Expedition 368X Preliminary Report: South China Sea Rifted Margin

International Ocean Discovery Program (IODP) Expedition 368X is the third of three cruises that form the South China Sea Rifted Margin program. Expeditions 367, 368, and 368X share the common key objectives of testing scientific hypotheses of breakup of the northern South China Sea (SCS) margin and comparing its rifting style and history to other nonvolcanic or magma-poor rifted margins. Four primary sites were selected for the overall program: one in the outer margin high (OMH) and three seaward of the OMH on distinct, margin-parallel basement ridges informally la-beled Ridges A, B, and C from north to south. The ridges are located in the continent–ocean transition (COT) zone ranging from the OMH to the interpreted steady-state oceanic crust (Ridge C) of the SCS. The main scientific objectives include In addition, sediment cores from the drill sites targeting primarily tectonic and basement objectives will provide information on the Cenozoic regional environmental development of the Southeast Asia margin.Expedition 368X a started after

1. Determining the nature of the basement in crustal units across the COT of the SCS that are critical to constrain style of rifting, 2. Constraining the time interval from initial crustal extension and plate rupture to the initial generation of igneous ocean crust, 3. Constraining vertical crustal movements during breakup, and 4. Examining the nature of igneous activity from rifting to seafloor spreading.
In addition, sediment cores from the drill sites targeting primarily tectonic and basement objectives will provide information on the Cenozoic regional environmental development of the Southeast Asia margin.
Expedition 368X was planned to reoccupy a site started during Expedition 368. Because of repeated breakdowns of the low clutch diaphragm in the drawworks, Hole U1503A was abandoned after installing casing to 991.5 m. Despite this setback to Expedition 368 and the South China Sea Rifted Margin program, Hole U1503A was completed during Expedition 368X. The overarching scientific goal of IODP Expeditions 367 and 368 was to unveil the mechanisms of continental breakup at the northern SCS margin from rifting through steady-state spreading. A key operational objective of Site U1503 was to sample the lowermost ~300 m of sediments on top of basement to constrain the age and subsidence history of the crust at this location, the timing of normal faulting, and the environment of the early half-graben fill. A second important goal was to sample at least 100 m of the igneous basement. Deep representative sampling of the igneous material at this site will provide an important reference frame for the modeling of breakup and early ocean spreading.
In Hole U1503A, the sediment sequence was cored with the rotary core barrel (RCB) system from 995.1 to 1597.84 m (602.74 m penetration; 128.01 m recovered; 21%) and then the underlying basement was continuously cored from 1597.84 to 1710.1 m (112.26 m penetration; 47.91 m recovered; 43%). Although logging deeper than 991.5 m (bottom of casing) was not possible because of unstable hole conditions, Hole U1503A was logged with the Vertical Seismic Imager in the cased portion of the hole. No days were lost to waiting on weather, and the only mechanical downtime was a 1.5 h period when an electrical malfunction caused the top drive to shut down for repairs.
Expedition 368X successfully completed the operational objectives in Hole U1503A that were abandoned during Expedition 368. In the SCS margin science program, material recovered during Expedition 368X will contribute toward meeting the four specific objectives of Expeditions 367 and 368. Postexpedition research on the sediments and basalt recovered from Hole U1503A will allow for determination of emplacement age and geochemical analyses of

Introduction
The South China Sea (SCS) margin (Figures F1, F2, F3) is an accessible and well-imaged location where drilling of synrift sediments and underlying basement will provide key constraints on the processes of rifting and eventual rupturing of the continental lithosphere during breakup at a highly extended rifted margin. International Ocean Discovery Program (IODP) Expeditions 367, 368, and 368X were based on drilling Proposals 878-CPP, 878-Add, 878-Add2, and 878-Add3. This project was implemented as a single science program, initially with 114 days of drilling operations spread across two IODP expeditions, as outlined in the Expedition 367/368 Scientific Prospectus (Sun et al., 2016b). Deep representative sampling of the basaltic material at Site U1503 would have provided an important reference frame for the modeling of breakup; however, planned operational objectives in Hole U1503A were not achieved during Expedition 368.
Expedition 368 installed a reentry system and 991.5 m of casing (10¾ inch) in Hole U1503A but the hole could not be deepened because of mechanical problems with the drawworks, including a fourth clutch diaphragm failure and a breakdown of the aft shaft bearing on one of the Elmagco brakes. The continuing failures of the low clutch diaphragm and the lack of available spares to last until the end of the expedition limited the operation of the ship's drilling equipment to shallower than 3400 m. Therefore, Site U1503 was abandoned without achieving the scientific objectives. Before departing Site U1503 during Expedition 368, the reentry system in Hole U1503A was inspected by subsea camera. The cone appeared to be 1-2 m below the seafloor but was clearly visible and expected to be available for reentry at a later date. Expedition 368X was able to take advantage of a unique opportunity in the R/V JOIDES Resolution schedule that resulted from forced equipment repairs. Expedition 368X adds an additional 20 days of drilling operations to the South China Sea margin science program. The drilling strategy for Expedition 368X was to finish the Expedition 368 Hole U1503A operational objectives, including coring the lowermost sediment and basement and logging the hole. During Expedition 368X, the sediment sequence was cored with the rotary core barrel (RCB) system from 995.1 to 1597.84 m (602.74 m penetration; 128.01 m recovered; 21%) and then the underlying basalt was continuously cored from 1597.84 to 1710.1 m (112.26 m penetration; 47.91 m recovered; 43%). Although logging deeper than 991.5 m (bottom of casing) was not possible because of unstable hole conditions, Hole U1503A was logged with the Vertical Seismic Imager (VSI) in the cased portion of hole.

