Seismic stratigraphy and structural evolution of the South Korea Plateau, East Sea (Sea of Japan)

The South Korea Plateau (SKP), a typical submarine plateau, preserves an important tectono‐sedimentary evolutionary record and represents a major frontier area for petroleum exploration in the East Sea (Sea of Japan). However, its tectonic mechanisms and their controls on sedimentary fill are underexplored. Here, we present the first integrated tectonostratigraphic framework of the SKP using reprocessed, two‐dimensional, seismic‐reflection profiles and borehole data. Four regional megasequence boundaries are interpreted, delineating four tectonostratigraphic packages: the syn‐rift (MS1), post‐rift phase 1 (MS2), post‐rift phase 2 (MS3) and syn‐compression (MS4) megasequences. We propose a four‐stage structural and sedimentary evolution model for the SKP based on the megasequences and structural development. Stage‐1 (latest Late Oligocene−Early Miocene): the SKP was rifted and extended through block faulting, resulting in the formation of rift basins dominated by fan‐delta and shallow‐lacustrine depositional systems. Stage‐2 (late Early Miocene−Middle Miocene): hemipelagic sedimentation prevailed with gravity‐controlled slope failures under a tectonically stable environment associated with slow thermal subsidence. Stage‐3 (late Middle Miocene−Late Miocene): continued thermal subsidence allowed the predominance of hemipelagic biogenic deposits accompanied by intermittent mass‐wasting‐induced turbidites and resulted in the development of a polygonal fault system. Stage‐4 (Early Pliocene−present): the SKP was influenced by E−W compression caused by an eastward movement of the Eurasian plate. Turbiditic and hemipelagic sedimentation was predominant with turbidity‐flow‐leveed channels derived from direct riverine input or through slope failures. Based on this tectonostratigraphic analysis, we reveal the variation in depositional systems and sand‐dispersal patterns for the SKP, highlighting potential targets for sandstone reservoirs: MS1, fan‐deltas and lacustrine‐fan turbidites; MS3, deepwater fan turbidites; and MS4, deepwater fan turbidites, channel‐levee complexes and turbidite frontal‐splay deposits. This study proposes a structural and sedimentary evolution model for the SKP that could enhance our understanding of reservoir potential for petroleum‐exploration in the future.


| INTRODUCTION
The South Korea Plateau (SKP) is a typical submarine plateau, located in a tectonically crucial zone under the influence of relative motions between the Eurasian, Pacific and Philippine Sea plates (Figure 1a).The plateau is located at the eastern continental margin of the Korea Peninsula and covers an area of approximately ca. 11,300 km 2 , characterised by a flat-topped feature lying a few hundred metres above the surrounding seafloor (Chough et al., 2018; Figure 1b).Since the early 2000s, the SKP has attracted attention in terms of gas-hydrate exploration, and extensive two-dimensional (2D), multi-channel, seismic data were acquired by the Korea Institute of Geoscience and Mineral Resources (KIGAM) between 2000 and 2004 (Figure 1c).Based on these geophysical data, numerous studies have been carried out in the fields of seismic stratigraphy (Cukur et al., 2015(Cukur et al., , 2018;;Horozal et al., 2017;Kwon et al., 2009), structural evolution (Chough et al., 2018;Kim et al., 2007;Kim, Lee, Choi, et al., 2015;Kim, Lee, Park, et al., 2015), volcano-tectonic evolution (Kim et al., 2011) and geohazard assessment (Cukur et al., 2018;Kim et al., 2022).Moreover, the Integrated Ocean Drilling Program (IODP) Expedition 346, which collected data at Site U1430 in the Eastern South Korea Plateau (ESKP), played an important role in enhancing our understanding of palaeoclimatic and palaeoceanographic variability (Kim et al., 2019;Kim, Yi, et al., 2020;Kim, Yoo, et al., 2020).In addition to these previous studies, gas-hydrate research has been actively conducted in the SKP.For that purpose, seismic data reprocessing was completed in 2021, offering an opportunity to show subsurface structures and features in outstanding quality.In the present study, we interpret the reprocessed seismic data for the first time to establish a tectonostratigraphic framework and identify major tectonic events and phases of basin infill in the SKP, based on seismostratigraphy concepts (Vail et al., 1977).
The early Exxon-derived seismic-and sequencestratigraphic approach has provided a basin foundation and conceptual key to analyse basin fills in hydrocarbon exploration (Posamentier et al., 1988;Vail, 1987;Van Wagoner et al., 1988).Seismic-stratigraphy is a general methodology for reconstructing stratigraphic frameworks using seismic data (Vail, 1987).In the late 1980s, this concept evolved into sequence-stratigraphy with the incorporation of outcrop, well and seismic data (Posamentier et al., 1988;Vail, 1987).In this approach, depositional systems and sandbody distribution are interpreted based on stratal stacking patterns and facies distribution, controlled by tectonic subsidence, sea-level change, sediment supply and climatic changes (Jervey, 1988;Posamentier et al., 1988;Vail et al., 1977;Van Wagoner et al., 1988).The early seismic-and sequencestratigraphic concepts were focused on sedimentation, mainly controlled by eustatic sea-level changes over tectonically stable continental margins (Shanley & McCabe, 1994).Since the 2000s, however, tectonostratigraphic or tectonosequence stratigraphic analysis has increasingly attracted the attention of researchers as a means to interpret sequence architecture and tectonic controls on sedimentation in tectonically active continental rift-basins (Zhou et al., 2014).In the previous tectonostratigraphic studies, a tectonostratigraphic unit (i.e.megasequence) is defined as a secondorder sequence developed over timescales of 10-100 Myr (Catuneanu, 2006).Megasequence represents a group of superposed sequences, equivalent to a Sloss-scale sequence (Catuneanu, 2006;Mitchum et al., 1977).Because the formation of a tectonostratigraphic unit is mainly controlled by basin-forming tectonism, a tectonostratigraphic unit provides valuable information about basin evolution processes and regional basin-infill history (Roberts & Bally, 2012).The SKP is composed of several rift basins (half-grabens), including the Bandal (BB), Okgye (OB) and Onnuri (ON) basins (Kim et al., 2022;Figure 1c).In the Cenozoic, the SKP was deformed by extension associated with back-arc rifting and separation of the southwestern Japan Arc (Kim, Lee, Choi, et al., 2015;Kim, Lee, Park, et al., 2015).In the present study, we decipher the basin morphology and sediment fill of the SKP as controlled by highly active tectonism.
During the past few years, several extensive seismicsequence analyses have been conducted in the SKP, and

Highlights
• Tectonostratigaphy of the South Korea Plateau (SKP) was investigated using seismic reflection data.• The SKP consists of four tectonostratigraphic packages (2nd-order megasequences).• Four-stage structural and sedimentary evolution model of the SKP was suggested.• Three potential targets for promising sandstone reservoirs were highlighted.
the findings have highlighted that tectonics influenced the sedimentary architecture and depositional history (Cukur et al., 2015(Cukur et al., , 2018;;Horozal et al., 2017;Kim, Lee, Choi, et al., 2015;Kim, Lee, Park, et al., 2015;Kwon et al., 2009; Figure 2).In the Western South Korea Plateau (WSKP), Kwon et al. (2009) identified four depositional-sequence units and suggested an evolutionary history model influenced by rifting-related volcanism (Figure 2).Cukur et al. (2015) also defined the seismic stratigraphy of four seismic units and presented a basin-extension mechanism based on faultderived stretching factors in the northern part of the WSKP (Figure 2).In the ESKP, Horozal et al. (2017) divided the total sediment fill into three megasequences and suggested sedimentary and structural evolution mainly controlled by regional/local tectonics (Figure 2).However, because previous seismic sequence analyses have focused on local or limited areas within the SKP, each scattered stratigraphic column cannot be correlated with each other, resulting in the absence of an integrated stratigraphic framework (Figure 2).Moreover, the existing understanding of the spatio-temporal characteristics and mechanisms of formation for the geological structures of the SKP, as well as their stratigraphic correlation with the seismic megasequences, is still limited, causing much controversy about the tectonic evolution (e.g.timing of compressional stress) of the SKP (Cukur et al., 2015(Cukur et al., , 2018;;Horozal et al., 2017;Kim, Lee, Choi, et al., 2015;Kim, Lee, Park, et al., 2015;Kwon et al., 2009; Figure 2).Consequently, it is essential to undertake systematic analyses based on the tectonostratigraphy to explore the controlling effects of tectonics on sequence architecture and establish an integrated tectonostratigraphic framework in the SKP.
In this paper, we present new structural and stratigraphic interpretations across the entire SKP based on newly reprocessed seismic-reflection data.The interpretations were derived by (1) defining a seismo-stratigraphic framework and identifying main tectonic events influencing the sedimentary succession, (2) presenting time-structure and -thickness maps to understand spatiotemporal variations of sediment fill, (3) assessing depositional systems and processes from seismic-facies analysis and (4) investigating and classifying geologic structures to reveal their relationship with the stratigraphic framework.
Finally, we provide an avenue for understanding sedimentdistribution and -dispersal patterns to propose potential reservoirs, which represent significant future petroleumexploration sites in the SKP.

