Cenozoic tectonic subsidence in deepwater sags in the Pearl River Mouth Basin, northern South China Sea

Cenozoic tectonic subsidence in deepwater sags in the Pearl River Mouth Basin, northern South China Sea

Tectonophysics 615–616 (2014) 182–198 Contents lists available at ScienceDirect Tectonophysics journal homepage: www.elsevier.com/locate/tecto Ceno...

5MB Sizes 1 Downloads 42 Views

Tectonophysics 615–616 (2014) 182–198

Contents lists available at ScienceDirect

Tectonophysics journal homepage: www.elsevier.com/locate/tecto

Cenozoic tectonic subsidence in deepwater sags in the Pearl River Mouth Basin, northern South China Sea Xie Hui a,b, Zhou Di a, Li Yuanping c, Pang Xiong c, Li Pengchun a, Chen Guanghao a, Li Fucheng a,b, Cao Jinghe a,b a b c

CAS Key Laboratory of Marginal Sea Geology, South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou 510301, China University of Chinese Academy of Sciences, Beijing 100049, China Institute of Science and Technology, Shenzhen Branch, China National Offshore Oil Corporation, Guangzhou 510420, China

a r t i c l e

i n f o

Article history: Received 5 September 2013 Received in revised form 10 January 2014 Accepted 12 January 2014 Available online 22 January 2014 Keywords: Tectonic subsidence Backstripping Deepwater sags Pearl River Mouth Basin Baiyun Sag South China Sea

a b s t r a c t The Cenozoic tectonic subsidence of the deepwater area in the Pearl River Mouth Basin, northern South China Sea is studied by subsidence analysis via backstripping calculations based on data of newly interpreted sequence boundaries. In the subsidence analysis local porosity–depth parameters were estimated based on well data, the paleo-water depth parameters and lithology were estimated for all grid nodes based on well data and sedimentary system maps, and ages of sequence boundaries were adjusted according to the International Chronostratigraphic Chart v2013/01. Sensitivity analysis shows that these are essential to ensure the quality of the subsidence analysis in the slope areas. Maps of tectonic subsidence rates of 18 Cenozoic sequences were constructed and spatial–temporal variations of the tectonic subsidence were discussed. The subsidence was restricted in the Baiyun and Liwan Sags before 30 Ma and extended to uplifts after 30 Ma. This indicates that the 30 Ma unconformity is the breakup unconformity that separated the syn-rift and post-rift sequences. Four substages of post-rift subsidence were identified by cyclic changes in the rate and pattern of tectonic subsidence at boundaries of 23.3 Ma, 17.2 Ma and 11.9 Ma, respectively. Anomalous post-rifting tectonic subsidence was significant and totaled to ~1200 m at the center of the Baiyun Sag, and punctuated by rapid subsidence and uplift events. The Baiyun Event was manifested in the study area by a short-lived uplift event in 23.9–23.3 Ma followed by a short-lived rapid subsidence in 23.3–19.8 Ma in the central study area. This localized rapid subsidence caused northward jump and anti-clockwise rotation of the shelf break. Two-order teeterboard (seesaw)-like subsidence was observed in the study area in the post-rifting stage, suggesting elastic deformations of the crust and lithosphere. © 2014 Elsevier B.V. All rights reserved.

1. Introduction The South China Sea is one of the largest marginal seas in western Pacific. Its northern continental margin is dominated by extensional basins, among which the Pearl River Mouth Basin is the largest one. The Baiyun Sag and its neighboring Liwan Sag are two deepwater sags of the Pearl River Mouth Basin, with a total area of over 30,000 km2, sediment thickness over 14 km in the center, and water depth ranging from 200 to 3000 m (Fig. 1). Exploration in the last decade revealed high potential for hydrocarbon in these deepwater sags (Chen et al., 2003; Pang et al., 2004, 2006, 2007a, Zhu et al., 2008, 2010). Six gas fields have been discovered by the China National Offshore Oil Corporation (CNOOC) in the Baiyun Sag. Among these the LW3-1 field discovered in 2006 has verified geological reserve of (1000–1700) × 108 m3 natural gas (Zhu et al., 2010). The Baiyun Sag and the Liwan Sag (Fig. 1) have attracted attention from both oil companies and marine scientists. The studies on these sags marked the beginning of the research in China on offshore deep water sags. In this paper we studied the tectonic subsidence of the deepwater area in the central segment of the Pearl River Mouth Basin, including the Baiyun Sag, the Liwan Sag, and a part of the Panyu Low uplift, the Dongsha 0040-1951/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.tecto.2014.01.010

Uplift and the Southern Uplift, with a total area of over 40,000 km2 (Fig. 1). Throughout the paper the term “subsidence” is used as synonym of the term of “tectonic subsidence”, which is defined as the vertical motion (positive downward) of basement at a site that is induced by tectonic forces such as thermal contraction, tectonic deformation, and dynamic topography. In order to obtain subsidence information, corrections need to be made for the effects of compaction and sediment loading, changes in paleo-water depth, and global sea level changes. This process is called as “subsidence analysis”, which is made through backstripping of individual sedimentary layers for calculating balanced basement depths (Sclater and Christie, 1980; Stam et al., 1987; Steckler and Watts, 1978). Previous studies on the Cenozoic subsidence in the Pearl River Mouth Basin have been reported in a number of papers (Clift and Lin, 2001; Clift et al., 2002a; Liao et al., 2011; Ru and Pigott, 1986; Xie et al., 2006; Zhao et al., 2011). Most studies were based on seismic and well data from continental shelf areas. With the proceeding of hydrocarbon exploration to deepwater areas, new 2D and 3D multi-channel seismic data have been acquired. In 2009 to 2010 the CNOOC launched a new round of sequence

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

183

Fig. 1. Map of structural divisions of the Pearl River Mouth Basin. Zhu 1, Zhu 2, and Zhu 3 are names of depressions; CSD—Chaoshan depression; SHU—Shenhu uplift; PYLU—Panyu Low uplift; DSU—Dongsha uplift; BYS—Baiyun sag; LWS—Liwan sag. Black box indicates the study area of this paper. The small halved dots are the locations of existing wells within the study area. The large dots with black triangles indicate the area of fastest tectonic subsidence for sags or uplifts (P3 and P4 are also the location of wells). The straight line AB shows the location of the seismic profile in Fig. 3. Inferred faults are after Sun et al. (2008).

stratigraphy analysis for the Baiyun Sag and the Liwan Sag, which resulted in high-quality isobath maps in the area of 44,500 km2 for 19 sequence boundaries (including seabed), as well as maps of sedimentary systems of selected sequences. In order to update our understanding on the tectonic subsidence and geodynamics of these deepwater sags, we conducted subsidence analysis via backstripping calculations based on these newly acquired data. Parameters for the calculation were carefully derived based on local well data, and the computer program of Stam et al. (1987) was revised to take account of the significant variation in lithology and paleo-bathymetry in deepwater areas. These actions have improved the reliability of the calculation as shown by sensitivity analysis. Maps of tectonic subsidence and its rates of the 18 sequences were constructed. Spatial and temporal variations of tectonic subsidence were revealed, and several rapid subsidence events and uplift events were identified. These have provided new information on the formation and evolution of the deepwater sags and of the Pearl River Mouth Basin, even of the South China Sea, as discussed in this paper. 2. Geological setting The South China Sea was formed by seafloor spreading during Late Oligocene to Miocene (Briais et al., 1993; Taylor and Hayes, 1983). The extension in the South China Block started in the Late Cretaceous – Early Paleocene and resulted finally in the opening of the South China Sea (Zhou et al., 1995). The origin of the extensional forces is controversial (Flower et al., 1998; Holloway, 1982; Tapponnier et al., 1982, 1986; Taylor and Hayes, 1983). The date for the onset of seafloor spreading in the South China Sea is also unsettled. A widely accepted pattern of magnetic anomalies in the South China Sea indicates that the seafloor spreading started first in the eastern and northwestern subbasins at linear magnetic anomaly C11 and propagated to the southwestern subbasin at the time of anomalies C6b–C7 and anomaly C6. Accordingly the onset of seafloor spreading is dated as 32 Ma by the date of anomaly C11 according to the magnetic time scale of Patriat (1987) (Briais et al., 1993; Taylor

and Hayes, 1983). But this date should be changed to 30 Ma, which is the new date of anomaly C11 according to the revised geomagnetic polarity timescale of Cande and Kent (1995). The ODP Site 1148 drilled at water depth of 3294 m of the southeast corner of the study area (see Fig. 1 for the location of the Site) terminated at 861 m below seafloor in marine sedimentary rocks dated as Early Oligocene 32.8 Ma with no unconformity penetrated at Early Oligocene ~ 30 Ma (Shipboard Scientific Party, 2000). Thus the breakup of the South China Sea was thought to be earlier (Li et al., 2005). This was echoed by Hsu et al. (2004), who identified anomaly C17 in the northeastern South China Sea and proposed the breakup of the South China Sea at 37 Ma. However, the 37 Ma onset model has not been widely recognized because the magnetic anomalies are weak and the correlations are unconvincing. In addition, Barckhausen and Roeser (2004) identified the anomaly C12 in the northwestern subbasin of the South China Sea and thus proposed the onset of seafloor spreading at 31 Ma. In this study we adapt 30 Ma as the age of the breakup unconformity, bearing in mind the uncertainty in dating. The Pearl River Mouth Basin lies in the northern continental margin of the eastern and northwestern subbasins of the South China Sea, with a total area of 200,000 km2. The basin is composed of two depression zones separated by three uplift zones, namely from the north to the south, the northern terrace, the northern depression zone consists of Zhu 3 and Zhu 1 depressions, the central uplift zone consists of Shenhu, Panyu, and Dongsha uplifts, the southern depression zone consists of the Zhu 2 and Chaoshan depressions, and the southern uplift zone along the margin of the deep-sea basin (Fig. 1). The basement of the Pearl River Mouth Basin consists of Jurassic and Cretaceous granites in its central and northern portions, un-metamorphosed Mesozoic sedimentary rocks in the east, and Paleozoic quartzite and other metamorphic rocks in the west (Zhou et al., 2008a). The crust that underlain the basin thins from ~ 30 km near the coast in the north to ~ 11 km near the deep-sea basin in the south (Huang et al., 2005; Kido et al., 2001; Nissen et al., 1995). Beneath the centers of sedimentary depressions, the Moho surface shoals significantly, for example, the crust

