Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

Contents lists available at SciVerse ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology

j ourna l homepage: www.e lsev ie r .com/ locate /pa laeo

Astrochronology of the Early Turonian–Early Campanian terrestrial succession in theSongliao Basin, northeastern China and its implication for long-period behavior of theSolar System

Huaichun Wu a,b,⁎, Shihong Zhang a, Ganqing Jiang c, Linda Hinnov d, Tianshui Yang a, Haiyan Li a,Xiaoqiao Wan a, Chengshan Wang a

a State Key Laboratory of Biogeology and Environmental Geology, China University of Geosciences, Beijing, 100083, Chinab School of Ocean Sciences, China University of Geosciences (Beijing), Beijing 100083, Chinac Department of Geoscience, University of Nevada, Las Vegas, NV 89154, USAd Department of Earth and Planetary Sciences, Johns Hopkins University, Baltimore, MD 21218, USA

⁎ Corresponding author at: State Key Laboratory of BGeology, China University of Geosciences, Beijing, 182334699; fax: +86 10 82320065.

E-mail address: whcgeo@cugb.edu.cn (H. Wu).

0031-0182/$ – see front matter © 2012 Elsevier B.V. Allhttp://dx.doi.org/10.1016/j.palaeo.2012.09.004

a b s t r a c t

a r t i c l e i n f o

Article history:Received 14 November 2011Received in revised form 13 August 2012Accepted 11 September 2012Available online 18 September 2012

Keywords:Late CretaceousCyclostratigraphyAstronomical time scale (ATS)Lacustrine paleoenvironmentsLong-period orbital forcingSongliao BasinChina

The first complete Early Turonian–Early Campanian lacustrine succession has been recovered from the SK-Isouth (SK-Is) borehole in the Songliao Basin (SLB), northeastern China. We conducted a detailedcyclostratigraphic study of natural gamma-ray (GR) log, thorium (Th) log, and magnetic susceptibility(MS) data from this core. Spectral analysis of the upper Quantou Formation (K2q3+4), Qingshankou Forma-tion (K2qn), Yaojia Formation (K2y), and lower Nenjiang Formation (K2n1+2) reveals a hierarchy of meter- todecameter-scale cycling in the data. The wavelength ratios of the cycles in these stratigraphic units are~20:5:2:1, corresponding with those of Milankovitch cycle periods of 405 kyr (long eccentricity):100 kyr(short eccentricity):37 kyr (obliquity):20 kyr (precession), indicating astronomical control on sedimenta-tion. An astronomical time scale (ATS) was established by tuning interpreted 405 kyr cycles to a 405 kyrorbital eccentricity target curve, and to four SIMS U–Pb zircon radioisotope ages. This ‘absolute’ ATS providesprecise numerical ages for stratigraphic boundaries, biozones, geological and geophysical events, and servesas a basis for correlation of strata and events between marine and terrestrial systems. The ages of the C33r/C34n geomagnetic polarity boundary in K2n

2 and three short reversal events in K2y are estimated as83.633 Ma, 84.819–84.862 Ma, 84.982–85.092 Ma and 85.240–85.629 Ma, respectively. Long-period ampli-tude modulations in the obliquity and eccentricity bands of the 405-kyr-tuned GR-Th series provide strongevidence that long-period orbital forcing influenced climate change and depositional processes in the SLB.The extracted amplitude modulations provide evidence that the orbits of Earth and Mars were not in secularresonance, and were undergoing chaotic interactions during this time, although the modulations do notmatch those of recent astronomical models.

© 2012 Elsevier B.V. All rights reserved.

1. Introduction

The Cretaceous Period is characterized by greenhouse climatesand a sequence of remarkable geological and geophysical eventsthat includes widespread oceanic anoxic events (OAEs), intense vol-canic activity, high sea level and the Cretaceous normal superchron(CNS). The evidence for these conditions and events mostly originatesfrom marine stratigraphy. To improve our understanding of changesin terrestrial environments of the Cretaceous, two boreholes (SK-ISouth and North) with a total length of 2485.89 m, recovering strata

iogeology and Environmental00083, China. Tel.: +86 10

rights reserved.

from Early Turonian to Early Paleogene, were drilled in the SongliaoBasin (SLB), northeastern China (Fig. 1) (Wang et al., 2009a).

Establishing a precise age framework for the drill cores has been acritical task for the SK-I Drilling Project. Recently, a new time frameworkwas established for the SK-I boreholes with a complete geomagneticpolarity chron sequence from upper C34n to C29n and four SIMS U–Pbzircon ages (Deng et al., 2013–this issue; He et al., 2012). The SK-Iscore contains the four U–Pb ages; the uppermost age confirms thatthe polarity reversal at 985.95 m is the C33r/C34n boundary (Fig. 2).Despite this important advance, the time resolution for the SK-Is core(Early Turonian–Early Campanian) remains too low to accuratelyestimate the ages and durations of geological and geophysical events.

It is well known that the Earth's astronomical parameters influ-ence the global climate (e.g., Hinnov and Ogg, 2007). Astronomicaltime calibration through cyclostratigraphic studies provides a tool toestablish an ultra-high resolution (0.02–0.4 Myr) astronomical time

122 Eo

Nenjiang

TongliaoJilin

47 No

46 No

45 No

44 No

43 No

124 Eo 126 Eo 128 Eo

SouthwestUplift ZoneSouthwest

Uplift ZoneSouthwest

Uplift Zone Changchun

Yushu

Harbin

Harbin

Bei'an

Daqing

(b)

Zha

nggu

angc

ai M

ount

ains

sniatnuoMna'gni

XrasseL

Gre

at X

ing'

an M

ount

ains

NorthPlunge Zone

NorthPlunge Zone

NorthPlunge Zone

NortheastUplift ZoneNortheast

Uplift ZoneNortheast

Uplift Zone

Suihua

Daqing

Qiqihar

West SlopeZone

West SlopeZone

West SlopeZone

SK-I s

YushuSoutheast

Uplift ZoneSoutheast

Uplift ZoneSoutheast

Uplift Zone

CentralDepression

Zone

CentralDepression

Zone

CentralDepression

Zone

(a)

Beijing

Lhasa Shanghai

Hongkong

Songliao Basin

Fig.1b

Fig. 1. (a) Location of the Songliao Basin in northeastern China. (b) Distribution offirst-order tectonic units and site of the SK-I south core.

56 H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

scale (ATS) (Hinnov and Hilgen, 2012). Recent cyclostratigraphicstudies have led to significant revisions and refinements of the Creta-ceous geological time scale (Ogg and Hinnov, 2012).

Besides short period astronomical forcing, the long period ampli-tude modulations of the eccentricity (2.34 Myr in the Cenozoic) andobliquity (1.2 Myr in the Cenozoic) may also play an important rolein major climate events recorded in both marine and continental set-tings (e.g., Lourens and Hilgen, 1997; Olsen and Kent, 1999;Shackleton et al., 1999; Mitchell et al., 2008; Abels et al., 2010). Therole of these long-period components in Cretaceous climate change,especially in terrestrial environments, is largely unknown.

In this study we take advantage of the great length and continuityof the lacustrine record provided by SK-Is in the SLB to analyze astro-nomical signal in the sedimentary record using gamma ray logging(GR), thorium logging (Th) and magnetic susceptibility (MS) data.An ATS was established by tuning to 405 kyr long eccentricity cyclesinterpreted in these sedimentary records to the La2010a astronomicalmodel (Laskar et al., 2011); U–Pb zircon ages (He et al., 2012) provideabsolute age tie points. On the basis of this new ATS, we discussthe sedimentary cycle response to long period (>405 kyr) orbital-forced climate change and evolution of the SLB.

2. Geological setting

The SLB is one of the largest and long-lived Cretaceous continentalbasins in the world. It is 750 km long, 330–370 km wide with an area

Fig. 2. Compilation of key data showing the stratigraphy, polarity chron, paleoenvironmentsedimentary accumulation rates of the SK-Is borehole in the SLB. Magnetostratigraphy resufrom Cheng et al. (2009), Gao et al. (2009) and Wang et al. (2009b,c). The “e” (blue) andThe interpreted 100 kyr and 405 kyr eccentricity cycles were extracted with Gaussian bandpfor 405 kyr eccentricity cycles: K2n1+2—0.0156±0.0027 cycles/m, K2y—0.021±0.004 cypassband for 100 kyr eccentricity cycles: K2n1+2—0.0625±0.0189 cycles/m, K2y—0.0859±

of 260,000 km2. It is surrounded by the Great Xing'an Mountains tothe west, the Lesser Xing'an Mountains to the north and theZhangguangcai Mountains to the east (Fig. 1). The basin was formedand filled during four tectonic episodes: mantle upwelling, rifting,post-rift thermal subsidence and structural inversion (Wang et al.,2007). The basement of the SLB is composed of Precambrian to Paleo-zoic metamorphic and igneous rocks and Paleozoic to Mesozoic gran-ites (Wang et al., 2006; Pei et al., 2007). The formation and sedimentfill of the SLB were closely related to the tectonic evolution stages,and was broadly controlled by the Mongolia–Okhotsk collision belt,and the westward subduction of the Pacific Plate underneath the Asiacontinent (Wang et al., 2007; Feng et al., 2010).

Borehole SK-Is is located at the central depression zone of the SLB(Fig. 1). A total of 944.23 m of core was collected with a recovery ratioof 99.73% (Wang et al., 2009a). The drill core covers four sedimentaryformations, from lower member 2 of the Nenjiang Formation (K2n2)to upper member 3 of the Quantou Formation (K2q3), that were de-posited during the post-rift thermal subsidence stage (Wang et al.,2007; Feng et al., 2010)(Fig. 2). The K2n1+2 (960 m to 1128.17 m)consists of dark mudstones intercalated with thin carbonate layers,black shales and oil shales that were deposited in deep lake environ-ments (Gao et al., 2011). The Yaojia Formation (K2y) (1128.17 m to1285.91 m) is composed of brownish, greenish and grayish mudstoneand greenish muddy siltstone that formed in deltaic and shallow la-custrine environments (Cheng et al., 2009). The Qingshankou Forma-tion (K2qn) (1285.91 m to 1782.93 m) consists of gray, dark gray andblack mudstone inter-bedded with marlstone, with oil shale in thelower and thin bedded gray siltstone in the upper part. It has ashallowing‐upward trend from deep lacustrine deposits in K2qn1

(Gao et al., 2009) to moderately deep and shallow lacustrine depositsin K2qn2+3 (Wang et al., 2009b). Finally, K2q3+4 (1782.93 m to1915.0 m) is characterized by grayish green or brown mudstone, silt-stone and sandstone interpreted as meandering river and deltaic de-posits (Wang et al., 2009c).

