability and sediment deformation [29]. Secondly, continuous gas accumulation induces a buoyancy increase, and a mud diapir initially forms in the weakness zone [116,153]. Thirdly, when the mud diapir extrudes into the seafloor, the gas and fluid eject into the seawater, and as muddy sediment is deposited around the ejection site, a mud volcano forms a conic-like shape [146].
The mud volcano formation is associated with the muddy sediment eruption [1,134]. Previous researchers found related evidence by AUV surveys in the Nile Deep Sea Fan [154]. Tectonic activity is supposed to be the primary trigger force driving the mud extrusion under the pre-condition of overpressure [5,29,140,141,143,145,150]. In other cases, the muddy flow driven by overpressure migrates upward through hydraulic fractures to form a mud volcano [130].
Deep heat flows are a possible driving force and fluid source for mud volcanoes [143]. In offshore SW Taiwan, heat flows are measured around the mud diapirs area. The results reveal that higher heat flow values exist above the mud diapirs, indicating that the heat flows migrate upward in the mud diapirs [155]. In the Barbados accretionary wedge, heat fluid roots from the underlying accretionary wedge or adjacent oceanic crust. With the temperature increase, the smectite-illite transformation is accelerated and releases quantities of water [156]. This process possibly promoted overpressure and undercompaction.
Periodic activities such as tectonic activity, overpressure, and heat flow eruption possibly result in biconic mud volcanoes. A buried mud volcano with biconic geometry was discovered in the South Caspian Basin. The South Caspian crust subducts into North Caspian crust, inducing the upper compressed mud layer migrates upward episodically, leading to the biconic geometry [157].
Gas hydrates are discovered within mud volcano areas, indicating that large quantities of methane gas and mud flow migrate upward to the shallow sediments [156]. When the gas and fluid migrate through the GHSZ, free gas is supposed to percolate into the surrounding sediments and form a gas hydrate [145]. Mud diapirs extrude through the Previous researchers have proposed the following mud volcano formation mechanism. Firstly, the deep gas percolates upward to the upper plastic clay layer, and the pore spaces are filled with gas, resulting in gravity instability and sediment deformation [29]. Secondly, continuous gas accumulation induces a buoyancy increase, and a mud diapir initially forms in the weakness zone [116,153]. Thirdly, when the mud diapir extrudes into the seafloor, the gas and fluid eject into the seawater, and as muddy sediment is deposited around the ejection site, a mud volcano forms a conic-like shape [146].
The mud volcano formation is associated with the muddy sediment eruption [1,134]. Previous researchers found related evidence by AUV surveys in the Nile Deep Sea Fan [154]. Tectonic activity is supposed to be the primary trigger force driving the mud extrusion under the pre-condition of overpressure [5,29,140,141,143,145,150]. In other cases, the muddy flow driven by overpressure migrates upward through hydraulic fractures to form a mud volcano [130].
Deep heat flows are a possible driving force and fluid source for mud volcanoes [143]. In offshore SW Taiwan, heat flows are measured around the mud diapirs area. The results reveal that higher heat flow values exist above the mud diapirs, indicating that the heat flows migrate upward in the mud diapirs [155]. In the Barbados accretionary wedge, heat fluid roots from the underlying accretionary wedge or adjacent oceanic crust. With the temperature increase, the smectite-illite transformation is accelerated and releases quantities of water [156]. This process possibly promoted overpressure and undercompaction.
Periodic activities such as tectonic activity, overpressure, and heat flow eruption possibly result in biconic mud volcanoes. A buried mud volcano with biconic geometry was discovered in the South Caspian Basin. The South Caspian crust subducts into North Caspian crust, inducing the upper compressed mud layer migrates upward episodically, leading to the biconic geometry [157].
Gas hydrates are discovered within mud volcano areas, indicating that large quantities of methane gas and mud flow migrate upward to the shallow sediments [156]. When the gas and fluid migrate through the GHSZ, free gas is supposed to percolate into the surrounding sediments and form a gas hydrate [145]. Mud diapirs extrude through the gas hydrate zone, and this phenomenon is discovered in offshore Norway [158], the Barbados accretionary wedge [156], SW Taiwan [35,134], and the Indian Andaman islands [150], which show a good relationship between mud volcanos and gas hydrates. The gas and fluid flow are also identified above the mud diapirs.