| Physiography
The East Sea (Sea of Japan) is located between the Asian continent and Japan (Figure 1a).The sea is a semi-enclosed back-arc basin, including three major sub-basins (i.e.Japan, Ulleung and Yamato Basins; Figure 1b).The basins are bordered by submarine structural highs (e.g. the Korea Plateau, Oki Bank and Yamato Ridge), originated from multi-axial, back-arc extension since early Oligocene time (ca. 32 Ma;Tamaki, 1988; Figure 1b).The East Sea (Sea of Japan) is connected to the Pacific Ocean, Okhotsk Sea and East China Sea through four shallow and narrow straits (Kim, Lee, Choi, et al., 2015;Kim, Lee, Park, et al., 2015).
The SKP, a submerged topographic high, is located in the northwestern corner of the Ulleung Basin (Figure 1c).It can be divided into two parts: the ESKP and WSKP, which are separated by the North Ulleung Trough (NUT; Figure 1c).The water depth of the WSKP ranges from less than 900 m to more than 2000 m, and its area is approximately ca. 7800 km 2 .The acoustic basement of the WSKP consists of shallow-seated dome structures, including rifted continental crustal blocks (e.g.Northwest Ulleung Ridge) and volcanic edifices (e.g.Kiminu Seamount).Between the dome structures, three major depressions (i.e. the BB, OB and ON) are present, which reach depths below 2000 m.In the WSKP, most of the geological structures (i.e.dome structures, depressions and basement-related faults) are aligned in the N−S or NE-SW directions.The ESKP, located to the east of the NUT, is smaller than the WSKP and exhibits a relatively monotonous geomorphology.The water depth of the ESKP ranges from less than 700 m to more than 2000 m, and its area is approximately 2800 km 2 .The ESKP is bound by the steep slope of the North Ulleung Escarpment (NUE) to the west and gradually deepens southward; it is connected to the deep Ulleung Interplain Gap (UIG; Figure 1c).In contrast to the WSKP, the ESKP is characterised by E−W or EEN−WWS trending geological structures.

| Tectonic setting
The East Sea (Sea of Japan) is a typical back-arc basin originating from complex extensional motion between the Eurasian, Pacific and Philippine Sea plates (ca.32 Ma; Lallemand & Jolivet, 1986;Tamaki & Honza, 1985;Tamaki et al., 1992; Figure 1a).In Early Oligocene time, the East Sea (Sea of Japan) started to form by extension and thinning of the proto-Japan on the margin of the Eurasian plate (Tamaki, 1995;Tamaki et al., 1992).Since Early Miocene time, three major sub-basins (i.e. the Japan, Yamato and Ulleung basins) in the East Sea (Sea of Japan) experienced different tectonic evolutionary processes.Seafloor spreading prevailed in the Japan Basin, resulting in smooth topography and a symmetric pattern of magnetic anomalies (Tamaki et al., 1992).Conversely, the southern East Sea (Sea of Japan) was dominated by crustal extension accompanied through southward drift and rotation of the northeastern and southwestern Japan, forming the Ulleung and Yamato basins (Tamaki et al., 1992).In Middle Miocene time, compressional stress, caused by northward collision of the Izu-Bonin Arc against the southwestern Japan, prevailed throughout the East Sea (Sea of Japan; Figure 1a), generating a variety of compressional structures in the southwestern Ulleung Basin (e.g.folding and faulting; Chough & Barg, 1987;Ingle Jr., 1992;Lee et al., 2011;Yoon & Chough, 1995).
The SKP is a remnant of continental crust formed by the separation of the southwestern Japan due to backarc rifting (Kim, Lee, Choi, et al., 2015;Kim, Lee, Park, et al., 2015).The acoustic basement of the SKP is characterised by extensional block faulting with intervening fault-controlled rift-basins (Chough et al., 2018;Kim et al., 2007Kim et al., , 2011;;Kim, Lee, Choi, et al., 2015;Kim, Lee, Park, et al., 2015).Kim, Lee, Choi, et al. (2015) and Kim, Lee, Park, et al. (2015) suggested a three-stage sequential episode of back-arc rifting to spreading processes and demonstrated a prominent breakup-unconformity separating syn-and post-rift sequences in the SKP (Figure 2).Afterward, Chough et al. (2018) presented a new tectonic evolution model of the SKP through a regional dextral strike-slip fault system.In addition, they highlighted the importance of interpreting structural elements with vertically unexaggerated seismic profiles and presented novel insights on basement-related fault distribution in the SKP.Since Middle Miocene time, the SKP has experienced contractional deformation (e.g.gentle anticlines and reverse faults) along the western margin triggered by the eastward movement of the Eurasian plate (Cukur et al., 2015).

| DATA SETS AND METHODS
We used a total of 7540 line-km of multi-channel, seismic data obtained from the SKP, UIG and the northern Ulleung Basin in 2001, 2002 and 2004 by the KIGAM's research vessel TAMHAE 2 (Figure 1c).In 2001, approximately 2270 line-km seismic data were acquired in the northwestern Ulleung Basin, using an 80-channel (1000 m-long) streamer and one 925-in 3 (2000 psi) air-gun array.The hydrophone group interval and shot spacing were 12.5 and 25 m respectively.The data were recorded at sampling intervals of 1 ms with a maximum recording length of 4 s two-way travel time (TWT).In 2002, approximately 2650 line-km seismic data were obtained in the WSKP, using an 80-channel (1000 mlong) streamer recording shots from one 925-in 3 (2000 psi) air-gun array.The record length was set to 6 s TWT.Lastly, in 2004, approximately 2530 line-km seismic data were acquired in the ESKP and UIG regions, using an 80-channel, 1000 m streamer recording shots from a 20.5-1/2000 psi source array.The shot interval was 25 m.Systematic data reprocessing in 2021 was performed by KIGAM and sequentially included the following: geometry definition, frequency filter, gain control, swell noise attenuation, designature, first demultiple, velocity analysis, pre-stack time migration, second demultiple and signal enhancement.
Seismic data from the SKP were interpreted using seismostratigraphic concepts.For this purpose, Schlumberger's Petrel software was used for seismic data interpretation and mapping.We defined a total of four megasequence boundaries (MB1-MB4), which are correlated with the tectonostratigraphic framework of adjacent areas (i.e. the Ulleung Basin and its southwestern margin; KIGAM, 2006;Kim, Yi, et al., 2020;Kim, Yoo, et al., 2020).Approximate geological ages of the megasequence boundaries were obtained from the tectonostratigaphic correlations with known geological ages in the Ulleung Basin (Kim, Yi, et al., 2020;Kim, Yoo, et al., 2020).In addition, palynological analysis at IODP Site U1430, situated in the ESKP and part of IODP Expedition 346, were used to support our chronostratigraphic interpretation (Horozal et al., 2017;Tada et al., 2015).
We used well-logs (e.g.natural gamma-ray, acoustic and density log), a depth-age relationship and lithostratigraphic data from IODP Site U1430 (Figure 1c; Tada et al., 2015).From the well-to-seismic tie, based on acoustic and density logs, we determined the depth of the seismic horizons and identified approximate geological ages of our stratigraphy, considering the depth-age relationship at IODP Site U1430.Furthermore, based on the reported lithology and sedimentation rates from the preliminary IODP report (Tada et al., 2015), we interpreted the depositional systems and sedimentation patterns in each megasequence.The time-structure map of the top of the acoustic basement in the SKP shows a range from less than 1.0 to over 3.5 s TWT, reflecting a complex structural development (Figure 3a).Based on the seismic-reflection facies and external forms, three acoustic basement types are identified: rifted continental crustal blocks (BH-1-BH-5), volcanic edifices (BH-6 and BH-7) and basement lows (BL-1-BL-6; e.g.Kim et al., 2011;Figure 3a; Table 1).The SKP can be subdivided into three segments: the WSKP, ESKP and Eastern Korean Continental Margin (EKCM), based on the segments' structural style and geomorphological characteristics.The WSKP is the largest area and has a depth range of the acoustic basement from less than 1.5 to over 3.5 s TWT, characterised by N−S or NE-SW trending structures (Figure 3a).Four basement lows (BL-1-BL-4) are present between the intervening rifted continental crustal blocks (BH-1-BH-3) and exhibit a depth of over 3.2 s TWT.Basement-related normal faults appear to dominate along the flank of the basement lows, forming halfgraben structures.Moreover, the volcanic edifices (BH-6 and BH-7) are located in the northwestern and southern parts of the WSKP respectively.The crest of the volcanic edifices is located at depth of less than 1.5 s TWT, and they are characterised by a conical-shaped external form.The ESKP is smaller and shallower than the WSKP, and has a depth range from 1.0 to 3.5 s TWT (Figure 3a).Most of the ESKP is occupied by an elongated, EEN−WWS oriented continental crustal block (BH-4).One basement low (BL-5) is located in the southwestern ESKP, which is bound by an EEN−WWS trending basement-related normal fault to the north.It exhibits a typical half-graben style influenced by block faulting (Figure 3a).The EKCM, the westernmost part, is the shallowest basement, ranging from less than 1.0 s to over 2.0 s TWT (Figure 3a).The segment is characterised by structurally less complex continental crust, while an NNW-SSE trending reverse fault with a length of 40 km is developed.The EKCM is connected with a small-scale basement low (BL-6) reaching 3.2 s TWT in the northwestern corner of the study area (Figure 3a).The basement low (BL-6) is bound by a N−S trending basement-related normal fault to the east.
The total sediment thickness in the SKP ranges from less than 0.1 s TWT in the basement highs to over 1.0 s TWT in the basement lows (Figure 3b).In general, the sedimentary succession has a tendency to infill the basement lows (BL-1-BL-6) of the SKP.In the WSKP, the main depocentres are located in the three basement lows (BL-1-BL-3) which range in thickness from 1.0 to 1.4 s TWT (Figure 3b).In the BL-1 and BL-3, the thickest deposits are present near the basement-related normal faults and tend to thin out gradually moving away from the faults, exhibiting wedge-like geometry.Similar to the WSKP, the ESKP exhibits a main depocentre in the southwestern basement low (BL-5; Figure 3b).Its sediment thickness reaches up to 1.2 s TWT and increases towards the fault plane.Excluding the main depocentre, most of the ESKP is filled with very thin deposits less than 0.1 s TWT thick.The topographic high in the EKCM is also covered with a thin layer, less than 0.2 s TWT thick (Figure 3b).However, towards the northwest, an NNW-SSE-trending elongated depocentre is present and its maximum thickness reaches over 1.6 s TWT in the northwestern corner of the study area (BL-6).In addition to the three segments, the NUT is filled with sediment more than 0.8 s TWT thick (Figure 3b).It shows thickening towards the eastern basement-related fault (i.e.NUE).