Global questions regarding formation of rifted margins
The Ocean Drilling Program (ODP;1985-2003 made a major effort to understand the processes of continental breakup along the rifted margins of the North Atlantic (ODP Legs 103,104,149,152,163,173,and 210). This effort resulted in the recognition of two 6 end-members of rifted margins (see summary of observations in Sun et al., 2016aSun et al., , 2016b. The first recognized end-member is volcanic rifted margins, examples of which are characterized by massive igneous activity in a relatively short period of time (~1-3 million years) during breakup and initial seafloor spreading; the pair of conjugate margins of Greenland and northwest Europe is an example of this type. In these locations, the asthenospheric mantle may have been anomalously hot (e.g., close to mantle plumes), which led to thermal weakening of the continental lithosphere followed by rapid plate rupture.
The second recognized end-member is magma-poor rifted margins, which are interpreted to endure hyperextension of the continental crust, with tectonic extension at the distal margin eventually exhuming the subcontinental mantle lithosphere and leading to serpentinization of the mantle. The Newfoundland and Iberia pair of conjugate margins, where serpentinite occupies a broad zone in the COT zone, is an example of this type of margin and is the only conjugate margin pair where this interpretation has been confirmed by scientific drilling. The introduction of water into the subcontinental lithospheric mantle is interpreted to have taken place through deep, crust-cutting faults, causing serpentinization that profoundly weakens the mantle lithosphere and facilitates plate rupture. The subsequent ultraslow spreading led to the formation of additional serpentinite on the seafloor (e.g., Dick et al., 2003) until sufficient magma production allowed for the formation of normal oceanic crust.
Elsewhere, other examples of highly extended rifted margins have been identified in seismic reflection data (e.g., Brune et al., 2017;Doré and Lundin, 2015), but it is not known if serpentinized mantle plays a critical role in all cases. Modeling by Beaumont (2008, 2011) suggests several scenarios for the formation of rifted margins in the absence of anomalously hot asthenospheric mantle. One scenario (Type I of Huismans and Beaumont, 2011) is the Iberia-Newfoundland-type margin described above. In this case, lithospheric thinning initially occurs in the (upper) crust, and extensional faults profoundly thin the continental crust (hyperextension) and eventually reach the mantle and cause serpentinization (Whitmarsh et al., 2001;Pérez-Gussinyé and Reston, 2001;Pérez-Gussinyé et al., 2006;Reston, 2009;Sutra and Manatschal, 2012). The schematic model of this type of margin development ( Figure  F4) guided the drilling strategy of Expeditions 367, 368, and 368X. Beaumont (2008, 2011) also suggest, however, that final plate rupture can occur without exhumation of the subcontinental mantle and can be followed rather quickly by igneous ocean crust formation, a scenario that our chosen drilling strategy also tested. Therefore, the highly extended northern margin of the SCS is an excellent location to examine whether this margin endured magmatism during breakup or whether its development is closer to the Iberia-type, amagmatic margin.

