1. Introduction
The concept of the Milankovitch cycles was put forward on account of global climate change, and they are caused by periodical variations of Sun-Earth orbital parameters including eccentricity, obliquity (ecliptic and equator), and precession (Lever H, 2004; Dobrovolskis AR, 2013; Laya JC et al., 2013; Meyers SR, 2019). Milankovitch cycle is an astronomical orbital force with global directional distribution and isochronism. Climate change and glacio-eustatic sea-level fluctuations caused by Milankovitch cycles drive the formation of high-frequency sequences of 4th−6th order (Table 1; Li HC et al., 2004; Berger WH, 2013; Elderbak K et al., 2014; Marra F et al., 2014; Laurin J et al., 2019). These changes are recorded in the sedimentary strata and can be reflected by the changes of sedimentary structure, lithology, and lithofacies as well as corresponding physical parameters and log curves (Li QM et al., 1996; Eriksson PG et al., 2013; Liu T et al., 2013). Milankovitch cycles serve as a criterion of sequence division free from human factors.
There are many methods for stratigraphic dating, among which paleontology and isotopic dating are commonly used (Guo HX et al., 2015; Qiao PJ et al., 2015). However, these two methods have shortcomings such as relatively low resolution, discontinuity in vertical direction, and possible sedimentary discontinuities between sampling points. Therefore, the sedimentation rates calculated by these two methods are overall less than or equal to actual sedimentation rates (Zhang ZS et al., 1999). Currently, stratigraphic dating based on Milankovitch cycles has specifically become the only useful tool that provides continuous and high-resolution geologic time scales (Hinnov LA and Hilgen FJ, 2012; Li MS et al., 2017), and has been revolutionarily applied in the Earth and planetary sciences (Hinnov LA, 2013). Well log curves have high resolution and continuity in vertical direction and keep clear records of paleoclimate change, which is the foundation for identifying Milankovitch cycles. The information used to identify Milankovitch cycles mainly includes natural gamma-ray (GR) and spontaneous potential (SP) data sensitive to depositional environment. They can effectively reflect changes in shale content (Yuan X et al., 2013; Zhang R et al., 2019), which is directly related to climate change. Therefore, GR and SP log data can be used as alternative indicators of climate change in the inversion of paleoclimate change. In this study, they were employed to analyze the Eocene sequences in the eastern depression of the North Yellow Sea Basin (NYSB). Meanwhile, spectral analysis of GR and SP data was conducted to find out the consistency between the sedimentary cycle ratio and the Milankovitch cycle ratio in the frequency domain, thus providing a basis for the identification of Milankovitch cycles. Then the sedimentation duration and sedimentation rates of Eocene strata discovered by wells were calculated. Guided by tectonic movement and sequence stratigraphy theory, continuous wavelet transform (CWT) was employed to determine high-resolution subsequences in the Eocene strata, and spectral analysis of GR curves was carried out with an independent sedimentary sequence as a separate window. Accordingly, the sedimentation duration and sedimentation rate of each sedimentary stage were calculated separately, followed by the high-resolution analysis of the sedimentary environment. Moreover, this paper analyzed the relatively stable relationship between sedimentary cycles and astronomical cycles in the study area and discussed the changes in sedimentary cycles and factors controlling the evolution of high-frequency sequences.
2. Geological setting
NYSB is located in the northern part of the Yellow Sea, covering an area of about 2210 km2 (Fig. 1a). It is surrounded by Liaodong-Haiyang Island Uplift in the north and the Jiaobei-Liugong Island Uplift in the south. Meanwhile, it faces the Bohai Bay Basin across the estuary of Bohai Sea in the west and is adjacent to the Anzhou Basin in the east, with a tectonic strike in NE trending (Cai QZ, 2002; Chen L et al., 2008; Wang R et al., 2017a). It can be divided into four depressions and two uplifts from east to west and from north to south, namely eastern depression, eastern uplift, central depression, central-western uplift, western depression, and southern depression group (Fig. 1b). NYSB is a Mesozoic-Cenozoic intracontinental fault basin developing in the background of the uplifting of the eastern part of the Sino-Korean Plate (the eastern part of the North China Block) during the Indosinian-Yanshan movement (Killops SD et al., 1991a, 1991b; Massoud MS et al., 1993; Zhang Z et al., 2018). Its basement is mainly comprised of Archean‒Lower Proterozoic metamorphic rocks and Middle and Upper Proterozoic‒Cambrian carbonate rocks and slate (Tian SF et al., 2004; Liu JP et al., 2015). The basin begun to subside at the start of the Middle Jurassic, and thus its sedimentary sequences are composed of Middle-Upper Jurassic, Lower Cretaceous, Eocene, Oligocene, and Neogene‒Quaternary sediments (Fig. 2). Twenty-four wells have been drilled in the eastern depression since the 1960s, discovering industrial oil flow from Mesozoic and Cenozoic sandstone intervals. This indicates that the eastern depression has great potential for the formation of oil and gas (Massoud MS et al., 1993; Kim IS et al., 2006; Cao D et al., 2007; Liu JP et al., 2015).