184

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

Fig. 2. Stratigraphic formations, sequence boundaries of the study area, and global and local sea level curves with ages adjusted according to the International Chronostratigraphic Chart v2013/01 (Cohen et al., 2013). The sequence boundaries and their ages are listed in the 7th and 5th columns, respectively.

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

thins rapidly to b 7 km under the center of the Baiyun Sag and b9 km under the center of the Liwan Sag (Hu et al., 2009; Wu et al., 2005). The general stratigraphic column of the Pearl River Mouth Basin is presented in Fig. 2. The Paleogene to Lower Oligocene strata consist of fluvial–lacustrine sediments in discrete rifts, among which the Eocene Wenchang and Enping Formations contain lacustrine dark mudstones that are major source rocks for hydrocarbons in this area. Sediments in the Upper Oligocene Zhuhai Formation are transitional (alternatively coastal and littoral) and contain both source rocks and reservoirs. Neogene strata consists of marine formations that constitute a generally transgressional sequence, indicating a trend of sea level rise in the basin that is opposite to the contemporaneous global trend of sea level drop (Haq et al., 1987) (Fig. 2). The Baiyun Sag is the largest and deepest sag in the Pearl River Mouth Basin, with an area of 14,000 km2 and the maximum thickness of Cenozoic sediments over 14 km (Fig. 3) (Huang et al., 2005). The Liwan Sag discovered recently lies in the southeast of the Baiyun Sag and very close to the continent–ocean boundary of the South China Sea. This small sag has an area of 4000 km2. The maximum sediment thickness is up to 8 km in the center of the Liwan Sag. Regional faults F1 and F2 crossing the study area (Fig. 1) are identified on gravity and magnetic maps and are not clearly seen in seismic sections (Chen et al., 2005; Sun et al., 2008; Zhou et al., 2009). A large number of faults are widely developed on the northern and southern flanks of Baiyun and Liwan Sags but mostly as small NWW-running faults. The structure style of the Liwan Sag is quite different from other sags in the basin. It is composed of multi-oriented folds and scattered domes complicated by intrusions and basement topography (CNOOC unpublished seismic data). Several episodes of Cenozoic tectonic events have been recognized in the Pearl River Mouth Basin (Fig. 2). These are the Shenhu Event (onset of rifting) at the earliest Cenozoic, the first and second episode of Zhuqiong Event (rifting episodes) at Early/Middle Eocene boundary and Middle/Late Eocene boundary, respectively, the Nanhai Event

185

(breakup) at ~ 30 Ma, the Baiyun Event (possibly related to ridge jump) at the Paleogene/Neogene boundary, and the Dongsha Event (uplifting possibly related to the docking of Philippine archipelago) in Late Miocene (Chen et al., 2003; Gong and Li, 1997; Pang et al., 2007b). 3. Data and methods 3.1. Data 3.1.1. Depth of Sequence boundaries The CNOOC project of sequence stratigraphy analysis for the Baiyun Sag and Liwan Sag was conducted in 2009–2010 based on 1 km by 1 km or denser multi-channel seismic lines, several 3D seismic patches, and 20 wells in the study area. 19 third-order sequences boundaries (including seabed) were identified and the depths of these boundaries were obtained by time–depth conversion using a cubic polynomial function. These data are the basis for our research. Because the sedimentary thickness in the Baiyun and Liwan Sags is very large, the depths obtained by using a cubic polynomial time– depth function would underestimate the depths of deeper sequence boundaries. This is because a cubic polynomial time–depth function has a quadratic velocity function, which indicates a decreasing velocity at the depths deeper than the inflection point, while in reality the velocity will increase towards larger depths unless the section is overpressured. In this study we used the depth data of seabed and upper 16 sequence boundaries derived from the seismic sequence study using a cubic polynomial function, but re-converted the depths of the Tg and T8 from the travel time data using a power function in the form of D = atb + C with 1 b b b 2 for time–depth conversion. Previous study has shown that such a power time–depth function has velocity increases towards larger depths with a decreasing rate of velocity increase. Thus the power time–depth function may avoid underestimation problem at greater depths meanwhile keep high quality of time–depth conversion at shallow depths (Zhou et al., 2008b).

Fig. 3. Seismic profile AB with interpretation across the study area. (a) The original seismic profile; (b) the seismic profile with interpretations; (c) the depth profile. The column in the right shows the interpretations of sequence boundaries. For location of the profile see Fig. 1; for the ages of the sequence boundaries see Table 1.

186

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

We gridded these depth data into 2 km by 2 km cells and obtained a total of 15,158 nodal points. Points with crossover data (older boundary shallower than younger boundary) or missing data were adjusted properly. Then the “cleaned” data set of seabed and sequence boundary depths was subjected to backstripping calculation. 3.1.2. Dates of sequence boundaries The ages of the 19 third-order sequence boundaries were given by CNOOC and are shown in the second column of Table 1. The dating methodology is briefed below. The sequence boundaries were firstly dated by nannofossil and foraminiferal zones for the marine and transitional sequences in Neogene and later Paleogene. However, the resolution was not high enough for the third order sequences. It has been recognized that the spectrum of the 3rd-order Neogene sea level curve of the Pearl River Mouth Basin has a reasonable correspondence with the 3rd-order global eustatic sea level curve of Haq et al. (1987) (see Fig. 2). Thus CNOOC used the ages of the 3rd-order global cycle boundaries of Haq et al. (1987) to date the local 3rd-order Neogene sequence boundaries (Pang et al., 2005; Xu et al., 1995). However, the ages of Haq et al. (1987) were based on the chronostratigraphic chart they used in 1987. Since then the chronostratigraphic chart has been updated several times. The dates of the 19 sequence boundaries as used by CNOOC have been adjusted to the latest International Chronostratigraphic Chart v2013/01 (Cohen et al., 2013) in this study using the method described in Appendix A. The upper Paleogene Zhuhai Formation is of alternative coastal and littoral facies. The top of the formation is regarded as the Paleogene/ Neogene boundary according to fossil records. However the corresponding sequence boundary in Haq et al. (1987) is slightly below the Paleogene/Neogene boundary. After the adjustment this sequence boundary is dated as 23.3 Ma, which is used as the age of the sequence boundary SB23.3. Five sequence boundaries are recognized within the Zhuhai Formation and their dates are determined by fossil records. In this paper these dates are adjusted by interpolation between the dates of the upper and lower boundaries (SB23.3 and T7 respectively) of the Zhuhai Formation. Except for the Zhuhai Formation the Paleogene strata in the Pearl River Mouth Basin are continental. The boundaries of T7, T8, and Tg are stratigraphic boundaries. CNOOC recognized the T7 boundary as

Table 1 Age of the sequence boundaries in the study area. The ages used by CNOOC and the ages used in this paper are listed in the second and third columns, respectively. The later ages are adjusted according to the International Chronostratigraphic Chart v2013/01. Sea level changes are according to Haq et al. (1987). Sequence boundary

Age (Ma) used in CNOOC

Age (Ma) adjusted in this paper

Maximum sea level change (m)

Minimum sea level change (m)

Seabed SB5.5 SB10.5 SB12.5 SB13.8 SB15.5 SB16.5 SB17.5 MFS18.5 SB21 SB23.3 ZHSB6 ZHSB5 ZHSB4 ZHSB3 ZHSB2 T7 T8 Tg

0 5.5 10.5 12.5 13.8 15.5 16.5 17.5 18.5 21 23.3 24.4 26 27 27.8 29 32 39 65

0 6.0 11.9 13.2 14.3 15.5 16.3 17.2 17.9 19.8 23.3 23.9 25.4 26.2 27.0 28.0 30.0 38.0 66.0