| Seismic facies
Seven seismic facies (SF1-SF7) were identified based on seismic features, including reflection amplitude, continuity, frequency and internal/external configurations (Mitchum et al., 1977;Sangree & Widmier, 1977; Table 2).We interpreted the depositional systems and environmental settings of each seismic facies based on previously published and widely cited seismic-based studies, and these are described as follows.
SF1 is characterised by moderate-to-high amplitude, high frequency and good continuity.It is also internally parallel to divergent configurations.This seismic facies is interpreted as lacustrine deposits (Cukur et al., 2015;Zhou et al., 2014).SF2 exhibits variable amplitude and poorto-moderate continuity with deformed and/or disrupted configurations.Based on these internal/external seismic features and its spatial distribution, SF2 is interpreted to represent fan-deltas and slope-aprons deposited in a nonmarine to shallow-water environment (Park et al., 2022;Zhou et al., 2014).SF3 displays low-to-moderate amplitude and poor continuity with a lens-shaped external form.Internally, SF3 exhibits deformed and chaotic configurations, indicating stratal heterogeneity.It is interpreted to represent mass-transport deposits (MTDs), including slide/slump and debris-flow deposits (Cukur et al., 2015;Kim, Yi, et al., 2020;Kim, Yoo, et al., 2020;Lee & Suk, 1998).SF4, Lee et al., 2001 characterised by high amplitude, poor continuity and complex and irregular configurations, is interpreted to represent volcanic sills and lava flows (Cukur et al., 2015;Kim, Yi, et al., 2020;Kim, Yoo, et al., 2020;Lee et al., 2011).SF5 exhibits low amplitude and good continuity and consists of parallel to divergent reflection packages.Based on its seismic characteristics, SF5 is interpreted to reflect hemipelagic sediments deposited under a quiescent, settling environment (Kim, Yi, et al., 2020;Kim, Yoo, et al., 2020;Sangree & Widmier, 1977).SF6 is characterised by moderate-tohigh amplitude and good continuity, with well-stratified configurations, interpreted as hemipelagic sediments interbedded with sandy turbidites (Joh & Yoo, 2009;Park et al., 2022).Finally, SF7 displays an intermediate seismic T A B L E 2 Seismic facies recognised in the study area and their geological interpretations.

| Megasequence boundaries
In the SKP, four megasequence boundaries (MB1-MB4) were identified based on stratal termination patterns (e.g.erosional truncation, onlap and/or downlap) and abrupt change in seismic facies (Figures 4-6).The megasequence boundaries are characterised by good continuity and strong amplitude throughout the SKP.Based on a regional stratigraphic correlation from our stratigraphy to a generalised tectonostratigraphic column with known geological ages in the Ulleung Basin, the four megasequences are correlated with ages as follows: MB1 (latest Late Oligocene); MB2 (late Early Miocene); MB3 (late Middle Miocene) and MB4 (Early Pliocene; Kim, Yi, et al., 2020;Kim, Yoo, et al., 2020).MB2 represents a regional unconformity and separates a lower wedge-shaped sedimentary succession from upper draping deposits (Figures 4-6).The time-depth of MB2 ranges from 0.8 to 3.7 s TWT and the maximum depth occurs within the NUT (Figure 7b).MB2 is generally present in the basement lows (BL-1-BL-6), reaching a depth of over 3.0 s TWT.Along the basement lows, MB2 shows a bowl-shaped geometry and onlaps onto the underlying MB1 or basement-related faults.Overall, MB2 is absent along the basement highs, while it is locally scattered along the ESKP and EKCM (BH-4 and BH-5; Figure 7b).MB2 exhibits a distribution pattern of the basementrelated normal and reverse faults similar to that of MB1.
The seismic section indicates that upward fault propagation from the acoustic basement largely continues to MB2, whereas some basement-related normal faults die out directly below MB2 (Figures 4 and 6).

| MB3 (late Middle Miocene)
Overall, the time-structure map of MB3 is similar to that of MB2; however, MB3 is characterised by an abrupt change in seismic features (Figure 7c).The lower sedimentary succession of MB3 displays very low amplitude and poor-to-moderate continuity, while the upper sedimentary succession of MB3 is dominated by moderate amplitude and a relatively continuous reflection package (Figures 4-6).The time-depth of MB3 ranges from less than 0.8 s to more than 3.5 s TWT (Figure 7c).The deepest part of the NUT reaches a depth of over 3.3 s TWT, and the major depressions (BL-1-BL-6) have similar depths of less than 3.0 s TWT.Along MB3, the number of basement-related normal faults is smaller Selected seismic profile in a N-S direction and its stratigraphic and structural interpretations on the WSKP.The profile displays MS1-MS4 separated by MB1-MB4 (for location, see Figure 1c).Major depressions are present between the intervening rugged ridges, which are devoid of basement-involved faults in the WSKP.BL, basement low; MB, megasequence boundary; MS, megasequence; WSKP, Western South Korea Plateau.
In particular, the basement-related normal faults, including two synthetic faults in BL-1 (ON) and one fault in the western flank of BL-3 (OB), disappear distinctly (Figure 7c).

| MB4 (Early Pliocene)
MB4 is characterised by high amplitude and good continuity and is distributed more broadly compared to MB3 (Figures 4-6).The sedimentary succession below MB4 shows moderate amplitude, while the overlying strata above MB4 are characterised by very high amplitude.
Furthermore, a layer-bound normal fault array (black lines in Figures 4 and 5) dies out immediately below MB4.The time-depth of MB4 ranges from 0.6 to 3.3 s TWT (Figure 7d).The NUT, as the deepest area, reaches depths of 3.3 s TWT, similar to that of MB3.Major depressions of MB4 are located along BL-1-BL-6, showing a similar distribution pattern with those from MB1 to MB3, and reaching a maximum depth of approximately 2.6 s TWT ( Figures 4-6 and 7d).The distribution of the basementrelated faults also exhibits a pattern similar to that of MB3.

| Megasequences
Four seismic megasequences (MS1-MS4) have been identified in the study area based on their approximate age (tied to the generalised tectonostratigraphic column with known geological ages in the Ulleung Basin) and seismic signatures.The megasequences MS1-MS4 are separated by the four megasequence boundaries (MB1-MB4) and the seafloor.Based on the time-thickness maps of MS1-MS4 (Figure 8), they show lateral variations in terms of thickness and distribution, and also in internal seismicreflection patterns.Figure 9 depicts the areal distribution maps of the seismic facies for MS1-MS4, using representative seismic facies, to infer the depositional systems and environmental settings.

| MS1
MS1 corresponds to a lowermost sedimentary succession overlying the acoustic basement and is bounded by MB1 and MB2.The sediment thickness of MS1 ranges from 0.04 to 0.40 s TWT (Figure 8a).In general, MS1 is thick in the basement lows (BL-1-BL-6), while sediment fill is absent or very thin shallower than 0.08 s TWT, on the basement highs (BH-1-BH-7).Most depocentres are located in half-graben-style rift basins (BL-1, BL-3, BL-5 and BL-6) associated with block faulting.Within this area, MS1 is as much as 0.40 s TWT thick and exhibits general thickening towards basement-related fault planes, resulting in wedge-shaped geometries.In addition to these rift basins, a thick sediment fill greater than 0.35 s TWT thick is also present in the basement low south of BL-1 (Figure 8a).(Figure 9a).The EKCM and ESKP, occupying basement highs, are composed of SF2 (Figure 9a).Locally, SF1 is restricted to the rift-basin (BL-5) directly south of the EEN−WWS-trending basement-related fault in the ESKP (Figure 9a).The WSKP and NUT, consisting of several basement lows (e.g.BL-1-BL-4), are dominated by SF1.SF2 is locally present along the eastern slope of BH-1 and the western slope of BH-2 respectively.Furthermore, interpreted volcanic edifices in the WSKP (BH-6 and BH-7) are surrounded by SF4, which show outward-dipping geometry.