Geological setting
The SCS is a modestly sized young ocean basin that formed along the eastern boundary of the Eurasian plate during the mid-to late Cenozoic ( Figure F1). Expeditions 367, 368, and 368X cored and logged a transect of drill sites across the COT in the northern SCS ( Figure F5).
The continental crust that was rifted to form the SCS was accreted to the Asian margin during the Mesozoic (Zhou and Li, 2000;Zhou et al., 2008;Li et al., 2012aLi et al., , 2012b. Starting about 80 million years later, this relatively young continental lithosphere underwent extensive rifting during the Paleogene, in the Eocene to early Oligocene ( Figure F3). Seafloor spreading in the eastern part of the SCS started during the Oligocene, and the oldest magnetic anomaly in the area of the drilling transect is interpreted to be Anomaly C11 (~29.5 Ma) or C12n (~31 Ma) (Briais et al., 1993;Li et al., 2013Franke et al., 2013). Seafloor spreading in the southwest SCS started at ~23 Ma, and even later in the most southwestern basin (Briais et al., 1993;Barckhausen and Roeser, 2004;Li et al., 2012aLi et al., , 2012bFranke et al., 2013).
The initial half-spreading rate was ~3.6 cm/y. It later slowed to 1.2 cm/y, and seafloor spreading eventually ceased by ~15 Ma (e.g., Li et al., 2014). The initial spreading rate in the SCS basin therefore appears to be higher than the ultraslow spreading off the Iberia-Newfoundland margin (Dick, 2003). Subduction of the eastern part of the SCS basin started at or before ~15 Ma along the Manila Trench (Li et al., 2013). For a more complete review of the regional setting and tectonic development of the SCS, see Shi and Li (2012), Li et al. (2013), Sun et al. (2014), and Franke et al. (2013).
The Expedition 367, 368, and 368X drilling transect is located ~50 km west of IODP Site U1435 (Figures F1, F4, F5) (Li et al., 2015a(Li et al., , 2015b, along a segment of the northern SCS margin bounded to the west by a transform fault. This margin segment exhibits a broad zone of extended crust (COT) (Figure F5) that may end to the east somewhere between IODP Sites U1432 and U1435. East of this segment, continental crust seems to thin into ocean crust in a narrower COT.
The segment of the SCS margin addressed during Expeditions 367, 368, and 368X is therefore characterized by a broad COT (~150 km) resulting from intense crustal stretching and extension prior to breakup. A deep sag basin (midslope basin [MSB]) of presumed Eocene to Oligocene age is present in the midslope area (Figure F5) and is bounded seaward by an outer margin high (OMH) forming a quite persistent structure along the margin to the east. Three distinct ridges (A, B, and C from north to south in Figures F2,  F5) are found seaward of the OMH in the more distal margin, representing progressively thinner continental crust in the COT or ocean crust. We refer to the continent/ocean boundary (COB) as the narrow zone in which the outermost, highly thinned continental lithosphere is replaced seaward by new crust that formed at a narrow spreading ridge in a steady-state fashion. The latter can include continuous tectonic exhumation of lithospheric mantle (e.g., Dick et al., 2003), accretion of normal igneous oceanic crust, or a mixture of these two processes. The nature and precise location of the COB at the SCS cannot be interpreted with confidence from the seismic data, and interpretation therefore requires drilling control.
The clear seismic reflections from the Mohorovičić seismic discontinuity (Moho) show distinct thinning of the continental crust ( Figure F5) across the COT with a thickness of ~6 km around its seaward end. Separate layers hypothesized to be upper, middle, and lower crust are distinguished in the landward part of the seismic profiles. The lower crust is acoustically transparent and may be as thin as ~6 km in places. Lower crust with a similar thickness and seismic appearance is reported from the northeastern SCS margin (McIntosh et al., , 2014Lester et al., 2013). The seaward continuation of this crustal layering into the COB zone is ambiguous, however, and prevents us from interpreting the exact location and detailed nature of the COB.
The upper crust shows numerous extensional faults soling out at low-angle detachment faults in the midcrustal level. This fault system generated a number of deep half-grabens filled with synrift sediments which were subsequently covered by postrift sediments. The boundary between synrift sediments and postrift sediments most 7 often follows the T70 seismic stratigraphic unconformity ( Figure  F5). Results from distant industry wells, as well as those of IODP Site U1435 suggest a breakup unconformity age of ~34 Ma. However, the time of crustal extension is not necessarily synchronous across the margin and could be younger toward the outer margin. A younger, widely distributed unconformity (T60) is also observed ( Figure F5). The T60 unconformity corresponds to a hiatus at ~23 Ma found at ODP Site 1148 (Wang, Prell, Blum, et al., 2000) and IODP Site U1435 and is approximately synchronous with a southward jump and change in kinematics of the SCS spreading axis, and with the breakup in the southwest part of the basin (Briais et al., 1993).
The OMH hosts a number of relatively shallow half-graben basins on top of this broad basement high. The stratigraphy of these smaller basins can be traced seismically into the deeper, central basin sag below the MSB ( Figure F5). The normal faults bounding the OMH basins are clearly imaged and, for the main part, dip landward. The small rift basins therefore offer an opportunity to sample the stratigraphy covering the entire period of rifting and postrift subsidence. The MSB itself is bounded landward by major, seawarddipping normal fault(s) that seemingly form major detachments soling out at middle to lower crustal levels but not penetrating through the lower crust. If true, this suggests that decoupling between the upper and lower crust, and that at least in this more landward part of the margin, faults never penetrated the lower crust (remained ductile?).
The interpretation of some seismic profiles suggests that the lower crust in the COT may thicken seaward but this observation is not well constrained ( Figure F5). Likewise, seismic imaging of the low-angle faults and detachments in the landward part of the COT cannot, with confidence, be traced into the distal margin regime. One possibility is that the main detachment zone was located above what later became Ridge A, effectively implying that Ridge A is a core complex consisting of lower continental crust or subcontinental mantle depending on how deeply detachments exhumed the lithosphere in the distal margin. Alternatively, if the main detachment underlies Ridge A, the latter would represent upper plate material of upper crustal origin.
Ridge A is domelike, for the most part, and shows neither normal faults nor clearly developed synrift half-grabens like the OMH. Excluding sediment and using the ocean-bottom seismometer (OBS) velocity constraints of Yan et al. (2001), Wang et al. (2006), and Wei et al. (2011), the crust below this outermost basement high is estimated to be only ~6-8 km thick. Seaward of Ridge A, the crust has a fairly uniform thickness of ~6 km, which could be consistent with oceanic crust (Yan et al., 2001;Li et al., 2014). The projection on the seismic profile of the magnetic lineation interpreted to be Anomaly C11 (Briais et al., 1993) coincides with the seaward part of Ridge A (Figures F2, F3, F5).
Both Ridges B and C consist of fault blocks rotated landward along seaward-dipping normal faults, some of which may be seismically traced to near the base of the crust. Ridge B shows seismic features along strike and in the uppermost crust that could be consistent with a volcanic origin. However, these features could also be consistent with a rotated fault block of upper continental crust (i.e., a distal extensional rider of upper plate origin), in which case prerift deposits could be present beneath the seismic unconformity defining the top of acoustic basement at Ridge B. The seismic layered structure of Ridge B makes it less likely to consist of lower crust or serpentinized mantle. Ridge C, in many ways, is seismically similar to Ridge B. However, the significant amplitude of the magnetic anomalies strongly suggests that Ridge C is indeed partial, if not full, igneous ocean crust.
Sampling the basement at Ridges A, B, and C was therefore essential for the SCS margin science program to distinguish between different tectonic models for breakup along highly extended margins. Ridges A and B help constrain the style of rifting. In contrast, Ridge C is assumed to represent the early igneous crust, and the material recovered here will address another key objective of the SCS margin science program: to constrain the nature of early oceanic crust formation, specifically to determine how quickly a robust igneous system was established, what mantle source is involved (e.g., composition, temperature), what conditions of mantle melting (degree and depth of melting) were present, and what continental crustal contamination of the igneous material, if any, derived from the asthenospheric mantle.

Expedition objectives
By drilling a transect across the SCS margin, expedition scientists set out to understand the timing and process of rifting, eventual rupturing of the continental crust, and onset of igneous oceanic crust at a highly extended rifted margin ( Figure F4). Four primary and sixteen alternate drill sites across a ~150 km wide COT zone were defined in the Scientific Prospectus (Sun et al., 2016b). The four primary sites were planned to target the four main tectonic features: the OMH (and its small rift basins) and the three basement ridges (A, B, and C) in the distal margin. At each of these sites, the nature of the acoustic basement and the record of synrift and postrift deposits were key targets.
As part of the SCS margin science program, Expedition 368X was planned to complete operations at Hole U1503A. The primary operational objective at Site U1503 was to sample the lowermost ~300 m of sediments on top of basement and at least 100 m of the igneous basement. The sedimentary target will constrain the age and subsidence history of the crust at this location, the timing of normal faulting, and the environment of the early half-graben fill. The basement objective will provide deep representative sampling of the basaltic material at this site and an important reference frame for the modeling of breakup. Although logging (triple combination [triple combo], Formation MicroScanner [FMS]-sonic, and VSI tool strings) into the basement was originally planned, the condition of Hole U1503A deeper than ~1300 m was unstable and not suitable for logging. Logging deeper than 991.5 m (bottom of casing) was not possible because of unstable hole conditions, but Hole U1503A was logged with the VSI in the cased portion of hole.