0 25 40 90 125 135 120 115 120 105 90 65 75 60 65 130 180 180 190

0 −40 −75 40 90 55 40 115 120 70 45 65 75 40 65 20 180 100 190

the breakup unconformity and dated by the onset of South China Sea opening, 32 Ma according to the earlier magnetic anomaly C11. In this study T7 is dated to 30 Ma according to the geomagnetic polarity timescale of Cande and Kent (1995). As discussed in Section 2, the time of onset of South China Sea opening and thus the age of T7 remain unsettled. The T8 and Tg are regarded by CNOOC as the base of Upper Eocene and Paleocene, respectively according spore and pollen records (Chen et al., 2003; Gong and Li, 1997). Their dates were 39 Ma and 65 Ma, respectively and in this paper adjusted to 38 Ma and 66 Ma, respectively. The T9 boundary has not been recognized in the study area (Pang et al., 2007a). Because there are no well penetrated T8 to Tg in the study area as yet, the ages of T8 and Tg are highly uncertain. The comparison of old and adjusted ages are listed in Table 1. The calculations in this paper were performed using the adjusted ages of sequence boundaries, but the names of sequence boundaries are retained to be consistent with the CNOOC system. 3.2. Method and parameters for backstripping For the deepwater sags in the Pearl River Mouth Basin, we used 1D backstripping method to calculate the tectonic subsidence and tectonic subsidence rate. The algorithm of 1D backstripping implies the assumption that the lithosphere strength is zero and the Airy isostasy is applicable. Previous studies have shown that this is a valid approximation for the study area where the lithosphere is weak (Clift et al., 2002b; Shi et al., 2005; Sun et al., 2008; Zhou et al., 2009). The tectonic subsidence Y of water-loaded basement at time t is given by (Steckler and Watts, 1978):  Y¼

   ρm −ρs ρm S ¼ W d− ΔSL ρm −ρw ρm −ρw

ð1Þ

where S is the sediment thickness at the former time; ρm, ρw and ρs are the mean densities of the mantle, water and decompacted sediment, respectively; ΔSL is the height of paleo-global sea level above the present day global sea level; and Wd is the paleo-water depth at the time of sedimentation. The reason for selecting 1D rather than 2D backstripping is that in 1D calculation we could use parameters that are variable with nodal points. Since Early Neogene the study area has resided on the area from outer continental shelf to lower continental slope (Pang et al., 2007b), the spatial variations in sedimentary lithology and paleo-water depths are significant and should not be overlooked. We modified the 1D backstripping program BURSUB (Stam et al., 1987) so that it may perform a matrix of 1D backstripping using the lithologic coding and paleo-water depths variable with grid nodes. Sensitivity analysis indicated that this might avoid large biases in the estimation of tectonic subsidence rate caused by using uniform parameters. 3.2.1. Porosity parameters In the study area, drillings revealed that the sediments are mainly mudstone, siltstone, and sandstone. The small occurrences of limestone and intrusive rocks are ignored in this research. We chose 17 wells (for locations see Fig. 1) in the Baiyun Sag and adjacent areas to estimate porosity parameters. The lithology and their porosity were determined from gamma ray, sonic and density curves. Then these data were used to fit the porosity–depth relation of each lithology by exponential function ϕ = ϕ0e−cZ where ϕ is the porosity at depth of Z, and ϕ0 and c are parameters of initial porosity and compaction coefficient, respectively. In this way we obtained ϕ0 and c parameters for mudstone and sandstone. The porosity parameters we obtained for the study area are significantly different from the widely used North Sea basin parameters of Sclater and Christie (1980) (Fig. 4). Sensitivity analysis has shown that a discrepancy of − 102%–+ 89% in the calculated tectonic subsidence at the DW1 well may have resulted from using the porosity

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

187

Table 3 Sedimentary lithology of sequences in the DW1 well.

Fig. 4. Comparison of the porosity–depth curves for sandstone and mudstone of the North Sea Basin (Sclater, and Christie, 1980) and of the study area estimated in this paper.

parameters of the study area rather than of the North Sea basin (Table 2). This indicates that by using local porosity parameters the quality of backstripping calculations are improved.

3.2.2. Lithology of each layer Lithological data from 17 wells and the contour lines of sand percentage deduced from seismic facies analysis in part of shelf regions are used to determine the lithology of the nodal points of the layers. The lithology of a layer at a point is coded by its sand/clay ratio. For example, the code of pure sandstone is 1 and pure mudstone is 2. The code of 1.8 indicates a layer at a point with 20% sandstone and 80% mudstone (example well is shown in Table 3). Then its porosity parameters are interpolated with these percentages from the parameters of sandstone and mudstone determined in Section 3.2.1. With this method, we coded the lithology of each grid nodal point on each layer. For the deepest five sequences wherein well data are sparse, we use the Table 2 Comparison of tectonic subsidence rate of the DW1 well given by backstripping using the North Sea Basin porosity parameters and using the Baiyun Sag porosity parameters. Sequences

Subsidence rate by Baiyun Sag parameters

Subsidence rate by North Sea parameters

Discrepancy

SB5.5–Seabed SB10.5–SB5.5 SB12.5–SB10.5 SB13.8–SB12.5 SB15.5–SB13.8 SB16.5–SB15.5 SB17.5–SB16.5 MFS18.5–SB17.5 SB21–MFS18.5 SB23.3–SB21 ZHSB6–SB23.3 ZHSB5–ZHSB6 ZHSB4–ZHSB5 ZHSB3–ZHSB4 ZHSB2–ZHSB3 T7–ZHSB2 T8–T7 Tg–T8

9.98 10.47 223.06 183.74 158.96 −66.36 183.78 −64.56 123.34 213.11 11.43 12.1 −10.62 84.42 107.12 175.64 24.51 25.66

12.71 12.7 227.04 199.58 165.98 −61.93 187.34 −59.9 128.43 212.81 18.63 22.91 0.27 99.66 119.53 172.04 30.45 28.04

27.35% 21.30% 1.78% 8.62% 4.42% −6.68% 1.94% −7.22% 4.13% −0.14% 62.99% 89.34% −102.54% 18.05% 11.59% −2.05% 24.24% 9.28%

Sequences

Thickness (m)

Sandstone (%)

Mudstone (%)

Litho code

SB5.5–Seabed SB10.5–SB5.5 SB12.5–SB10.5 SB13.8–SB12.5 SB15.5–SB13.8 SB16.5–SB15.5 SB17.5–SB16.5 MFS18.5–SB17.5 SB21–MFS18.5 SB23.3–SB21 ZHSB6–SB23.3 ZHSB5–ZHSB6 ZHSB4–ZHSB5 ZHSB3–ZHSB4 ZHSB2–ZHSB3 T7–ZHSB2 T8–T7 Tg–T8

336 278 104 318 150 65 61 92 179 50.8 38.4 177.8 166.7 122.5 106 96.3 746 742

0 0 0 0.63 0 0 0 0 10.06 0 41.02 28.57 12.60 31.32 25.55 3.63 75.77 61.82

100 100 100 99.37 100 100 100 100 89.94 100 58.98 71.43 87.40 68.68 74.45 96.37 24.23 38.18

2 2 2 1.99 2 2 2 2 1.90 2 1.60 1.71 1.87 1.69 1.74 1.96 1.24 1.38

arithmetic mean of the available wells to code the lithology for all nodal points. Sensitivity analysis showed that the discrepancy caused by using variable lithologic codes may be as large as 53% in the calculation of the tectonic subsidence. This indicates that by using variable rather than uniform lithological coding the reliability of backstripping results are improved. 3.2.3. Paleo-water depth and eustatic sea-level change The data on paleo-water depths from wells are used to constrain the paleo-water depth of each sequence. However there are merely 17 õindustrial wells available within the study area (Fig. 1). These wells are distributed in the shelf and upper slope areas, no well so far in the southern areas including the Liwan Sag. Most wells did not penetrate deep sequences. The data about paleo-water depths from these wells were obtained by paleontological studies and given in ranges, which is often very large (for example, 0–500 m for “delta front to bathyal”, and 500–2000 m for “lower slope”). Such paleo-water depth data cannot fit the needs of our subsidence study. Thus we used the wells only as control points as the estimated water depths should not be outside the given bathymetric ranges. The determination of paleo-water depth in this study was mainly based on the maps of depositional systems, which was compiled by the CNOOC based on well data and seismic sequence studies. In these maps the sedimentary facies, locations of shorelines, shelf breaks, canyons, slope fans, basin-floor fans, and sometimes the sand content contours for sequence boundaries are mapped. As an example the map of sedimentary facies at SB10.5 is shown in Fig. 5. We assumed that the paleo-water depth of the paleo-shorelines is zero, and the slope angle of the paleo-continental shelf was approximately the same as that of the present shelf (0.09°). Thus the paleo-water depth of the paleo-continental shelf in the study area could be determined by linear interpolation on the distance from the paleo-shoreline. For example, the paleo-water depth of the shelf break was around 60 m at SB10.5 by such calculation; this is possible as there was a significant seawater regression at this boundary. The estimation of paleo-water depth for the slope areas was more complex. The ODP Site 1148 is located b30 km off the SE corner of the study area (Fig. 5). At this site the present water depth is 3294 m (Shipboard Scientific Party, 2000), and paleobathymetry was determined by Zhao (2005), who pointed out that the water depth was increasing from 32.8 Ma to the present, and the water depth was 2500 m to 3500 m after the late middle Miocene (14 Ma). Thus we set the paleo-water depth of this site at SB10.5 to 2700 m assuming a gradual increase of water depth. The paleo-water depths in the area between the site and the shelf break are determined by nonlinear interpolation, assuming a smooth seafloor topography