| MS2
MS2, bound by MB2 and MB3, generally ranges from 0.01 to 0.40 s TWT in thickness (Figure 8b).Similar to MS1, MS2 mostly occupies the basement lows (BL-1-BL-6), the basement highs are covered with very thin deposits of MS2 (BH-1-BH-7).Although the deepest depression is located in the NUT along the lower boundary (MB2), two major depocentres are present on the eastern slope of the EKCM and at the northwestern corner of the study area (BL-6; Figure 8b).The thickness of the depocentre in the eastern slope of the EKCM reaches 0.40 s TWT and decreases eastward (Figure 8b).The depocentre in the northwestern margin of the study area shows gradual thickening northward, with a maximum thickness of 0.38 s TWT.In the case of the half-grabenstyle rift basins (BL-1, BL-3 and BL-5), MS2 shows an approximately uniform thickness of less than 0.25 s TWT, in contrast to the thickness of MS1, characterised by gradual thickening towards fault planes with wedgeshaped geometries (Figure 8a,b).
MS2 exhibits low amplitude and poor-to-moderate continuity with an internally chaotic reflection package.MS2 generally exhibits SF5, SF6 and SF7, excluding the eastern ESKP where the seismic-facies patterns and distributions are difficult to interpret due to the very thin thickness or absence of the sediment fill (Figure 9b).Overall, SF7 predominates throughout MS2, while SF5 is present in BL-1 and on both sides of BH-2 (Figure 9b).Moreover, the EKCM consists locally of two elongated sediment fills showing SF6.

| MS3
MS3 is the thickest sedimentary succession and is bound by MB3 and MB4.It ranges in thickness from less than 0.01 s to over 0.40 s TWT (Figure 8c).In general, MS3 thins eastward from the ESKP to WSKP, except for topographic highs (BH-1, BH-2, BH-3, BH-5, BH-6 and BH-7).The thickest sediment fills are located in the northwestern EKCM and along the eastern slope of the EKCM, with a maximum thickness of over 0.40 s TWT (Figure 8c).Moreover, several depressions (BL-1, BL-2 and BL-4) in the WSKP are filled with sediment greater than 0.30 s TWT thick (Figure 8c).In the NUT, MS3 is relatively thick, more than 0.35 s TWT, which is in contrast to the thickness of MS1 and MS2 in the NUT, and gradually thins westward as the distance from the NUE increases.Overall, the deposit is thin in the ESKP less than 0.16 s TWT in thickness, and the thick sediment is restricted to the southern depression (BL-5) in the ESKP (Figure 8c).
MS3 is characterised by a reflection package showing low-to-moderate amplitude and good continuity.It comprises SF3, SF5, SF6 and SF7 (Figure 9c).MS3 is generally dominated by SF5 in the WSKP and ESKP, whereas SF6 is mostly distributed in the EKCM (Figure 9c).Typically, the NUT is entirely occupied by SF7, but SF3 is only local with the N-S elongated shape.

| MS4
MS4 is an uppermost sedimentary succession bound by MB4 and the seafloor.The sediment thickness of MS4 ranges from 0.04 to 0.40 s TWT (Figure 8d).Three major depocentres are present in BL-1, BL-6 and the NUT, and their thickness is over 0.35 s TWT.Commonly, these depocentres are characterised by ponded sediment fill.Above an arbitrary seismic horizon in the MS4 (white dotted line, H1 in Figure 10), the ponded sediment fill exhibits a flat-lying and onlapping reflection pattern.In addition, thick sediment fills, over 0.24 s TWT in thickness, are also present in several depressions (BL-2 and BL-3) in the WSKP (Figure 8d).The EKCM and ESKP are filled with very thin sediments of less than 0.08 s TWT thick (Figure 8d).In the northwestern margin of the study area, N−S-trending channel systems can be identified (Figure 10).These channel systems are connected with the Gangneung and Donghae canyons from the eastern continental margin of the Korea Peninsula (Cukur et al., 2021).The western channel system connected to the Gangneung canyon is characterised by stacked channels and channel-levee complexes in the ponded accommodation space (i.e.small isolated basin or depression created by local tectonic activities; Figure 10b−e).The stacked channels are approximately 0.4-0.8km in width, with a time thickness of 0.05 s TWT (Figure 10b,c).The channel fills are composed of a very high-amplitude reflection package.Some channels build up with gull-wing style levees showing abrupt lateral pinch-out geometries (Figure 10d,e).The eastern channel system connected to the Donghae canyon comprises several stacked channels with approximately 0.8 km width (Figure 10f,g).
MS4 is subdivided into upper and lower parts exhibiting different seismic features (Figures 4 and 5).The lower part directly above MB4 is dominated by moderate-tohigh amplitude reflections with good continuity, while the upper part consists of a low-amplitude, continuous reflection package.MS4 is generally composed of a combination of SF3, SF5, SF6 and SF7 (Figure 9d).Overall, SF6 predominates in MS4, excluding in the ESKP, which is occupied by SF5.SF5 is also present at the northern slope of BH-3, western slope of BH-7 and at the southwestern corner of the study area.Moreover, SF3 is widely distributed with NNW-SSE-or N−S-trending geometries in the WSKP.The sediment fills showing SF3 lie directly below the seafloor, and their strike is perpendicular to the headwall scarps developed along the EKCM (e.g.Cukur et al., 2021; Figure 9d).In the southwestern margin of the study area, SF7 is present with N−S-trending geometry.

| Geological structures
The SKP accommodates a variety of geological structures, including normal and reverse faults, folds and syntectonic growth strata formed during highly active tectonic movement (Figures 11-13; Table 3).Based on the geometry and distribution patterns of the geological structures, we investigated their mechanisms of formation, timing and deformation patterns to infer tectonic controls on the sedimentation.

| Fault geometry and classification
The present study recognises various normal and reverse faults from seismic interpretation (Figure 11; Table 3).The faults were classified into four types (Type-A to Type-D) based on geometry, distribution and how the faults propagate and grow within the sedimentary succession.
Type-A faults are basement-related normal faults in the central WSKP (BL-1 and BL-3), southern ESKP (BL-5) and northwestern EKCM (BL-6; Figure 11a; Table 3).This fault type is characterised by large-offset normal faults crosscutting the acoustic basement.They are tall (ca.1.5 s TWT), long (ca.60 km), linear, and have a maximum heave of approximately 6 km.The EKCM and WSKP are dominated by N−S-or NE-SW-striking Type-A faults, while EEN-WWSstriking Type-A faults prevail in the ESKP.In the SKP, Type-A faults are generally located along the flank of the basement highs (i.e.rifted continental crustal blocks), forming half-graben-style rift basins (BL-1, BL-3, BL-5 and BL-6; Figure 11a).Particularly, in the case of BL-1 (i.e.Onnuri Basin), the large-offset basement-related normal fault accommodates two synthetic faults, resulting in dominostyle block faulting (Figure 11a).In general, Type-A faults propagate from MB1 to MB4, while some faults terminate below MB2.Moreover, MS1 has a tendency to increase in thickness towards Type-A fault planes, forming a wedgeshaped sedimentary succession (Figure 11a).
Type-B faults are layer-bound normal faults located in the central WSKP (BL-1, BL-2 and BL-3) and southwestern ESKP (BL-5; Figure 11a; Table 3).This type of fault is confined within a discrete stratigraphic unit (i.e.MS3), forming a layer-bound normal-fault array (Figure 12).In general, the lower tips of the Type-B faults terminate at MB3, whereas their upper tips die out at MB4.In the seismic profile, the layer-bound normal faults do not reach the upper sediment fill characterised by high amplitude and good continuity with well-stratified configurations (i.e.SF6; Figure 12).In contrast, the Type-A basementrelated faults offset the upper sediment fill characterised by SF6 as well as MB4.Although it is difficult to interpret a strike and distribution pattern in detail due to the absence of time-slice-based seismic-attribute maps, it is likely that Type-B faults do not have a consistent direction.
Type-C faults are dome-related normal faults in the northwestern WSKP (BL-1 and BH-6; Figure 11b,c; Table 3).This type of fault is present above the basement highs (e.g.submerged continental ridges and volcanic edifices).Type-C faults are characterised by high-angle normal faults and show a concentric distribution pattern above the crest of the basement highs (Figure 11b,c).Moreover, overburden strata above the basement highs, cut by several Type-C faults, display antiformal structures and growth-stratal architecture.
Type-D faults are large-scale reverse faults developed in the northwestern EKCM (Figure 11d; Table 3).This type of fault is tall (ca.0.9 s TWT), long (ca.50 km), has a NW−SE strike, and a heave of approximately 2 km.Type-D faults are developed from MB1 to MB4 and offset the seafloor (Figure 11d).To the west of Type-D faults, the sedimentary succession is highly deformed and folded in the form of faultpropagation folds.Typically, the reverse-fault surface shows no characteristics of bending by folding.Moreover, MS4 has different thicknesses across Type-D faults (Figure 11d).