Detailed core descriptions indicate that meter-scale sedimentary cy-cles are well developed throughout the SK-Is borehole (Cheng et al.,2008, 2009; Gao et al., 2009; Wang et al., 2009b,c). The meter-scale cy-cles are expressed by variations in rock color, lithology and composition,with sedimentary structures that include cross-bedding and normalgrading (indicative of shallow lake andmeandering river facies), reversegrading (primarily deltaic facies) and non-graded, fine-grained deposits(deep lacustrine facies) (Cheng et al., 2008). Cycle thicknesses rangefrom 0.5 m to 1.5 m, and in K2q3+4 are up to 6.51 m. Fischer plots indi-cate that the cycles are bundled into fifth and fourth order cycles with1:3 to 1:4 and 1:6 to 1:12 ratios, respectively (Cheng et al., 2008).These lower order cycles were attributed to Milankovitch cycles of pre-cession and eccentricity. This interpretation is consistent with spectralanalysis of the natural gamma-ray logs of K2n and K2qn throughoutthe SLB (Wu et al., 2007, 2009).

Recently, He et al. (2012) obtained four SIMS zircon U–Pb radioiso-tope ages for bentonite beds in the SK-Is borehole, providing ages of83.7±0.5 Ma at 1019 m, 90.4±0.4 Ma at 1673 m, 90.1±0.6 Ma at1705 m, and 91.4±0.5 Ma at 1780 m (Fig. 2). On the basis of theseages, they interpreted the magnetic reversal at 985.95 m as the bound-ary of C33r/C34n, and three magnetic reversals (1144.4 m–1149.8 m,1163.85 m–1175.9 m and 1193.15 m–1239.9 m) as short reversalevents (Fig. 2). This age framework of SK-Is provides a solid basis forestablishing the ATS in this study.

s, lithology, SIMS U–Pb zircon ages, cyclostratigraphic interpretation and variations inlts and SIMS U–Pb zircon ages are from He et al. (2012). Paleoenvironmental data are“E” (red) represent 100 kyr short and 405 kyr long eccentricity cycles, respectively.

ass filters using free software of Analyseries 2.0.4.2 (Paillard et al., 1996). Filter passbandcles/m, K2qn—0.0261±0.005 cycles/m, and K2q3+4—0.0127±0.0036 cycles/m. Filter0.025 cycles/m, K2qn—0.113±0.025 cycles/m, and K2q3+4—0.0508±0.0116 cycles/m.

LAE2

LAE1

0 5 10 15 20 25

1900

1850

1800

1750

1700

1650

1600

1550

1500

1450

1400

1350

1300

1250

1200

1150

1100

1050

1000

Normalized GR (Th for K y)2 ( )10 m /kg-8 3Filter outputSediment accumulation

rates (cm/kyr)egatS

revirgnir edn ae

m

DEL

ekalpeed

ekalpeed

yletare dom

SL

DEL

ekalwolla hs

MDL

ekalpeed

Fine sandstone muddy siltstone silty mudstone mudstone marl oil shaleLegends

nainotnaS

naicainoC

nainoruT

1e3e

r33C

n43C

)nK(

gnaijneN

2)y

K(aijoa

Y2

)nqK(

uoknahsgniQ

2)q

K(uotnau

Q2 q

K2

4q

K2

3n q

K2

1nq

K2

3+2

yK

2

1y

K2

3+2

nK

2

1n

K2

2

noitamro

F

rebme

M Pol

arit

ych

ron

±5.0

4 .1 9±

6.01. 09

±4.0

4. 09±

5.07.38

E 0405

E 1405

E 2405

E 3405

E 4405

E 5405

E 6405

E 7405

E 8405

E 9405

E 10405

E 11405

E 12405

E 13405

E 14405

E 15405

E 16405

E 17405

E 18405

E 19405

E 20405

E 21405

e0

e1

e2

e3

e4

e5

e6

e7

e8

e9

e10

e11e12

e13e14

e15e16

e17e18e19

e20e21

e22e23

e24e25

e26e27

e28e29 e30

e31e32

e33e34

e35e36

e37e38e39e40e41e42

e43e44

e45e46e47

e48e49e50

e51e52e53

e54e55

e56e57e58

e59e60e61e62e63

e64e65e66

e67e68

e69e70

e71e72

e73e74e75e76e77

e78e79e80

e81e82

e83

e84

e85

e86

e87

lanoitisopeD

tnemno rivne

Lithology

U-Pb age(Ma)

e88

E 0405

E 1405

E 2405

E 3405

E 4405

E 5405

E 6405

E 7405

E 8405

E 9405

E 10405

E 11405

E 12405

E 13405

E 14405

E 15405

E 16405

E 17405

E 18405

E 19405

E 21405

E 20405

- +

Cam.

2e

-4 -2 0 2 419

0018

5018

0017

5017

0016

5016

0015

5015

0014

5014

0013

5013

0012

5012

0011

5011

0010

5010

00

Dep

th(m

)

4 8 12 16 20 24×

57H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

GR(API)

1816

1814

1812

1810

1808

1806

1804

1802

1800

1798

1796

1794

1792

1790

1788

1786

1784

Dep

th(m

)

ekaL SL

tnorF

atleD

revirgni redna

M

nialpdoolF

tnemknab

mE

nialpdoolF

EB

FL

EB

FLRB

EB

FL

EB

FL

EB

RBEB

RB

EB

FL

EB

atleD

)qK(

noitamro

Fuotnau

Q2

SL: Shallow lakeFL: Floodplain

Fig. 3. Variations of lithological, magnetic susceptibility (MS) and gamma ray loggingdata (GR) of the upper Quantou Formation (K2q), showing higher MS and GR valuesin mudstone.

58 H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

3. Data and methods

3.1. Magnetic susceptibility

Magnetic susceptibility (MS) refers to the capacity of a substanceto become magnetized when subjected to an external magneticfield. It has been proven to be an effective paleoenvironmentalproxy and has been used in cyclostratigraphic research of both ma-rine and continental sequences (e.g., Heller et al., 1991; Nádor et al.,2003; Boulila et al., 2008; Weber et al., 2010; Wu et al., 2012). Inthis study, a total of 2870 samples were collected with an averagespacing of ~0.3 m from the SK-Is core. MS was measured on aKappabridge KLY-3 in the Paleomagnetism and Geochronology Labo-ratory of the Institute of Geology and Geophysics, Chinese Academy ofSciences. The SK-Is core is characterized by relatively high MS valuesranging from 0.5×10−8 m3/kg to 28×10−8 m3/kg. Rock magneticstudies indicate a mix of magnetite, hematite and goethite contribut-ing to the core sediment (He et al., 2012). The MS series is stronglycyclic throughout the core (Figs. 2 and 3).

3.2. Natural gamma-ray (GR) and thorium (Th) logging data

GR logging data, which records the intensity of gamma rays emit-ted during the decay of radioactive atomic nuclei, additionally relatesto the amount of 40K, 232Th and 238U in the rock, and reflects theamount of clay and organic material in sediment (ten Veen andPostma, 1996; Schnyder et al., 2006). The GR logging data of theSK-Is were collected at a 0.125 m resolution in the borehole (Figs. 2and 3). Eight sandstone beds in K2y have abnormally high GR values,which is due to an unusually high concentration of U (Cheng et al.,2009; Wu et al., 2011). This may lead to misinterpretation of cyclicityin K2y, and so the GR series in K2y was not used for Milankovitch cycleanalysis. In contrast, the thorium (Th) logging data of K2y was foundto closely and consistently track lithological change, with high Th inthe mudstones and low Th in the sandstones (Fig. 2). In this study,the GR series was chosen for cyclostratigraphic analysis of the K2n,K2qn and K2q, and the Th series for K2y. The GR and Th series werestandardized and joined into a single “composite” series for the astro-nomical calibration (Fig. 2).

3.3. Time-series methods

The multitaper method (MTM) of spectral analysis and harmonicanalysis (Thomson, 1982) and wavelet analysis (Torrence andCompo, 1998) were conducted on the MS and GR-Th series tosearch for Milankovitch cycles. Prior to analysis, the series werepre-whitened in Kaleidagraph™ by subtracting a 35% weighted av-erage (Cleveland, 1979). The cycle length ratio method (Mayerand Appel, 1999; Weedon, 2003) was applied to investigate thelink between the sequence of sedimentary cycles and theoretical as-tronomical cycles. MTM power spectral analysis was conductedusing the SSA-MTM toolkit (Ghil et al., 2002), with robust rednoise estimation reported at the 90%, 95% and 99% confidence levelsfor interpretation of spectral peak significance (Mann and Lees,1996). Wavelet analysis was used to compute evolutionary spectra,for tracking changes in cycle frequencies due to changes in sedi-mentation rate.

The interpreted Milankovitch cycles were extracted using Gauss-ian (e.g., Dziewonski et al., 1969) as implemented in the freewareAnalyseries 2.0.4.2 (Paillard et al., 1996), and Taner bandpass filtering(Taner, 2000). Amplitude modulation envelopes of the filtered serieswere obtained using Hilbert transformation (Hinnov, 2000), in orderto isolate long-period components of the interpreted eccentricity andobliquity.

59H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

3.4. Late Cretaceous astronomical solutions

The calibration of cyclostratigraphy to astronomical models hasbecome a fundamental tool for refining the geologic time scale(Hinnov and Hilgen, 2012). La2004 was compiled by numerical inte-gration of the equations of motion of the Solar System's planetaryorbits, and includes a model of the Earth's precession that accountsfor tidal dissipation of the Earth–Moon system (Laskar et al., 2004).Recently, Laskar et al. (2011) formulated a new astronomical solution,La2010, which has an improved precision in the initial conditionsprovided by a special version of the ephemeris INPOP08. La2010aprovides the Earth's orbital elements only; precession and obliquitymust be reconstructed using the precession model given in Laskaret al. (1993) (details in the Appendix). While there are four solutions(La2010a–d), Laskar et al. (2011) recommend La2010a, with its smallintegration step (1×10−3 yr) and low error through 60 Ma.