NYSB has undergone several stages of tectonic activities and contains a multicycle structural-sedimentary sequence (Li NS, 1995; Feng ZQ et al., 2002; Lin CS et al., 2004). From the Late Jurassic to the Early Cretaceous, the Pacific and Izanagi plates subducted towards the Eurasian Plate in SW and NNW directions (Moore GW, 1990; Lapierre H et al., 1997; Zhu G et al., 2004). This, along with magmatism, affected the structures of the NYSB near the boundary of the Pacific and Eurasian plates (Wang R et al., 2017a, 2017b). In the Middle Paleogene, the subduction of the eastern Pacific Plate towards the Eurasian Plate was gradually transformed to NW from the original NNW, which changed the stress field in the NYSB. As a result, the depressions of NYSB remained transtensional. During the Eocene, extrusion action stopped, tensional faults became active, and lacustrine basins rapidly expanded, followed by the development of a thick sedimentary system of lacustrine facies (Wang YJ et al., 2007). During this period, reservoirs were well preserved, interbeds consisting of sandy shale and glutenite developed, and the strata had strong stratigraphic cycles and were sensitive to the change of climate in the NYSB.
3. Methods
Previous studies have proved that the geological records of Milankovitch cycles exist in deep-sea, lacustrine, and fluvial sediments (Hays JD et al., 1976). Only indirect methods can be used to determine the presence of Milankovitch cycles in strata in the case of unavailable accurate stratigraphic dating data. Assuming that the ratios among Milankovitch cycles are stable in a certain geological period, the presence of Milankovitch cycles can be confirmed by revealing that ratios among some various stratigraphic cycles are similar to the ratios among Milankovitch cycles (Li QM et al., 1996). This paper analyzed the dominant frequency of each peak in spectrum curves and calculated the thickness of stratigraphic cycles and the ratios among the thickness. If the thickness ratios of sedimentary cycles are similar or equal to the ratios among astronomical cycles in the strata within a certain depth range, it can be concluded that the paleoclimate change is preserved in the strata (Wang YJ et al., 2007).
Prior to the signal analysis, the log curves were processed to eliminate high-frequency noise and low-frequency background using the programs developed with Matlab. Then the signals generated were standardized by wavelet reconstruction, and then white noise was removed from the signals by the method of median filtering (Wu SY and Liu J, 2015). Spectral analysis was conducted for peak frequencies with confidence levels of 99%, 95%, and 90% using the program Redfits developed based on Matlab by Boris Priehs from the University of Bremen in Germany. The analytical results were plotted with Matlab.
3.1 Astronomical cycles
In this paper, the software Analyseries 2.0.8 was employed to calculate the theoretical values of insolation, eccentricity (E), obliquity (O), and precession (P) at 38°N (the eastern depression of NYSB) from 36.5 Ma to 53 Ma in the Eocene according to Laskar J et al. (2011). Then spectral analysis was carried out for the calculated values with a sampling interval of 1 kyr, obtaining four eccentricity cycles, namely 402 ka (E4), 125 ka (E3), 99 ka (E2), and 95 ka (E1), two obliquity cycles, namely 51 ka (O2) and 39 ka (O1), and three precession cycles, i.e., 23 ka (P3), 22 ka (P2), and 19 ka (P1) (Fig. 3). The stable proportional relationship among these theoretical orbital cycles can be used as a reference for the determination of astronomical cycles in this study.
Most researchers believe that the sedimentation occurred at a low rate and was stable in geological history, and thus the ratios among the sedimentary cycles are stable in a certain geological historical period. Therefore, strata can be considered to have been affected by Milankovitch cycles if the ratios among some cycles in the strata are equal to the ratios among Milankovitch cycles. Otherwise, the stratigraphic cyclicity may be due to changes irrelevant to Milankovitch cycles (such as the periodicity of diagenesis, frequencies of turbidity current, or other sedimentation phenomena; Brescia M et al., 1996). Different methods have been proposed to verify the presence of Milankovitch cycles. For instance, Bailey RJ (2001) considered that Milankovitch cycles are recorded in the study area if the strata at the same depth in adjacent areas have peaks with the same frequencies. In contrast, Lever H (2004) and Poletti L (2004) believed that it is necessary to analyze different alternative indicators of the strata at the same depth or the same alternative indicators of strata in different areas to improve the accuracy.