188

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

Fig. 5. The map of depositional system for the sequence boundary SB10.5. The black five-sided polygon is the study area. The inset shows the theoretical seafloor topography from the shelf break to deep-sea basin.

with slope angle larger in the upper slope but gently decreases towards lower slope (see the curve in the inset of Fig. 5). By present-day analog, the slope angle was set about 1.86° in average for the upper slope where canyons developed, while that was about 0.17° for the lower slope where basin-floor fans developed. In this way the 1300 m, 2000 m, and 2400 m bathymetric contours were drawn along the bottom ends of the canyons and the starts and ends of basin-floor fans, respectively (Fig. 5). With these data points and lines, we estimated the paleowater depth at SB10.5 by interpolation and extrapolation and resulted in Fig. 6. This laborsome process was performed for other sequences

one by one. The paleo-water depth is set to zero at Tg (basement), and the paleo-water depths of T8 and T7 were interpolated between the paleo-water depth of ZHSB2 (lower Zhuhai Formation) and Tg. The estimated paleo-water depth was given as one number for one point at one sequence, not as ranges. The values of global eustatic sea-level change at sequence boundaries are read from the curve of Haq et al. (1987). Their maximum and minimum values are listed in the last two columns of Table 1. In the backstripping calculation average values are used as the eustatic sea level.

Fig. 6. The map of the paleowater depth for the sequence boundary SB10.5. The black five-sided polygon is the study area.

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

3.3. Uncertainties in the backstripping Results of backstripping are subject to uncertainties in methodology, data, and parameters. As we have used classical methodology, here we discuss only the uncertainty caused by data and parameters. As discussed in the previous text, this study has made great effort to reduce the uncertainty by using the newest and corrected data of sequence boundary depths and their ages, using local porosity parameters, and using lithological and paleo-bathymetry codes that are variable with grid nodes. Here we give a general review on two most principal sources of uncertainties: the boundary age and paleo-water depth. The uncertainty in boundary age will be passed on to the estimation of subsidence rate, but the effect is uniform over the entire area for individual sequences. The age uncertainty is relatively large for the Paleogene sequences which are less penetrated and mostly consist of continental sediments with poor paleontological constrain. The calculation of the tectonic subsidence is highly sensitive to paleo-bathymetry. In order to show the sensitivity intuitively we assign the values of ρm, ρw and ρs in formula (1) as 3.33 g/m3, 1.03 g/m3 and 2.33 g/m3, respectively, then the formula (1) becomes 



Y ¼ 0:43 S þ W d −1:45 ΔSL:

ð2Þ

In formula (2) the coefficient of the term S (the sediment thickness) is less than 1/2 of the coefficient of Wd (paleo-bathymetry) and less than 1/3 of the coefficient of ΔSL (increment of eustatic sea level). The value of S is derived from wells and seismic sections with relatively less uncertainty. The value of ΔSL is uniform in the study area and less than 190 m in Cenozoic (Haq et al., 1987). Thus largest uncertainty lies in Wd, and this uncertainty is 100% inherited by the calculated tectonic subsidence (Y). In this study the paleo-bathymetry for each sequence boundaries was carefully estimated as described in the previous section. Paleobathymetry maps have been adjusted many times to make the spatial and temporal variations of paleo-bathymetry smoother and to avoid unreasonable variations in resulted subsidence rate maps. The uncertainty in paleo-bathymetry estimation could not be quantified, but should be relatively small in the areas where sedimentary features have been mapped. As the paleo-water depth estimated in this paper is based on the newly compiled depositional system maps that we have never had before, it is reasonable to say that these are the best estimates that we might have made so far. However, the abovementioned efforts could not guarantee the freedom from significant uncertainties in the estimated tectonic subsidence, especially in the northeast and southwest portions of the study area where the sedimentary system is not mapped. This needs to be kept in mind in the interpretation of the backstripping results. 4. Results The backstripping calculation for the study area has yielded two sets of informative results: the sedimentation rates and the subsidence rates for the 18 sequences in all grid nodes. The spatial and temporal variations of sedimentation rate have been analyzed in another paper (Xie et al., 2013). The present paper focuses on the subsidence rate, which is the subsidence divided by corresponding time span. Contour maps of subsidence rates in the study area are drawn (Fig. 7), and the subsidence rates along the profile AB are mapped (Fig. 8). Five data points (P1 to P5 in Fig. 1, among these P3 and P4 are well sites) are selected in the centers of sags and uplifts, whose locations are indicated in Fig. 1. Curves of cumulative subsidence and histograms of subsidence rate are drawn for these points (Figs. 9 and 10). Based on these figures, the Cenozoic subsidence history and spatial variations of the deepwater sags in the Pearl River Mouth Basin are analyzed. In consideration of the unevenly distributed uncertainty in paleowater depth estimations which would be inherited in the estimation

189

of tectonic subsidence, we discuss only the variations in general patterns. Although the uncertainty in identifying sequence boundaries is relatively large for Paleogene sequences, they are more reliable for Neogene sequences. The uncertainty in the ages of sequence boundaries influences the estimation of subsidence rate, but does not bias the spatial comparison of subsidence features. 4.1. Temporal variations of Cenozoic subsidence The Cenozoic subsidence history of the study area may be divided into two distinct stages by the 30 Ma (T7) boundary. In this section we will discuss the detailed variations in these stages. 4.1.1. The pre-30 Ma stage The subsidence in this stage was characterized by restricted subsidence in sags only (Figs. 7a–b, 8a). The time-averaged subsidence rate was ~100 m/My in the Baiyun Sag and ~40 m/My in the Liwan Sag. Outside these two sags the subsidence rate was near zero (Figs. 7a–b, 9c–e). The subsidence was more diverse during 38–30 Ma (Late Eocene to Early Oligocene Enping Stage) relative to that in 66–38 Ma (Paleogene to Middle Eocene Shenhu and Wenchang Stages). In the central Baiyun Sag the subsidence rate was higher in 38–30 Ma than that in 66–38 Ma, but in the Liwan Sag and surrounding regions the subsidence rate in 38–30 Ma was lower than that in 66–38 Ma or even slightly uplifted (Figs. 7a–b, 8a). This might be an indication of an enhanced rifting in 38–30 Ma with stronger hanging wall subsidence and footwall uplift. However, the subsidence rates in the syn-rift stage are of high uncertainty because the poor age control of the sequences. These rates are certainly underestimated because the erosions that have been observed in seismic sections are not estimated in this study. 4.1.2. The post-30 Ma stage This stage was characterized by relatively strong subsidence in the entire study area with cyclical changes in subsidence rate. It could be subdivided into four substages, each substage was terminated by a period of uplift or slow subsidence. The first substage lasted from 30 Ma (T7) to 23.3 Ma (SB23.3) and corresponding to the deposition of the Zhuhai Formation (mostly Upper Oligocene). It was characterized by moderate subsidence rate (b 150 m/My in general) with the area of fastest tectonic subsidence shifted frequently (Figs. 7c–h, 8b). In the central Baiyun Sag the subsidence slowed down gradually (Fig. 10a), which is normal for the post-rifting subsidence. But in the Liwan Sag the subsidence rate increased in 30–26.2 Ma and exceeded the subsidence rate in the Baiyun Sag (Fig. 7c–e). Later in 26.2–25.4 Ma a NE–SW elongated uplift belt appeared across the NW Liwan Sag, and the area of fastest tectonic subsidence shifted back to the Baiyun Sag (Fig. 7f). After then the subsidence rate increased again in the Liwan Sag (Figs. 7g–h, 10b) in contrast with the decreasing rate in the Baiyun Sag (Fig. 10a). The first substage ended in 23.9–23.3 Ma (latest Oligocene) when the subsidence decreased significantly in the study area except in the Liwan Sag and the Panyu Low Uplift. A NE–SW running belt of low uplift appeared in the center of the study area (Fig. 7h). In contrast a rapid subsidence occurred in the Panyu Low Uplift to the north (Fig. 7h), which caused the rapid sedimentation in the Panyu Low Uplift in this period (Xie et al., 2013). However it needs to be checked if the subsidence rate at the P3 point in the Panyu Low Uplift may be as high as nearly 600 m/My (Fig. 10c). The second substage lasted from 23.3 Ma (SB23.3) to 17.2 Ma (SB17.5) in the Early Miocene and is characterized by rapidly changing subsidence from strong to weak in the Baiyun and Liwan Sags. This substage started with the most rapid subsidence in 23.3 Ma (SB23.3) to 19.8 Ma (SB21) with the maximum subsidence rate N300 m/My in the southern Baiyun Sag and northwestern Liwan Sag (Figs. 7i and 8c), which caused a dramatic change in paleo-topography. The line of shelf break, traced by sediment progradation identified on seismic lines,