Coring and logging strategy
Drilling operations were based on the original plans for Expedition 368, which were designed to core and log through thick sediment sections and, significantly, into underlying basement using casing in the upper, unstable part of the sedimentary section. The operational approach was to reenter Hole U1503A using the reentry system and 991.5 m of 10¾ inch casing installed during Expedition 368 ( Figure F6).
Hole U1503A was designed for coring using the RCB system to extend from the base of the casing through the sediments and into the underlying basement. Multiple pipe trips to replace hard rock RCB bits would be required by the depth of target within basement. Upon completion of the coring objectives, the RCB bit was to be dropped either in the bottom of the hole or on the seafloor before 8 downhole wireline logging data could be collected. For this deeper logging, we planned to use the triple combo and FMS-sonic tool strings and the VSI tool string to conduct check shots.
During Expedition 368X, we had to modify this general operational plan in response to borehole conditions and the need to focus our operations time to achieve our highest priority basement objective. The reentry cone installed during Expedition 368 was observed to be slightly below the level of the seafloor and full of debris from previous drilling in Hole U1503A. Thus, we installed a secondary free-fall funnel (FFF) with a 2.7 m extension into the reentry system to ensure our ability to reenter Hole U1503A during multiple pipe trips to replace RCB bits (see Operations). RCB coring from 995.1 m (below the casing) to 1710.1 m penetrated 602.74 m of sediments and 112.26 m of basement. We then tripped pipe out of the hole, dropped the bit on the seafloor, and reentered Hole U1503A. Logging deeper than 991.5 m was not possible because of unstable hole conditions, so Hole U1503A was logged with the VSI in the cased portion of hole (end of casing at 991.5 m).

Site summary
Site U1503

Background and objectives
Site U1503 (proposed Site SCSII-9B) is located at 3867.7 m water depth near the top of the structural high named Ridge C ( Table T1). Ridge C is the most seaward ridge of the three margin-parallel basement ridges that characterize the lower continental slope underlain by thin (5-7 km) crust and possibly the oldest oceanic crust. Ridge C is believed to represent at least partial, if not full, igneous crust and therefore to have formed after the completion of continental breakup along this margin segment of the northern SCS.
A key operational objective of Site U1503 was to sample the lowermost ~300 m of sediments on top of basement to constrain the age and subsidence history of the crust at this location, the timing of normal faulting, and the environment of the early sediment infill. The other critical goal was to sample the igneous stratigraphy to at least 100 m below the acoustic basement. Deep representative sampling of the basaltic material at this site will provide an important reference frame for the modeling of breakup. With an estimated sediment thickness of 1640 m overlying basement, obtaining basement samples and log data at this site represented a challenging operation.

Operations
Operations were conducted in one hole at Site U1503 (Table  T1). In Hole U1503A, the reentry system and 10¾ inch casing (to 991.5 m) installed during Expedition 368 were used to reenter the existing hole. The RCB system cored from 995.1 to 1597.84 m (602.74 m penetration; 128.01 m recovered; 21%) and then continuously cored 112.26 m into the underlying basalt from 1597.84 to 1710.1 m (47.91 m recovered; 43%). Logging with the VSI tool string was conducted within the casing only.

Lithostratigraphy, igneous and metamorphic petrology, and structural geology
Site U1503 is divided into four lithostratigraphic units ( Figure  F8); three units are composed of sediments and one unit consists of igneous basement. The uppermost 995.1 m of sediments were drilled without coring during Expedition 368. Lithostratigraphic Unit I (995.10-1484.74 m) is a 489.6 m thick sequence of well con-solidated to lithified brownish gray, moderately bioturbated claystone with greenish gray sandstone and siltstone interbeds ( Figure  F9). Some of the coarser intervals have high carbonate content and multiple sedimentary structures (mud clasts, a fining-upward sequence, and parallel and convolute laminations) that are potentially related to deep-sea turbiditic flows. The relatively low recovery (21%) of Unit I is attributed to the presence of thick sandstone layers that are inferred from the high-amplitude reflectivity of seismic section. Unit II (1484.74-1542.77 m) is divided into Subunits IIA (1484.74-1533.60 m) and IIB (1533.60-1542.77 m) based on carbonate and nannofossil contents. Subunit IIA has a recovery of 36% and consists of lithified dark reddish brown massive claystones with greenish gray intervals containing heavier bioturbation. Subunit IIB has extremely low recovery (3%) and consists of lithified reddish brown clay-rich chalk. The separation of Subunits IIA and IIB is also based on geochemical data. Subunit IIA has low carbonate content and high Sr, Fe, Ni, Zn, and Al concentrations, whereas Subunit IIB is carbonate rich and low in Sr, Fe, Ni, Zn, and Al. Unit III (1542.77-1597.84 m) contains heavily bioturbated lithified greenish gray nannofossil-rich claystone, greenish gray claystone, light greenish gray siltstone, and dark gray banded claystone. Recovery in Unit III is extremely low (5%). A 4 cm thick greenish black interval in Unit III is likely composed of highly altered volcanoclastic material.
Lithostratigraphic Unit IV (1597.84-1710.10 m) is an igneous unit that samples the uppermost part of the SCS basement. The boundary between the sediment of Unit III and the underlying basalt of Unit IV is at 1597.84 m, but it is unfortunately disturbed by drilling and corresponds to the separation between claystone and basalt rubble at the bottom of the core. Unit IV continues through 112.26 m of basement, of which 47.91 m was recovered. Unit IV is composed primarily of sparsely plagioclase to plagioclase phyric basalt with no vesicles to high vesicle content ( Figure F10). Basalts have mostly an ophitic to subophitic texture with euhedral phenocrysts of plagioclase. This basalt unit contains chilled margins with preserved fresh glass and occasional hyaloclastites with brecciated glass fragments imbricated with clayey sediments or recrystallized carbonate. Veins occur throughout Unit IV and are predominantly filled with carbonates, Fe oxides, chlorites, and zeolites. Alteration of these basalts remains generally low, as evidenced by the minimal alteration of interstitial glass and the good preservation of plagioclase. The textures, contacts, and structures of Unit IV suggest an emplacement as pillow or lobate lava flows in subaqueous environment.

