Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Detecting Sedimentary Cycles using Autocorrelation of Grain size

Detecting Sedimentary Cycles using Autocorrelation of Grain size Detecting Sedimentary Cycles using Autocorrelation of Grain size SUBJECT AREAS: 1,2 1 2 Shangbin Xiao , Rui Li & Muhong Chen GEOMORPHOLOGY PALAEOCLIMATE 1 2 China Three Gorges University, Yichang 443002, China, South China Sea Institute of Oceanology, Chinese Academy of ATTRIBUTION Sciences, Guangzhou 510301, China. GEOLOGY Detection of sedimentary cycles is difficult in fine-grained or homogenous sediments but is a prerequisite for the interpretation of depositional environments. Here we use a new autocorrelation analysis to detect Received cycles in a homogenous sediment core, E602, from the northern shelf of the South China Sea. 2 December 2012 Autocorrelation coefficients were calculated for different mean grain sizes at various depths. The results Accepted show that sediments derived from rapid depositional events have a better autocorrelation. Analysis of two 8 March 2013 other cores confirms this result. Cores composed of sediments deposited quickly under stable and/or gradually changing hydrodynamic conditions, have higher autocorrelation coefficients, whereas, those Published composed of sediments deposited during calm periods have relatively low autocorrelation coefficients. It 11 April 2013 shows that abrupt changes in autocorrelation coefficients usually indicate the existence of a boundary between adjacent sedimentary cycles, with each cycle beginning with a high positive autocorrelation coefficient of grain size and ending with a low negative one. Correspondence and requests for materials he detection of sedimentary cycles in cores is key to understanding their depositional environments, and should be addressed to determining if such data can be used to reconstruct high-resolution palaeoclimatic records. Sedimentary S.B.X. (shangbinx@ T cycles are commonly identified using qualitative assessments of changes in, for example, grain size, sedi- 1–8 mentary structure, and the stacked patterns of the sedimentary layers . However, the precise identification of 163.com) 4,9,10 sedimentary cycles can sometimes be rather difficult, especially within relatively homogeneous sediments . For example, Core NS97-13, a sedimentary sample collected from the South China Sea at a water depth of 2120 m (8u20.389N, 115u55.389E), is lithologically homogeneous and composed of mud . Analysis of the anisotropy of the magnetic susceptibility and the results of AMS C dating indicate that the sediments may have been deposited from turbidity currents. However, it is very difficult to identify any sedimentary cycles within the core, either visually or by measuring and analysing the grain size. This is a common problem specifically noted in other 9,10,12,13 studies . Correlation is a mathematical tool frequently used in signal processing for analysing functions or series of values, and it describes the mutual relationship between two or more random variables. Autocorrelation describes 14,15 the correlation of a set of data with itself . This is not the same as cross-correlation, which is the correlation between two different signals . A time-series data set can be separated into three components, namely the overall trend, the noise, and the cyclicity. Many researchers are interested in removing the trend and use autocorrelation to reveal important information about the temporal behaviour of the system. Autocorrelation may be also 16,17 pervasive in sedimentary profiles , and has been used between pixels in digital sediment images to measure 16,18–21 average grain-size . In the present study, we analyze Core E602, a sample collected from the shelf of the northern South China Sea (Fig. 1), develop a new mathematical method that involves the autocorrelation analysis of the grain size and discuss the mechanisms of sediments transportation involved. Our new method enables the detection of sedi- mentary cycles within any core, but within lithologically homogeneous cores in particular. Moreover, the method is quantitative and eliminates subjective interpretation that can be difficult to replicate between investigators and studies. Results Core E602 has two lithostratigraphic units determined by changes in lithology and color (Fig. 2). The lower unit (37 to 368 cm) is composed of dark-gray silty sand and is homogeneous with little variation in