190

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

was running south of the Baiyun Sag in a SW–NE strike prior to 23.3 Ma (SB23.3), but shifted to north of the Baiyun Sag and rotated clockwisely to a SWW–NEE strike after 19.8 Ma (SB21). This indicates that a rapid deepening of paleo-water depth occurred in the time from 23.3 Ma to 19.8 Ma and in the area between the earlier and later shelf breaks (especially in the green area in Fig. 7i). Our analysis indicates that this rapid deepening of paleo-water depth has resulted from a rapid tectonic subsidence in this restricted area, not from sedimentary loading because the sedimentation rate in this period was low (Xie et al., 2013) and the loading effect was removed during backstripping. Between 19.8 Ma (SB21) and 17.2 Ma (SB17.5) the study area experienced continued subsidence. The subsidence was larger in the Liwan Sag than in the Baiyun Sag during this period (Fig. 7j). Then a rapid uplift occurred from 17.9 Ma (MFS18.5) to 17.2 Ma (SB17.5), which terminated the continued sea level rise in the area and made the 17.9 Ma boundary a maximum flooding surface (Figs. 7k and 8c). In this transient uplift event the Liwan Sag, the central Baiyun Sag, and especially the southern study area uplifted. In contrast again, a moderate subsidence occurred in the Panyu Low Uplift to the north. The third substage lasted from 17.2 Ma (SB17.5) to 11.9 Ma (SB10.5), the period from later Early Miocene to entire Middle Miocene. It is

characterized by the subsidence centered along the shelf break that struck SW–NE across the Baiyun Sag (Fig. 7l–p). This substage started with moderate subsidence rate, then a period of low subsidence rate from 16.3 Ma to 15.5 Ma, and then a gradually increasing subsidence rate until 11.9 Ma (SB10.5), with the maximum subsidence rate up to 400 m/My in the northern Baiyun Sag along the shelf break in a NE–SW orientation (Figs. 7p, 8d). The fourth substage from 11.9 Ma (Late Miocene) to the present was characterized by very slow subsidence rate (generally b25 m/My) (Figs. 7q–r, 8d). The study area was basically in a sedimentary hunger (Fig. 10f). 4.2. Spatial variations of Cenozoic subsidence The Cenozoic subsidence in the study area varied significantly also in spatial domain. The subsidence was distinct in deepwater sags with respect to the surrounding uplifts. In deepwater sags the total Cenozoic tectonic subsidence is very large, up to 6000 m in the Baiyun Sag (Fig. 9a) and up to 5000 m in the Liwan Sag (Fig. 9b). The surrounding uplifts (the Panyun Low Uplift, Dongsha Uplift and the Southern Uplift) are characterized by much smaller tectonic subsidence. The total

Fig. 7. Maps of tectonic subsidence rate for the sequences above Tg in the study area. Fine purple lines indicate sag boundaries. The red line AB shows the location of the profile in Figs. 3 and 7. Symbols of structural units are explained in Fig. 1.

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

191

Fig. 7 (continued).

Cenozoic tectonic subsidence in each of these uplifts was no more than 2000 m (Fig. 9c, d, e). The discrepancy in total subsidence between sags and uplifts was generated in the pre-30 Ma stage. The curves of tectonic subsidence reveal that the pre-30 Ma tectonic subsidence was large in the Baiyun and Liwan Sags but very small or near zero in surrounding uplift areas (Fig. 9f). In the post-30 Ma stage subsidence occurred in both sags and uplifted areas, with the subsidence rate in the sag centers (90–110 m/My in average) only less than double higher than that in surrounding uplifts (50–60 m/My in average) (Fig. 10a–e). The subsidence in the Baiyun and Liwan Sags was also different. As after 11.9 Ma (SB10.5) the subsidence was very small in the entire study area, we are to discuss the subsidence from 66 Ma to 11.9 Ma. In this period the subsidence rate was relatively stable in the Baiyun Sag with ~ 100 m/My in average, but more fluctuated in the Liwan Sag with an average rate of ~ 80 m/My (Figs. 9a, b and 10a, b). In general the total subsidence and subsidence rate in the Baiyun Sag were larger than those in the Liwan Sag. This difference is mostly due to the differentiated subsidence in the pre-30 Ma stage. In the early post-30 Ma stage from 30 Ma to 23.3 Ma the area of the fastest tectonic subsidence shifted frequently between the two sags like a teeterboard or seesaw. The Baiyun Sag led the subsidence in most

time, but the subsidence rate in the Liwan Sag exceeded that in the Baiyun Sag in periods of 28–26.2 Ma, 25.4–23.9 Ma, and 19.8–17.9 Ma (Figs. 7d, e, g, j and 10a, b). The teeterboard-like subsidence was also observed between the Panyu Low Uplift in the north with respect to other areas in the central and south of the study area. The contour maps and histograms of subsidence rates (Figs. 7 and 10) show that the Baiyun and Liwan Sags and other areas experienced two significant uplift in two periods of 23.9– 23.3 Ma and 17.9–17.2 Ma, which was exactly the period of rapid subsidence in the Panyu Low Uplift. These were also the two most rapid sedimentation periods in the study area and the only two periods that the rapid sedimentation extended northwards to the Panyu Low Uplift (Xie et al., 2013). 5. Tectonic implications 5.1. The date of breakup unconformity The breakup unconformity, as indicated literally, is the unconformity generated by the lithosphere breakup as a consequence of stress release. Thus the breakup unconformity is generally considered as the

192

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

Fig. 7 (continued).

unconformity between syn-rift and post-rift sequences (Falvey, 1974), and is synchronous with the onset of seafloor spreading. As discussed in the previous section, the T7 seismic interface has been recognized as the breakup unconformity based on structural style identified in numerous industrial seismic profiles in the Pearl River Mouth Basin. Because the fossil dating is poor for the continental Paleogene in the basin, the T7 interface has been dated as 30 Ma (Early Oligocene) mainly according to the age of the magnetic anomaly C11 in the South China Sea (Chen et al., 2003; Gong and Li, 1997). After the ODP Leg 184, younger date for the breakup unconformity in the northern South China Sea margin was proposed. While recognizing that the seafloor spreading started at ~30 Ma, Clift and Lin (2001) proposed that the extension in the Pearl River Mouth Basin continued ~ 5 My after the onset of seafloor spreading. That means the age of the cessation of rifting is ~25 Ma (Late Oligocene). However, in Clift et al. (2001) the active extension was said to be completed by 28 Ma based on the report of Shipboard Scientific Party (2000). Pertaining reference in the shipboard report is found only on page 11 of Chapter 9, where the authors proposed that “the Unit VI/Unit V boundary is equivalent to the breakup unconformity of Falvey (1974)”, because (above the top of Unit VI) “no significant normal faulting is noted”, thus “the top of Unit VI may

be interpreted as marking the end of active extension”. According to the biostratigraphic age–depth plot in the same report, the Unit VI/Unit V boundary lies at the hiatus dated from 28.5 Ma to 23.5 Ma. From these discussions we may see that the proposed date of end-of-rifting, either 25 Ma or 28 Ma, is in fact poorly constrained and evidenced. The top of Unit VI is not “the end of active extension”, as significant post-rifting subsidence has been observed above the top of Unit VI (see Section 5.3 for discussions), part of which were caused by extension. At ODP Site 1148, four hiatus are observed in the interval between 28.5 Ma to 23.5 Ma, among which the later two hiatus lie below and above the slumped interval (Li et al., 2006). Thus these hiatus might possibly be caused by local event such as washout and slumping at continental slope. The subsidence pattern revealed in this study supports T7 as the breakup unconformity, which is dated 30 Ma according to the age of the C11 magnetic anomaly. As discussed in Section 4.2, in the pre-30 Ma stage subsidence was restricted in sags only, while the subsidence rate was near zero in surrounding uplifts. But after 30 Ma subsidence extended to uplifts with rate only slightly smaller than that in the sags (Fig. 9f). These features agree with the classical model of extensional basin that in syn-rift stage subsidence occurs in discrete rifts while rift

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

193

Fig. 8. Curves of tectonic subsidence rate for the 18 sequences along the profile AB. For location of AB see Fig. 1.

shoulders are stable or uplifted, and in post-rift stage thermal subsidence spreads over the entire area (Braun and Beaumont, 1989). This indicates that the unconformity at 30 Ma that separated syn-rift and post-rift sequences is the breakup unconformity of Falvey (1974). Our results do not support the proposal of a breakup unconformity later than 30 Ma (e.g., at 28 Ma, 25 Ma, or 23.3 Ma). Although in the study area the 23.3 Ma unconformity is widely observed, no clear contrast between syn-rift and post-rift structures is observed across the 23.3 Ma unconformity.