3.2 Analysis of spectral power
Firstly, the GR and SP log curves of Well NYS3 drilled in the eastern depression of the NYSB were preprocessed. One-dimensional discrete wavelet multi-scale decomposition was adopted to eliminate high-frequency noise and low-frequency background. Then wavelet reconstruction was employed to standardize the log curves and the method of median filtering was used to remove white noise. The sampling interval of every log curve was set at 0.125 m. For ease of comparison, continuous wavelet transform (CWT) and fast fourier transform (FFT) were carried out for the GR and the SP log curves of wells within the same depth range. High frequencies in the spectrum curves represent their relative importance along depth. Therefore, the harmonic frequencies of the high frequencies in the curves correspond to the dominant frequencies of the curves, which were converted into corresponding wavelength. Then the cycle thickness was obtained based on sampling intervals.
4.1 Spectral analysis of log curves of Well NYS3
Spectral analysis was conducted for GR and the SP log curves of Well NYS3 within a depth range of 1342−2593 m, revealing the presence of Milankovitch cycles (Figs. 4a, b). In these curves with vertical axes representing relative strength of spectra and the horizontal axes denoting frequencies (number of cycles per meter), the peak frequencies at points A, A’, B, C, E, and F in the GR log curves are 0.00820, 0.00960, 0.02558, 0.02718, 0.04400, and 0.05760, respectively, and the peak frequencies at points A, A’, B, C, E, and F in the SP log curves are 0.00800, 0.00960, 0.02400, 0.02718, 0.04400, and 0.05760, respectively. Accordingly, the cycle thicknesses at points A, A’, B, C, E, and F in the GR log curves are 15.26, 13.03, 4.89, 4.60, 2.84, and 2.17, respectively, and the cycle thicknesses at points A, A’, B, C, E, and F in the SP log curves are 15.64, 13.03, 5.21, 4.60, 2.84, and 2.17, respectively. Therefore, the ratio among the cycle thickness at points A, A’, B, C, E, and F according to the GR log curves is 1.00∶0.85∶0.32∶0.30∶0.19∶0.14, and the ratio among peak frequencies at points A, A’, B, C, E, and F in the SP log curves is 1.00∶0.83∶0.33∶0.29∶0.18∶0.14. Meanwhile, the ratio among 125 ka (E3), 99 ka (E2), 51 ka (O2), 39 ka (O1), 23 ka (P3), and 19 ka (P1) calculated by the method proposed by J Laskar is 1.000∶0.792∶0.312∶0.304∶0.184∶0.152. It can be discovered that ratios among peak frequencies at points A, A’, B, C, E, and F corresponds well to the ratio among change periods of the Earth orbit. Furthermore, it can be concluded that the stratigraphic cycles with a thickness of 15.26–15.64 m and 13.03 m correspond to the eccentricity cycles of 125 ka and 99 ka, respectively, the cycles with a thickness of 4.89–5.21 m and 4.60 m correspond to the obliquity cycles of 51 ka and 39 ka, respectively, and the cycles with a thickness of 2.84 m and 2.17 m correspond to the precession cycles of 39 ka and 23 ka, respectively.
4.2 Spectral analysis of log curves of wells NYS2 and NYS1
Similarly, spectral analysis was performed for the GR and the SP log curves of well-preserved Eocene strata discovered in wells NYS1 and NYS2 in the eastern depression (Figs. 4c–f). The curve section corresponding to a depth of 2178–2277 m was removed from the SP curve of Well NYS2 due to serious deformation. The thickness of cycles at different depths was calculated, and the ratios among the thickness were calculated and compared with the ratio among Milankovitch cycles. Then the sedimentation duration and sedimentation rates at different depths were calculated, as shown in Table 2.