Biostratigraphy
Between Sections 368X-U1503A-2R-1 and 368X-U1503A-48R-CC, a 5 cm whole-round sample from the core catcher was collected on the catwalk by JOIDES Resolution Science Operator (JRSO) IODP technical staff. The sample was vacuum sealed and stored for shipment to the Gulf Coast Repository (GCR) in College Station, TX, at the end of the expedition. Exceptions to this sampling strategy occurred in particularly low-recovery cores. Deeper than 1441.7 m in lithified sediments and basement, no samples were collected for biostratigraphy.

Paleomagnetism
The intensity of the natural remanent magnetization in sediments at Site U1503 is higher in the reddish claystone (10 −2 A/m) of Unit II than in the brownish, greenish, and gray sediments of both lithostratigraphic Units I (10 −3 A/m) and III (10 −4 A/m); in the 9 basalts (Unit IV), average initial intensity is at least two to four orders of magnitude higher than in sediments (1 A/m).
Both sediments and basalts can show two components of magnetization: a component isolated at lower fields that, at least in sediments, shows steep positive inclination and can be correlated to drilling overprinting and a characteristic component of magnetization (ChRM) that shows both reversed and normal polarities and can be isolated in different field ranges depending on the coercivity of the magnetic carries. In sediments, the presence of both normal and reverse polarities and a mean ChRM inclination of 26.4° ± 8.0° (close to the 25° inclination expected at the coordinates of Hole U1503A) support a primary nature of the magnetization. However, the observed inclinations ( Figure F11) cannot be correlated with a reference geomagnetic polarity timescale because of the extremely low recovery rate.
The anisotropy of magnetic susceptibility shows strong oblate shape and horizontal planar fabric in sediments, which is consistent with deposition in a calm pelagic environment. An inclined planar fabric is present in lithostratigraphic Units II and III and shows a foliation inclined by ~30°. In basalts, a prolate shape of the ellipsoid and an intermediate fabric indicate the presence of a flow; however, the flow direction cannot be determined because samples are not oriented.

Geochemistry
Geochemical analyses were conducted for headspace gas safety monitoring in all sediment cores and in two basement cores. Methane content is low (<30 ppmv) in the sediment sections shallower than 1394 m. Deeper than 1404 m, methane content in sediments gradually increases (average = 1249 ppmv) before reaching a maximum of 5066 ppmv very near the sediment/basement boundary. Quantification of sediment CaCO 3 , inorganic and organic carbon, nitrogen, and sulfur contents was made for 43 samples (Figure F12). Samples with carbonate content >30 wt%, from near the top of the cored sediments, correspond to sandstones in lithostratigraphic Unit I. With three exceptions, total organic carbon (TOC)/total nitrogen (TN) ratios at Site U1503 range from <1 to 6.4 (average = 4.3; Figure F12), suggesting that the majority of the organic matter is likely from a marine source. Source rock analysis (SRA) was performed on three sedimentary samples. Samples collected for X-ray diffraction (XRD) analysis (Sun et al., 2018) were retained for shorebased analysis of major element oxides and several trace elements using inductively coupled plasma-atomic emission spectroscopy (ICP-AES).

Physical properties
We measured physical properties on whole-round cores, working-half sections, and discrete samples. These measurements included gamma ray attenuation (GRA) bulk density, magnetic susceptibility, natural gamma radiation (NGR), P-wave velocity, moisture and density (MAD) and porosity, and thermal conductivity. The variations in physical property values led us to define physical properties (PP) Units 1-9, each of which has specific characteristics ( Figure F13). Lithostratigraphic Unit I includes PP Units 1-5. Subunit IIA includes PP Units 6 and 7. Subunit IIB and Unit III correspond to PP Unit 8. Unit IV corresponds to PP Unit 9. In PP Unit I, variations of physical properties are mostly related to the nature of the sediment; NGR, magnetic susceptibility, and porosity are higher, and density, P-wave velocity, and thermal conductivity are lower in the claystone and siltstone than in the sandstone. P-wave velocity of the claystone and siltstone also increases slightly with depth because of lithification. In PP Unit 7, magnetic susceptibility increases markedly in the red clay of Subunit IIA. The nannofossilrich claystone of Unit III, which corresponds to PP Unit 8, displays a clear decrease in bulk density, magnetic susceptibility, and P-wave velocity with depth. Such variations, which are opposite to those expected from lithification or compaction effects, are possibly due to the abundance of nannofossils in the sediment. Magnetic susceptibility in the basalts is very high but shows some variations that might correspond to changes in the nature of the magnetized minerals or the grain size. Compared to the sedimentary rocks above the basement, P-wave velocity in the basalts is quite high. NGR (<10 counts/s) and porosity are quite low in the basalts, whereas density is much higher than in sediment. The high magnetic susceptibility values in Cores 368X-U1503A-87R and 88R correspond to basalts that display more massive textures than those from the other cores and are likely to have distinct magnetized minerals or grain size.

Downhole measurements and seismic correlation
A VSI tool string was deployed to collect a vertical seismic profile at 18 stations with intervals of 50 m from 974.9 to 124.9 m within Hole U1503A. To avoid potentially deteriorated conditions in the open hole, the VSI tool string did not pass into the open hole; all check shots were conducted inside the casing. The VSI tool string was first lowered to near the end of the casing at ~991 m. The logging string was then pulled up and stopped at 18 stations. The VSI tool string was combined with telemetry and gamma ray tools. Following data collection, logging specialists at the Lamont-Doherty Earth Observatory provided corrected two-way traveltimes. The combined VSI and P-wave measurements were used to calculate the velocity-depth relationship for Hole U1503A using the interval velocity from the VSI experiment for the top part and the average P-wave velocity of sediment cores. Seismic waveforms in Hole U1503A were then extracted from seismic data and converted to depth using the composite depth-velocity relationship for the hole (Figure F14). The boundaries of multiple PP and Lithostratigraphic Units are connected to the prominent reflectors in the seismic image.