| Fold geometry
The fold geometry and stratal patterns around the fold were analysed to infer the formation timing and its influence on sedimentation in the SKP (Figure 13).As discussed in Section 4.5.1, the fault-propagation fold is located in the northwestern corner of the study area, where all the megasequences exhibit folding (Figure 11d).The fault-propagation fold is mainly oriented in NNW-SSE or N−S directions and extends approximately 30 km in length (Figure 13a).Although it is difficult to interpret the entire fold axis due to limited seismic data, it is likely that the structural elevation decreases gradually from over 0.6 s to less than 0.3 s TWT towards the North.Around the fold, a growth-strata-thinning feature is observed in the upper MS4 (Figure 13b,c).Sediment thickness between arbitrary lines H1 and H2 thins across the fold limb towards the fold axis, indicating a typical growth-stratal pattern.In contrast, sediment fills below H1 show a parallel stratal pattern and have similar thickness along the fold limb.Moreover, H1 acts as an onlapping surface in the ponded accommodation space caused by folding (Figure 13b

| Structural and sedimentary evolution of the SKP
Based on our stratigraphic and structural analysis, we suggest that the structural and sedimentary evolution of the SKP has been mainly influenced by tectonic activity.The basement lows, developed by extensional faulting and associated tectonic subsidence during the initial basin-rift phase, played an important role in locating the main depocentres for all basin-evolution phases.In addition, the F I G U R E 1 2 Seismic profile and its stratigraphic and structural interpretations, showing a layer-bound normal-fault array (i.e.Type-C fault) in MS4 (for location, see Figure 1c).The faults are characterised by regional extent, layer-bound geometry and concentration within one discrete stratigraphic unit, interpreted as a PFS (Cartwright, 2011;Cartwright et al., 2003;Ghalayini & Eid, 2020).The PFS terminates just above MS3 and generally does not penetrate the sandy turbiditic layer characterised by very high amplitude and well-stratified seismic reflections in MS4.MB, megasequence boundary; MS, megasequence.basin fill within rift-basins is strongly influenced by the displacement geometry of a variety of geological structures (Gibson et al., 1989;Gupta et al., 1998).In the SKP, several fault-and-fold structures developed under the influence of a combination of tectonic movement, volcanism and sediment loading (Chough et al., 2018;Cukur et al., 2015;Kim, Lee, Choi, et al., 2015;Kim, Lee, Park, et al., 2015).Based on our study, the geometric characteristics and formational timing of the geological structures provide a new perspective from which to understand basin evolution and sedimentary history of the SKP.Considering these controlling factors (i.e.tectonic activities and associated geological structures) as mentioned above, we herein present a model of tectonic and sedimentary evolution of the SKP with four stages: Stage-1 (syn-rift), Stage-2 (postrift phase 1), Stage-3 (post-rift phase 2) and Stage-4 (syncompression; Figure 14).
5.1.1 | Stage-1 (syn-rift; latest Late Oligocene to Early Miocene) The SKP formed by extension was associated with back-arc rifting in the NW Pacific subduction zone and separation of the southwestern Japan since Late Oligocene time (Kim, Lee, Choi, et al., 2015;Kim, Lee, Park, et al., 2015, Kim et al., 2022;Tamaki et al., 1992).At the time, the plateau was extended through dominostyle block faulting, resulting in rapid tectonic subsidence along the fault planes (Figure 14a).The block faults are interpreted as syn-extensional faults associated with continental rifting, forming a wedge-shaped syn-rift sedimentary succession characterised by gradual thickening towards the fault plane (e.g.Cukur et al., 2015;Kwon et al., 2009;Zhang et al., 2020; Table 3).In previous studies, most of the syn-extensional faults (i.e.basementrelated faults) were interpreted using low-quality and vertically exaggerated seismic profiles, resulting in misleading interpretations of fault distribution in the SKP (Chough et al., 2018).For example, the slopes of the continental ridges were misinterpreted as basement-related faults, suggesting the formation of graben-style rift basins (Cukur et al., 2015;Kwon et al., 2009).Based on our structural interpretation, however, syn-extensional faults were actually developed along one side or flank of the rugged continental ridges, resulting in asymmetric half-graben-style rift-basin development (Chough et al., 2018; Figure 14a).Syn-extension faults associated with continental rifting; rapid subsidence by block faulting caused wedge-shaped syn-rift sedimentary succession (Cukur et al., 2015;Kwon et al., 2009 Post-sedimentary normal faults caused by differential compaction (Dasgupta, 2018;Kim, Yi, et al., 2020;Kim, Yoo, et al., 2020;Misra, 2018 -4), the SKP was influenced by E-W compression due to eastward movement of the Eurasian plate.As a result, the contractional tectonism resulted in reverse faulting, folding and associated ponded depressions.During this time, the SKP was dominated by turbiditic and hemipelagic sedimentation, with intermittent masstransport deposits.From the EKCM, large volumes of sediment were supplied through slope channels connected to submarine canyons.It is likely that turbidity currents were triggered from direct riverine input or from the shelf margin in the EKCM.The tectonic evolution of the East Sea (Sea of Japan; insets in the bottom left of each panel) is modified after Yoon and Chough (1995).Blue and red arrows indicate the direction of seafloor spreading and the relative tectonic motion of adjacent plates respectively.JB, Japan Basin, UB, Ulleung Basin, YB, Yamato Basin.During Stage-1, the lowermost parts of the rift-basin were filled with volcanic sills and lava flows above the acoustic basement (Figure 14a).In the WSKP and NUT, several volcanic mounds are present, and their flanks are covered with volcanic materials (e.g.volcanoclatics).These observations indicate that intensive volcanism was active during early-stage of continental rifting, and that volcanic material was supplied via fault fractures and eruption centres of the volcanic edifices (e.g.Kim et al., 2011).In the incipient rifting stage, it is likely that the entire plateau had yet to be submerged below sea level and that sedimentation was predominant in basement lows (i.e.half-grabenstyle rift-basins) caused by extensional block faulting (e.g.Kim et al., 2013; Figure 14a).These rift-basins were filled with wedge-shaped syn-rift sedimentary layers, thickening towards the fault planes in the form of syntectonic growth strata.The syn-rift sediments show continuous, high-frequency, parallel reflections, indicating the predominance of an anoxic, lacustrine environment similar to interpretations made in the Tamtsag Basin, Mongolia (Zhou et al., 2014; Figure 14a).Meanwhile, the crestal parts of the continental ridges were sub-aerially exposed as evidenced by water incursion between the intervening rifted continental ridges.Therefore, terrestrial sediment was supplied by sub-aerial erosion and/or denudation of the surrounding highlands, resulting in the deposition of stacked fan deltas and sublacustrine fans (e.g.Zhou et al., 2014).Kim et al. (2013) recognised wave-planation surfaces originating from erosion of substratal landmasses and erosion-induced sediments on the adjoining rifted trough-and-terrace margins in the SKP.Similarly, the rift basins in our study were also possibly influenced by insitu deposition of the autochthonous detritus from the adjacent rifted continental fragments.Furthermore, the steep apron deposits and/or slope aprons, characterised by moderate-to-high amplitude and lateral-accretion patterns, occurred along the eastern slope of the EKCM.This observation implies that sediments were supplied from the EKCM, forming the erosion remnants of the continental ridge (e.g.Chough et al., 2018).
5.1.2| Stage-2 (post-rift phase 1; late Early Miocene to Middle Miocene) For Stage-2, basin extension due to the block faulting ceased on the SKP, and slow thermal subsidence predominated, resulting in termination of vertical-tip propagation of syn-extension faults within MS2 and uniform thickness of MS2 (Figure 14b).Previous studies reported that basin rifting continued until Middle Miocene time and that the post-rift phase in the SKP began afterwards (Cukur et al., 2018;Horozal et al., 2017).However, based on the termination of basement-related faults and the geometry of sediment fill in the MS2 presented in this study, it is suggested that tectonic activity was quiescent during the post-rift period from late Early Miocene time.While it has been previously reported that the basin could have subsided rapidly owing to the high sedimentation rate during the post-rift phase (Dupré, Bertotti, & Cloetingh, 2007;Xie et al., 2006), MS2 is relatively thin across the entire study area, and a low sedimentation rate below 3 m/myr was identified at Site U1430 in the ESKP based on IODP Expedition 346 (Horozal et al., 2017;Tada et al., 2015; Figure 15).This interpretation suggests that rapid tectonic subsidence and/or extension ended by the late Early Miocene time and that sedimentation across the SKP occurred under a tectonically stable environment with slow thermal subsidence during Stage-2.
During Stage-2, an open-marine environment predominated, which is reflected in changes in sedimentation (Kim et al., 2013; Figure 14b).It is likely that the depositional system changed from lacustrine and fan-delta under the non-marine/shallow-water setting (Stage-1) to turbiditic, hemipelagic and mass-transport processes under the open-marine setting (Stage-2).In general, based on seismic facies analysis, SF7 predominats in MS2, and SF5 locally occurs in the northern WSKP (e.g.BL-1; Figure 9b).This observation, along with the inferred change of sedimentary environment, suggests MTDs were supplied from adjacent highlands due to slope failures during Stage-2 (Figure 14b).Because the continental ridges in the SKP are characterised by the over-steepening of slopes, it is reasonable that large-scale submarine mass-wasting could have occurred in a similar fashion to the slope failures defined by Gong et al. (2011).Based on the results of the present study, in the case of BL-1 (i.e.ON) in the WSKP, MS2 comprises hemipelagic sediments, which drape the basement highs, suggesting the predominance of a pelagic settling under a tectonically quiescent environment without basement-related fault activation (Figure 14b).Furthermore, during this period, thick sediments characterised by SF7 were deposited on the northern and eastern slope of the EKCM, which indicates high sediment supply sourced from the EKCM.In contrast, the ESKP is draped with very thin pelagic sediments, which cause the absence of seismic reflections or the presence of only a few seismic reflections.