4.3 Orbital tuning in the Eocene
According to the data in Table 2, the ratios between the thickness of Eocene stratigraphic cycles are very close to the ratios between Milankovitch cycles. The thickness ratio between the cycles at points A’ and A is 0.72–0.86, and the difference between it and corresponding Milankovitch cycle ratio (0.792) is not greater than 8.8%. The thickness ratio between cycles at points B and A is 0.29–0.32, and the difference between it and corresponding Milankovitch cycle ratio (0.312) is not greater than 7.05%. The thickness ratio between cycles at points C and A is 0.28–0.3, and the difference between it and corresponding Milankovitch cycle ratio (0.304) is not greater than 7.89%. The thickness ratio between cycles at points E and A is 0.18–0.19, and the difference between it and corresponding Milankovitch cycle ratio (0.184) is not greater than 3.2%. The thickness ratio between cycles at points F and A is 0.14–0.15, and the difference between it and corresponding Milankovitch cycle ratio (0.152) is not greater than 7.89%. According to comprehensive analysis, Milankovitch cycles are well preserved in the Eocene strata, and the thickness of stratigraphic cycles caused by eccentricity cycles, obliquity cycles, and precession cycles is 13.03–15.89 m, 3.70–5.21 m, and 2.17–2.94 m, respectively. The sedimentation rates of stratigraphic cycles can be calculated by dividing the thickness of the cycles by corresponding Milankovitch cycle duration. Meanwhile, the sedimentation rates calculated based on different cycle thicknesses have small errors. In this study, the sedimentation rates of stratigraphic cycles were calculated to be 121.2−127.12 m/Ma. The thickness of strata is jointly determined by sedimentation duration and sedimentation rate of the strata. On the premise that strata are relatively stable and assuming that the strata denudated and currently preserved share the same sedimentation rate, the sedimentation rate of strata currently preserved can be calculated by dividing the thickness of the strata by the sedimentation rate of the strata. Among them, the thickness of the strata currently preserved can be determined by drilling, and the sedimentation rate can be calculated based on Milankovitch cycles, as mentioned above. According to the calculation based on GR and SP log curves, the sedimentation duration of the Eocene strata at Well NYS3 is 9.98 Ma and 10.247 Ma, respectively, with a relative error of 2.49%, and the sedimentation duration of Well NYS1 is 5.998 Ma and 5.727 Ma, respectively, with a relative error of 4.73%. As for the sedimentation duration of Eocene strata of Well NYS2, it is 7.865 Ma at a depth of 1330–2277 m as calculated based on the GR curve, and is 6.839 Ma at a depth of 1330–2178 m as calculated according to the SP curve. Therefore, the sedimentation duration of Eocene strata at Well NYS2 is 1.026 Ma at a depth of 2178–2277 m. There are three types of stratigraphic cycles in the Eocene strata in the eastern depression of NYSB, namely long-, medium-, and short-term cycles, with a depth of 13.03–15.89 m, 3.7–5.21 m, and 2.17–2.94 m, respectively.
4.4 Sedimentation duration and sedimentation rates
As shown in Fig. 5 and Table 3, Well NYS1 is located in the uplift area of the eastern depression, revealing Eocene strata with a thickness of 728 m, sedimentation duration of 5.998 Ma, and a sedimentation rate of 121.36 m/Ma. Well NYS2 is located in the slope area, showing Eocene strata with a thickness of 947 m, sedimentation duration of 7.865 Ma, and a sedimentation rate is 121.2 m/Ma. Well NYS3 is located near the sedimentary center, revealing Eocene strata with a thickness of 1251 m, sedimentation duration of 10.247 Ma, and a sedimentation rate of 122.08 m/Ma. Therefore, it can be concluded that from the uplift area to the center of the lacustrine basin along the slope, the sedimentation duration becomes longer and the sedimentary strata grow thicker, while the sedimentation rate is relatively stable (Fig. 6). In the Early Eocene, the location of Well NYS3 firstly received sediments, while the locations of wells NYS1 and NYS2 suffered sedimentary bypassing. As the lacustrine basin expanded and the lake level rose, the strata located in the slope and uplift areas began to received sediments. Factors affecting sedimentation rates mainly include tectonic subsidence, climate, lake level change, and provenance. In terms of tectonic movement, the study area falls into the same depression and thus is slightly different in structure and relatively stable. This indicates that tectonic subsidence is an important factor contributing to the stable sedimentation rate. During the Eocene, the study area had a humid climate, relatively high precipitation, sufficient supply of provenance, and relatively shallow water overall, with shallow lacustrine sedimentation being widely distributed. This indicates that climate, lake level, and provenance are also major factors of the relatively stable sedimentation rate in the areas interacting with each other.