Brine pools are discovered located on the mud volcanoes in the Mediterranean Ridge, indicating salt occurrence [69,70,159]. When a mud flow migrates through a salt layer the salt dissolves in the fluid and migrates to the seafloor. Due to its larger density, salt-rich fluid deposits in a crater created by the mud volcano and forms a brine pool. Authigenic carbonate is found around the brine pool, indicating that the hydrocarbon gas and fluid possibly share the mud volcano pathways [70]. Carbonate outcrops are observed in the Nile Deep Sea Fan [154] and eastern Mediterranean mud volcanoes [160]. Carbonate chimneys filled with soft clay in the pipes are discovered in the South China Sea [82] and Mediterranean Ridge [70], indicating that the carbonate forms during the upward migrations of the soft clay.

Salt Diapirs
Salt diapirs is salt tectonics forms that generally present an up-warp geometry caused by the inherent instability of the salt ( Figure 6A). Salt is mostly composed of halite which may contain various amounts of evaporites [161]. Salt layers are considered to be mechanically weak, incompressible, less dense, and fluid-like [162]. These characteristics make salt layers inherently unstable and are usually deformed by differential sediment loading and gravity spread [30][31][32]. The salt diapir generally induces many extensional faults that the gas and fluid can migrate through ( Figure 6B) [36]. In the Ionian Basin, salt-tectonic structures play a critical role in creating migration pathways (both lateral and vertical) and remobilizing overpressured fluids [163]. The low sediment porosity and permeability in the Ionian zone imply fluid migration through "fracture macro-permeability conduits" in the vicinity of faulted and highly deformed zones [44,45,163]. gas hydrate zone, and this phenomenon is discovered in offshore Norway [158], the Bar bados accretionary wedge [156], SW Taiwan [35,134], and the Indian Andaman island [150], which show a good relationship between mud volcanos and gas hydrates. The ga and fluid flow are also identified above the mud diapirs.
Brine pools are discovered located on the mud volcanoes in the Mediterranean Ridge indicating salt occurrence [69,70,159]. When a mud flow migrates through a salt layer th salt dissolves in the fluid and migrates to the seafloor. Due to its larger density, salt-ric fluid deposits in a crater created by the mud volcano and forms a brine pool. Authigeni carbonate is found around the brine pool, indicating that the hydrocarbon gas and flui possibly share the mud volcano pathways [70]. Carbonate outcrops are observed in th Nile Deep Sea Fan [154] and eastern Mediterranean mud volcanoes [160]. Carbonate chim neys filled with soft clay in the pipes are discovered in the South China Sea [82] and Med iterranean Ridge [70], indicating that the carbonate forms during the upward migration of the soft clay.

Salt Diapirs
Salt diapirs is salt tectonics forms that generally present an up-warp geometry cause by the inherent instability of the salt ( Figure 6A). Salt is mostly composed of halite whic may contain various amounts of evaporites [161]. Salt layers are considered to be mechan ically weak, incompressible, less dense, and fluid-like [162]. These characteristics mak salt layers inherently unstable and are usually deformed by differential sediment loadin and gravity spread [30][31][32]. The salt diapir generally induces many extensional faults tha the gas and fluid can migrate through ( Figure 6B) [36]. In the Ionian Basin, salt-tectoni structures play a critical role in creating migration pathways (both lateral and vertica and remobilizing overpressured fluids [163]. The low sediment porosity and permeabilit in the Ionian zone imply fluid migration through "fracture macro-permeability conduits in the vicinity of faulted and highly deformed zones [44,45,163]. Critical slope angle is a significant factor that triggers salt tectonic activity [165,166 In an oblique slope, salt layers are supposed to occur as a brittle-ductile layer which gen erally leads to differential loading, which is thought to initiate the diapir reactivatio [32,167]. Salt layers generally exist as detachment layers. The overlying sediments ca slide along the salt layers under the influence of gravity and then form salt diapir [168,169]. Intense transform-fault activity can also lead to salt diapirs [170]. In that case the basement faults and the basement steps are thought to control the salt evolution [171 173]. In addition, episodic sedimentation in the depocenter is considered to change loca pressure gradients that generally lead to salt diapir [174].