5.2. The Baiyun Event The event that formed the 23.3 Ma unconformity was named as the Baiyun Event by Pang et al. (2009, originally dated as 23.8 Ma), which was expressed not only by a widespread unconformity, but also by an abrupt landward retreat and clockwise rotation of paleo-shelf break, a rapid change in Nd isotope and many other geochemical indexes (Li et al., 2003; Pang et al., 2009). Our study showed that the Baiyun Event was also expressed as local rapid uplift in 23.9–23.3 Ma followed by local rapid subsidence in 23.3– 19.8 Ma. In 23.9–23.3 Ma a SW–NE running uplift belt appeared in the study area while the Panyu Low Uplift in the north and the southeastern Liwan Sag in the south were subsiding (Figs. 7h, 10a–c). In contrast, from 23.3 Ma to 19.8 Ma the central portion of the study area turned to subsidence at a very high rate (exceeded 300 m/My in the center, see Fig. 7i). This sudden change in subsidence pattern was the most striking one in the Cenozoic tectonic evolution of the study area.

The origin of the Baiyun Event is not clear so far. It was correlated to southward ridge jump in the South China Sea (Pang et al., 2009; Sun et al., 2008). The ridge jump occurred at 7–6b magnetic anomalies (Briais et al., 1993), which are 25 Ma to 23 Ma according to the magnetic time scale of Cande and Kent (1995), and thus ended at the Baiyun Event. However, the revised correlation of magnetic anomalies by Barckhausen and Roeser (2004) suggests the ridge jump at 25 Ma. The relation between the ridge jump and the Baiyun Event has not yet been clarified. The newly discovered localized rapid variations in subsidence rate seem to be indicative of magmatic intrusion activity in the area. Through gravimetric and thermal modeling Shi et al. (2005) showed that a dense intrusion, for example basaltic dikes, into crust could cause surface uplift followed immediately by rapid subsidence, and the subsidence might last for no more than 5 My and then slow down. Indeed, the occurrence of igneous intrusions or extrusions in the abovementioned restricted area was interpreted from a seismic profile (e.g. Fig. 5 in Sun et al., 2008). However the size of interpreted intrusions is not as large as that required by Shi's model (one third of a 90 km wide belt). Other authors proposed that the Baiyun Event has widespread influences (Shao et al., 2004; Pang et al., 2009), and suspected that the Baiyun Event might be the most significant Cenozoic tectonic event that caused the topographic reversal in China (Wang, 2005). There might be other high-order agents that caused this event. 5.3. Anomalous post-rift subsidence The classical stretching model of extensional basins includes a rapid or finite initial syn-rift subsidence followed by an exponentially

194

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

Fig. 9. Curves of tectonic subsidence for (a) the central Baiyun Sag (P1), (b) the central Liwan Sag (P2), (c) the Panyu Low Uplift (P3), (d) the Dongsha Uplift (P4), (e) Southern Uplift (P5), and (f) all the five points (P1–5). For location of the points see Fig. 1.

decaying post-rift thermal subsidence (Jarvis and McKenzie, 1980; McKenzie, 1978). However, it has been observed in many basins worldwide that the post-rift subsidence often excess the exponentially decay law (Kusznir and Karner, 2007; Morley and Westaway, 2006). The excess portion of the subsidence is called as “anomalous post-rift subsidence” and has been discovered in deepwater sags as well as shelf areas in the Pearl River Mouth Basin (Clift and Lin, 2001; Liao et al., 2011; Shipboard Scientific Party, 2000; Zhao et al., 2010) and in other basins of the northern South China Sea (Cui et al., 2008; Xie et al., 2006). Our study based on newly defined sequence boundaries and using localized backstripping parameters confirmed the existence of significant anomalous post-rift subsidence. In Figs. 11a and 11b we compared the backstripped subsidence curve and rate at the point P1 in central Baiyun Sag with the theoretical post-rifting thermal subsidence calculated using the formulation of McKenzie (1978). The stretching factor of 3.5 (Liao et al., 2011) and initial crust and lithosphere thickness of 32 km and 95 km, respectively are used together with other parameters in McKenzie (1978) in the calculation, because with these parameters the calculated initial (syn-rift) subsidence is close to the observed one. Results show a total of ~ 1200 m of anomalous post-rift subsidence

from 30 Ma to the present, which is N40% of the total post-rift subsidence observed (Fig. 11a). Because of the anomalous post-rift subsidence, the average post-rift subsidence rate is similar to the syn-rift subsidence rate in the Baiyun Sag and even higher than the syn-rift subsidence rate in the Liwan Sag (Figs. 9a, b and 10a, b).

5.4. Style of post-rift basin deformation The histograms in Fig. 10 reveal that the subsidence rate in the study area was not declining exponentially in the post-rift stage as predicted by classical thermal contraction model for the post-rift subsidence (McKenzie, 1978). The post-rift subsidence was featured by three cycles of variable subsidence rate as discussed in Section 4.1.2. The cycle period was about 6–7 My. Two rapid subsidence events are observed in periods of 23.3–19.8 Ma and 14.3–11.9 Ma, while two uplift events in periods of 23.9–23.3 Ma and 17.9–17.2 Ma. The episodic fluctuations of subsidence rate with time indicate that in addition to thermal contraction there were other episodic agents that generated the anomalous post-rift subsidence.

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

195

Fig. 10. Histograms of tectonic subsidence for (a) the central Baiyun Sag (P1), (b) the central Liwan Sag (P2), (c) the Panyu Low Uplift (P3), (d) the Dongsha Uplift (P4), and (e) the Southern Uplift (P5), and (f) all the five points (P1–5). The numbers within the histograms are the ages (Ma) at the bases of the boxes. For location of the points see Fig. 1.

Another feature of the post-rift subsidence in the study area is the teeterboard-like subsidence in the period approximately from 30 Ma to 16.3 Ma. A teeterboard is a board above a central pivot so that when one end is going down the other end will go up. As discussed in Section 4.2 and shown in Figs. 7 and 10, this includes two orders of teeterboard: One consists of the Baiyun Sag and the Liwan Sag as two ends with ~100 km wavelength (the distance between the centers of the two sags). Another higher order teeterboard consists of the Baiyun and Liwan Sags at one end and the Panyu Low Uplift at the other end, with a wavelength of ~300 km or longer (two times of the distance between centers of sags and uplift). The episodic and teeterboard-like subsidence in the study area looks like the deformation of a weak elastic board over a liquid bed: when one part is pressed downward, the other part will go up. Clift et al. (2002b) suggested the low viscosities of 1018 to 1019 Pa s for the shelf area of the central and eastern South China margin based on basement topography in the shelf area. The spatial and temporal variations of subsidence patterns revealed in our study may place constrain for resolving the

lithospheric flexural rigidity of the slope area of the northern South China Sea. 6. Conclusions The paper provided for the first time the maps of tectonic subsidence rate for 18 Cenozoic sequences in the 44,500 km2 deepwater areas of the Pearl River Mouth Basin in the northern South China Sea. The quality of these backstripping results was controlled not only by using newest data set from systematic sequence stratigraphic studies, but also by carefully selecting parameters. Because the paleo-water depth and sedimentary lithology vary significantly in the deepwater area, we estimated the paleo-water depth for all sequences based on wells and sedimentary system maps; we also estimated local porosity parameters. The computer program was revised to accommodate lithological and bathymetrical parameters that vary with grid nodes. Because rate depends on ages, we adjusted the ages of sequence boundaries according to the newest International Chronostratigraphic Chart v2013/01.

196

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

Fig. 11. Comparison of observed and theoretical post-rift subsidence (a) and subsidence rate (b) in central Baiyun Sag (point P1 in Fig. 1).

Thus the reliability of the estimated tectonic subsidence has been significantly improved. Our results show that before 30 Ma (T7 boundary) the tectonic subsidence was restricted in the Baiyun and Liwan Sags, but after 30 Ma the subsidence extended over the entire area including the uplifts. This indicates that the 30 Ma unconformity (T7) in late Early Oligocene is the breakup unconformity that separates syn-rift and post-rift sequences. The average subsidence rate in the syn-rift stage is about the same as in the post-rift stage in the Baiyun and Liwan Sags. Four substages of post-rift subsidence were identified by cyclic changes in the rate and pattern of tectonic subsidence at boundaries of 23.3 Ma, 17.2 Ma and 11.9 Ma, respectively. Compared with the theoretical post-rift thermal subsidence which has an exponentially decaying subsidence rate, the observed post-rift subsidence is much higher. The portion of observed total post-rift subsidence that exceeds the theoretical subsidence, which is called as the anomalous post-rift subsidence, was totaled ~1200 m in the center of the Baiyun Sag. The Baiyun Event at 23.3 Ma near the Oligocene/Miocene boundary had a clear expression in the tectonic subsidence as a localized rapid uplift in 23.9–23.3 Ma followed by a localized rapid subsidence in 23.3– 19.8 Ma. This might be caused by a dense and hot intrusive activity in the area between the central Baiyun Sag and the central Liwan Sag, but more works are needed to verify this speculation. Two orders of teeterboard-like subsidence were observed in the study area. One consists of the Baiyun and Liwan Sags as the two ends that experienced faster tectonic subsidence alternatively, forming a teeterboard of shorter-wavelength (~ 100 km). A longer wavelength (300 km or longer) teeterboard consists of the Baiyun and Liwan Sags at one end and the Panyu Low Uplift at the other end. The rapid subsidence occurred in the Panyu low Uplift only when the Baiyun and Liwan Sags were experiencing rapid uplift. This phenomenon might be an indication of a deformation of a low-strength elastic board over a soft ductile bed. More works are needed to explain the highly variable spatial and temporal pattern of tectonic subsidence in the deepwater area of the Pearl River Mouth Basin, especially the origin and tectonic implications