The La2004 and La2010a solutions are considered to be valid over~42 Ma and 60 Ma, respectively (Laskar et al., 2004, 2011). Beyondthis time, the precision of both solutions decrease rapidly due to cha-otic behavior of Solar System and an inability of the models to predictthis behavior accurately. However, the 405 kyr eccentricity variationis stable over much longer times, and is reliable for astronomicalcalibration of Mesozoic stratigraphy.

Spectral analysis of ETP realizations of La2004 and La2010a for the82–93 Ma interval (Fig. 4) shows small differences in the short

0 0.01 0.02 0.03

Frequency

0

20

40

60

80

100

Pow

er

0

20

40

60

80

100

Pow

er

2340129

99

9548.1

2

37.9

36.8

2340405

12299

95 48.1

37.9

36.8

130

(a)

(b)

123.5405

0 0.01 0.02 0.03

Frequency

Fig. 4. MTM (multitaper method) power spectra of (a) ETP signal of the La2010a solution (Lover the interval 83–93 Ma. Peaks are labeled in kyr. Taner filter passbands used in Figs. 8 a(flow, fcenter, fhigh, in cycles/kyr): wide-band long eccentricity at 405 kyr (0.00146858, 0.00line), narrow-band short eccentricity at 95 kyr (0.0095, 0.01025, 0.0110), and obliquity at

eccentricity band; both solutions include a faster Cretaceous Earth ro-tation rate according to the tidal model in Laskar et al. (2004). Wewillcompare these astronomical periodicities and their long-termmodulations with those in the MS, GR and Th series to confirm astro-nomical drivers in SLB sedimentary cyclicity.

4. Cyclostratigraphy

4.1. Cycle analysis and interpretation

Spectral analysis of the untuned MS stratigraphic series of theK2n1+2 reveals significant wavelengths at 20 m, 15 m and 3.1 m(Fig. 5a). The K2n1+2 GR spectrum shows prominent peaks at 65 m,16 m, 6.7 m, 4.5 m, 3.3 m and 3 m above 95% confidence (Fig. 6a);wavelet analysis indicates relatively continuous cycles with wave-lengths of 50 m, 16 m, 6.7 m, 4.5 m and 3 m (Fig. 6b). The ratios ofthe major wavelengths at 65 m, 20–16 m, 6.7 m, 3.3 m is ~20:5:2:1,whichmatches well with those of the Late Cretaceous astronomical pa-rameters shown in Fig. 4. Thus we interpret the 65 m, 16 m, 6.7 m,4.5 m, 3.3 m and 3 m lithological cycles as corresponding to the eccen-tricity (405 kyr and 100 kyr), obliquity (42 kyr and 28 kyr), and pre-cession (21 kyr and 19 kyr), respectively.

The power spectra of the untuned MS and Th series of K2y indicatesignificant peaks with 46 m, 34 m, 11.6 m, ~4.8 m, 3.3 m, 2.6 m and1.9 m wavelengths (Figs. 5b and 6c). The presence of parallel bands

0.04 0.05 0.06 0.07

(cycles/kyr)

8.3

22.721.5

18.5

22.6

21.418.3

28.2

0.04 0.05 0.06 0.07

(cycles/kyr)

askar et al., 2011), and (b) ETP signal of the La2004 solution (Laskar et al., 2004), bothnd 9 are shown as shaded areas and dashed lines, defined as follows, from left to right246858, 0.00346858), wide-band short eccentricity (0.0065, 0.0090, 0.0115) (dashed37.4 kyr (0.0259, 0.0267, 0.0275).

Fig. 5. The 2πMTM (multitaper method) power spectra and wavelet scalograms of untuned MS series of (a) first member of Nenjiang Formation (K2n1), (b) Yaojia Formation (K2y),(c) Qingshankou Formation (K2qn) and (d) third and fourth members of the Quantou Formation (K2q3+4) in the depth domain. The purple curve ‘M’ indicates the smoothed, fittedred-noise spectrum. The blue, green and red curves indicate 99%, 95% and 90% confidence limits. The shaded contours in wavelet scalograms are normalized linear variances, withblue representing low spectral power and red representing high spectral power. Regions below curves on both ends indicate the cone of influence where edge effects becomesignificant.

60 H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

in the wavelet spectra indicates that spectral power is confined to thedistinct frequency bands shown in the power spectrum (Figs. 5c and6d). Spectral wavelength ratios of 46 m, 11.6 m, 4.8 m, 3.3 m, 2.6 m

and 1.9 m are very close to orbital period ratios (Figs. 5b and 6c), andthey may represent eccentricity (405 kyr and 100 kyr), obliquity(41 kyr and 31 kyr), and the precession index (22.5 kyr and 16.5 kyr).

Fig. 6. The 2π MTM (multitaper method) power spectra and wavelet scalograms of (a, b) the GR logging data of first and second members of the Nenjiang Formation (K2n1+2),(c, d) the Th logging data of Yaojia Formation (K2y), (e, f) the GR logging data of Qingshankou Formation (K2qn) and (g, h) the GR logging data of third and fourth members ofthe Quantou Formation (K2q3+4) in the depth domain. See legends and notes in Fig. 5.

61H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

A hierarchy of significant peaks is present in the power spectrum ofthe untuned MS and GR series of K2qn with wavelengths of 195 m,107 m, 73 m, 60 m, 36 m, 10.8 m, 8.8 m, 4.8 m, 4 m, 3.2 m, 2.56 m,2.1 m, 1.8 m, 1.6 m and 1.54 m (Figs. 5d and 6e). Wavelet analysisshows significant power in the same frequency bands as the power

spectrum (Figs. 5e and 6f). According to the cycle length ratio method,these sedimentary cycles may correspond to ~2.2 Myr, 1.2 Myr,824 kyr, 680 kyr, 123 kyr, 100 kyr, 54 kyr, 45 kyr, 36 kyr, 29 kyr,24 kyr, 21 kyr, 18 kyr and 17.5 kyr Milankovitch cycles (Figs. 5d and6e).

Fig. 7. MTM power spectra and wavelet scalograms of the normalized 405 kyr-tuned GR (a, b) and MS (c, d) time series of the SK-Is borehole. Significant peaks are labeled in kyr.See legends and notes in Fig. 5. In (a) Taner filter passbands used in Figs. 8 and 9 are shown as shaded areas and dashed lines, defined as follows, from left to right (flow, fcenter, fhigh, incycles/kyr): wide-band long eccentricity at 405 kyr (0.00146858, 0.00246858, 0.00346858), wide-band short eccentricity (0.0065, 0.0090, 0.0115) (dashed line), narrow bandeccentricity at 95 kyr (0.009, 0.010, 0.011), and obliquity at 37.4 kyr (0.0259, 0.0267, 0.0275).

62 H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

63H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

Power spectra of the untuned MS and GR series of K2q3+4 indicatethe presence of significant peaks of 20.5 m, 16 m, 8.3 m, 7.9 m, 3.8 m,3.2 m and 2.6 m (Figs. 5f and 6g). Wavelet analysis reveals continu-ous cycles with periods of 16–24 m, 6.5–8 m and 2.5–4 m (Figs. 5gand 6h), which may represent short eccentricity, obliquity and theprecession index.

4.2. Astronomical calibration of the SK-I south core

Direct cycle counting method was used to establish SK-Is corecyclostratigraphic framework following the method of Westerholdet al. (2007). Bandpass filters were designed to isolate the 405 kyr(“long”) and ~100 kyr (“short”) eccentricity cycles in sedimentary re-cords based on our interpretation in Section 4.1. The filtered long andshort eccentricity cycles are numbered according to their position rel-ative to the top of the SK-Is core in K2n2 (Fig. 2). The 405 kyr (long ec-centricity) cycles are referred to as E4050, E4051, E4052, etc., while the100 kyr (short eccentricity) cycles were labeled as e0, e1, e2, etc.

The untuned MS, GR and Th series display similar E405 cycles(Fig. 2). The K2n1+2 member records ~12 e cycles and ~3 E405 cycles.The K2y member has 15 e cycles and ~3.5 E405 cycles. For K2qn andK2q3+4, 55 and ~7 e cycles, 13.5 and ~1.7 E405 cycles were identified,respectively.

A major objective of this study is to establish an ATS for the SK-Iscore so that the durations of major geological events can be preciselyestimated and their correlation with events in marine sequences canbe fine-tuned. The short eccentricity cycles are unstable due to thechaotic motion of planets in the inner Solar System during the Meso-zoic, and are not suitable for astronomical calibration. Therefore, wetuned the interpreted 405 kyr eccentricity cycles to the La2010asolution.

A successful astronomical calibration is further based on interpre-tation of the phase relationship between sedimentary cycles and as-tronomical target curves. The identification of long and shorteccentricity, obliquity and precession cycles in the untuned GR, Thand MS series of the SK-Is core suggests strong astronomical climateforcing in the development of the SLB sedimentary cycles. HigherGR (Th) values in lacustrine sediments in the SK-Is core are relatedto higher content of clay minerals and organic carbon, which mayhave resulted from wetter and warmer climate conditions (Wu etal., 2011). Wetter and warmer periods could have enhanced chemicalweathering, clay mineral input, nutrient input and higher productiv-ity in the paleolake, resulting in positive GR excursions. Thus, it is an-ticipated that wetter and warmer climate conditions prevailed duringthe deposition of sediments with higher eccentricity. Therefore,tuning was done in this study by assigning ages of 405 kyr (long ec-centricity) maxima of the La2010a solution to corresponding positivepeaks of E405 cycles in the GR (Th for the K2y) series of the SK-Is. Thetuning was conducted using the Linage function in Analyseries 2.0.4.2(Paillard et al., 1996).