sand, silt and clay content. The upper unit (0 to 37 cm) changes gradually up-core from silty-sand to sandy-silt. The clay content also appears to increase gradually in the same direction. SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 1 www.nature.com/scientificreports 36 37 Figure 1 | Location of Cores E602 (A) and DD2 and PC-6 (B). Maps of sediment distribution (A) and (B) are reproduced from Liu et al. and Qin et al. , respectively. Figure 2 | Grain size, chronology, and granulometery of Core E60. SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 2 www.nature.com/scientificreports Figure 3 | Distribution of accumulation probability (a) and skewness vs. standard deviation of grain size distribution (b) for Core E602. As shown in Fig. 3(a), all of the 886 sediment samples from Core segment dominated the CM patterns for the lower unit, whereas E602 are composed of saltation and suspension fractions, the bound- the RS segment was episodic. The RS segment determined the CM ary of which is located at approximately 3.8 PHI. The suspension patterns for the upper unit, which implies that a uniform suspension fraction ranges in content from 15 to 35%. The slope of the salutation during the deposition of the unit was controlled by traction currents fraction, which reflects a sorting trend, remains almost the same (Fig. 7). Furthermore, all of the sediment samples fall within a small throughout both core units. region in the CM plot (Fig. 5a), which indicates that they are well The standard deviation of grain-size distributions of the sediment sorted. It has already been noted that the upper unit of core E602 samples obtained from the lower core unit ranges from 1.3 to 1.6, and exhibits an increase in fines sediment from top to bottom, as well as their skewness ranges from 1.4 to 2.5. Samples obtained from the decreases in grain-size and sorting. However, also evident from the upper unit exhibit little change in sediment sorting, with standard CM plot (Fig. 5b) are the relatively larger C values of sediment from deviations in the range 1.6 – 2.2. However, the skewness of grain-size 0-25 cm depth, signifying that the sediments are coarser, than those distributions of the samples decreases to 0.6 to 1.6, perhaps as a result obtained from below 25 cm. of the higher fine-sediment fraction in the upper unit. Given that the The foraminifera samples at the base of core E602 (348–350 cm) bimodal frequency curves shown in Fig. 4 are typical of grain-size gave a C age of 10289-10584 a BP (Table 1), which suggests that the distributions for all the samples from the core, the higher content of deposition began during the early Holocene. According to a recent fine sediment fraction did not obviously affect sediment sorting in report on sea level changes in the South China Sea , sea level rose the samples, but did result in a reduced degree of the skewness. from approximately 235 m to 220 m during this period (,10 to The CM patterns in core sediments (Fig. 5(a)) plotted on the QR ,9 ka BP). Combining water depth at the coring site, and the core segment, which represents graded suspension, and the RS segment, length therefore gives a paleowater depth of ,30–45 m, placing it 22,23 which represents uniform suspension (Fig. 6(b)). The QR within the shallow continental shelf environment. The uniform content of sand, silt and clay, and granulometery in the lower unit of E602, implies reworking of late Pleistocene sedi- ments composed of medium and fine sand with high textural matur- ity. By contrast, fine-grained terrigenous sediment dominates the upper unit, with the coarsest sediment being deposited at the top of the core (Fig. 5b). The characteristics of Core E602, in terms of grain size, sediment- ary structure (Fig. 7b) and CM patterns, indicate that its sediments were, for the most part, deposited under stormy conditions, which is consistent with the modern reports of frequent tropical storms . Resuspension occurs when deep water waves enter water shallower than one-half the wave length (i.e. the wave base) . The wave base under a fair-weather is less than 10–15 m . However, the storm deposit extends about 40 m , i.e., bottom sediments below the fair-weather wave base would be reworked under storm condition . Given that each storm can gradually strengthen, reduce, or maintain its intensity, means that it could produce sedimentary units with Figure 4 | Grain size distribution of several samples obtained from