5.1.3| Stage-3 (post-rift phase 2; late Middle Miocene to Late Miocene) Since late Middle Miocene time, the East Sea (Sea of Japan) entered into a compressional regime in response to plate-tectonic movements along the Japan (Chough et al., 2000;Yoon & Chough, 1995).At that time, the sea was influenced by NW−SE-trending compression parallel to the northwest drift direction of the Philippine Sea plate (Figure 14c).The southern shelf area of the Ulleung Basin was rapidly uplifted and deformed, creating compressional structures including thrust belts (e.g.Dolgorae Thrust Belt (DTB) in Figure 1b) and anticline (e.g.Dolgorae-I Anticline within the DTB; Lee et al., 2011;Yoon et al., 2014).The Hupo Basin located in the southwest of the SKP was also affected by NW−SE compressional stress, resulting in reactivation of inherited extensional faults and flexural bending associated with inversion uplift (Park et al., 2022).In contrast to these tectonically active settings in adjoining areas, the SKP does not appear to be affected by NW−SE trending compression, as evidenced by the absence of compression-related structures.
In that regard, it is interpreted that tectonic stress, caused by the northwest drift of the Philippine Sea plate, did not extend to the SKP because several intervening ocean basins south of the SKP (i.e.Hupo and Ulleung basins) suppressed the transfer of tectonic stress.As a result, the SKP continually experienced a tectonically stable environment during Stage-3, and slow thermal subsidence prevailed throughout the plateau (Figure 14c).
During Stage-3, hemipelagic sedimentation largely prevailed across the SKP, except for in the EKCM and NUT (Figure 14c).In the ESKP and WKSP, hemipelagic sediments were deposited under a quiescent environment with intermittent turbidite deposits.Compared to Stage-2, MTD input decreased sharply, which indicates a relative sea-level rise (e.g.Cukur et al., 2018) during Stage-3.It is generally accepted that rising sea-level increases slope stability, weakening mass-transport processes (e.g.Hiscott & Aksu, 1996).However, because the continental ridges in the SKP were still characterised by steep slope gradients, it is likely that small-scale mass-wasting occurred which generated turbidity current during periods of punctuated, high-frequency sea-level lowering (e.g.Cukur et al., 2018).Seismic profiles show that the portion of moderate-to-high amplitude reflections increases in MS3, in contrast to the lower sedimentary succession (i.e.MS2).These seismic characteristics imply that the volume of sandy turbidites increased at the time, thus increasing seismic amplitude due to the impedance contrast between the sandy turbidites and interbedded muddy sediments.A similar seismic pattern has been found in the Tamtsag Basin, Mongolia and has been attributed to the impedance contrast between sandy and muddy sediments (Zhou et al., 2014).Moreover, from the EKCM to western WSKP, MS3 is characterised by a gradual thinning eastward, suggesting sediment input from the EKCM.Although feeder systems such as submarine canyons and slope channels cannot be identified due to limited seismic data, it is highly probable that enormous volumes of sedimentary layers were supplied by either direct riverine input or shelf-margintriggered turbidity currents from the EKCM (e.g.Cukur F I G U R E 1 5 Seismic profile crossing of IODP Site U1430 on the ESKP.From the well-to-seismic tie based on acoustic and density logs, the megasequence boundaries in our study are correlated with the depth-age relationship from IODP Site U1430.Geological information, including geological age, lithologic units, sedimentation rate and lithologies, is summarised for each megasequence based on Tada et al. (2015).MB, megasequence boundary; MS, megasequence. et al., 2018).In the NUT, relatively thick sediments were deposited through mixed mass-transport processes and turbiditic/hemipelagic sedimentation (Figure 14c).Moreover, syn-tectonic growth strata, thickening towards the basement-related fault (i.e.NUE), are shown, which represent rapid basin subsidence caused by block faulting.
MS3, deposited during Stage-3, is deformed by a number of layer-bound normal faults, interpreted as a polygonal fault system (PFS), based on their spatial distribution and geometry similar to interpretations made in the Nankai Trough and Levant Basin offshore Lebanon (Cartwright, 2011;Ghalayini & Eid, 2020; Figure 14c).These faults are characterised by regional extent and are confined to a specific stratigraphic interval (i.e.MS3).The concentration of the faults within a discrete stratigraphic unit, referred to as 'tiers', is typical in PFS (Cartwright et al., 2003).Furthermore, PFS develops in host rock consisting mainly of hemipelagites dominated by claystones or biogenic mudstones (Cartwright, 2011).In the current study, the SKP was also dominated by hemipelagic sedimentation during Stage-3 and the presence of 50%-80% biogenic sediments was confirmed in MS3 from the IODP Site U1430 (Horozal et al., 2017;Tada et al., 2015; Figure 15).Therefore as explained above, it is likely that the faults nucleated and grew as syn-depositional faults in the mudstone-dominated succession during Stage-3 in response to a range of diagenetic processes, including compaction and water expulsion, as suggested by Cartwright (2011) and King and Cartwright (2020; Figure 14c). 5.1.4| Stage-4 (syn-compression; Early   Pliocene to present) Since Pliocene time, the SKP was influenced by E−W compression due to eastward-movement of the Eurasian plate (Figure 14d).In the adjacent areas (i.e.Hupo and Ulleung basins), a variety of contractional structures (e.g.Gorae-V anticline and inversion structures in Figure 1b) were formed under the E−W contractile regime (Lee et al., 2011;Park et al., 2022).Similarly, we identified fold structures and reverse faults (i.e.Type-D fault) on the northwestern margin of the study area.Because compressional stress influenced the sedimentary evolution of the SKP during Stage-4, it is critical to reveal the timing and style of contractional deformation.However, there is still controversy about the timing of compressional stress in the SKP (Cukur et al., 2015(Cukur et al., , 2018;;Horozal et al., 2017;Kim et al., 2007;Kwon et al., 2009).Based on the results of the present study, growth strata between the arbitrary horizons H1 and H2 typically thin across the fold limbs towards the structural highs (Figure 13b,c).This observation indicates that the compressional stress continued until the time of H2 (e.g.Salomon-Mora et al., 2009;Yarbuh & Contreras, 2017).This interpretation is also supported by the stratal onlapping pattern in the upper layers onto the H2 surface.A similar interpretation related to timing of structural development has been made in the western Gulf of Mexico basin (Yarbuh & Contreras, 2017).Based on the characteristics of these growth-strata and stratal-termination patterns around the fold, we interpret that the SKP experienced E−W compression from the early part of Stage-4, and that this tectonic force continued until the end of Stage-4 (Figure 14d).In addition, growth-strata-thinning across the fold crest indicates that sediment accumulation exceeded the uplift due to compressional folding during that period (e.g.Anastasio et al., 2021).Furthermore, the fold is cut by a large-scale reverse fault, forming a fault-propagation fold.The reverse fault shows a sharp fault plane without bending characteristics, and it locally intersects near the seabed.This implies that the reverse fault postdated the folding and developed more recently similar to interpretations by Boyer (1978).In addition to these contractional structures, dome-related normal faults were developed, and overburden strata were uplifted above the dome-structures on the northwestern WSKP (e.g.submerged continental ridge and volcanic edifice; Figure 11b,c).These faults are interpreted to be post-sedimentary normal faults caused by differential compaction, which commonly occurs above buried mounds (e.g.Dasgupta, 2018;Misra, 2018;Sun et al., 2020;Zhang et al., 2020;Zhao et al., 2014).As sedimentation was concentrated along the flank of the dome-structures as shown by growth-stratal patterns, the overburden above the crest of the dome-structures was lengthened in the form of an antiformal structure and cut by normal faults during Stage-4 in a similar fashion to the differential compaction-derived deformation defined by Dasgupta (2018) and Fossen (2016).
During Stage-4, the SKP was mainly dominated by turbiditic and hemipelagic sedimentation and MTDs (Figure 14d).In seismic profiles, MS4 consists of very high-amplitude and well-stratified reflections, indicating the presence of a sand-dominant succession.It is very likely that sandy turbidites were supplied into the SKP by submarine-slope failures (e.g.Cukur et al., 2018).In the Ulleung Basin, turbidite sediments have been deposited in the distal basin-floor by mass-wasting generated along the southern slope (Yoo et al., 2017).Based on core and welllog data from drilling site UBGH2-6 in the Ulleung Basin, these turbidite sediments are interbedded with hemipelagic layers attaining a thickness of 45 m, and the maximum thickness of an individual sand bed is over 50 cm (Bahk et al., 2013).The PFS terminates below MB4, which generally does not penetrate the sandy turbiditic layer, as evidenced by very high-amplitude and well-stratified seismic reflections in MS4.Previous studies reported that turbiditic sandstones act as a mechanical barrier to the propagation of the PFS (Ghalayini & Eid, 2020).Therefore, the PFS-terminating feature just below MB4 also supports the predominance of sandy turbidites during Stage-4.Furthermore, the apparent predominance of large-scale MTDs (i.e.debris-flow deposits) in the WSKP suggests that mass-wasting was relatively active during Stage-4 (Figure 14d).Along the upslope of the EKCM, several headwall scarps were identified, caused by submarine landslides (e.g.Cukur et al., 2021).The mass-wasting is interpreted to have been actively generated due to slope failures of the upper headwall, resulting in the deposition of MTDs (Figure 14d).In the case of the NUT and ESKP, it is also likely that turbidite was developed through masstransport processes without a specific feeder system.
The ponded depressions in BL-1, BL-6 and the NUT play an important role in the history of sedimentation, as main depocentres during Stage-4 (Figure 8d).BL-1 and BL-6 consist of high-amplitude, parallel reflections, characterised by the ponded-sediment fill showing an onlap pattern (Figure 10b,d).The ponded depressions were probably filled with large volumes of sediment provided from the EKCM after they were developed by compressional stress in early Stage-4.In BL-6, we also identified several stacked channels connected through submarine canyons from the Korean Peninsula (Figure 10).We interpret that these stacked channels acted as a feeder system with sediments supplied through turbidity currents, infilling the ponded depressions (e.g.Posamentier & Kolla, 2003).Furthermore, levee-overbank development on both sides of the channels indicates high sediment supply from the EKCM during Stage-4.A similar leveeoverbank developed by intensive sediment supply was reported in the Bengal Basin, offshore Myanmar (e.g.Yang & Kim, 2014).