Spectral and wavelet analysis of the 405 kyr-tuned GR (Th forK2y) series shows significant spectral peaks with periodicities of~2340 kyr, 1360 kyr, 960 kyr, 680 kyr, 405 kyr (tuned), 140 kyr,131 kyr, 107 kyr, 99 kyr, 44 kyr, 40 kyr, 37.4 kyr, 33 kyr, 26.8 kyr,23.5 kyr, 22.2 kyr, 21 kyr and 17.6 kyr above the 99% confidencelevel (Fig. 7a). The power spectrum of the 405 kyr-tuned MS serieshas significant peaks at 1200 kyr, 680 kyr, 405 kyr (tuned), 250 kyr,227 kyr, 101 kyr, 95 kyr, 37.4 kyr, 28 kyr, 25.5 kyr, 20.5 kyr and18.4 kyr (Fig. 7c). Wavelet analysis of the 405-kyr tuned GR (Th forK2y) and MS series show comparable spectral bands (Fig. 7b, d).

These spectra compare well with those of the La2010a and La2004solutions (Fig. 4) and support our cyclostratigraphic interpretationand proposed ATS for the SK-Is core. The long-period components inthe spectra, especially those at 2340 kyr and 1200 kyr, may be man-ifestations of eccentricity and obliquity amplitude modulation cycles,and are discussed further below (Section 5.5).

SIMS zircon U–Pb ages of four bentonite beds were recentlyobtained from SK-Is (He et al., 2012). Here we used the age of91.4±0.5 Ma at 1780 m in K2qn1 as the initial anchor point, whichallows us to tie the maximum of E40519 at 1770.125 m to the maxi-mum of the 405 kyr eccentricity cycle of La2010a at 91.231 Ma(Fig. 8). The established ATS gives ages for 1780 m, 1705 m,1673 m, and 1019 m at 91.341 Ma, 90.472 Ma, 90.080 Ma and83.834 Ma, which are very close (within dating uncertainties) to thecorresponding SIMS U–Pb zircon ages of 91.4±0.5 Ma, 90.1±0.6 Ma, 90.4±0.4 Ma and 83.7±0.5 Ma (Fig. 8a–d).

5. Discussion

5.1. Astrochronology of the SK-Is core

The new ATS provides a continuous ‘absolute time framework’from 83.40 Ma–92.08 Ma (Early Turonian–Early Campanian) andage assignments for stratigraphic units and biological and geologicalevents. The astronomically calibrated boundary ages of K2n2/K2n1,K2n1/K2y2+3, K2y2+3/K2y1, K2y1/K2qn2+3, K2qn2+3/K2qn1, K2qn1/K2q4 and K2q4/K2q3 are 83.882 Ma, 84.673 Ma, 85.748 Ma,86.010 Ma, 90.435 Ma, 91.371 Ma and 91.867 Ma, respectively(Fig. 8b). On the basis of these age constraints, the Campanian/Santonian, Santonian/Coniacian and Coniacian/Turonian boundariesare assigned to 971.1 m, 1259.6 m and 1542.0 m (Figs. 2 and 8a), re-spectively, in the SK-Is borehole. This ATS also serves as a basis forcorrelating the major Later Cretaceous geological events betweenma-rine and continental sequences.

Fifteen ostracode biozones were identified in the SK-Is borehole(Wan et al., 2013–this issue). The age of the top boundary and dura-tions of these biozones were estimated using the ATS as shown inTable 1. For example, the durations of ~0.22–0.27 Ma for fourostracode biozones are consistent with the 0.22–0.25 Ma cycles inthe MS series from 85 to 86 Ma (visible in the scalogram, Fig. 7d),suggesting that long-period components of obliquity forcing affectedostracode evolution. Previously, van Dam et al. (2006) found thatlong-period astronomical climate forcing during the Cenozoic was amajor determinant of species turnover in small mammals in CentralSpain. Our results may indicate that ostracode turnover in theSongliao paleolake was also related to long-period astronomicalclimate forcing.

5.2. Age and duration of magnetic polarity events

The C-sequence of marine magnetic anomalies was assembled byCande and Kent (1992, 1995) from a composite of South Atlantic pro-files with additional resolution from selected Pacific surveys(Gradstein et al., 2004). They calculated absolute ages for theirsynthetic magnetic anomaly pattern by applying a cubic-spline fit toselected radiometric ages. Gradstein et al. (2004) recalibrated theLate Cretaceous C-sequence with a spline fit of revised and additionalradiometric ages. Recently, Husson et al. (2011) estimated the agesand durations of each magnetochron from C32r.2r to C29n throughcyclostratigraphic study of Late Campanian–Maastrichtian sedimen-tary successions from four ODP (Ocean Drilling Program) and DSDP(Deep Sea Drilling Program) holes.

The age of the end of chron C34n was assigned as 83.0 Ma byCande and Kent (1992, 1995), 84.0 Ma by Gradstein et al. (2004),83.3 Ma by Ogg et al. (2008), ~83.4 Ma by He et al. (2012), and83.64 Ma by Ogg (2012). According to our new ATS (Fig. 8), the ageof the polarity boundary of C33r/C34n is estimated as 83.633 Ma.

Three short polarity reversal events identified at the upper part ofchron C34n in the K2y have not been observed globally (He et al.,2012). Further studies are needed to confirm their global significance.If their presence is corroborated, the ATS ages for these three eventsare estimated as 84.819 Ma (upper boundary) to 84.862 Ma (lower

Age(Ma)

-0.6

0

0.6

405

kyr

filte

rof

GR

-0.02

0

0.02

100

kyr

filte

rof

Ecc

La1

0a

-0.010

0.01

405

kyr

filte

rof

Ecc

La0

4

0

0.004

0.008

95 k

yr A

Mof

Ecc

La0

4

0

0.04

0.08

Ecc

La0

4

-0.010

0.01

405

kyr

filte

rof

Ecc

La1

0a

0.0040.0060.0080.01

405

kyr

AM

of E

ccL

a04

0

0.2

0.4

0.6

405

kyr

AM

of G

R

0

0.04

0.08

Ecc

La1

0a

-0.4

0

0.4

95 k

yr f

ilter

of G

R

-0.8-0.4

00.40.8

100

kyr

filte

rof

GR

-0.02-0.01

00.010.02

100

kyr

filte

rof

Ecc

La0

4

-0.01

0

0.01

95 k

yr A

Mof

Ecc

La0

4

00.0040.0080.012

95 k

yr f

ilter

of E

ccL

a10a

-0.01

0

0.01

95 k

yr f

ilter

of E

cc10

a

00.10.20.30.4

95 k

yr A

Mof

GR -4

-2

0

2

4

Nor

mal

ized

GR

0.0040.0060.0080.01

405

kyr

AM

0f E

ccL

a10a

83.5 84 84.5 85 85.5 86 86.5 87 87.5 88 88.5 89 89.5 90 90.5 91 91.5 92

83.5 84 84.5 85 85.5 86 86.5 87 87.5 88 88.5 89 89.5 90 90.5 91 91.5 92

TuronianConiacianSantonian.

maC

06.38

03 .68

08.98

K n22 K n2

1 K y22+3 y

K2

1

K qn2

2+3 qK

2

4

K qn2

1 K q23Member

(age, Ma)288.38 85

.748

376.48

010.68

534.09

173.19

768.19

Stage

e3 C34n2ee1C33r

336.38

918.482 68.48

042.58

926.58

289.48290.58

(91.4 ± 0.4)91.341

(90.1 ± 0.4)90.472

(90.4 ± 0.6)90.080

(83.7 ± 0.5)83.834

Polaritychron(Ma)

ATS age(U-Pb age)

(Ma)

E 0405

E 1405

E 2405

E 3405 E 4405E 5405

E 6405E 7405

E 8405E 9405 E 10405

E 11405 E 12405E 13405

E 14405 E 15405E 16405 E 17405

E 18405

E 19405 E 20405 E 21405

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

(p)

(q)

(r)

(s)

(t)

(u)

(v)

(w)

LAE1LAE2

deep lake shallow lake MDL

DEL SL

moderately deep lake deep lake meanderingriver

DEL

64 H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

65H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

boundary), 84.982 Ma (upper boundary) to 85.092 Ma (lowerboundary), and 85.240 Ma (upper boundary) to 85.629 Ma (lowerboundary), with durations of 43 kyr, 110 kyr and 389 kyr, respective-ly (Figs. 2 and 8). Our estimates of the ages and durations of theseshort polarity reversals provide constraints for future identificationand global correlation of such events within the upper CretaceousNormal Superchron (CNS) in other successions.

5.3. Sediment accumulation rates (SAR)

Our ATS for the SK-Is core provides a strong basis for the estima-tion of sediment accumulation rate (SAR) in the SLB. SAR is deter-mined using the e cycles but assuming 405-kyr (E) tuned ages attheir boundaries. The SAR variation through the core shows signifi-cant 405 kyr cycles with SAR maxima corresponding to long eccen-tricity maxima (Fig. 2). It indicates that changes in SAR are directlyrelated to astronomically driven enhanced siliciclastic input duringwet and warm periods.

The SAR variations in SK-Is core can be divided into four stages,which coincide with major environmental changes related to SLB tec-tonic evolution (Feng et al., 2010) (Fig. 2):

(1) The SARs of the K2q3+4 decrease from 22.1 cm/kyr in thelower part to 15.9 cm/kyr in the upper part, with an averageof 18.8 cm/kyr. SARs in the K2q3+4 are related to the deposi-tional environments, which are mainly fluvial channel sand-stones in the lower part, and mudstone, siltstone andsandstone in the upper part.

(2) K2qn consists of mainly mudstone, with the lower part depos-ited in deep lake environments and the upper part in moder-ately deep to shallow lake environments. The SARs of theK2qn vary from 5.6 cm/kyr to 14.4 cm/kyr, with an average of9.27 cm/kyr.

(3) The average SAR of the K2y is 11.8 cm/kyr and varies from8.3 cm/kyr to 13.9 cm/kyr.

(4) The SAR increases abruptly in the K2n1+2, with values from10.7 cm/kyr to 17.5 cm/kyr and an average of 13.2 cm/kyr.The average SAR in the deep lacustrine K2n1+2 is higher thanthat of the underlying shallow lacustrine K2y. This phenome-non seems to conflict with the common depositional modeland may have resulted from tectonic uplift during K2n1+2 de-position, in response to an episode of accelerating uplift fromPacific Plate subduction.