of the abnormal, episodic, and teeterboard-like post-rift subsidence observed in the study area. Acknowledgments The study was funded by the High-Tech R&D Program of China (No. 2008AA093001), National Science and Technology Key Project of the Ministry of Science and Technology of China (No. 2011ZX05025003-005), and Chinese National Science Foundation project (No. 40976033). We thank CNOOC Shenzhen Branch for providing valuable new data for this study. We also thank two anonymous reviewers whose valuable comments helped in improving the paper. Appendix A The adjustment of boundary ages of Haq et al. (1987) according to the International Chronostratigraphic Chart v2013/01 (Cohen et al., 2013) was performed. The basic idea of the correction is to replace the old ages of stratigraphic boundaries with new ages, and then to find the ages of sequence boundaries and downlap surfaces via interpolation. As the correlation between stratigraphic boundaries and magnetic chrons varies slightly between the International Chronostratigraphic Chart issued in different years, we adjusted the ages of old stratigraphic boundaries according to magnetic chrons. Because the International Chronostratigraphic Chart v2013/01 does not provide magnetic chrons, we use the magnetic chrons in the 2009 International Chronostratigraphic Chart. Thus the correction was proceeded in the following steps: 1) In Haq et al. (1987), we can find the ages and corresponding magnetic chrons of the stratigraphic boundaries. These ages and chrons are listed in columns 2 and 3, respectively in Table A1. 2) In the 2009 Chronostratigraphic Chart, we can find the ages and corresponding magnetic chrons of the stratigraphic boundaries. These ages and chrons are listed respectively in columns 5 and 6 in

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

197

Table A1 Adjustment of ages of stratigraphic boundaries in Haq et al. (1987). Bold numbers are the boundary ages with N2 My difference after the adjustment. 1

2

3

4

5

6

7

Boundary (base)

Age (Haq et al., 1987)

Haq magnetic chron.a

New age of Haq chron

Age (2009)

2009 magnet. chron.

Age (2012)

Pleistocene Pliocene U. Miocene M. Miocene L. Miocene U. Oligocene L. Oligocene U. Eocene M. Eocene L. Eocene U. Paleocene L. Paleocene

1.65 5.2 10.2 16.2 25.2 30 36 39.4 49 54 60.2 66.5

C1 C3, 9/10 C5, 1/2 C5B C6C, 2/5 C10, 1/5 C13, 1/2 C17, 1/2 C21 C24, 2/3 C26, 1/4 C29

1.8 5.4 11.6 16 23.1 28.0 33.8 38.0 47.8 56.0 59.3 66.0

2.6 5.3 11.6 16.0 23.0 28.4 33.9 37.2 48.6 55.8 61.7 65.5

C2 C3, 1/2 C5, 2/3 C5B C6C, 1/4 C10, 1/2 C13, 2/5 C17, 1/2 C21 C24, 2/3 C27, 1/3 C29, 4/5

2.59 5.33 11.62 15.97 23.03 28.1 33.9 38.0 47.8 56.0 61.6 66.0

a

The fraction number shows the location of the boundary from the base of the magnetic chron.

Table A1. Note that the chrons of the stratigraphic boundaries in the 2009 chart are not all the same as those in the 1987 chart. 3) Assuming that the ages of the magnetic chrons are the same for the 2009 and 2013 charts, we replace the ages in 2009 chart (column 5) with the ages in 2012 chart (column 7). 4) Find the ages of Haq's magnetic chrons (column 3) by interpolation on column 7. These are the new ages of the stratigraphic boundaries of Haq et al. (1987), which are listed in column 4 in Table A1. The numbers in bold are the new ages with larger (N2 m.y.) difference from the old ages. 5) Using the new ages of the stratigraphic boundaries (column 4 in Table A1) as control points, we find the new ages of all the sequence tracts in Haq's curve via graphic interpolation. The results of corrected boundary ages of Haq et al. (1987) are shown in Fig. 2 in the main text. The comparison of the CNOOC boundary ages and the corrected boundary ages are shown in Table 1 of the main text. References Barckhausen, U., Roeser, H.A., 2004. Seafloor spreading anomalies in South China Sea revisited. In: Clift, P., P.W., Kuhnt, W., Hayes, D. (Eds.), Continent–Ocean Interactions Within East Asian Marginal Seas. American Geophysical Union, Washington, DC, pp. 121–125. Braun, J., Beaumont, C., 1989. A physical explanation of the relation between flank uplifts and the breakup unconformity at rifted continental margins. Geology 17, 760–764. Briais, A., Patriat, P., Tapponnier, P., 1993. Updated interpretation of magnetic anomalies and seafloor spreading stages in the South China Sea: implication for the Tertiary tectonics of Southeast Asia. J. Geophys. Res. 98, 6299–6328. Cande, S.C., Kent, D.V., 1995. Revised calibration of the geomagnetic polarity timescale for the Late Cretaceous and Cenozoic. J. Geophys. Res. 100, 6093–6095. Chen, C., Shi, H., Xu, S., Chen, X., 2003. Formation Conditions of Tertiary Oil/Gas Reservoirs in Pearl River Mouth Basin (East)Science Press, Beijing. Chen, H.-Z., Wu, X., Zhou, D., Wang, W.-Y., Hao, H., 2005. Meso-Cenozoic faults in Zhujiang River Mouth Basin and their geodynamic background. J. Trop. Oceanogr. 2, 52–61. Clift, P., Lin, J., 2001. Preferential mantle lithospheric extension under the South China margin. Mar. Pet. Geol. 18, 929–945. Clift, P.D., Lin, J., ODP Leg 184 Scientific Party, 2001. Patterns of extension and magmatism along the continent–ocean boundary, South China margin. In: Wilson, R.C., Whitmarsh, R.B., Taylor, B., Froitzheim, N. (Eds.), Non-Volcanic Rifting of Continental Margins: A Comparison of Evidence From Land and Sea. Geological Society, London, Special Publication, 187, pp. 489–510. Clift, P., Lee, J.I., Clark, M.K., Blusztajn, J., 2002a. Erosional response of South China to arc rifting and monsoonal strengthening; a record from the South China Sea. Mar. Geol. 184, 207–226. Clift, P., Lin, J., Barckhausen, U., 2002b. Evidence of low flexural rigidity and low viscosity lower continental crust during continental break-up in the South China Sea. Mar. Pet. Geol. 19, 951–970. Cohen, K.M., Finney, S., Gibbard, P.L., 2013. International Chronostratigraphic Chart v2013/01International Commission on Stratigraphy. Cui, T., Xie, X., Jianye, R., Zhang, C., 2008. Dynamic mechanism of anomalous post-rift subsidence in the Yinggehai Basin. Earth Sci. J. China Univ. Geosci. 33, 349–357. Falvey, D.A., 1974. The development of continental margins in plate tectonic theory. Journal of Australian Petroleum Exploration Association 14, 95–106.