| Implications for oil and gas reservoir potential
The SKP includes several frontier basins (e.g.OB, ON and BB) with petroleum-exploration potential.As previous work on sediment-dispersal patterns and sand distribution in the SKP is extremely limited, there is great uncertainty regarding reservoir potential.In the present study, we reveal the distribution trends of inferred sandy reservoirs through a detailed interpretation of the structural evolution, filling processes and depositional systems in a second-order megasequence framework.Based on our examination, we predict the favourable sandstone reservoirs in each megasequence as listed below: 1. MS1: Fan-delta deposit, lacustrine-fan turbidites.2. MS2: None. 3. MS3: Deepwater-fan turbidites.4. MS4: Deepwater-fan turbidites, channel-levee complexes, turbidite frontal-splay deposits.
Reservoir sandstones in MS1 (syn-rift) are associated with thick successions of fan-delta and lacustrine-fan turbidites (Figure 16).First, fan-delta sandstones have the potential to act as oil and gas reservoirs within the early syn-rift succession which was similarly found in the north Falkland Basin (Richards & Hillier, 2000).Because these sandstones are point-sourced from the nearby footwall of the basin-bounding faults, it is likely that they consist of coarse-grained conglomerates, gravelly sandstones and fine-to coarse-grained sandstones (e.g.Wang et al., 2021).Second, gravity-flow-dominated turbiditic sands were deposited over the hangingwall of the fault in a deep lacustrine-turbidite-fan setting (e.g.Gawthorpe et al., 2018;Schwarz & Wood, 2016; Figure 16).As such, these fan-sandstones could be viable reservoir targets associated with high sediment-supply rates, and their reservoir quality should be good because light muddy sediments escape from the feeder channel during long sedimenttransport, interpreted as Wang et al. (2021).In seismic profile, MS1 is composed of high amplitude and sheetlike seismic reflections, interpreted as lacustrine turbidite sandstones similar to interpretations made in the Falkland Basin (Jones et al., 2019).However, in the case of syn-rift successions, the distribution of sandstone reservoirs is typically influenced by syn-depositional faulting, which causes depositional complexity and hampers exploration of prospective targets (Leila et al., 2022).Therefore, the sandstones associated with fan-deltas and lacustrine-fan turbidites in MS1 can be promising petroleum reservoirs if geological risk factors, such as reservoir compartmentalisation and gas leakage, caused by syn-depositional faulting, are addressed or avoided (Figure 16).
Reservoir sandstones in MS3 (post-rift phase 2) are representative of deepwater-fan turbidites (Figure 16).MS3 consists of hemipelagites partly interbedded with turbiditic sandy layers showing moderate-to-high amplitude and sheet-like geometry.Although small portions of deepwater fan turbidites are included, it is expected that these fan sandstones could serve as good sandstone reservoirs.Furthermore, the mud-dominant setting of MS3 is more suitable for defining a petroleum-system, because the fan sandstones in MS3 are surrounded by thick successions of marine mudstones acting both as intraformational seals and the source rocks (e.g.Dam & Sønderholm, 2021).
Reservoir sandstones in MS4 (syn-compression) are associated with deepwater turbidite sandstones, channel-levee complexes and turbidite frontal-splay deposits (Figure 16).First, the lower MS4 is composed of mass-wasting-derived sandy turbidites.Because these turbidite sandstones are typically sourced from continental ridges with long transport distances, they are characterised by relatively clean, better sorted and more mature sediments, which have relatively greater reservoir potential (Schwarz & Wood, 2016;Wang et al., 2021).Second, in the northwestern corner of the study area, we identified several channel systems connected by submarine canyons (i.e.Gangneung and Donghae canyons; Cukur et al., 2021; Figure 10a).The stacked channels consist of high-amplitude reflections, indicating sandprone channel fills in a similar to sand-prone channel fills with strong amplitude in the offshore eastern Borneo, Kalimantan, Indonesia (Pirmez & Flood, 1997;Posamentier & Kolla, 2003).Moreover, some channels are associated with levee deposits in the form of levee-channel complexes.The levee construction implies that the rate of sediment supply was high and turbidity flow was sufficiently reinforced to build the levees.Previous studies reported that the levee deposits contain reservoir-quality thin-bedded sandstone (e.g.Clemenceau et al., 2000;Piper & Savoye, 1993).In the Bay of Bengal, offshore Myanmar, F I G U R E 1 6 Schematic model showing the proposed reservoir potential in the SKP.Good reservoir sandstones are identified based on structural and sedimentary evolution, sediment-filling processes and depositional systems in the megasequence framework of the SKP.In MS1, reservoir sandstones are associated with a thick succession of fan-delta and lacustrine-fan turbidites.In MS3, reservoir sandstones may be present in deepwater-fan turbidites caused by slope failures.Lastly, in MS4, deepwater-fan turbidites, channel-levee complexes and turbiditic frontal-splay deposits could be potential reservoirs.With additional seismic acquisition and petroleum-system analysis, these potential reservoirs could provide significant future petroleum-exploration targets on the SKP. the Shwe field reservoir consists of levee overbank deposits, which are thin-bedded sandstones with a net-to-gross ratio of 33% (Yang & Kim, 2014).Moreover, it has been observed that these leveed channels directly feed lobes or lobe complexes, referred to as frontal-splay complexes (Posamentier & Kolla, 2003).When the slope gradient decreases and turbidity flows decelerate, frontalsplay complexes form at the terminus of the channels (Weimer et al., 2007).In our study area, we interpret that the canyons and slope channels from the Korean Peninsula sourced large, sandy, frontal-splay complexes from the North Korea Plateau (Figure 16).Therefore, acquiring seismic data in the western part of the North Korea Plateau is necessary for identifying the presence of the frontal-splay complexes as potential reservoirs.

| CONCLUSIONS
The tectonic and sedimentary evolution of the SKP was investigated to establish an integrated tectonostratigraphic framework of the SKP using newly reprocessed 2D seismic-reflection data.The sedimentary succession of the SKP consists of four megasequences: MS1 (syn-rift), MS2 (post-rift phase 1), MS3 (post-rift phase 2) and MS4 (syn-compression).Based on stratigraphic and structural analyses of the four megasequences, tectonic activities were the dominant control on the depositional history of the SKP.The SKP evolved structurally and sedimentologically according to the following four stages: (1) Stage-1 (syn-rift; latest Late Oligocene to Early Miocene) was mainly controlled by extensional block faulting.Fan-delta and shallow-lacustrine depositional systems prevailed in half-graben-style rift basins; (2) Stage-2 (post-rift phase 1; late Early Miocene to Middle Miocene) was dominated by hemipelagic sedimentation with gravity-controlled slope failures under a tectonically stable environment with thermal subsidence; (3) Stage-3 (post-rift phase 2; late Middle Miocene to Late Miocene) was still affected by slow thermal subsidence, while hemipelagic biogenic sediments were deposited, resulting in PFS development.At the same time, mass-wasting was intermittently generated, which provided sandy turbidites into the plateau; (4) Stage-4 (syn-compression; Early Pliocene to present) entered into an E−W compressional regime under the influence of the eastward-drifting Eurasian plate.During that period, turbiditic and hemipelagic sedimentation was predominantly accompanied by episodic slope failures.Meanwhile, turbidity currents, sourced from direct riverine input or slope failures in the EKCM, debouched the lower-slope and basin-floor environments, thereby producing a predominance of leveed channel deposits.Tectonostratigraphic investigation of the SKP has identified promising sandstone reservoirs for petroleum exploration: (1) reservoir sandstones associated with fandelta and lacustrine-fan turbidites in MS1; (2) reservoir sandstones associated with deepwater-fan turbidites in MS3; (3) reservoir sandstones associated with deepwaterfan turbidites, channel-levee complexes, and turbidite frontal-splay deposits in MS4.The tectonostratigraphic framework proposed in this study provides sedimentarybasin evolutionary frameworks based on the influence of tectonics on both structure and depositional environment, and an avenue for understanding reservoir potential with regard to petroleum exploration.