5.4. Songliao Lake anoxic events

Two lacustrine anoxic events (LAEs) in the SLB have been docu-mented in K2qn1 and K2n2 (Fig. 2) (Gao et al., 1994; Li and Pang,2004; Wu et al., 2009). The cause of the LAEs remains uncertain, al-though some researchers believe that they are related to great marineincursions that occurred in the SLB (Wang et al., 2001; Li and Pang,2004; Xi et al., 2011). Based on our new ATS for the SK-Is core, theages of LAE1 and LAE2 are estimated as 91.3 Ma and 83.8 Ma(Fig. 8). This points to global coastal onlap events Tu4 and Cam1, at

Fig. 8. Amplitude modulations of the interpreted eccentricity in the 405 kyr-tuned GR-Th timeal., 2012); (b) SLB formation member, with boundary ages based on the ATS; (c) magnetic po(2012) comparedwith ATS ages (red text); (e) depositional facies of the SLB (LAE=lake anoxiculations of the interpreted 405-kyr cycles in the GR-Th time series; (g) filtered 405-kyr cycles(i) astronomically calibrated GR-Th time series used to obtain results in f–h and j–k; (j) amp95-kyr band of the GR-Th time series; (l) amplitude modulations of the 405-kyr band ofLa2010a short eccentricity; (o) La2010a eccentricity; (p) amplitude modulations of the 95-(r) amplitude modulations of the 405-kyr band of La2004 eccentricity; (s) filtered 405-kyr ban(v) amplitude modulations of the 95-kyr band of La2004 eccentricity; (w) filtered 95-kyr baLa2010a eccentricity filtered series, and in Fig. 7a for the GR-Th filtered time series.

89.54 Ma and 83.1 Ma, respectively, based on the Ogg et al. (2008)timescale (Snedden and Liu, 2010), as possible origins of the incursions.

It is also possible that the LAEs are terrestrial expressions of OceanicAnoxic Event 3 (OAE3), during which a series of carbon cycle eventshave been found throughout late Turonian–early Campanian time, al-though they are restricted to the Atlantic Provinces (Wagreich, 2012).

5.5. Implications for Solar System behavior during the Late Cretaceous

The secular frequencies of planetary motion reflect longitude ofperihelia (gi) and inclination (si) of the orbits for each planet in theSolar System, e.g., g1, g2, g3, g4 and g5 are for Mercury, Venus, Earth,Mars and Jupiter, respectively. Earth's orbital eccentricity variationhas four long-term cycles with modeled periods of ~2.34 Myr,960 kyr, 680 kyr and 405 kyr (Fig. 4). These long-period cycles corre-spond to the interactions between the orbital perihelia of the planets,g4–g3, g1–g5, g2–g1 and g2–g5, respectively. The major terms are the2.34 Myr (g4–g3) and 405 kyr (g2–g5) cycles, both of which stronglymodulate the short eccentricity variation, e.g., (g4–g5)−(g4–g2) and(g3–g5)−(g3–g2). The 2.34 Myr cycle also modulates the 405-kyrcycle (g2–g5)−(g4–g3) and (g2–g5)+(g4–g3) (see Fig. 5b in Laskaret al., 2011). Earth's obliquity variation has a long-period modulationcycle of ~1.2 Myr related to inclination variations of the Earth's orbit-al plane and its interaction with that of Mars, s4–s3, appearing as abeat frequency in the obliquity variation, i.e., (k+s4)−(k+s3),where k is Earth's precession constant (e.g., Hinnov, 2000).

The two orbital motions of Earth and Mars described by g4–g3 ands4–s3 are presently in 2:1 secular resonance (i.e., 2.4 Myr:1.2 Myr),which has persisted as far back as the Oligocene, and documentedin cyclostratigraphy (Pälike et al., 2004; van Dam et al., 2006; Abelset al., 2010).

It is hypothesized that the Solar System experienced chaotic be-havior between 60 Ma and 100 Ma (Laskar et al., 2004, 2011). Be-cause it is impossible to numerically compute the exact motion ofthe planets for times prior to ~50 Ma, Cretaceous cyclostratigraphyuniquely provides an opportunity to detect possible chaotic signalsin the astronomical parameters to pinpoint chaotic intervals.

Thus far in Cretaceous stratigraphy, putative g4–g3 cycleshave been reported in multi-million year long cyclostratigraphicsequences: 2.5 Myr cycles in South Atlantic DSDP cores of theLate Cretaceous (Herbert, 1999), and 1.6 Myr cycles in TethyanAptian strata of central Italy (Huang et al., 2010a). Further back intime, ~2 Myr cycles are present in the Kimmeridgian–TithonianKimmeridge Clay of Dorset, UK (Huang et al., 2010b), 2.0 Myr cy-cles in the Oxfordian Terres Noires Formation, Vocontian Basin,France (Boulila et al., 2010), 1.7 Myr cycles in the Late TriassicNewark Basin, USA (Olsen and Kent, 1999), and 1.8 Myr cycles inthe Triassic bedded chert sequence of Inuyama, Japan (Ikeda etal., 2010). These variable periodicities all point to (although havenot yet confirmed) instability in the Earth–Mars orbits throughoutthe Mesozoic Era. There is also increasing evidence that manyso-called “third order eustatic sequences” are related to s4–s3 forc-ing during the Cenozoic icehouse, and to g4–g3 forcing in the Meso-zoic greenhouses (Boulila et al., 2011).

series. From the top: (a) geologic stage with boundary ages from GTS2012 (Gradstein etlarity chrons with boundary ages according to the ATS; (d) radioisotope ages of He et al.event; MDL=moderately deep lake; DEL=delta, SL=shallow lake); (f) amplitudemod-from the GR-Th time series; (h) filtered short eccentricity cycles in the GR-Th time series;litude modulations of the interpreted 95-kyr band of the GR-Th time series; (k) filteredLa2010a eccentricity; (m) filtered 405-kyr band of La2010a eccentricity; (n) filteredkyr band of La2010a eccentricity; (q) filtered 95-kyr band of the La2010a eccentricity;d of La2004 eccentricity; (t) filtered La2010a short eccentricity; (u) La2004 eccentricity;nd of the La2004 eccentricity. Taner passbands are defined in Fig. 4 for the La2004 and

Table 1The depth and ages of the top boundaries of the ostracode biozones identified in theSK-Is core (Wan et al., 2013–this issue) and their durations.

Formation Ostracode biozones Depth Age Duration

(m) (Ma) (Myr)

Nenjiang C. liaukhenensis–Ilyocyprimorpha 959.55M. magna–M. heiluntszianensis 1015.76 83.81 0.11C. gracila–C. gunsulinensis 1029.33 83.92 0.23C. anonyma–C. fabiforma 1063.05 84.15 0.55

Yaojia C. formosa–C. sunghuajiangensis 1130.38 84.70 0.77C. favosa–M. tabulate 1221.26 85.47 0.27C. exornata–C. dongfangensis 1252.25 85.74 0.22

Qingshankou C. panda–M. obscura 1279.41 85.96 0.25T. vestilus–T. fusiformis–T. pumilis 1310.05 86.21 1.25C. fuyuensis–T. symmetrica 1417.73 87.46 0.43C. edentula–Lycopterocypris grandis 1465.18 87.89 0.39C. nota–S. tumida 1511.09 88.28 1.1C. gibbosa–C. dekhoinensis 1611.99 89.38 0.83T. torsuosus–T. torsuosus var. nota 1683.03 90.21 1.14

Quantou M. longicaudata–C. subtuberculisperga 1781.20 91.35

66 H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

Empirical estimates of the g4–g3 and s4–s3 terms may be recoveredfrom the amplitude modulation (AM) series of the interpreted longeccentricity and obliquity variations in the GR-Th records, and com-pared with those of the theoretical astronomical solutions. TheGR-Th series is longer and more highly resolved than the MS series(Figs. 2 and 3), and so GR-Th was selected for the amplitude

Age

37 k

yr A

Mof

obl

i La0

4

37 k

yr A

Mof

GR

37 k

yr A

Mof

obl

i La1

0a

83.5 84 84.5 85 85.5 86 86.5 87 87.5

83.5 84 84.5 85 85.5 86 86.5 87 87.5

Santonian

.ma

C

06.38

03.68

K n22 K n2

1 K y22+3 y

K2

1

Member(age, Ma)

288.38

376.48

376.48

010.68

Stage

e32ee1C33r

336.38

918.482 68.48

042.58

926.58

289.48290.58

+_(83.7 0.5)83.834

Polaritychron(Ma)

ATS age(U-Pb age)

(Ma)

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

LAE2

deep lake shallow lake MDL

DEL SL

Fig. 9. Amplitude modulations of the interpreted obliquity variation in the SLB, compared wmodulations of the 37.4-kyr band of the GR-Th time series; (g) filtered 37.4-kyr band of(37.4 kyr band); (i) filtered 37.4 kyr band of the La2010a obliquity variation; (j) amplitudband of the La2004 obliquity variation. Bandpass filter definitions are given in Figs. 4 and 7

modulation analysis. The results (Figs. 8 and 9) indicate some similar-ities between data and theory, but also many notable differences.

Amplitude modulations of the interpreted 405 kyr and 95 kyrbands of the GR-Th time series show long period cycling that are par-tially in phase in the younger part of the series, and are decoupled inthe older part (Fig. 8f, j). There is an extended interval of low ampli-tudes associated with the deep and moderately deep lake facies,where carbonate contribution to the sediment may dilute contribu-tions from radioactive carriers (organic carbon, clay, igneous grains).In theory the AM series of the 405-kyr and 95-kyr bands should beanti-phased, as in Fig. 8l, p for La2010a eccentricity and Fig. 8r, v forLa2004 eccentricity. There are numerous possible reasons for themismatching, ranging from amplitude disruptions caused by facieschanges, to the possibility that an eccentricity signal is not actuallypresent in the GR-Th series. On the other hand, the shorteccentricity-filtered GR-Th series (Fig. 8h) shows bundling patternsthat match closely with the direct 405-kyr filtered series, as doesthe same filtering in the Laskar solutions (Fig. 8m, n, s, and t). This, to-gether with the power spectrum of the GR-Th time series (Fig. 7a),constitutes strong evidence for a robust orbital eccentricity signal inthe GR-Th time series.