Core decreasing (segment (21) in Figure 8), increasing (segment (9) in E602. Figure 8), or invariable mean sediment-size with depth respectively. SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 3 www.nature.com/scientificreports Figure 5 | CM plots of Core E602. Passega’s CM image technique, where C is the coarsest one percentile (the value at 99% on a cumulative curve) and M 23,34 is the median grain-size, has been widely used to relate the grain-size characteristics of sediments to the processes of their deposition. Passega and Passega & Byramjee made a distinction between the following types of sediment movement: (1) particles rolled on the bed, even when there is no turbulence. (2) Graded suspension (the term is preferred to ‘saltation’), composed of sand and coarseness decrease from the bottom of the river to an elevation of 2 m upwards. The size of the coarsest particles in the graded suspension (Cs) fluctuates according to the maximum turbulence when settling begins. (3) Overlying graded suspension, Passega described uniform suspension, defined by a constant concentration of particles at the top of the water column. Uniform suspension may be in direct contact with the bed when turbulence decreases below a threshold, allowing the graded suspension to settle on the bed. Slow-flowing rivers display uniform suspension, even during floods . Figure 6 | Different CM patterns within Core E602. Figure 7 | CM plot of the upper unit of Core E602 (a) and a typical sedimentary pattern resulting from a storm current (b). SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 4 www.nature.com/scientificreports Table 1 | AMS C data of Core E602 Calendar age/a BP Depth(cm) Material C age/a BP (1 sigma) (2 sigma) 48,50 Nummulites 8790 6 40 9068,9597 9019,9756 148,150 8850 6 45 9097,9742 9050,9807 248,250 8740 6 50 9038,9576 8988,9751 348,350 Nummulites, Puedorotaria, Elphidum 9690 6 50 10289,10584 10201,11119 First-order autocorrelation coefficients of mean grain-size vary periods of calm weather. Assuming sediment provenance is invariable with sample size (5–30) at different depths, as shown in Figure 8. and water depth does not change, calm hydrodynamic conditions Autocorrelation coefficients vary considerably at some depths, for produce little variation in grain size, and thus no abrupt changes in example from 20.59 to 0.50 at ,27 cm, but change little at other the autocorrelation coefficients of mean grain-size. However, storms or depths, for example at,70 cm. Figure 9 illustrates how the points of tsunamis can alter the hydrodynamic conditions considerably, and can abrupt change in autocorrelation coefficient fit remarkably well with cause rapid accumulation of sediment with significant changes in grain 1,4,8 changes in mean grain-size of the sediment samples. size . A sedimentary cycle may be formed during a single storm event, and depending on changes in current intensity, can result in either a graded or ungraded sequence. Thus, although there are some- Discussion The accumulation of sediment can be either rapid, for example during times no discernable changes in grain size within a particular sequence storms as a result of turbidity currents, or it can be much slower during of storms, the degree of autocorrelation may be high due to the inertia Figure 8 | Variation in the first-order autocorrelation coefficients of mean grain-size with sample size for Core E602. The numbers 1-24 in parentheses refer to different sedimentary layers/cycles. SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 5 www.nature.com/scientificreports Figure 9 | Variation in the first-order autocorrelation coefficients of mean grain-size with sample size for Core DD2. A vertical abrupt change (C3, which recorded a storm event) in the gradient of the contour of autocorrelation coefficients of the core is indicated by an arrow. or internal coherence that is inherent in the rapid accumulation of base of a new cycle, which results in a high degree of autocorrelation sediment. Figure 8 shows that sharp changes in autocorrelation coef- of grain size. This may in turn result in a degree of inertia of grain size ficient tend to be found at the boundaries between different cycles that within a sedimentary profile. Thus, a cycle begins with high positive may have resulted from large differences in the energy of successive values of autocorrelation coefficient and ends with low negative ones. storm currents. It is this inertia that