4.5 Astronomical forcing of Eocene strata
It can be concluded by combining Fig. 3 and Table 3 that the average relative strength of Eocene eccentricity, obliquity, and precession at points A, A’, B, C, E, and F is 2560.67, 3283.20, 1683.58, 1016.40, 1883.30, and 1210.47, respectively. Meanwhile, the ratios of the relative strength between each peak and the peak at point A is 1 (long eccentricity), 1.28 (short eccentricity), 0.66 (long obliquity), 0.40 (short obliquity), 0.74 (long precession), and 0.47 (short precession). Therefore, it can be considered that eccentricity and precession are the main astronomical forcing affecting the stratigraphic cycles in the study area. Furthermore, eccentricity and obliquity had the most and least effects on the stratigraphic cycles, respectively, indicating that the sedimentary cycles caused by the eccentricity (thickness: 13.03−15.89 m) are the most distinct ones. Eccentricity determines the distance between the Sun and the Earth. Meanwhile, quasi-periodic variations of the Earth’s orbital parameters impact the total solar radiation on Earth’s surface, which is a strong driver of climatic signatures. During glacial periods when the minimum eccentricity leads to the decrease in the insolation reaching the Earth, coarse-grained sandstone is produced. On the contrary, the maximum eccentricity leads to the increase in the insolation on Earth, and the warm and humid climate contributes to abundant provenance. In this case, fine-grained mud shale is formed (Tian SF, 2012).
5. Discussion
During the Early Paleogene, the Himalayan movement started and the mantle continued bulging upwards in general, just as it did in the Late Cretaceous. In this case, the NYSB underwent stable uplift and denudation, thus lacking Paleocene sediments. In the Eocene, the extrusion basically stopped, and thus basins began to subside again and the lacustrine basin began to receive sediments in a larger scope (Zhang L et al., 2009). During this period, the NYSB boasted high precipitation, humid climate, and sufficient supply of provenance. Meanwhile, water was relatively shallow overall, and shallow lacustrine facies was widely distributed in NYSB, with semi-deep lacustrine facies locally. Therefore, the Eocene strata in NYSB are characterized by interbeds consisting of sand and mud and bear thick organic-rich lacustrine mudstone. The organic carbon content was up to 7% and mixtures of Type I amorphous kerogen and Type III kerogen developed (Wang LF et al., 2010). At the end of the Eocene, the third tectonic reversal and sedimentary discontinuity caused by tectonic extrusion occurred in the North Yellow Sea due to the effects of the first episode of the Himalayan movement (Gong CL et al., 2009). As a result, unconformity was formed between the Eocene strata and the overlying Oligocene strata. In this study, under the guidance of sequence stratigraphy, the Eocene strata were divided using CWT based on log curves, core, and seismic data of wells NYS1, NYS2, and NYS3, obtaining six stages of high-resolution four-order sequences, namely E1−E6 (Fig. 6). Spectral analysis of the GR log curves was carried out with each sequence as a separate window, achieving sedimentation duration and sedimentation rates corresponding to Milankovitch cycles (Table 4, Table 5, Table 6; Fig. 5). The cycle thickness obtained corresponds to obliquity and precession due to the relatively small thickness of the strata studied, and each window shows clear characteristics of Milankovitch cycles.
For a close lacustrine basin, its lake level is lower than its base level, and thus the change of its relative lake level is not affected by the tectonic subsidence of its basement (Deino A et al., 2006; Yang YQ et al., 2016). Therefore, the effects of tectonic subsidence mode on the four-order sequences can be ignored, while only the effects of climate change are discussed. The climate change is cyclical. In the case of a dry climate, the atmospheric precipitation is small, and thus the water injected into a lacustrine basin is less than the water evaporated. As a result, the lake level decreases, so does the accommodation space of the lacustrine basin (Maselli V et al., 2014). The situation in the case of a humid climate is completely contrary to that in a dry climate. Meanwhile in this case, a large number of sediments are brought into lacustrine basins by rivers, which further improves the relative lake level (Ji YL et al., 2005).
The Eocene strata developed from highstand system tracts to lowstand system tracts, with the maximum lacustrine flooding surface developing at the interfaces of E3 and E4 (Fig. 5). The location of Well NYS1 was in a paleo-uplift in the Early Paleogene and suffered sedimentary bypassing or discontinuity in the Early Eocene.
E6 was deposited in the early stage of the formation of Eocene lacustrine basins. During this stage, the climate was arid - semi-arid, the provenance originated from the northern and western highlands, and only Well NYS3 received sediments. According to the core data analysis, E6 consists of gray coarse-grained pebbly sandstone, medium-grained conglomerate, light gray siltstone, and mudstone. Among them, the mudstone is mainly reddish-brown in the lower part and grayish-green in the upper part. E6 has a box- or bell-shaped GR curve. It is dominated by shallow lacustrine sedimentary facies, with a sedimentation rate of 117.146 m/Ma, sedimentation duration of 2.808 Ma, and a thickness of 329 m. The bottom of E6 is composed of thick-laminated coarse-grained clastic rocks such as coarse-grained sandstone and glutenite interbedded with thin-laminated reddish-brown mudstone. It shows a seismic facies of continuous and strong in-phase axial reflection, and thus is a sublacustrine fan formed for the reason that the gravity flow system as a mixture of the sand, mud, and gravel produced in case of flooding or sliding was directly inserted into the lake bottom.