The obliquity AM series of the GR-Th time series (Fig. 9f) is character-ized by ~1 Myr cycles, along with a long term trend that could be relatedto facies changes (as with the AM eccentricity series). Both La2004 andLa2010a AM obliquity series (Fig. 9h, j) have strong 1.2 Myr cycles;La2004 hasmoreminima that coincidewith GR-Th AMobliquityminima.

(Ma)

37 k

yr f

ilter

of G

R

22

23

24

Obl

iqui

ty(

O)

of L

a04

22

23

24

25

Obl

iqui

ty(O

)of

La1

0a

88 88.5 89 89.5 90 90.5 91 91.5 92

88 88.5 89 89.5 90 90.5 91 91.5 92

TuronianConiacian

08.98K qn2

2+3 qK

2

4

K qn2

1 K q23

534.09

173. 19

768.19

C34n

+_(91.4 0.4)91.341

+_(90.1 0.4)90.472

+_(90.4 0.6)90.080

LAE1

moderately deep lake deep lake meanderingriver

DEL

ith those of the La2010a and La2004 solutions. (a)–(e) are as in Fig. 8; (f) amplitudethe GR-Th time series; (h) amplitude modulations of the La2010a obliquity variatione modulations of the La2004 obliquity variation (37.4 kyr band); (k) filtered 37.4 kyra. La2010a obliquity was obtained using the procedure given in the Appendix.

0 0.001 0.002 0.003 0.004Frequency (cycles/kyr)

0

0.2

0.4

0.6

0.8

10.4

0.57

0.75

1.421.97

2.8

(8)2.9

1.95

0.99

0.75

0.9

1.44

Fig. 11. 2πmultitapered amplitude spectra of the SAR time series (black solid line) andthe AM obliquity (37.4) of the GR-Th time series (red dashed line). Underlined labelsindicate SAR spectrum periods in Myr; plain labels indicate GR-Th AM obliquity pe-riods in Myr.

67H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

Closer comparison of the AM time series for the GR-Th series,La2010a and La2004 eccentricity and obliquity and their spectra(Fig. 10) is, however, non-conclusive: in the time domain, theGR-Th AM series resembles that of the La2004 solution. The twoGR-Th AM spectra (Fig. 10g) are very noisy: the bracketed 8 Myrpeak, which appears in both spectra, reflects the extended suppres-sion of GR amplitudes through the middle of the series related to fa-cies changes. Discounting this frequency from the GR-Th spectrumleaves the following modulation periodicities for consideration: forthe AM eccentricity band, 2.6 Myr and 1.05 Myr; for the AM obliquityband, 2.9 Myr, 1.95 Myr, 1.44 Myr, and 0.99 Myr (Fig. 10g). From thisperspective, the GR-Th series is more comparable to La2010a. In par-ticular, the AM obliquity of La2010a has large 2.25 Myr and 1.17 Myrcomponents (Fig. 10g), compared with GR-Th components at1.95 Myr and 1.44 Myr (Fig. 10h). AM eccentricity-AM obliquitycross-phase studies (not shown) indicate that none of the frequenciesare phase-locked for either the data or the models. Thus it is impossi-ble to conclude confidently whether the 1 Myr to 2 Myr scale modu-lations in the SLB are of astronomical origin.

A final surprise in this work is the spectrum of the SAR time series(Fig. 11). The 405 kyr-tuned SAR time series was evaluated at a100-kyr sampling rate. Its amplitude spectrum shows a strong400-kyr cycle, influenced in part by the 405-kyr tuning. In addition,there are longer, million-year scale variations with periodicitiesof 2.8 Myr, 1.97 Myr, 1.42 Myr, 0.99 Myr and 0.75 Myr. This is re-markably similar to the AM obliquity amplitude spectrum (fromFig. 10g). These results imply that SLB sedimentation rates wereinfluenced by obliquity forcing. Direct evidence for obliquity variationin SLB is relatively weak in the GR-Th series, but relatively strong inthe MS time series (compare Fig. 7a and c). The MS series tracks theconcentration of magnetic minerals in the sediment, all of which aredetrital materials eroded from surrounding areas. The MS spectrumthus suggests a significant component of obliquity forcing in climateprocesses leading to erosion, transport and deposition of SLBsediment.

Age (Ma)84 86 88 90 92

0.6

0.40.2

0

0.012

0.008

0.004

0

0.012

0.008

0.004

0

AM

of

EA

M o

f E

AM

of

E

GR-T h log

La2010a

La2004

(a)

(b)

(c)

(d)

(e)

(f)

Age (Ma)84 86 88 90 92

Fig. 10. Long-period amplitude modulations (AM) of the eccentricity (E) and obliquity (T) baof the obliquity (37.4 kyr) in GR-Th time series, (c) AM of the 405 kyr La2010a eccentricity,(f) AM of the La2004 obliquity (37.4 kyr); and corresponding 2π multitapered amplitude spthe obliquity (37.4 kyr) in GR-Th time series (dashed line), (h) AM of the 405 kyr La2010a ec(h) AM of the 405 kyr La2004 eccentricity (solid line) and AM of the La2004 obliquity (37.4labels indicate AM obliquity periods in Myr.

6. Conclusions

Spectrum and wavelet analysis of the GR (Th for K2y) and MS timeseries measured in the SK-Is borehole and core from the Late Creta-ceous Songliao Basin, China, show strong astronomical signals withfrequency components indicative of 405 kyr long orbital eccentricity,the multiple ~100 kyr terms of short orbital eccentricity, 37.4 kyrobliquity and 20 kyr precession index cycles. An astronomical timescale (ATS) was established by calibrating extracted 405-kyr cyclesto the La2010a astronomical solution. This ATS provides high-resolution estimates of the ages and durations of Late Cretaceous geo-logical and geophysical events in the SLB, as follows:

• The SK-Is borehole covers 8.68 Myr from 83.40 Ma–92.08 Ma(Early Turonian to Early Campanian).

0.250.20.150.10.050

0.002

0.001

0

0.01

0.005

0

0.12

0.08

0.04

000012

0.0008

0.0004

0

0.0015

0.001

0.005

00 0.0005 0.001 0.0015 0.002

Frequency (cycles/kyr)

0.002

0.001

0

0

0.0001

0.0002

0.0003

0

0.001

0.002

0.003 AM

ofT

AM

ofT

(8)

2.9 1.44

0.99

0.75

1.952.6

1.05

5.14

1.173.87

2.341.78

2.16 1.27

AM of EAM of T

(g)

(h)

(i )

0 0.0005 0.001 0.0015 0.002Frequency (cycles/kyr)

AM

ofT

AM

ofT

AM

ofT

AM

ofE

AM

ofE

AM

ofE

AM

ofT

AM of EAM of T

AM of EAM of T

nds, from Figs. 8 and 9. (a) AM of the 405 kyr eccentricity in GR-Th time series, (b) AM(d) AM of the La2010a obliquity (37.4 kyr), (e) AM of the 405 kyr La2004 eccentricity,ectra of (g) AM of the 405 kyr eccentricity in GR-Th time series (solid line) and AM ofcentricity (solid line) and AM of the La2010a obliquity (37.4 kyr) (dashed line), and (i)kyr) (dashed line). Underlined labels indicate AM eccentricity periods in Myr; italicized

68 H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

• The ATS boundary ages for K2n2/K2n1, K2n1/K2y2+3, K2y2+3/K2y1,K2y1/K2qn2+3, K2qn2+3/K2qn1, K2qn1/K2q4 and K2q4/K2q3 are83.882 Ma, 84.673 Ma, 85.748 Ma, 86.010 Ma, 90.435 Ma, 91.371 Maand 91.867 Ma, respectively.

• The durations of ostracode biozones suggest that ostracode turn-over may have been related to astronomical forcing.

• The ATS age of the polarity boundary of C33r/C34n is 83.633 Ma,and three short reversal events at K2y are estimated as 84.819 Ma(top)–84.862 Ma (base), 84.982 Ma (top)–85.092 Ma (base) and85.240 Ma (top)–85.629 Ma (base), respectively.

• Sediment accumulation rates (SAR) of the SK-Is core successionrange from 5.6 cm/kyr to 22.1 cm/kyr with significant ~400 kyrand ~2 Myr periodic variations.

• Two lacustrine anoxic events in the SLB occur at 91.341 Ma (LAE1)and at 83.834 Ma (LAE3), which correspond to times of major globalsea level transgressions (Tu4 and Cam1).

Inner Solar System dynamics can be represented in part by the in-teractions of the orbits of Earth and Mars. This information is embed-ded in the amplitude modulations of Earth's orbital eccentricity andobliquity variations. In the SLB GR-Th series, million-year scale cyclicityis present in amplitude modulations filtered out from the eccentricityand obliquity bands. The extracted modulations do not maintain astrict 2:1 period ratio between g4–g3 and s4–s3 that would be requiredfor Earth–Mars secular resonance. This suggests that during the EarlyTuronian–Early Campanian, the Earth and Mars orbits were deviatedfrom 2:1 secular resonance, and experienced chaotic interactions.There is significant interference from long-term facies changes relatedto the tectonic evolution of the SLB, and so in the future this conclusionneeds to be tested in coeval pelagic marine sequences, where tectonicinfluences are minimal, as for example the Western Interior Seaway(Locklair and Sageman, 2008).

Acknowledgments

We express our sincere appreciation to the editor (Prof. DaveBottjer) and two anonymous reviewers for their careful review andconstructive suggestions that significantly improved the paper. Themagnetic susceptibility data from the SK-Is core was kindly providedby Professor Chenglong Deng from the Institute of Geology and Geo-physics, Chinese Academy of Sciences. This work was jointlysupported by the National Key Basic Research Development Programof China (2012CB822002, 2006CB701400), the National ScienceFoundation of China (Grants 40802012, 91128102) and FundamentalResearch Funds for the Central Universities. LH was partiallysupported by US National Science Foundation Grant EAR-0718905.

Appendix A

Computation of obliquity and precession index for the La2010a so-lution was accomplished on a Linux platform as follows.