Gravity spread and gravity glide can induce salt diapirs [165]. In the Angolan mar gin, related salt diapirs are thought to be caused by compressional deformation of th Critical slope angle is a significant factor that triggers salt tectonic activity [165,166]. In an oblique slope, salt layers are supposed to occur as a brittle-ductile layer which generally leads to differential loading, which is thought to initiate the diapir reactivation [32,167]. Salt layers generally exist as detachment layers. The overlying sediments can slide along the salt layers under the influence of gravity and then form salt diapirs [168,169]. Intense transform-fault activity can also lead to salt diapirs [170]. In that case, the basement faults and the basement steps are thought to control the salt evolution [171][172][173]. In addition, episodic sedimentation in the depocenter is considered to change local pressure gradients that generally lead to salt diapir [174].
Gravity spread and gravity glide can induce salt diapirs [165]. In the Angolan margin, related salt diapirs are thought to be caused by compressional deformation of the gravity spreading system [175][176][177]. The salt diapirs in the northern Gulf of Mexico are dominated by gravity gliding [178].
The activities of the salt diapirs induce potential pathways for gas and fluid. Many pockmarks are found located above the faults in the eastern Mediterranean. The faults caused by salt diapirs possibly provide conduits for gas and fluid migration [179]. During the salt diapir extrusion, radial faults develop around the top of the diapir and dramatically modify the geometry of the polygonal faults [180]. When the fluid migrates through the salt layers, the salt is transported to the seafloor and then forms a brine pool [69,70,159].

Sediment Ruptures
Landslides, polygonal faults and tectonic faults are driven by gravity, compaction forces and tectonic forces. The structures have one or multiple prominent rupture surfaces that can provide a conduit for gas and fluid migration.

Landslides
Submarine landslides generally develop in shallow marine sediments on the continental slope, caused by sediments sliding along continental slopes under gravity, with sediments transported from the slope to the deep seafloor ( Figure 7A) [181]. The largest documented submarine slide is the Storegga Slide in Norway [182]. The landslide is composed of head-scarp, sliding surface (basal shear surface/basal glide surface), slide-body, and toe-scarp ( Figure 7B) [3]. Landslides are considered as mass-transport depositions, and the scales generally range from centimeters to hundreds of kilometers [183,184]. Submarine landslides are soft-sediment deformations with apparent slide surfaces and offsets, which develop in sediment layers with high sedimentation rates, low permeability, and low shear strength [185,186]. Landslide research can estimate the hazards and risks [187], including seabed infrastructure destruction, coastal area collapse, and destructive tsunamis [188][189][190]. The landslide deposits can help deduce paleo-geological conditions and tectonic activities. Stacked submarine landslide deposits seen in the northern South China Sea are probably induced by multi-period sliding and erosion [191,192]. The canyons are eroded by turbidity currents induced by a landslide, the turbidites deposit in the canyons [193], which reveal the northeastward bottom current. The multi-period landslide deposits are also recognized in the southern Scotia Sea, which are considered as evidence of paleo-earthquakes [194]. the salt diapir extrusion, radial faults develop around the top of the diapir and cally modify the geometry of the polygonal faults [180]. When the fluid migrates the salt layers, the salt is transported to the seafloor and then forms a br [69,70,159].

Sediment Ruptures
Landslides, polygonal faults and tectonic faults are driven by gravity, com forces and tectonic forces. The structures have one or multiple prominent rupture that can provide a conduit for gas and fluid migration.