allows the application of the autocorrelation In addition to storm cessation, a sediment cycle could terminate analysis of grain size to the determination of sedimentary cycles in for several other reasons, including abrupt changes in water depth or cores of this type. 4,8,28 sediment supply . However, in the studies cores, relatively rapid It is relatively easy to identify sedimentary sequences or cycles sediment accumulation or more stable sedimentation occurs at the those are composed of individual layers that are either normally Figure 10 | Variation in first-order autocorrelation coefficients of mean grain-size with sample size for Core PC-6. The two black bold vertical lines indicate the boundaries of neighbouring cycles; multiple black dot-dashed lines indicate the boundaries of neighbouring subcycles; and the red vertical line, which shows an abrupt change in the gradients of the contour of the autocorrelation coefficients of the core, indicates the base of a storm layer. Note that no obvious subcycles of sedimentation can be discerned in the muddy unit above 450 cm using traditional sedimentological methods. SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 6 www.nature.com/scientificreports shows a range of sedimentary dynamics associated with foreshore, nearshore and graded (Layers 19 & 20 in Fig. 8) or reverse-graded (Layers 8 & 9 in shallow continental shelf environments . Fig. 8), because any abrupt changes in grain size at the boundaries As is known ,the correlation coefficient of a pair of random variables (X, Y)is between these layers are clear. However, in some cases there are no given by the formula: clear changes in grain size within a sequence that is composed of n ðÞ x { xðÞ y {y i i sediments deposited during a number of different events (e.g. Layers i~1 r~sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð1Þ 10,13 & 17,18 in Fig. 8). Under such circumstances, it is difficult to n n P P 2 2 ðÞ x { x ðÞ y {y determine the number of sedimentary cycles with any accuracy. In i i i~1 i~1 Core E602, for example, the small variation in mean grain-size values from ,183 to ,100 cm (Layers 10,13 in Fig. 8) formed a cycle of Where x and y are the average values of X and Y, respectively; Xi and Yi are the i-th specific variable values of {(x , y ): i 5 1,.., n; i i sedimentation, similar to an aggradational stack. It was difficult to i 5 a particular case i; identify the sedimentary cycles or sequences of the core in terms of n 5 the number of cases. changes in mean grain-size alone. Fortunately, autocorrelation ana- 16 The autocorrelation coefficient of a series of data {x :i 5 1,.., n}, is also calculated lysis can reveal the hidden relationships that affect sediments at by the same formula above except y is replaced with x in a form lagged by one (i.e., i i11 first order autocorrelation) or with x in a form lagged by k (i.e., k order different depths. Abrupt changes in autocorrelation coefficient of i1k autocorrelation). grain size can provide a clear indication of the boundaries between Because the number of cases adopted has direct influence on the values of both the neighbouring sedimentary layers. For example, at least 24 deposi- autocorrelation coefficient of the variable X and correlation coefficient of a pair tional cycles have been identified in Core E602 (Fig. 8). variables (X, Y), varying number of cases (range: 3,20) from a successive grain size data are taken here to calculate the autocorrelation coefficients and disclose their In a further example, the muddy unit of Core DD2 was formed in changes with depth. a relatively stable sedimentary environment and during a period of calm weather . However, the sediments at ,107 cm in the core (C3 on Fig. 9) may be the result of the combined influence of a winter 1. Kortekaas, S. & Dawson, A. Distinguishing tsunami and storm deposits: an example from Martinhal, SW Portugal. Sedimentary Geology 200, 208–221 coastal current and a storm current, because there is also an abrupt (2007). change in the autocorrelation coefficients of the core at this point. 2. Horikawa, K. & Ito, M. Non-uniform across-shelf variations in thickness, grain Core PC-6 provides a third example (Fig. 10). In this core, sedi- size, and frequency of turbidites in a transgressive outer-shelf, the Middle ments deposited during different environmental conditions appear Pleistocene Kakinokidai Formation, Boso Peninsula, Japan. Sedimentary Geology to have different characteristics of grain size and stacked patterns. 