ACKNO WLE DGE MENTS
South Korea Plateau, structural and sedimentary evolution, structural development, tectonic activity

F
I G U R E 1 (a) Plate-tectonic boundaries of NE Asia modified from Taira (2001).Inserted box indicates the location of the map shown in (b).(b) Physiographic map of the East Sea (Sea of Japan; contour interval is 1 km).Inserted box indicates the study area, shown in (c).(c) Bathymetry of the SKP and locations of the two-dimensional, multi-channel, seismic data.Sky blue colour represents the study area defined as a closed polygon, drawn with a contour line of 3.3 s TWT in the time-structure map of the acoustic basement.Heavy lines with figure numbers indicate the locations of seismic profiles used in the present study.Bathymetric contours are in metres.AS, Anyongbok Seamount; BB, Bandal Basin; DI, Dok Island; DTB, Dolgorae Thrust Belt; EKCM, Eastern Korean Continental Margin; ESKP, Eastern South Korea Plateau; GS, Gorae-V structure; KS, Kiminu Seamount; NKP, North Korea Plateau; NUE, North Ulleung Escarpment; NUT, North Ulleung Trough; NWUR, Northwest Ulleung Ridge; OB, Okgye Basin; ON, Onnuri Basin; SKP, South Korea Plateau; UB, Ulleung Basin; UI, Ulleung Island; UIG, Ulleung Interplain Gap; WSKP, Western South Korea Plateau.

F
I G U R E 2 Summary of the structural evolution and stratigraphy of the SKP and adjacent areas.Red arrows indicate the timing of compressional stress, and blue dotted lines indicate the break-up unconformities in the SKP.EKCM, Eastern Korean Continental Margin; ESKP, Eastern South Korea Plateau; SKP, South Korea Plateau; UIG, Ulleung Interplain Gap; WSKP, Western South Korea Plateau.EAGE KIM et al.

4. 3
.1 |MB1 (latest Late Oligocene)   MB1 is the lowermost seismic reflection representing the top of the acoustic basement (Figures4-6).MB1 has high amplitude and nearly continuous seismic reflection, characterised by an internally chaotic and/or transparent reflection package.Typically, MB1 shows several depressions associated with basement-related faults (BL-1-BL-6) between intervening basement highs (BH-1-BH-7; Figure7a).The time-depth of MB1 ranges from less than 0.7 s to over 4.0 s TWT.The deepest BL-1 reaches depths of over 4.0 s TWT and is aligned in the N−S F I G U R E 4 Selected seismic profile in the E-W direction and its stratigraphic and structural interpretations.The profile displays four megasequences (MS1-MS4) separated by megasequence boundaries (MB1-MB4; for location, see Figure1c).South Korea Plateau consists of several basement lows (e.g.Onnuri and Okgye rift basins), bounded by basement-related normal faults.Contractional structures, such as reverse faults and folds, can be observed in the western part of the section.BH, basement high; BL, basement low; MB, megasequence boundary; MS, megasequence.direction.MB1 is generally traceable throughout the study area, while it is locally cut by several basement-related faults (Figure7a).Such basement-related faults comprise a series of N-or NNE-oriented normal faults and NWoriented reverse faults in the WSKP, while the ESKP is dominated by E-or EEN-oriented basement-related faults (Figure7a).The basement-related normal faults are developed along one side or flank of the basement lows in the SKP, resulting in the formation of rift basins (BL-1, BL-3 and BL-5) with half-graben style.Moreover, volcanic edifices are present in the northwestern and southern WSKP, forming the basement highs (BH-6 and BH-7; Figure7a).
4.3.2| MB2 (late Early Miocene) This area shows no rift-basin-style depression, and the sediment thickness increases eastward from the eastern slope of the EKCM to the deep basement low.MS1 is dominated by low-to-high amplitude internal reflections with moderate continuity.MS1 consists of several seismic-facies associations, while the three segments of the SKP (i.e.EKCM, WSKP and ESKP) are characterised by different seismic-facies distributions F I G U R E 6 Selected seismic profile in a N-S direction and its stratigraphic and structural interpretations on the ESKP.The profile displays MS1-MS4 separated by MB1-MB4 (for location, see Figure 1c).A rift basin associated with a basement-related fault is filled with a wedge-shaped sedimentary succession.BH, basement high; BL, basement low; MB, megasequence boundary; MS, megasequence.EAGE KIM et al.F I G U R E 7 Time-structure maps of MB1-MB4.AS, Anyongbok Seamount; BH, basement high; BL, basement low; MB, megasequence boundary; NUT, North Ulleung Trough; UI, Ulleung Island.F I G U R E 8 Time-thickness maps of MS1-MS4.AS, Anyongbok Seamount; BH, basement high; BL, basement low; MS, megasequence; NUT, North Ulleung Trough; UI, Ulleung Island.

F
I G U R E 1 0 (a) Time-structure map of MB4 showing likely sediment-supply directions from the Korean Peninsula.Channel systems identified in the present study are connected by submarine canyons (i.e.Gangneung and Donghae canyons from Cukur et al., 2021).(b, c) Seismic profile and its stratigraphic and structural interpretations.Several channels are stacked and show seismic reflections exhibiting high amplitude.(d, e) Seismic profile and its stratigraphic and structural interpretations showing stacked channels with levee construction above the horizon H1. (f, g) Seismic profile and its stratigraphic and structural interpretations showing stacked channels above the horizon H1.For the location of the seismic profiles, see (a).MB, megasequence boundary; MS, megasequence.

F
Seismic profiles showing various types of faults (for location, see Figure 1c).Faults are classified into four types: (a) Type-A (basement-related normal fault interpreted as syn-extension fault) and Type-B (layer-bound normal fault array interpreted as a polygonal fault system, or PFS), (b, c) Type-C (dome-related normal fault interpreted to be a post-sedimentary fault) and (d) Type-D (reverse fault).MB, megasequence boundary; MS, megasequence.
−e).Along the eastern flank of the fold, seismic reflections show clear onlap termination pattern towards H1 (Figure 13c,e).

F
I G U R E 1 3 (a) Time-structure map of MB4 showing an NNW-SSE-or N-S-striking fold structure.(b, c) Seismic profile and its stratigraphic and structural interpretations.The fold is cut by a reverse fault in the form of a fault-propagation fold.Growth strata between horizons H1 and H2 are thin across the fold limb towards the structural high, while shallow reflections show onlap termination against the H2.(d, e) Seismic profile and its stratigraphic and structural interpretations showing ponded sediment fill.In addition to the fold, a ponded depression is present, which illustrates onlapping strata.T A B L E 3 Description and interpretation of fault types.
conceptual model illustrating the structural and sedimentary evolution of the SKP, controlled mainly by tectonic activity.(a) From the latest Late Oligocene to Early Miocene time (Stage-1), the SKP was rifted and extended through block faulting, resulting in halfgraben-style rift-basin development.In the incipient rifting phase, intensive volcanism provided volcanic sills and lava flows from faulting and eruption centres at the volcanic edifices, occupying the lowermost part of the rift-basins.As the water invaded the rift valleys associated with rapid tectonic subsidence, the SKP was dominated by lacustrine deposition, forming a wedge-shaped syn-rift succession.At the same time, fan-delta and sublacustrine-fan deposits were supplied from subaerial erosion and/or denudation of adjoining continental ridges.(b) From the late Early Miocene to Middle Miocene time (Stage-2), slow thermal subsidence prevailed as basin extension and rifting ceased.During this time, hemipelagic sedimentation was predominant in tectonically quiescent environments.Intermittently, submarine masswasting occurred due to the over-steepening of slopes on the ridges, providing large amounts of MTDs.(c) From the late Middle Miocene to Late Miocene time (Stage-3), tectonically stable environments still predominated, resulting in thermal subsidence.During this period, hemipelagic sedimentation prevailed, causing polygonal-fault development in response to diagenetic processes associated with compaction and water expulsion.Meanwhile, the SKP was partly covered with sandy turbiditic deposits, as shown by the high-amplitude characteristics of the seismic reflections.(d) From the Early Pliocene time to the present (Stage This study is a contribution of the Basic Research Project of the Korea Institute of Geoscience and Mineral Resources (KIGAM) 'Technology development for storage efficiency improvement and safety assessment of CO 2 geological storage (23-3413)', as well as the gas-hydrate R&D (research and development) project 'Evaluation of gas hydrate characteristics in the southern part of Korea Plateau (23-1143-2)' of the Ministry of Trade, Industry and Energy.This research was also supported by the Korea Institute of Marine Science & Technology Promotion (KIMST) funded by the Ministry of Oceans and Fisheries ('Study on characterisation of submarine active faults in the Southwest China Sea; 23-9851').We thank the scientific parties, as well as the crews and staff of the research vessels who participated in data acquisition.ORCID Dong-Geun Yoo https://orcid.org/0000-0003-1179-4160 Description and classification of acoustic basement types.
T A B L E 1