Landslides
Submarine landslides generally develop in shallow marine sediments on t nental slope, caused by sediments sliding along continental slopes under grav sediments transported from the slope to the deep seafloor ( Figure 7A) [181]. Th documented submarine slide is the Storegga Slide in Norway [182]. The landslid posed of head-scarp, sliding surface (basal shear surface/basal glide surface), sli and toe-scarp ( Figure 7B) [3]. Landslides are considered as mass-transport dep and the scales generally range from centimeters to hundreds of kilometers [183,1 marine landslides are soft-sediment deformations with apparent slide surfaces sets, which develop in sediment layers with high sedimentation rates, low perm and low shear strength [185,186]. Landslide research can estimate the hazards a [187], including seabed infrastructure destruction, coastal area collapse, and de tsunamis [188][189][190]. The landslide deposits can help deduce paleo-geological co and tectonic activities. Stacked submarine landslide deposits seen in the northe China Sea are probably induced by multi-period sliding and erosion [191,192]. yons are eroded by turbidity currents induced by a landslide, the turbidites depo canyons [193], which reveal the northeastward bottom current. The multi-peri slide deposits are also recognized in the southern Scotia Sea, which are consider idence of paleo-earthquakes [194]. Most submarine slopes are inherently stable, and usually require external factors to trigger landslides [188,196]. Climate-induced stress variation, including glaciation succession, eustatic sea level, rapid sediment load, bottom flow temperature variation, and gas hydrate disassociation, significantly changes the slope stability [197,198]. The Storegga Slide located off the Norwegian margin is the largest submarine landslide in the world [158]. The landslide is thought to respond to climatic variability. The Storegga Slide is also a result of rapid loading with excess pore pressure, reducing the effective shear strength of the underlying clays [188,199]. Field measurements show that extra pore pressure still exists in the adjacent areas [200].
Salt layers are mechanically weak and flow like fluids [162], which leads to residual friction decreases. As a result, submarine landslides are readily initiated above salt layers. Submarine landslides are observed on the continental slope in the southeastern Mediterranean, whereb they are supposed to interact with a group of step-faults induced by extensional salt tectonic activity [201].
Submarine landslides are probably related to earthquakes and high fluid pressure [183,196,202], and excess pore pressure is considered the primary pre-condition, and earthquakes trigger the landslides [190,203]. High fluid pressure is generally caused by rapid sedimentation and gas and fluid migration. During lower sea level periods, the shallow gas exsolution induces buoyancy increases, leading to pore pressure increases and slope instability [193,204]. In addition, the fluid accumulation leads to sediment liquidation and pore pressure increases. As a result, shallow sediments flow more easily and form shear failures [26,205].
Submarine landslides are generally caused by slope instability, and many factors collectively make significant contributions to slope instability [189,206]. As presented above, the pre-condition including soft plastic layers (clay or salt), excess pore pressure, gas-rich sediments, and liquified sediments. The main trigger factors are considered to be earthquakes and storm waves. Submarine slope failures mostly occur on very gentle slopes with gradients of less than 5 • [205], indicating that slope gradients are probably not the dominant factor of the landslide.
Numerous submarine landslides are close to pockmarks and mud volcanoes linked by fluid accumulation [197]. Pockmarks are recognized on the seafloor induced by fluid escape that is probably derived from the underlying mass transport deposits [195]. In addition, pockmark fields are developed adjacent to landslide areas, indicating a close relationship between them in the US continental slope [207] and western Mediterranean Sea [26].
According to a comprehensive literature investigation, we summarize three formation mechanisms of polygonal faults [244].
Density inversion is a significant mechanism [211,215,216,229,236]. Rapid sedimentation generally forms fluid-rich clays. The pore fluid in the upper clay migrates into overlying coarse-grained layers. The pore fluid in the lower clay migrates into underlying coarse-grained layers sealing the pore fluid in the middle clay layer. As compaction continues, the middle pore fluid is difficult to extrude, so that the middle clay density becomes lower than that of the upper clay. This state is termed density inversion and generally leads to gravity instability and shear strength decreases. When the pore fluid pressure exceeds the limit, fractures form, and the pore fluid is released. The density inversion indicator is probably the wave-like lithological interfaces induced by differential overpressure and buoyancy accumulation. According to a comprehensive literature investigation, we summarize three formation mechanisms of polygonal faults [244].
Density inversion is a significant mechanism [211,215,216,229,236]. Rapid sedimentation generally forms fluid-rich clays. The pore fluid in the upper clay migrates into overlying coarse-grained layers. The pore fluid in the lower clay migrates into underlying coarse-grained layers sealing the pore fluid in the middle clay layer. As compaction continues, the middle pore fluid is difficult to extrude, so that the middle clay density becomes lower than that of the upper clay. This state is termed density inversion and generally leads to gravity instability and shear strength decreases. When the pore fluid pressure exceeds the limit, fractures form, and the pore fluid is released. The density inversion indicator is probably the wave-like lithological interfaces induced by differential overpressure and buoyancy accumulation.