Flower, M., Tamaki, K., Hoang, N., 1998. Mantle extrusion: a model for dispersed volcanism and DUPAL-like asthenosphere in East Asia and the western Pacific. Geodyn. Ser. 27, 67–88. Gong, Z., Li, S., 1997. Basin Analysis and Oil/Gas Accumulation in the Northern Margin of South China SeaScience in China Press, Beijing. Haq, B.U., Hardenbol, J., Vail, P.R., 1987. Chronology of fluctuating sea levels since the Triassic. Science 235, 1156–1167. Holloway, N., 1982. North Palawan Block, Philippines; its relation to Asian mainland and role in evolution of South China Sea. AAPG Bull. 66, 1355–1383. Hsu, S.-K., Yeh, Y.-C., Doo, W., Tsai, C., 2004. New bathymetry and magnetic lineations identifications in the northernmost South China Sea and their tectonic implications. Mar. Geophys. Res. 25, 29–44. Hu, D., Zhou, D., Wu, X., He, M., Pang, X., Wang, Y., 2009. Crustal structure and extension from slope to deepsea basin in the northern South China Sea. J. Earth Sci. 20, 27–37. Huang, C., Zhou, D., Sun, Z., Chen, C., Hao, H., 2005. Deep crustal structure of Baiyun Sag, northern South China Sea revealed from deep seismic reflection profile. Chin. Sci. Bull. 50, 1131–1138. Jarvis, G.T., McKenzie, D.P., 1980. Sedimentary basin formation with finite extension rates. Earth Planet. Sci. Lett. 48, 42–52. Kido, Y., Suyehiro, K., Kinoshita, H., 2001. Rifting to spreading process along the northern continental margin of the South China Sea. Mar. Geophys. Res. 1–15. Kusznir, N.J., Karner, G.D., 2007. Continental lithospheric thinning and breakup in response to upwelling divergent mantle flow: application to the Woodlark, Newfoundland and Iberia margins. In: Karner, G.D., Manatschal, G., Pinheiro, L.M. (Eds.), Imaging, Mapping and Modelling Continental Lithosphere Extension and Breakup. Geological Society, Special Publication, London, pp. 389–419. Li, X.-H., Wei, G., Shao, L., Liu, Y., Liang, X., Jian, Z., Sun, M., Wang, P., 2003. Geochemical and Nd isotopic variations in sediments of the South China Sea: a response to Cenozoic tectonism in SE Asia. Earth Planet. Sci. Lett. 6664, 1–14. Li, Q., Jian, Z., Su, X., 2005. Late Oligocene rapid transformations in the South China Sea. Mar. Micropaleontol. 54, 5–25. Li, Q., Jian, Z., Li, B., 2006. Oigocene–Miocene planktonic foraminifer biostratigraphy, Site 1148, northern South China Sea. In: Prell, W.L., Wang, P., Blum, P., Rea, D.K., Clemens, S.C. (Eds.), Proceedings of the Ocean Drilling Program, Scientific Results, vol. 184, pp. 1–26. Liao, J., Zhou, D., Zhao, Z., Zhang, Y., Xu, Z., 2011. Numerical modeling of the anomalous post-rift subsidence in the Baiyun Sag, Pearl River Mouth Basin. Sci. China Earth Sci. 54, 1156–1167. Mckenzie, D., 1978. Some remarks on the development of sedimentary basins. Earth Planet. Sci. Lett. 40, 25–32. Morley, C.K., Westaway, R., 2006. Subsidence in the super-deep Pattani and Malay basins of Southeast Asia: a coupled model incorporating lower-crustal flow in response to post-rift sediment loading. Basin Res. 18, 51–84. Nissen, S.S., Hayes, D.E., Buhl, P., Diebold, J., Bochu, Y., Weijun, Z., Yongqin, C., 1995. Deep penetration seismic soundings across the northern margin of the South China Sea. J. Geophys. Res. 100, 22407–22422 (22433). Pang, X., Yang, S., Zhu, M., Li, Q., 2004. Deep-water fan systems and petroleum resources on the northern slope of the South China Sea. Acta Geol. Sin. 78 (626623-631). Pang, X., Chen, C., Shi, H., Shu, Y., Shao, L., He, M., Shen, J., 2005. Response between relative sea-level change and the Pearl River deep water fan system in the South China Sea. Earth Sci. Front. 3, 167–177. Pang, X., Shen, J., Yuan, L., Lian, S., He, M., Shu, Y., 2006. Petroleum prospect in deep-water fan system of the Pearl River in the South China Sea (in Chinese with English abstract). Acta Petrol. Sin. 3, 11–16. Pang, X., Chen, C., Peng, D., Zhou, D., Chen, H., 2007a. The Pearl River Deep-water Fan System & Petroleum in South China SeaScience Press, Beijing. Pang, X., Chen, C., Peng, D., Zhu, M., Shu, Y., He, M., Shen, J., Liu, B., 2007b. Sequence stratigraphy of deep-water fan system of Pearl River, South China Sea. Earth Sci. Front. 14, 220–229. Pang, X., Chen, C., Zhu, M., He, M., Shen, J., Lian, S., Wu, X., Shao, L., 2009. Baiyun movement: a significant tectonic event on Oligocene/Miocene boundary in the northern South China Sea and its regional implications. J. Earth Sci. 20, 49–56.

198

H. Xie et al. / Tectonophysics 615–616 (2014) 182–198

Patriat, P., 1987. L'evolution du systeme de dorsales de l'Ocean Indien, Terres Australes et Antartique Francaises, Paris. Ru, K., Pigott, J.D., 1986. Episodic rifting and subsidence in the South China Sea. AAPG Bull. 70, 1136–1155. Sclater, J.G., Christie, P.A.F., 1980. Continental stretching: an explanation of the post-MidCretaceous subsidence of the Central North Sea Basin. J. Geophys. Res. 85, 3711–3739. Shao, L., Li, X., Wang, P., Jian, Z., Wei, G., Pang, X., Liu, Y., 2004. Sedimentary record of the tectonic evolution of the South China Sea since the Oligocene—Evidence from deep sea sediments of ODP Site 1148. Advances in Earth Science 19, 539–544 (In Chinese). Shi, X., Burov, E., Leroy, S., Qiu, X., Xia, B., 2005. Intrusion and its implication for subsidence: a case from the Baiyun Sag, on the northern margin of the South China Sea. Tectonophysics 407, 117–134. Shipboard Scientific Party, 2000. Proceedings of the Ocean Drilling Program, Initial Reports, vol. 184. Ocean Drilling Program, College Station, TX. Stam, B., Gradstein, F.M., Lloyd, P., Gillis, D., 1987. Algorithms for porosity and subsidence history. Comput. Geosci. 13, 317–349. Steckler, M.S., Watts, A.B., 1978. Subsidence of the Atlantic-type continental margin off New York. Earth Planet. Sci. Lett. 41, 1–13. Sun, Z., Zhong, Z., Zhou, D., Pang, X., Huang, C., Chen, C., He, M., Xu, H.-H., 2008. Dynamics analysis of the Baiyun Sag in the Pearl River Mouth Basin, north of the South China Sea. Acta Geol. Sin. 82, 73–83. Tapponnier, P., Peltzer, G., Le Dain, A., Armijo, R., Cobbold, P., 1982. Propagating extrusion tectonics in Asia: new insights from simple experiments with plasticine. Geology 10, 611–616. Tapponnier, P., Peltzer, G., Armijo, R., 1986. On the mechanics of the collision between India and Asia. Geol. Soc. Lond. Spec. Publ. 19, 113–157. Taylor, B., Hayes, D.E., 1983. Origin and history of the South China Sea basin. Geophys. Monogr. Ser. 27, 23–56. Wang, P.-X., 2005. Cenozoic deformation and history of sea-land interactions in Asia. Earth Science – Journal of China University of Geosciences 30, 1–18 (in Chinese with English abstract). Wu, X., Zhou, D., Pang, X., Hao, H., Shen, J., 2005. Geophysical field and deep structures of Baiyun Sag, Zhujiang River Mouth Basin. J. Trop. Oceanogr. 2, 62–69.

Xie, X., Müller, R.D., Li, S., Gong, Z., Steinberger, B., 2006. Origin of anomalous subsidence along the Northern South China Sea margin and its relationship to dynamic topography. Mar. Pet. Geol. 23, 745–765. Xie, H., Zhou, D., Pang, X., Li, Y., Wu, X., Qiu, N., Li, P., Chen, G., 2013. Cenozoic sedimentary evolution of deepwater sags in the Pearl River Mouth Basin, northern South China Sea. Mar. Geophys. Res. 34, 159–173. Xu, S., Yang, S., Huang, L., 1995. The application of sequence stratigraphy to stratigraphic correlation. Earth Sci. Front. 2, 115–123. Zhao, Q., 2005. Late Cainozoic ostracod faunas and paleoenvironmental changes at ODP Site 1148, South China Sea. Mar. Micropaleontol. 54, 27–47. Zhao, Z., Zhou, D., Liao, J., He, M., Guo, X., Zhang, Y., Xu, Z., 2010. Lithospheric stretching modeling of the continental shelf in the Pearl River Mouth Basin and analysis of post-breakup subsidence. Acta Geol. Sin. 84, 1135–1145. Zhao, Z., Sun, Z., Xie, H., Yan, C., Li, Y., 2011. Baiyun deepwater cenozoic subsidence and lithospheric stretching deformation (in Chinese). Chin. J. Geophys. 54, 3336–3343. Zhou, D., Ru, K., Chen, H.-Z., 1995. Kinematics of Cenozoic extension on the South China Sea continental margin and its implications for the tectonic evolution of the region. Tectonophysics 251, 161–177. Zhou, D., Sun, Z., Chen, H.-Z., Xu, H.-H., Wang, W.-Y., Pang, X., Cai, D.-S., Hu, D.-K., 2008a. Mesozoic paleogeography and tectonic evolution of South China Sea and adjacent areas in the context of Tethyan and Paleo-Pacific interconnections. Island Arc 17, 186–207. Zhou, D., Hu, D.-K., He, M., Lian, S.Y., 2008b. The selection of fitting curve in time-depth transformation of deep-seated strata and crust. Earth Science Journal of China University of Geosciences 33, 531–537 (in Chinese with English abstract). Zhou, D., Sun, Z., Liao, J., Zhao, Z., He, M., Wu, X., Pang, X., 2009. Filling history and postbreakup acceleration of sedimentation in Baiyun Sag, deepwater northern South China Sea. J. Earth Sci. 20, 160–171. Zhu, J., Shi, H., He, M., Pang, X., Yang, S., Li, Z., 2008. Origins and geochemical characteristics of gases in LW3-1-1 well in the deep sea region of Baiyun Sag, Pearl River Mouth Basin. Nat. Gas Geosci. 2, 229–233. Zhu, J., Shi, H., Pang, X., He, M., Yang, S., Li, Z., 2010. Oil filling phases and timing analyze based on fluid inclusion in LW3-1-1 well, Baiyun Sag. Explor. Tech. 1, 52–56.