E5 was deposited in the initial stage of the expansion of lacustrine basins, during which the arid climate was transformed into a humid climate, and the supply of provenance was sufficient. With an increase in lake level, the carrying capacity of water flow increased. Sediments were successively transported from the northern and western highlands to lacustrine basins by rivers, transported into lakes by rivers during flood expansion, and then migrated to lakesides by waves and lacustrine flow and were deposited. After long-distance transportation and repeated wash by lacustrine waves, rocks were highly mature and had high sorting and roundedness, and thus their sedimentation rates began to increase. E5 at Well NYS3 in the basin center consists of light-gray siltstone, fine-grained sandstone, medium-grained pebbly sandstone, and brown and greyish-green mudstone. It shows semi-deep lacustrine sedimentary facies and the seismic facies of a sub-parallel, continuous - relatively-continuous strong reflection structure. It has a sedimentation rate of 122.427 m/Ma, sedimentation duration of 1.56 Ma, and a stratum thickness of 191 m. Well NYS2 is located in the downthrow side of a contemporaneous fault on the slope and began to be deposited in this period. E5 revealed in Well NYS2 is comprised of gray and grayish-white fine-grained sandstone, brownish-yellow siltstone, and brown and reddish-brown mudstone. It shows the seismic facies of a disorderly and discontinuous-relatively continuous foreset reflection structure. Thus it is a nearshore subaqueous fan formed in the sedimentary environment of shore-shallow lacustrine facies with shallow water and high light transmission. It has a sedimentation rate of 122.52 m/Ma, sedimentation duration of 657 Ma, and a stratum thickness of 203 m.
E4 was deposited in a humid climate, during which the lake level reached the maximum flooding surface. With the expansion of lacustrine basins, the supply of provenance was short and clastic rocks were transported from highlands over a long distance. As a result, E4 revealed by Well NYS1 contains brown mudstone. In contrast, E4 revealed by the other two wells consists of light-gray fine-grained sandstone and greenish-gray mudstone, which are dominated by pure mudstone with a small grain size, dark colors, and high organic content. E4 revealed by Well NYS3 shows deep-lacustrine sedimentary facies and the sheet-shaped seismic facies with parallel reflection inside. Meanwhile, it contacts its overlying and underlying strata in a uniform way and is distinctly stratified, and shows high-frequency strong reflection with a medium - strong amplitude and high continuity in the seismic facies. E4 revealed by Well NYS3 runs through the hanging wall and footwall of faults due to tectonic effects. Therefore, it has a relatively high sedimentation rate of 104.468 m/Ma, sedimentation duration of 1.847 Ma, and a thickness of 193 m. E4 revealed by Well NYS2 on the slope shows a seismic facies of a broom-like foreset reflection structure, indicating that it is a fan delta front formed in the sedimentary environment of deep-semi-deep lacustrine facies. It has a sedimentation rate of 91 m/Ma, sedimentation duration of 1.978 Ma, and a thickness of 180 m. E4 revealed by Well NYS1 located in a paleo-highland shows the seismic facies with hilly and continuous - relatively continuous reflection, indicating fan delta plain facies in a semi-deep lacustrine sedimentary environment. It was formed due to the fast deposition of clastic rocks or volcaniclastics near the provenance, with a sedimentation rate of 107.878 m/Ma, sedimentation duration of 1.557 Ma, and a thickness of 169 m.
E3 was deposited when the humid climate became warm and humid. At this stage, lacustrine basins began to uplift and shrink due to the suffer extrusion caused by tectonic reversal. As a result, the lake level declined. E3 revealed by Well NYS1 is composed of light-gray medium-grained sandstone and grayish-brown mudstone. The seismic facies shows truncation and toplap at top and downlap at bottom in nearshore parts and is composed of divergent in-phase axes with poor continuity - medium continuity and medium - low amplitudes. Among them, non-systematic lateral in-phase axes terminate towards the provenance near the slope. All these indicate the sedimentary facies of a fan delta front. It has a sedimentation rate of 121.212 m/Ma, sedimentation duration of 1.865 Ma, and a thickness of 226 m. In contrast, E3 revealed by Well NYS2 show a seismic facies of a parallel reflection structure in which frequencies and phase number tend to be enhanced and increase towards the center of a lacustrine basin, with wavy sheet-shaped reflection locally. It is mainly composed of light-gray medium-grained sandstone, fine-grained sandstone, and greenish-gray and brownish-yellow mudstone, with sand poorly developing. It is dominated by the sedimentary facies of a fan delta front, with a sedimentation rate of 120.18 m/Ma, sedimentation duration of 2.028 Ma, and a thickness of 224 m. E3 revealed by Well NYS3 consists of light-gray fine-grained sandstone and brown and grayish-brown mudstone, with sand relatively developing. Its seismic facies shows medium-high amplitudes and distinct continuous-intermittent reflection. Owing to the development of sedimentary facies of a fan delta front and the effects of faults, it has a relatively high sedimentation rate of about 134.06 m/Ma, sedimentation duration of 1.103 Ma, but a small thickness of 148 m.