Syneresis fracture is another mechanism [218,226,232,[245][246][247][248]. During the early burial, the shallow marine sediments exist in a colloidal state. The fine-grained particles tend to arrange closer to each other under the influence of interparticle attractive forces. As a result, the pore fluid is extruded, and the sediments become more consolidated. The process is suggested to begin at the interface between the seawater and seafloor [243]. As the burial depth increases, the vertical sediments shrink showing normal compaction, and the lateral shrinkage apprears as fractures. The syneresis happens randomly and induces irregular trends and strikes of the polygonal faults. Syneresis fracture is another mechanism [218,226,232,[245][246][247][248]. During the early burial, the shallow marine sediments exist in a colloidal state. The fine-grained particles tend to arrange closer to each other under the influence of interparticle attractive forces. As a result, the pore fluid is extruded, and the sediments become more consolidated. The process is suggested to begin at the interface between the seawater and seafloor [243]. As the burial depth increases, the vertical sediments shrink showing normal compaction, and the lateral shrinkage apprears as fractures. The syneresis happens randomly and induces irregular trends and strikes of the polygonal faults.
Shear failure is another mechanism [208,210,214,217,220,226,246,249,250]. Researchers proposed to use the Mohr-Coulomb Law to explain the formation of polygonal faults. The shear strength of water-saturated sediments is weaker than that of rock; therefore, these sediments more comfortably form shear failures. In the Mohr Circle, shear failure forms when the Mohr Circle contacts the Mohr-Coulomb Failure Envelopes. Two factors make contributions to the location and size of the Mohr Circle. First, lateral extensional stress induced by compaction increases as the burial depth increases, inducing a minimum principal stress decrease and the radius of Mohr Circle increases. When the Mohr Circle contacts the Mohr-Coulomb Failure Envelopes, shear failures form. Second, continuous pore pressure increases lead to a decrease of the maximum principal stress and the minimum principal stress. The Mohr Circle shifts leftward and contacts the Mohr-Coulomb Failure Envelopes, and shear failures form. In addition, the coefficient of friction in finegrained sediments is thought to be exceptionally low, and fractures more easily form under the mechanical compaction induced by gravity [210,220,251].
Fine-grained sediments are considered as the sealing layers in the marine environment [223,246], but the occurrence of polygonal faults substantially improves the permeability of fine-grained layers [252]. Therefore, polygonal faults are the relief pathway of pore fluid pressure and the migration pathway of the deep hydrocarbon gases and fluids [212,222,239,253,254]. For one thing, polygonal faults improve the permeability of the fine-grained layers and have a negative influence on the burial of CO 2 and nuclear waste under seafloor [223,246,249]. For another thing, the permeability increase positively influences the upward migration of hydrocarbon gas and fluid that migrate through the polygonal faults to the GHSZ and form gas hydrates [208].
Pockmarks and gas pipes are recognized off the shore of Norway. Most of them are distributed above polygonal faults, indicating that the gas and fluid migrated through the polygonal faults and was released at the pockmarks [255]. Some polygonal faults extend to the seafloor with pockmarks above them, and some gas pipes are distributed at the top end of the polygonal faults [107,123,243].