220, 105–115 (2009). 3. Gonza´ lez-Alvarez, R. et al. Paleoclimatic evolution of the Galician continental Abrupt changes in the autocorrelation coefficients of the core shelf (NW of Spain) during the last 3000 years: from a storm regime to present occurred not only at these boundaries, but also at the boundaries conditions. Journal of Marine Systems 54, 245–260 (2005). of the subcycles of sedimentation. The latter are sometimes difficult 4. Morton, R. A., Gelfenbaum, G. & Jaffe, B. E. Physical criteria for distinguishing to identify within relatively homogeneous sedimentary profiles. At sandy tsunami and storm deposits using modern examples. Sedimentary Geology 200, 184–207 (2007). the same time, the storm deposit event reported at ,100 cm is 5. Sun, Y., Gao, S. & Li, J. Preliminary analysis of grain-size populations with clearly revealed by the abrupt change in the autocorrelation coeffi- environmentally sensitive terrigenous components in marginal sea setting. cients of grain size within the core (Fig. 10). Chinese Science Bulletin 48, 184–187 (2003). A degree of autocorrelation in sedimentary profiles is a widespread 6. Anthony, E. J. & He´ quette, A. The grain-size characterisation of coastal sand from the Somme estuary to Belgium: sediment sorting processes and mixing in a tide- phenomenon, and is a result of the similarity of the hydrodynamic and storm-dominated setting. Sedimentary Geology 202, 369–382 (2007). conditions present in the stable sedimentary environment that char- 7. Sedgwick, P. E. & Davis, R. A. Stratigraphy of washover deposits in Florida: acterized the sedimentation . Sediments derived from storms or implications for recognition in the stratigraphic record. Marine geology 200, other events have a higher degree of autocorrelation as a result of 31–48 (2003). 8. Bentley, S. J., Sheremet, A. & Jaeger, J. M. Event sedimentation, bioturbation, and their rapid deposition , and abrupt changes in autocorrelation coef- preserved sedimentary fabric: Field and model comparisons in three contrasting ficient usually indicate the boundaries between successive sediment- marine settings. Continental Shelf Research 26, 2108–2124 (2006). ary cycles. The analysis of autocorrelation coefficients is therefore a 9. Tuttle, M. P., Ruffman, A., Anderson, T. & Jeter, H. Distinguishing tsunami from useful tool in the identification of sedimentary sequences or cycles, storm deposits in eastern North America: the 1929 Grand Banks tsunami versus and our results confirm its potential of for detecting sedimentary the 1991 Halloween storm. Seismological Research Letters 75, 117–131 (2004). 10. Dawson, A. G. & Shi, S. Tsunami deposits. Pure and applied geophysics 157, cycles in homogeneous sediment sequences. Combined with the 875–897 (2000). 16,18–21 use of rapid techniques to determine grain-size from digital 11. Tang, X., Chen, M., Liu, J., Zhang, L. & Chen, Z. The anisotropy of magnetic sediment images, autocorrelation analysis should therefore improve susceptibility of Core NS97-13 sediments from the Nansha Islands sea area in the the accuracy of sedimentary environmental determination. southern South China Sea. Acta Oceanologica Sinica 31, 69–76 (in Chiness with English Abstract) (2009). 12. Allison, M. A., Sheremet, A., Gon˜ i, M. A. & Stone, G. W. Storm layer deposition on Methods the Mississippi–Atchafalaya subaqueous delta generated by Hurricane Lili in Core E602, with a length of 3.68 m, was collected with gravity sampling pipe in 2002. Continental Shelf Research 25, 2213–2232 (2005). October 2004 from a sandy area on the northern shelf of the South China Sea 13. Goff, J., McFadgen, B. & Chague´ -Goff, C. Sedimentary differences between the (112u14.889E, 20u44.939N), in a water depth of 65 m, and distant from any estuarine 2002 Easter storm and the 15th-century Okoropunga tsunami, southeastern influence. In the laboratory, the core was described in detail and then split into a total North Island, New Zealand. Marine geology 204, 235–250 (2004). of 886 samples (300 samples at 0.25 cm intervals from 0 to 75 cm and 586 samples at 14. Phillips, C., Parr, J. & Riskin, E. Signals, Systems, and Transforms. (Prentice Hall 0.5 cm intervals from 75 to 368 cm). The samples were initially pre-processed by the 1999). addition of excess H O (Q 5 30%), followed by the addition of HCl (3N). The 15. Herman, E. K., Toran, L. & White, W. B. Quantifying the place