E2 was deposited in a warm climate. During this stage, lake water grew shallow and lacustrine basins continued shrinking. E2 at Well NYS1 consists of large sets of variegated glutenite interbedded with brownish-yellow and grayish-green mudstone. Its seismic facies shows continuous - relatively continuous strong reflection, indicating an overwater distributary channel of a fan delta plain. It has a sedimentation rate of 111.105 m/Ma, sedimentation duration of 1.224 Ma, and a thickness of 136 m. E2 at Well NYS2 is comprised of light gray coarse-grained sandstone and medium-grained sandstone interbedded with greenish-gray mudstone, showing a reverse rhythm with a thin lower part and a thick upper part. Its seismic facies shows sheet-shaped strong reflection with good continuity, indicating the sheet-shaped sand sedimentary facies of a fan delta front, with a sedimentation rate of 99.88 m/Ma, sedimentation duration of 1.224 Ma, and a thickness of 122 m. E2 at Well NYS3 consists of light-gray coarse-fined sandstone, variegated glutenite, and greenish-gray mudstone, also showing a reverse rhythm with a thin lower part and a thick upper part. Its seismic facies shows a relatively continuous foreset reflection with downlap at the lower interface, indicating an estuarine dam sedimentary facies of a fan delta front. It has a high sedimentation rate of up to 145.455 m/Ma due to the sufficient supply of provenance, sedimentation duration of 1.369 Ma, and a thickness of 199 m.
E1 was deposited in a dry climate. During this stage, lacustrine basins contracted over large areas and lake water became gradually shallower. Accordingly, the lacustrine basins became a fan delta plain overall, where large sets of variegated glutenite interbedded with thin laminated greenish-gray mudstone were deposited. E1 revealed by Well NYS1 has the seismic facies with downlap at its bottom interface. Owing to the coarse-grained terrigenous clastic sediments and the lack of fine-grained sediments, it has a sedimentation rate of 112.936 m/Ma, sedimentation duration of 1.744 Ma, and a thickness of 197 m. E1 revealed by Well NYS2 shows the seismic facies of imbricate foreset reflection, with the foreset layer terminating at the top and bottom interfaces in the forms of toplap and downlap, respectively. It was formed due to the deposition of coarse-grained clastic materials in the period with a relatively static water level. It has a sedimentation rate of 105.85 m/Ma, sedimentation duration of 1.871 Ma, and a thickness of 191 m. E1 revealed by Well NYS3 has the seismic facies of a broom-like reflection structure that is divergent towards the basin center. Meanwhile, it has a sedimentation rate of 122.52 m/Ma, sedimentation duration of 1.559 Ma, and a thickness of 191 m.
Therefore, it can be inferred from the above analysis that Milankovitch cycles serve as an effective method to analyze stratigraphic cycles. Astronomical cycles affect climate change. Meanwhile, as discovered from the analysis of stratigraphic cyclicity, climate controls stratigraphic sequences by affecting precipitation and water evaporated, which further leads to lake level change. In the study area, the climate controls the development of stratigraphic sequences mainly by the following processes:
(i) During the periods of lowstand system tracts (E6 stage), the climate was dry, and the water evaporated was greater than precipitation, resulting in a low lake level and a relatively small basin area. In this case, only the strata in the center of the lacustrine basin received sediments dominated by coarse-grained materials transported by seasonal rivers. Therefore, fan deltas developed in the basin center. In contrast, strata in highlands suffered sedimentary bypassing or discontinuity.
(ii) During the periods of lacustrine expansion system tracts (E5 and E4 stages), the climate became gradually humid, and the supply of sediments and water amount kept increasing. Meanwhile, the lakes expanded, and the sedimentary strata regressively lapped out at the top towards the edges of the lacustrine basins. When the climate was humid, the lacustrine basins reached the maximum lacustrine flooding surface, and the supply of provenance was insufficient. As a result, the sedimentation rate decreased, and nearshore subaqueous fans and turbidite fans mainly developed.