(1) The Laskar et al. (1993) astronomical solution FORTRAN codeand input files were downloaded from the online VizieR Cata-logue Service:http://vizier.cfa.harvard.edu/ftp/cats/VI/63/prepa.f.http://vizier.cfa.harvard.edu/ftp/cats/VI/63/prepa.par.http://vizier.cfa.harvard.edu/ftp/cats/VI/63/integ.f.http://vizier.cfa.harvard.edu/ftp/cats/VI/63/integsub.f.http://vizier.cfa.harvard.edu/ftp/cats/VI/63/integ.par.http://vizier.cfa.harvard.edu/ftp/cats/VI/63/climavar.f.http://vizier.cfa.harvard.edu/ftp/cats/VI/63/climavar.par.Mac users will have to usemac2unix (one time only) to convertall files after downloading.

(2) Concatenate integ.f and integsub.f into a single file integall.f.

(3) Compile prepa.f, integall.f and climavar.f with gfortran.(4) Edit prepa.par as follows:

&NAMSTDnomfich='ELL.BIN',nomfichder='DER.BIN',nomascpos='ORBELP.ASC',nomascneg='ORBELN.ASC',datedebut=−100.D0,datefin=0.D0,statut='unknown'&END

(5) Edit integ.par as follows:

&NAMSTDnomfich='ELL.BIN',nomfichder='DER.BIN',pas=200,nechant=5,datefin=0.D0,datedebut=−100.D0,statut='unknown',ecritpos='oui',ecritneg='oui',fichrespos='PRECP.ASC',fichresneg='PRECN.ASC'&END

(6) Edit climavar.par as follows:

&NAMSTDnomascpos='ORBELP.ASC',nomascneg='ORBELN.ASC',nomprecpos='PRECP.ASC',nomprecneg='PRECN.ASC',nomsolpos='CLIVARP.ASC',nomsolneg='CLIVARN.ASC',datedebut=−100.D0,datefin=−0.D0,statut='unknown'&END

(7) Download La2010a orbital solution file La2010a_alkhqp3L.datfrom:http://www.imcce.fr/Equipes/ASD/insola/earth/La2010/index.htmlEdit this file to delete columns 2 and 3 (“a” and “l”); save asORBELN.ASC.(Mac users will have to run the output file throughmac2unix.)

(8) Run prepa, then run integall, choosing FGAM=1 and CMAR=1.3(to match output of La2004); finally, run climavar. The output fileCLIVARN.ASC has 4 columns for t, e, eps, and CP, where t=time inkiloyears before present (negative), e=eccentricity, eps=obliqui-ty (in radians), and CP=sin(longitude of perihelion from movingequinox)=precession index.

References

Abels, H.A., Abdul Aziz, H., Krijgsman, W., Smeets, S.J.B., Hilgen, F.J., 2010. Long-periodeccentricity control on sedimentary sequences in the continental Madrid Basin(middle Miocene, Spain). Earth and Planetary Science Letters 289, 220–231.

Boulila, S., Galbrun, B., Hinnov, L.A., Collin, P.Y., 2008. High-resolution cyclostratigraphicanalysis from magnetic susceptibility in an Upper Kimmeridgian (Upper Jurassic)marl–limestone succession (La Méouge, Vocontian Basin, France). SedimentaryGeology 203, 54–63.

Boulila, S., Galbrun, B., Hinnov, L.A., Collin, P.Y., Ogg, J.G., Fortwengler, D., Marchand, D.,2010. Milankovitch and sub-Milankovitch forcing of the Oxfordian (Late Jurassic)Terres Noires Formation (SE France) and global implications. Basin Research 22,717–732.

69H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

Boulila, S., Galbrun, B., Miller, K.G., Pekar, S.F., Browning, J.V., Laskar, J., Wright, J.D.,2011. On the origin of Cenozoic and Mesozoic “third order” eustatic sequences.Earth-Science Reviews 109, 94–112.

Cande, S.C., Kent, D.V., 1992. A new geomagnetic polarity time scale for the lateCretaceous and Cenozoic. Journal of Geophysical Research 97, 13917–13951.

Cande, S.C., Kent, D.V., 1995. Revised calibration of the geomagnetic polarity timescalefor the late Cretaceous and Cenozoic. Journal of Geophysical Research 100,6093–6095.

Cheng, R.H., Wang, G.D., Wang, P.J., 2008. Sedimentary Cycles of the CretaceousQuantou–Nenjiang Formations and Milankovitch Cycles of the South Hole of theSLCORE-I in the Songliao Basin. Acta Geologica Sinica (Chinese Edition) 82,55–64 (in Chinese with English abstract).

Cheng, R.H., Wang, G.D., Wang, P.J., Gao, Y.F., Ren, Y.G., Wang, C.S., Zhang, S.H., Wang,Q.Y., 2009. Description of Cretaceous sedimentary sequence of the Yaojia Forma-tion recovered by CCSD-SK-Is borehole in Songliao Basin: lithostratigraphy,sedimentary facies and cyclic stratigraphy. Earth Science Frontiers 16, 272–287(in Chinese with English abstract).

Cleveland, W.S., 1979. Robust locally weighted regression and smoothing scatterplots.Journal of the American Statistical Association 74, 829–836.

Deng, C.L., He, H.Y., Pan, Y.X., Zhu, R.X., 2013. Chronology of the terrestrial UpperCretaceous in northeast Asia. Palaeogeography, Palaeoclimatology, Palaeoecology.385, 44–54 (this issue).

Dziewonski, A., Bloch, S., Landisman, M., 1969. A technique for the analysis of transientseismic signals. Bulletin of the Seismological Society of America 59, 427–444.

Feng, Z.Q., Jia, C.Z., Xie, X.N., Zhang, S., Feng, Z.H., Cross, T.A., 2010. Tectonostratigraphicunits and stratigraphic sequences of the nonmarine Songliao basin, northeastChina. Basin Research 22, 79–95.

Gao, R.Q., Zhang, Y., Cui, T.C., 1994. Cretaceous Petroleum Bearing Strata in the SongliaoBasin. Petroleum Industry Press, Beijing, pp. 1–333 (in Chinese with Englishabstract).

Gao, Y.F., Wang, P.J., Cheng, R.H., Wang, G.D., Wan, X.Q., Wu, H.Y., Wang, S.X., Liang,W.L., 2009. Description of Cretaceous sedimentary sequence of the first memberof the Qingshankou Formation recovered by CCSD-SK-Is borehole in SongliaoBasin: Lithostratigraphy, sedimentary facies and cyclic stratigraphy. Earth ScienceFrontiers 16, 314–323 (in Chinese with English abstract).

Gao, Y.F., Wang, P.J., Cheng, R.H., Wang, G.D., Wan, X.Q., Wu, H.Y., Wang, S.X., Liang,W.L., 2011. Centimeter-scale sedimentary sequence description of UpperCretaceous Nenjiang Formation (Lower members 1&2): lithostratigraphy,facies and cyclostratigraphy, based on the scientific drilling borehole in theSongliao Basin. Earth Science Frontiers 18, 195–217 (in Chinese with Englishabstract).

Ghil, M., Allen, M.R., Dettinger, M.D., Ide, K., Kondra-shov, D., Mann, M.E., Robertson,A.W., Saunders, A., Tian, Y., Varadi, F., Yiou, P., 2002. Advanced spectral methodsfor climatic time series. Reviews of Geophysics 40 (1), 3.1–3.41.

Gradstein, F.M., Ogg, J.G., Smith, A.G., 2004. A Geologic Time Scale 2004. CambridgeUniversity Press, Cambridge, U.K. 589 pp.

Gradstein, F.M., Ogg, J.G., Schmitz, M., Ogg, G., 2012. The Geologic Time Scale 2012.Elsevier.

He, H.Y., Deng, C.L., Wang, P.J., Pan, Y.X., Zhu, R.X., 2012. Toward age determination ofthe termination of the Cretaceous Normal Superchron. Geochemistry, Geophysics,Geosystems 13, Q02002. http://dx.doi.org/10.1029/2011GC003901.

Heller, F., Liu, X.M., Liu, T.S., Xu, T.C., 1991. Magnetic susceptibility of loess in China.Earth and Planetary Science Letters 103, 301–310.

Herbert, T.D., 1999. Toward a composite orbital chronology for the Late Cretaceous andEarly Palaeocene GPTS. Philosophical Transactions of Royal Society of London A357, 1891–1905.

Hinnov, L.A., 2000. New perspectives on orbitally forced stratigraphy. Annual Review ofEarth and Planetary Science 28, 419–475.

Hinnov, L.A., Hilgen, F.J., 2012. Cyclostratigraphy and astrochronology. In: Gradstein,F.M., Ogg, J.G., Schmitz, M., Ogg, G. (Eds.), The Geologic Time Scale 2012. Elsevier,pp. 63–83.

Hinnov, L.A., Ogg, J.G., 2007. Cyclostratigraphy and the astronomical time scale. Stratig-raphy 4, 239–251.

Huang, C.J., Hesselbo, S.P., Hinnov, L.A., 2010a. Astrochronology of the late JurassicKimmeridge Clay (Dorset, England) and implications for Earth system processes.Earth and Planetary Science Letters 289, 242–255.

Huang, C.J., Hinnov, L., Fischer, A.G., Grippo, A., Herbert, T., 2010b. Astronomical tuningof the Aptian Stage from Italian reference sections. Geology 38, 899–902.

Husson, D., Galbrun, B., Laskar, J., Hinnov, L.A., Thibault, N., Gardin, S., Locklair, R.E.,2011. Astronomical calibration of the Maastrichtian (Late Cretaceous). Earth andPlanetary Science Letters 305, 328–340.

Ikeda, M., Tada, R., Sakuma, H., 2010. Astronomical cycle origin of bedded chert: a mid-dle Triassic bedded chert sequence, Inuyama, Japan. Earth and Planetary ScienceLetters 297, 369–378.

Laskar, J., Joutel, F., Boudin, F., 1993. Orbital, precessional, and insolation quantitaties forthe Earth from −20 Myr to +10 Myr. Astronomy and Astrophysics 270, 522–533.

Laskar, J., Robutel, P., Joutel, F., Gastineau, M., Correia, A.C.M., Levrard, B., 2004. A longterm numerical solution for the insolation quantities of the Earth. Astronomy andAstrophysics 428, 261–285.