of karst aquifers in 2 2 distribution of grain size was classified following Wentworth , and measured using a the groundwater to surface water continuum: A time series analysis study of storm Mastersizer 2000 from Malvern Instruments Ltd., with a range of measurement of behavior in Pennsylvania water resources. Journal of Hydrology 376, 307–317 0.02 to 2000 mm and a resolution of 0.01 PHI. The errors in mean grain size for the (2009). same sample obtained following repeated measurements were less than 3%. The mean 16. Rubin, D. M. A simple autocorrelation algorithm for determining grain size from grain size (unit: PHI) was used to calculate the autocorrelation coefficients of the grain digital images of sediment. Journal of Sedimentary Research 74, 160–165 (2004). size data. 17. Xiao, S., Liu, W., Li, A., Yang, S. & Lai, Z. Pervasive autocorrelation of the chemical Radiocarbon dating of the core sediments was carried out at the National Ocean index of alteration in sedimentary profiles and its palaeoenvironmental Sciences Accelerator Mass Spectrometry Facility (NOSAMS), Woods Hole implications. Sedimentology 57, 670–676 (2009). Oceanographic Institution. All the C ages (Table 1) were calibrated to give calendar 18. Barnard, P., Rubin, D., Harney, J. & Mustain, N. Field test of an autocorrelation ages using CALIB4.3 . technique for determining grain size using a digital camera. Sedimentary Geology The significance of the autocorrelation of grain size is herein discussed in the 201, 180–195 (2007). context of two other sedimentary cores obtained from the inner shelf of the East 19. Warrick, J. A. et al. Cobble Cam: Grain-size measurements of sand to boulder 29 30 China Sea that have previously been described, namely DD2 and PC-6. Core DD2 from digital photographs and autocorrelation analyses. Earth Surface Processes is composed of muddy sediments formed in a shallow continental shelf. Core PC-6 and Landforms 34, 1811–1821 (2009). SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 7 www.nature.com/scientificreports 20. Buscombe, D. & Masselink, G. Grain-size information from the statistical 33. Hughen, B. K., McCormac, G., van der Plicht, J. & Spurk, M. INTCAL98 properties of digital images of sediment. Sedimentology 56, 421–438 (2008). radiocarbon age calibration, 24,000-0 cal BP. Radiocarbon 40, 1041–1083 (1998). 21. Buscombe, D. Estimation of grain-size distributions and associated parameters 34. Passega, R. Texture as characteristic of clastic deposition. AAPG Bulletin 41, from digital images of sediment. Sedimentary Geology 210, 1–10 (2008). 1952–1984 (1957). 22. Bravard, J.-P. & Peiry, J.-L. The CM pattern as a tool for the classification of 35. Passega, R. & Byramjee, R. GRAIN-SIZE IMAGE OF CLASTIC DEPOSITS. alluvial suites and floodplains along the river continuum. Geological Society, Sedimentology 13, 233–252 (1969). London, Special Publications 163, 259–268 (1999). 36. Liu, Z. S., Zhao, H. T., Fan, S. Q. & Chen, S. Q. Geology of the East China Sea (in 23. Passega, R. Grain size representation by CM patterns as a geologic tool. Journal of Chinese). Beijing: Science Press (2002). Sedimentary Research 34, 830–847 (1964). 37. Qin, Y. S., Zhao, Y. Y., Chen, L. R. & Zhao, S. L. Geology of the East China Sea (in 24. Hori, K. et al. Delta initiation and Holocene sea-level change: example from the Chinese). Science Press, Beijing (1987). Song Hong (Red River) delta, Vietnam. Sedimentary Geology 164, 237–249 (2004). 25. Davies, J. Geographical variation in coastal development. (Longman Publishing Group, 1972). Acknowledgements 26. Evans, R. D. Empirical evidence of the importance of sediment resuspension in We thank Dr. Paul Blanchon, who is at Institute of Marine Sciences and Limnology, lakes. Hydrobiologia 284, 5–12 (1994). National University of Me´ xico, for giving highly constructive suggestions for the first 27. Wright, M. E. & Walker, R. G. Cardium Formation (U. Cretaceous) at Seebe, manuscript. This work was sponsored by National Science Foundation of China Alberta-storm-transported sandstones and conglomerates in shallow marine (No. 41273110) and the China Postdoctoral Special Science Foundation (No. 200801440). depositional environments below fair-weather wave base. Canadian Journal of Earth Sciences 18, 795–809 (1981). 