(iii) The period of highstand system tract (E3 stage) witnessed a transition from humid to arid climate. During this stage, the lacustrine basin area no longer expanded, and the lake level remained the same as it was before. The strata were dominated by aggradation-type parasequences, their top was mainly comprised of coarse-grained clastic rocks, and fan deltas developed.
(iv) The periods of lacustrine contraction system tracts (E2 and E1 stages) suffered a dry climate. As a result, the water evaporated was greater than precipitation, the lake level decreased, and large sets of coarse-grained clastic rocks were deposited until lake extinction.
6. Conclusions
(i) It was determined that Eocene strata in the eastern depression of the NYSB preserve Milankovitch cycles according to spectral analysis. Furthermore, the thickness of stratigraphic cycles affected by eccentricity cycles, obliquity cycles, and precession cycles is 13.03−15.89 m, 3.70−5.21 m, and 2.17−2.94 m, respectively. According to the peaks of average strength of Eocene eccentricity, obliquity, and precession calculated, the eccentricity and obliquity had the most and least effects on the Eocene strata in the eastern depression, respectively.
(ii) The sedimentation rates of Eocene strata in the eastern depression of the NYSB calculated based on the Milankovitch cycles range between 121.2 m/Ma and 127.12 m/Ma. From the uplift area to the center of the lacustrine basin along the slope, the sedimentation duration becomes longer and the sedimentary strata grow thicker, while the sedimentation rates are relatively stable.
(iii) Under the guidance of sequence stratigraphy, the Eocene strata were divided using CWT based on log curves, core, and seismic data of wells, obtaining six stages of sedimentary sequences. Spectral analysis was carried out with each sequence as a separate window, achieving the sedimentation rates and sedimentation duration of each sequence. Moreover, the sedimentary environment of each sequence was analyzed from the perspective of climate change according to vertical division and correlation of sequences.
(iv) The Eocene strata developed from highstand system tracts to lowstand system tracts. During the stage of E6, the climate was arid, and sedimentation of shore-shallow lacustrine facies began to develop in the lower parts of the lacustrine basins, with a sedimentation rate of 117.146 m/Ma. The stage of E5 underwent a transition from arid to humid climate. During this stage, the lacustrine basins expanded, and the sedimentation of shore-shallow lacustrine and semi-deep lacustrine facies developed, with a sedimentation rate of 112−123 m/Ma. During the stage of E4, the climate was humid, and the lacustrine basins expanded to the largest extent and reached maximum lacustrine flooding surface. At this stage, the sedimentation of semi-deep lacustrine and deep lacustrine facies developed, with a sedimentation rate of E4 of 91−108 m/Ma. The stage of E3 underwent a transition from humid to warm climate. During this stage, the lacustrine basins began to shrink due to tectonic reversal, but the supply of provenance was sufficient. Meanwhile, the sedimentation of fan delta front and fan delta facies mainly developed, with a sedimentation rate ranges from 120 m/Ma to 134 m/Ma. During the stage of E2, the climate became warm, and the sedimentation of coarse-grained fan-delta plain and fan-delta front developed. The sedimentation rate of E2 is relatively high in the center of the lacustrine basin, ranging from 100 m/Ma to 111 m/Ma. During the stage of E1, the climate was arid, the lacustrine basins contracted over a large area, and the sedimentation of coarse-grained fan-delta plain continued developing with a sedimentation rate of 105−123 m/Ma.
CRediT authorship contribution statement
Shu-yu Wu conceived the presented idea. Shu-yu Wu and Jun Liu prepared the manuscript in consultation. Jun Liu supervised the findings of this work. All authors discussed the results and contributed to the final manuscript.
Declaration of competing interest
The authors declare no conflict of interest.
Acknowledgment
This study was supported by projects of the China Geological Survey entitled “Integrated Observation Data Integration and Application Service of Natural Resource Elements” (DD20208067), “Petroleum geological survey in key areas of Yellow Sea” (DD2021353), “Geological survey on tectonic and sedimentary conditions of Laoshan Uplift” (DD20190818), and the project of National Natural Science Foundation of China entitled “Study on Hydrocarbon Accumulation Failure and Fluid Evolution Reduction of Permian Reservoir in Laoshan Uplift, South Yellow Sea” (42076220). The authors would like to express their gratitude to Qingdao Institute of Marine Geology, China Geological Survey for providing basic data. The authors’ thanks also go to the editors and reviewers for their sincere comments and assistance in the preparation of this manuscript.