Preliminary scientific assessment
In this section, we assess the achievements of Expedition 368X in terms of meeting the four specific objectives of Expeditions 367 and 368 as stated in the Scientific Prospectus (Sun et al., 2016b). Based on the reentry system and casing installation completed during Expedition 368 and the need to address uncompleted operational objectives, Expedition 368X operated exclusively in Hole U1503A. The drilling strategy for Expedition 368X was to finish the operational objectives of Expedition 368 in Hole U1503A, including coring the lowermost sediments and basement and logging the hole. This objective was partly achieved during Expedition 368X with the potential for completion following postcruise research.
Site U1503 was a priority deep site on distal Ridge C, as stated in the Scientific Prospectus (Sun et al., 2016b). Expedition 368X penetrated 112.26 m into the basalt basement of Hole U1503A and recovered 47.91 m (43%). Although the age of crustal material from Hole U1503A is not yet determined, onboard igneous and metamorphic petrology and structural geology observations were made. The textural and structural features of the basalts suggest an emplacement as pillow or lobate lava flows at seafloor. Basement in Hole U1503A is composed primarily of sparsely plagioclase to plagioclase phyric basalt. Vesicles are apparent in some intervals, and most samples are ophitic in texture. The basalts contain hypohyaline to holohyaline chilled margins with preserved fresh glass and occasional hyaloclastites with glass fragments imbricated with clayey sediments. No indication of subcontinental mantle or igneous continental rocks is apparent at Site U1503. Even if the emplacement age of these basalts is not determined yet, we infer that they represent the incipient or steady-state formation of an oceanic crust.

To determine the time lag between plate rupture and asthenospheric upwelling that allowed decompression melting to generate igneous ocean crust.
This objective was partly achieved during Expedition 368X, with the potential for completion following postcruise research.
The recovery of igneous basement rocks at Site U1503 is ideal for geochemical analyses to determine the rock composition, melting process, age of crystallization, or mantle fertility. The combination of such analyses and further geochemical or thermomechanical modeling will constrain the mantle origin and melting processes that lead to the formation of these basalts and potentially reveal if they were emplaced at a steady-state, mature seafloor-spreading ridge. Expeditions 367, 368, and 368X sampled basalts at three sites (U1500, U1502, and U1503) that form a 36 km transect across the COT. The accurate measurement of the absolute age of basement formation at each of these sites will be crucial to determine the velocity of continental breakup processes from the first occurrence of mid-ocean-ridge basalt (MORB) to the steady-state accretion. Comparison of the composition of these basalts with those sampled at Sites U1431 and U1433 during IODP Expedition 349 will also constrain the scenario of breakup to steady-state igneous crust emplacement. These observations and future studies form a unique data set representing a benchmark for a rifted margin with moderate magmatism that can be compared to the Iberia-Newfoundland magma-poor margin, where a time lag between crustal rupture and onset of seafloor spreading of >15 million years has been determined.

To constrain the rate of extension and vertical crustal movements.
This objective was partly achieved during Expedition 368X, with the potential for completion following postcruise research.
The inability to pursue coring at Site U1503 prevented Expedition 368 from extending subsidence studies seaward of Ridge B (Site U1500). With the recovered material of Expedition 368X, when the basalts are dated, we will be able to calibrate magnetic models and better identify magnetic isochrons and determine the spreading rate.

To improve the understanding of the Cenozoic regional tectonic and environmental development of the Southeast Asia margin and SCS by combining Expedition 367 and 368 results with existing ODP/IODP sediment records and regional seismic data.
This objective was partly achieved during Expedition 368X, with the potential for completion following postcruise research.

Contributions include
• The recovery of sandstone that may contain zircons useful for provenance studies to determine the sediment origin (China, Palawan, Taiwan and associated uplift, and Luzon). • The recovery of red clays that have been observed at most regional sites and record an episode of open-sea, deep-marine conditions. When the Luzon arc started to close the SCS in the east, the SCS was no longer connected with the Pacific Ocean and became a marginal oceanic basin with a different oceanic circulation and sedimentation. • The observation that the chalk layer of lithostratigraphic Subunit IIB may be regionally correlated. • The observation that sediments just above the basement are tilted, which can contribute to the dating of faulting at Ridge C. • Information regarding the composition and age of formation of the basalts at Site U1503 that, along with results of the previous IODP expeditions, provides unique information on the regional mantle processes during the formation of the SCS and therefore on the geodynamics of Southeast Asia.

Operations
Hong Kong port call On the first day of the port call, we loaded hazardous freight, 155 short tons of potable water, and 404 short tons of drill water, all of which were loaded from a barge on the starboard side of the vessel. On the second day of the port call, 71.9 metric tons of marine gas oil were pumped on board from a barge on the starboard side. The Schlumberger active heave wireline compensator hydraulic system was repaired by the Schlumberger mechanic with parts that were air freighted to the port call. The system was tested in port after the repairs were completed.
During the 2 day port call, a reentry funnel extension was prepared to extend the reentry system in Hole U1503A. Additionally, the outer core barrel was prepared and the RCB core barrels were spaced out. The coring assembly was laid out on the piperacker skate track for a quick deployment on arrival at Site U1503. The passage plan for the expedition was completed, and arrangements were made with the agent and immigration for a departure at 0800 h on 17 November. All port call-generated trash was offloaded to the pier, the pier was cleaned, and all equipment was secured for sailing. During the second day of port call, the science party continued expedition preparations that included an orientation to life at sea, general safety introductions and tours, and introductions to the laboratories. The Expedition Project Manager gave a presentation about expedition expectations, tasks, and reports. The science party also received shipboard safety introductions by the Captain, First Mate, and Physician.
At 0500 h on 17 November 2018, immigration personnel arrived on board and cleared the personnel and vessel for departure. Shortly before 0800 h, the pilot arrived on board. With the assistance from two harbor tugs, Hai Tong and Hai Hoi, the vessel was underway, and the last line was released at 0824 h. The vessel proceeded to the pilot station, and the pilot departed the vessel at 0854 h. While underway, the science party continued training in the laboratories, which included introductions to the handheld portable X-ray fluorescence spectrometer, curation, and description software. The science party also received a lifeboat safety orientation and tours of the bridge and dynamic positioning system. The transit was completed in 24.6 h and ended at the expedition's first and only hole, U1503A. The average vessel speed was 11.5 kt over the 284 nmi distance. The vessel arrived at Site U1503 at 0900 h on 18 November, and the thrusters were lowered.