Tectonic Faults
Tectonic faults are generally developed around the plate boundaries in both continents and oceans. An active submarine fault is one of the tectonic faults used to describe repeat fault activities in recent geological time, indicating periodic tectonic activities such as earthquakes [28,38,[256][257][258][259][260]. Tectonic faults are generally recognized by reflector deformation, offset, and divarication ( Figure 9A). Submarine active faults connect the deep underlying strata with the seafloor, providing effective pathways for deep thermogenic gases and fluids [28]. Therefore, the faults can help to estimate the hydrocarbon gas and fluid sources. Submarine active faults are mainly studied in the Sea of Marmara ( Figure  9B) [28,38,256,258,[261][262][263][264][265][266][267][268], the eastern Mediterranean Sea [269,270], the Aegean Sea [271] and the Japan Sea [272][273][274]. ers proposed to use the Mohr-Coulomb Law to explain the formation of polygonal faults. The shear strength of water-saturated sediments is weaker than that of rock; therefore, these sediments more comfortably form shear failures. In the Mohr Circle, shear failure forms when the Mohr Circle contacts the Mohr-Coulomb Failure Envelopes. Two factors make contributions to the location and size of the Mohr Circle. First, lateral extensional stress induced by compaction increases as the burial depth increases, inducing a minimum principal stress decrease and the radius of Mohr Circle increases. When the Mohr Circle contacts the Mohr-Coulomb Failure Envelopes, shear failures form. Second, continuous pore pressure increases lead to a decrease of the maximum principal stress and the minimum principal stress. The Mohr Circle shifts leftward and contacts the Mohr-Coulomb Failure Envelopes, and shear failures form. In addition, the coefficient of friction in fine-grained sediments is thought to be exceptionally low, and fractures more easily form under the mechanical compaction induced by gravity [210,220,251].
Fine-grained sediments are considered as the sealing layers in the marine environment [223,246], but the occurrence of polygonal faults substantially improves the permeability of fine-grained layers [252]. Therefore, polygonal faults are the relief pathway of pore fluid pressure and the migration pathway of the deep hydrocarbon gases and fluids [212,222,239,253,254]. For one thing, polygonal faults improve the permeability of the finegrained layers and have a negative influence on the burial of CO2 and nuclear waste under seafloor [223,246,249]. For another thing, the permeability increase positively influences the upward migration of hydrocarbon gas and fluid that migrate through the polygonal faults to the GHSZ and form gas hydrates [208].
Pockmarks and gas pipes are recognized off the shore of Norway. Most of them are distributed above polygonal faults, indicating that the gas and fluid migrated through the polygonal faults and was released at the pockmarks [255]. Some polygonal faults extend to the seafloor with pockmarks above them, and some gas pipes are distributed at the top end of the polygonal faults [107,123,243].

Tectonic Faults
Tectonic faults are generally developed around the plate boundaries in both continents and oceans. An active submarine fault is one of the tectonic faults used to describe repeat fault activities in recent geological time, indicating periodic tectonic activities such as earthquakes [28,38,[256][257][258][259][260]. Tectonic faults are generally recognized by reflector deformation, offset, and divarication ( Figure 9A). Submarine active faults connect the deep underlying strata with the seafloor, providing effective pathways for deep thermogenic gases and fluids [28]. Therefore, the faults can help to estimate the hydrocarbon gas and fluid sources. Submarine active faults are mainly studied in the Sea of Marmara ( Figure  9B) [28,38,256,258,[261][262][263][264][265][266][267][268], the eastern Mediterranean Sea [269,270], the Aegean Sea [271] and the Japan Sea [272][273][274]. The North-Anatolian Fault (NAF) is taken as an example. The NAF is a continental transform fault distributed along the boundary of the Anatolian plate and Eurasian plate, which is thought to induce devastating earthquakes. The fault splays into three branches, and two branches make up the boundary of the Sea of Marmara [28]. The NAF is induced by Anatolia-Aegean block westward escape because of the Arabia plate subduction into the Eurasia Plate [256]. As a result, deep thermogenic hydrocarbon gases and fluids migrate upward along the fault planes and form cold seepages on the seafloor [263]. Authigenic carbonate and microbial mats are also observed on the seafloor along the fault segment [28,263]. The faults are characterized by creep deformation and abrupt deformation, and any abrupt deformation is generally produced by earthquakes. Previous studies indicated that most earthquakes are reactive along the initial active faults [257][258][259].
Acoustic anomalies indicating the gas venting are recognized on acoustic profiles. The anomalies are distributed along NAF in the Sea of Marmara, revealing a close relationship between gas seepage and fault activities [264]. Active seepage is observed above the NAF, which is considered as a conduit for the gas seepage [28].

Summary and Perspectives
Shallow marine structures are developed in shallow sediments at depths of less than 1000 mbsf. According to their characteristics and formation mechanisms, these structures are classified into three groups: overpressure-associated structures, diapir structures and sediment ruptures.