Laskar, J., Fienga, A., Gastineau, M., Manche, H., 2011. La2010: a new orbital solution forthe long-term motion of the Earth. Astronomy and Astrophysics 532, A89.

Li, M.W., Pang, X.Q., 2004. Contentious petroleum geochemical issues in China's sedi-mentary basins. Petroleum Science 1, 4–22.

Locklair, R.E., Sageman, B.B., 2008. Cyclostratigraphy of the Upper Cretaceous NiobraraFormation, Western Interior, U.S.A.: a Coniacian–Santonian orbital timescale. Earthand Planetary Science Letters 269, 540–553.

Lourens, L.J., Hilgen, F.J., 1997. Long-periodic variations in the Earth's obliquity andtheir relation to third-order eustatic cycles and late Neogene glaciations. Quaterna-ry International 40, 43–52.

Mann, M.E., Lees, J.M., 1996. Robust estimation of background noise and signaldetection in climatic time series. Climate Change 33, 409–445.

Mayer, H., Appel, E., 1999. Milankovitch cyclicity and rock-magnetic signatures ofpalaeoclimatic change in the Early Cretaceous Biancone Formation of the SouthernAlps, Italy. Cretaceous Research 20, 189–214.

Mitchell, R.N., Bice, D.M., Montanari, A., Cleaveland, L.C., Christianson, K.T., Coccioni, R.,Hinnov, L.A., 2008. Ocean anoxic cycles? Prelude to the Livello Bonarelli (OAE 2).Earth and Planetary Science Letters 267, 1–16.

Nádor, A., Lantos, M., Toth-Makk, A., Thamo-Bozso, E., 2003. Milankovitch-scale multi-proxy records from fluvial sediments of the last 2.6 Ma, Pannonian Basin, Hungary.Quaternary Science Reviews 22, 2157–2175.

Ogg, J.G., 2012. Geomagnetic polarity time scale. In: Gradstein, F.M., Ogg, J.G., Schmitz,M., Ogg, G. (Eds.), The Geologic Time Scale 2012. Elsevier, pp. 85–113.

Ogg, J.G., Hinnov, L.A., 2012. Cretaceous. In: Gradstein, F.M., Ogg, J.G., Schmitz, M., Ogg,G. (Eds.), The Geologic Time Scale 2012. Elsevier, pp. 815–878.

Ogg, J.G., Ogg, G., Gradstein, F.M., 2008. The Concise Geologic Time Scale. CambridgeUniversity Press, pp. 1–177.

Olsen, P.E., Kent, D.V., 1999. Long-period Milankovitch cycles from the Late Trias-sic and Early Jurassic of eastern North America and their implications for thecalibration of the early Mesozoic time-scale and the long-term behaviour ofthe planets. Philosophical Transactions of the Royal Society of London 357,1761–1786.

Paillard, D., Labeyrie, L., Yiou, P., 1996. Macintosh program performs time-series anal-ysis. Eos 77, 379.

Pälike, H., Laskar, J., Shackleton, N.J., 2004. Geologic constraints on the chaotic diffusionof the solar system. Geology 32, 929–932.

Pei, F.P., Xu, W.L., Yang, D.B., Zhao, Q.G., Liu, X.M., Hu, Z.C., 2007. Zircon U–Pb geochro-nology of basement metamorphic rocks in the Songliao Basin. Chinese ScienceBulletin 52, 942–948.

Schnyder, J., Ruffell, A., Deconinck, J.-F., Baudin, F., 2006. Conjunctive use of spectralgamma-ray logs and clay mineralogy in defining late Jurassic–early Cretaceouspalaeoclimate change (Dorset, U.K.). Palaeogeography, Palaeoclimatology,Palaeoecology 229, 303–320.

Shackleton, N.J., Crowhurst, S.J., Weedon, G.P., Laskar, J., 1999. Astronomical calibrationof Oligocene–Miocene time. Philosophical Transactions of the Royal Society ofLondon A 357, 1909–1927.

Snedden, J.W., Liu, C., 2010. A compilation of Phanerozoic sea-level change, coastal onlapsand recommended sequence designations, Search and Discovery Article #40594.Available at http://www.searchanddiscovery.com/documents/2010/40594snedden/ndx_snedden.pdf.

Taner, M.T., 2000. Attributes Revisited, Technical Publication. Rock Solid Images, Inc.,Houston, Tex (Available at http://rocksolidimages.com/pdf/attrib_revisited.htm).

Ten Veen, J.H., Postma, G., 1996. Astronomically forced variations in gamma-ray intensity:Late Miocene hemipelagic successions in the eastern Mediterranean basin as a testcase. Geology 24, 15–18.

Thomson, D.J., 1982. Spectrum estimation and harmonic analysis. Proceedings of theIEEE 70, 1055–1096.

Torrence, C., Compo, G.P., 1998. A practical guide towavelet analysis. Bulletin of theAmerican Meteorological Society 79, 61–78.

van Dam, J.A., Abdul Aziz, H., Ángeles Álvarez Sierra, M., Hilgen, F.J., van den HoekOstende, L.W., Jourens, L.J., Mein, P., van der Meulen, A.J., Pelaez-Campomanes, P.,2006. Long-period astronomical forcing of mammal turnover. Nature 443,687–691.

Wagreich, M., 2012. "OAE 3" – regional Atlantic organic carbon burial during theConiacian–Santonian. Climate of the Past 8, 1447–1455.

Wan, X., Zhao, J., Scott, R.W.,Wang, P., Feng, Z., Huang, Q., Xi, D., 2013. Late Cretaceous stra-tigraphy, Songliao Basin, NE China: SK1 cores. Palaeogeography, Palaeoclimatology,Palaeoecology. 385, 31–43 (this issue).

Wang, P.J., Liu, W.Z., Shan, X.L., Bian, W.H., Ren, Y.G., Du, X.D., Wan, C.B., Cheng,R.H., 2001. Depositional Events: Introduction, Example, Application. JilinScience and Technology Press, Changchun, pp. 1–182 (in Chinese with Englishabstract).

Wang, Y., Zhang, F.Q., Zhang, D.W., Miao, L.C., Li, T.S., Jie, W.Q., Meng, Q.R., Liu, D.Y.,2006. Zircon SHRIMP U–Pb dating of meta‐diorite from the basement of theSongliao Basin and its geological significance. Chinese Science Bulletin 51,1877–1883.

Wang, P.J., Xie, X.A., Frank, M., Ren, Y.G., Zhu, D.F., Sun, X.M., 2007. The CretaceousSongliao Basin: valcanogenic succession, sedimentary sequence and tectonicevolution, NE China. Acta Geologica Sinica 81, 1002–1011.

Wang, C.S., Huang, Y.J., Zhao, X.X., 2009a. Unlocking a Cretaceous geologic and geo-physical puzzle: scientific drilling of Songliao Basin in northeast China. The LeadingEdge 28, 340–344.

Wang, G.D., Cheng, R.H., Wang, P.J., Gao, Y.F., Wang, C.S., Ren, Y.G., Huang, Q.H., 2009b.Description of Cretaceous sedimentary sequence of the Quantou Formationreovered by CCSD-SK-I s borehole in Songliao Basin: lithostratigraphy, sedimenta-ry facies and cyclic stratigraphy. Earth Science Frontiers 16, 324–338 (in Chinesewith English abstract).

Wang, P.J., Gao, Y.F., Cheng, R.H., Wang, G.D., Wu, H.Y., Wan, X.Q., Yang, G.S.,Wang, Z.X., 2009c. Description of Cretaceous sedimentary sequence of thesecond and third member of the Qingshankou Formation recovered byCCSD-SK-I s borehole in Songliao Basin: lithostratigraphy, sedimentary faciesand cyclic stratigraphy. Earth Science Frontiers 16, 288–313 (in Chinesewith English abstract).

70 H. Wu et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 385 (2013) 55–70

Weber, M.E., Tougiannidis, N., Kleineder, M., Bertram, N., Ricken, W., Rolf, C.,Reinsch, T., Antoniadis, P., 2010. Lacustrine sediments document millennial-scale climate variability in northern Greece prior to the onset of the northernhemisphere glaciation. Palaeogeography, Palaeoclimatology, Palaeoecology291, 360–370.

Weedon, G., 2003. Time-Series Analysis and Cyclostratigraphy. Cambridge UniversityPress, Cambridge, pp. 1–259.

Westerhold, T., Röhl, U., Laskar, J., Raffi, I., Bowles, J., Lourens, L.J., Zachos, J.C., 2007. Onthe duration of magnetochrons C24r and C25n and the timing of early Eoceneglobal warming events: implications from the Ocean Drilling Program Leg 208Walvis Ridge depth transect. Paleoceanography 22, PA2201.

Wu, H.C., Zhang, S.H., Sui, S.W., Huang, Q.H., 2007. Recognition of Milankovitch cyclesin the natural gamma‐ray logging of Upper Cretaceous terrestrial strata in theSongliao Basin. Acta Geologica Sinica 81, 996–1001.

Wu, H.C., Zhang, S.H., Jiang, G.Q., Huang, Q.H., 2009. The floating astronomical timescale for the terrestrial Late Cretaceous Qingshankou Formation from the Songliao

Basin of Northeast China and its stratigraphic and paleoclimate implications. Earthand Planetary Science Letters 278, 308–323.

Wu, X.S., Guo, J.J., Huang, Y.J., Fu, J.W., 2011. Well logging proxy of the LateCretaceous palaeoclimate change in Songliao Basin. Journal of Palaeogeography13, 103–110.

Wu, H.C., Zhang, S.H., Feng, Q.L., Jiang, G.Q., Li, H.Y., Yang, T.S., 2012. Milankovitch andsub-Milankovitch cycles of the Early Triassic Daye Formation, South China andtheir geochronological and paleoclimatic implications. Gondwana Research 22,748–759.

Xi, D.P., Wan, X.Q., Feng, Z.Q., Li, S., Feng, Z.H., Jia, J.Z., Jing, X., Si, W.M., 2011. Discoveryof Late Cretaceous foraminifera in the Songliao Basin: evidence from SK-I andimplications for identifying seawater incursions. Chinese Science Bulletin 56,253–256.