Hole U1503A
After arriving at Site U1503, the vessel shifted to dynamic positioning mode at 0953 h and the drill floor was cleared to begin operations in Hole U1503A. We did not deploy an acoustic positioning beacon but prepared one for immediate deployment, if required.
The upper guide horn was removed, and a reentry funnel extension was lowered and secured in the moonpool. A four-stand RCB bottom-hole assembly was assembled and lowered to 2990 meters below rig floor (mbrf ). The subsea camera was deployed and run down the drill string while the drill string was deployed to 3874 mbrf. The search for the reentry system was quickly successful. The reentry cone was observed to be slightly below the level of the seafloor and full of debris from the previous drilling in Hole U1503A. The vessel was positioned for reentry, and the bit was lowered into the reentry cone at 0014 h on 19 November 2018. After several attempts to reenter the casing, the bit passed into the casing at 0039 h. The subsea camera system was recovered, and a FFF was welded to the top of the reentry funnel extension. The reentry FFF and extension were deployed down the outside of the drill pipe ( Figure F15).
The upper guide horn was reinstalled, and the bit was lowered into the 10¾ inch casing. The drill string was filled every 20 stands as the string was lowered toward the bottom of the casing. The driller tagged hard fill inside the casing at 956 m. The top drive was picked up, and the driller washed down while pumping with slow rotation to 995.1 m. Total depth was reached at 1600 h on 19 November. A 30 bbl high-viscosity sweep was circulated, and a nonmagnetic core barrel was deployed to begin RCB coring in Hole U1503A. Cores 2R through 40R (995.1-1364.9 m) were cored at a rate of just over 100 m/day, and recovery was generally poor (17%). Recovery began to improve in Cores 41R through 51R (1364.9-1470.5 m; 38%), and the penetration rate dropped to <4 m/h.
Coring continued until 0400 h on 26 November and reached a depth of 1528.0 m with Core 57R. With 50.8 h on the bit, the hole was circulated clean with a 30 bbl sweep of high-viscosity mud. The bit was raised to 1383 m in the open hole (casing extends to 991.5 m), and 200 bbl of 10.5 lb/gal heavy mud was displaced into the borehole. The heavy mud was used to increase the hydrostatic pressure on a very unstable section of hole. The top drive was set back, and the drill pipe was pulled back to just above the casing shoe (953.0 mbsf ). The upper guide horn was pulled up, and the subsea camera system was deployed.
While running the subsea camera system to just above the seafloor, a slip and cut of the drilling line was performed. The bit was pulled out of the hole at 1247 h, and it cleared the secondary funnel without issue. The measurement to the top of the cone was confirmed at 3879.7 mbrf. The subsea camera system was pulled back to the surface and secured on board. The remainder of the drill string was recovered, and the bit cleared the rig floor at 2045 h on 26 November. The outer core barrel was disassembled and in-spected, and a new C-4 bit was picked up and installed with a mechanical bit release (MBR).
The drill pipe was lowered to 3855 mbrf, and at 0545 h the subsea camera was deployed. Hole U1503A was reentered at 0830 h on 27 November after 30 min of maneuvering. The subsea camera system was pulled back to the surface, and the trip back to bottom continued until the driller encountered a bridge at 1345.0 m. The top drive was picked up, and the driller washed and reamed from 1345.0 m back to bottom at 1528.0 m. After reaching total depth, the hole was circulated clean with high-viscosity mud. A nonmagnetic core barrel was dropped, and coring resumed at 2100 h on 27 November.
Coring with the second bit began with Core 58R and continued through Core 75R (1528.0-1633.0 m). Basement was reached at 1597.84 m. After reaching basement, half cores were used to improve recovery. Basement recovery from 1605.0 to 1633.0 m was 33%. With 42.2 h on the bit, the hole was circulated clean with three 50 bbl sweeps of high-viscosity mud. The bit was raised to 1355.7 m in the open hole, and 200 bbl of 10.5 lb/gal heavy mud was displaced into the borehole with the intent of increasing the hydrostatic pressure on the unstable section of the hole. The top drive was set back, and the drill pipe was recovered and cleared the seafloor at 1050 h on 1 December. After recovering the remaining drill pipe, the bit cleared the rig floor at 1725 h. The outer core barrel was disassembled, cleaned, and inspected, and a new C-7 bit was picked up and installed with an MBR. The drill pipe was lowered to the seafloor, and the subsea camera system was deployed. After 60 min of maneuvering, Hole U1503A was reentered at 0439 h on 2 December, and the subsea camera system was pulled back to the surface.
The trip back to the bottom of Hole U1503A continued until 1375.0 m, where the driller experienced a hard tag of the bit on a bridge. The top drive was picked up, and the driller washed and reamed from 1375.0 to 1517.0 mbsf, where an electrical problem caused the top drive to stop operating. The problem was resolved, and 1.5 h later we began washing back to total depth. A wash core barrel was deployed at 1615 m, and washing and reaming continued. Total depth (1633.0 m) was reached, and the hole was circulated clean with 40 bbl of high-viscosity mud. The wash barrel was retrieved by wireline, and a nonmagnetic core barrel was dropped. Coring began again at 2100 h on 2 December. Basement coring with the third bit began with Core 76R from 1633.0 m and continued through Core 77R to 1657.3 m before high torque and excessive overpull forced the driller to work the drill string back to 1566.0 m to regain good circulation and rotation. The core barrel, which had been dropped for Core 78R, was retrieved, and the driller washed and reamed back to bottom (1657.3 m). A 40 bbl mud sweep was pumped down the drill string and circulated out to the seafloor. A core barrel was dropped, and half-coring began at 1657.3 m and continued through Core 88R to a final total depth of 1710.1 m. All cores after Core 85R were full cores. High-viscosity mud sweeps were pumped frequently to help improve the hole conditions. The last core on deck was recorded at 0625 h on 5 December.