The overpressure-associated structures are mainly induced by pore pressure changes related to gas and fluid accumulation, migration and release, including seepages, pockmarks and gas chimneys. Diapir structures are primarily driven by differential loading and gravity spread in plastic layers and include mud diapirs and salt diapirs. The sediment ruptures are associated with gravity, compaction forces and tectonic forces, and including landslides, polygonal faults and tectonic faults. In general, each structure would not be induced by a single mechanism, and many external factors might have been involved.
These structures can be paths for gas and fluid migration. The tectonic faults ( Figure 10H) generally cut into deep hydrocarbon reservoirs and extend laterally a long distance, being the primary pathways for deep thermogenic gases. Some tectonic faults vertically extend into shallow horizons and act as shallow gas pathways. The polygonal faults ( Figure 10G) are generally formed in deep and shallow fine-grained sediment layers that enhance the sediment permeability. The deep polygonal faults are suggested to play a role as pathways for the gas and fluid migration and the shallow ones form as reservoirs. Abundant gas and fluid accumulation in fine-grained sediments can decrease the sediment density and shear strength, making the layers easy to deform and offset by differential forces. As a result, the deeply buried layers form mud diapirs ( Figure 10D), and the shallow layers form landslides ( Figure 10F). The mud diapirs with high gas and fluid saturation erupt at the seafloor and form mud volcanoes ( Figure 10D), releasing gases and fluids into the seawater. When the mud diapirs extrude through a salt diapir ( Figure 10E) area, the salt dissolves and is carried to the seafloor to form a brine pool. However, the mud diapirs with low gas and fluid saturation extrude to the seafloor and form mud mounds ( Figure 10D). When the gas and fluid accumulate in shallow layers and form an overpressure, pre-existing or after-born gas chimneys are the preferential migration pathways. Some gas chimneys ( Figure 10C) extend vertically into the seafloor to release the overpressured gases and fluids. A strong overpressure can form a vigorous seepage (Figure 10A), and a weak overpressure forms a gentle seepage. The sediments around the seepage site are disturbed by the gas and fluid flow, leaving a pockmark ( Figure 10B) on the seafloor. A regular and continuous flow can form a circular pockmark, and the irregular and episodic ones form irregular pockmarks.
These shallow structures have significant impacts on the gas hydrate and authigenic carbonate distribution. Gas hydrates are formed at relatively low temperatures and high pressures by hydrocarbon gases and water that is widely found in shallow marine sediments. The gas source is generally composed of deep thermogenic gases and shallow biogenic gases. The gases migrate upward along the pathways provided by the structures to the GHSZ and sulfate-methane interface (SMI) and form gas hydrates and authigenic carbonates, respectively. However, gas hydrates and authigenic carbonates are tight layers that generally seal the pathway and induce overpressures beneath them. With the increase in overpressure, a new pathway breaks through or bypasses the sealing area driven by pressure. The new pathway is generally a new gas chimney that extends to the seafloor and forms a seepage. In some cases, the sealing layers are developed at a very shallow depth, and the gas hydrates and authigenic carbonates are blown out to the seafloor and deposit around the seepage site. Some bacterial species can feed on methane and take part in the methane and sulfate geochemical cycle, so sometimes gas hydrates, authigenic carbonates and bacterial communities occur associatively around a seepage site.
pressures by hydrocarbon gases and water that is widely found in shallow marine ments. The gas source is generally composed of deep thermogenic gases and shallo ogenic gases. The gases migrate upward along the pathways provided by the struc to the GHSZ and sulfate-methane interface (SMI) and form gas hydrates and authi carbonates, respectively. However, gas hydrates and authigenic carbonates are tigh ers that generally seal the pathway and induce overpressures beneath them. With th crease in overpressure, a new pathway breaks through or bypasses the sealing area d by pressure. The new pathway is generally a new gas chimney that extends to the sea and forms a seepage. In some cases, the sealing layers are developed at a very sh depth, and the gas hydrates and authigenic carbonates are blown out to the seafloo deposit around the seepage site. Some bacterial species can feed on methane and take in the methane and sulfate geochemical cycle, so sometimes gas hydrates, authigeni bonates and bacterial communities occur associatively around a seepage site.