The hole was circulated clean, and we began pulling back with the top drive installed. High torque and overpull were observed while pulling back and reaming to ~1600 m. Backreaming continued to 1383 m while we continued to experience high torque and overpull. The hole was displaced with 200 bbl of high-viscosity mud, and the bit was pulled back to 1355 m. The top drive was set back, and the drill pipe was pulled back to the casing shoe and occasionally experienced overpull as the hole collapsed around the drill string. After reaching the casing shoe at 991.5 m, the remainder of 12 the pipe trip to the seafloor was normal. The bit cleared the seafloor at 1450 h and was pulled 12.7 m above the reentry cone. The subsea camera system was installed, and while running the camera system to bottom, the drilling line was slipped and 115 ft of line was cut off. The top drive was picked up, and the sinker bars were inserted through the blocks. The rotary shifting tool was run to the bit and engaged to shift the retaining sleeve, which activated the MBR. The subsea video camera recorded the bit release at 1920 h on 5 December. The wireline was pulled back to the surface, and the top drive was racked.
Hole U1503A was reentered for the fourth time at 2155 h on 5 December. The end of the pipe was set at 71.0 m in preparation for logging. After reentry, the subsea camera system was recovered to the surface at 0000 on 6 December. A short delay was required to time the VSI experiment for daylight hours. Logging activities began at 0400 h on 6 December. The rig floor had been prepared for logging during the down period needed for seismic work to be done in daylight hours. The tools were rigged up, tested, and lowered into the pipe at 0430 h.
Because the cable on the winch was new, it was necessary to conduct standard cable seasoning/detorquing by pausing every 500 m while the tools were lowered into the pipe and pulling up a short distance before continuing running in the hole. Upon reaching 3850 mbrf, a downlog was started for a rough depth reference. No other logs exist to tie into, so there was marginal uncertainty about the exact depth at which a gamma ray signal might begin to register, given that the original casing/cone sunk into the sediment ~5.1 m and the cuttings stacked up to almost the top of the secondary cone. Lacking any better reference and prior to shooting any seismic stations, the log depth was adjusted in such a way that the top gamma ray reading occurred at the driller's seafloor depth of 3879 mbrf. The only logging run for this expedition was with the VSI tool string in the cased section of Hole U1503A. Open-hole conditions were highly adverse, so no logs were conducted in the open-hole portion of this site. The work of taking seismic measurements through the casing began, as always, with a toolbox talk/safety meeting on the drill floor with the rig crew on duty. It was also noted that highpressure air would be in use near the fantail for the air guns and that the fantail should therefore be avoided during seismic shooting.
Just prior to logging, the average heave was estimated at 0.3 m. The active heave compensator was turned on when the VSI tool string reached 3965 mbrf and was utilized during each VSI station. The protective species watch began at 0630 h. After an hour of observation, the soft start began. The seismic air gun cluster (2 × 250 inch 3 G-guns in a horizontal array) was turned over to the Schlumberger wireline engineer at 0820 h, and the seismic survey began.
The single tool string consisted of the following tools: • VSI, • Enhanced Digital Telemetry Cartridge (EDTC), and • Logging equipment head-Q tension (LEH-QT).
The initial shots were taken at 975 m, and additional stations were shot every 50 m thereafter to 125 m. No further stations could be recorded because the pipe was set inside casing at 70 m. All 18 viable stations were successfully shot and recorded. Moderate noise was present in some areas, and most of the shots required manual transit time picking. However, overall the data quality was quite reasonable and fit for interpretation.
A continuous gamma ray log was recorded from the bottom of the casing upward past the seafloor. This log is the only depth reference for the expedition, and it was planned and recorded in real time so that the uppermost gamma ray activity corresponds with the driller's seafloor, thus tying the station measurements to the coring data as closely as possible under the circumstances. Once the gamma ray log was completed at ~3860 m at 1055 h on 6 December, the tools were pulled out of the hole and reached the surface at 1330 h. The tools were rigged down by 1400 h, thus concluding the logging run.
The drill string was pulled out of the hole to the surface, and it cleared the seafloor at 1450 h on 6 December. By 2345 h, all rig floor equipment was secured, ending operations in Hole U1503A and at Site U1503. The thrusters and hydrophones were raised, and the vessel began the transit to Hong Kong at 0030 h on 7 December. A total of 444.25 h (18.5 days) were recorded during Expedition 368X while in Hole U1503A. One downtime incident occurred; an electrical malfunction caused the top drive to shut down for 1.5 h while the problem was repaired. After the 284 nmi transit to Hong Kong averaging 9.4 kt, the vessel arrived at the pilot station. The pilot arrived on board at 0654 h, and the first line ashore was at 0754 h on 8 December, thus ending Expedition 368X.
14 Table T1. Expedition 368X hole summary. * = reentry system and casing installed to 991.5 m during Expedition 368.  Figure F1. Seismic data coverage and magnetic anomalies of the South China Sea basin, Expeditions 367, 368, and 368X. Black lines = ocean-bottom seismometer refraction data. Other seismic lines are mostly multichannel seismic reflection data. Yellow lines = magnetic isochrons from Briais et al. (1993). White stars = Expedition 367, 368, and 368X drill sites, red squares = ODP Leg 184 sites, red circles = IODP Expedition 349 sites. For more details, see Figure F2.