28. Kakinoki, T., Tsujjimoto, G., Yamada, F., Sakai, D. & Uno, K. Beach profile and Author contributions sediment characteristics of a mixed sand beach under diurnal sea level variations. Xiao, S. B. carried out grain-size analysis and prepared the primary manuscript. Li, R. Journal of Coastal Research SI 64, 765–770 (2011). prepared the figures. All authors reviewed and discussed the manuscript. 29. Xiao, S. et al. Recent 2000-year geological records of mud in the inner shelf of the East China Sea and their climatic implications. Chinese Science Bulletin 50, 466–471 (2005). Additional information 30. Xiao, S. et al. Coherence between solar activity and the East Asian winter monsoon Competing financial interests: The authors declare no competing financial interests. variability in the past 8000 years from Yangtze River-derived mud in the East License: This work is licensed under a Creative Commons China Sea. Palaeogeography, Palaeoclimatology, Palaeoecology 237, 293–304 Attribution-NonCommercial-NoDerivs 3.0 Unported License. To view a copy of this (2006). license, visit http://creativecommons.org/licenses/by-nc-nd/3.0/ 31. Martino, R. L. & Sanderson, D. D. Fourier and autocorrelation analysis of estuarine tidal rhythmites, lower Breathitt Formation (Pennsylvanian), eastern How to cite this article: Xiao, S., Li, R. & Chen, M. Detecting Sedimentary Cycles using Kentucky, USA. Journal of Sedimentary Research 63, 105–119 (1993). Autocorrelation of Grain size. Sci. Rep. 3, 1653; DOI:10.1038/srep01653 (2013). 32. Wentworth, C. K. A scale of grade and class terms for clastic sediments. The Journal of Geology, 377–392 (1922). SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 8 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Scientific Reports Springer Journals

Detecting Sedimentary Cycles using Autocorrelation of Grain size

Scientific Reports , Volume 3 (1) – Apr 11, 2013

Loading next page...
 
/lp/springer-journals/detecting-sedimentary-cycles-using-autocorrelation-of-grain-size-LsT0hlUfPm

References (45)

Publisher
Springer Journals
Copyright
Copyright © 2013 by The Author(s)
Subject
Science, Humanities and Social Sciences, multidisciplinary; Science, Humanities and Social Sciences, multidisciplinary; Science, multidisciplinary
eISSN
2045-2322
DOI
10.1038/srep01653
Publisher site
See Article on Publisher Site

Abstract

Detecting Sedimentary Cycles using Autocorrelation of Grain size SUBJECT AREAS: 1,2 1 2 Shangbin Xiao , Rui Li & Muhong Chen GEOMORPHOLOGY PALAEOCLIMATE 1 2 China Three Gorges University, Yichang 443002, China, South China Sea Institute of Oceanology, Chinese Academy of ATTRIBUTION Sciences, Guangzhou 510301, China. GEOLOGY Detection of sedimentary cycles is difficult in fine-grained or homogenous sediments but is a prerequisite for the interpretation of depositional environments. Here we use a new autocorrelation analysis to detect Received cycles in a homogenous sediment core, E602, from the northern shelf of the South China Sea. 2 December 2012 Autocorrelation coefficients were calculated for different mean grain sizes at various depths. The results Accepted show that sediments derived from rapid depositional events have a better autocorrelation. Analysis of two 8 March 2013 other cores confirms this result. Cores composed of sediments deposited quickly under stable and/or gradually changing hydrodynamic conditions, have higher autocorrelation coefficients, whereas, those Published composed of sediments deposited during calm periods have relatively low autocorrelation coefficients. It 11 April 2013 shows that abrupt changes in autocorrelation coefficients usually indicate the existence of a boundary between adjacent sedimentary cycles, with each cycle beginning with a high positive autocorrelation coefficient of grain size and ending with a low negative one. Correspondence and requests for materials he detection of sedimentary cycles in cores is key to understanding their depositional environments, and should be addressed to determining if such data can be used to reconstruct high-resolution palaeoclimatic records. Sedimentary S.B.X. (shangbinx@ T cycles are commonly identified using qualitative assessments of changes in, for example, grain size, sedi- 1–8 mentary structure, and the stacked patterns of the sedimentary layers . However, the precise identification of 163.com) 4,9,10 sedimentary cycles can sometimes be rather difficult, especially within relatively homogeneous sediments . For example, Core NS97-13, a sedimentary sample collected from the South China Sea at a water depth of 2120 m (8u20.389N, 115u55.389E), is lithologically homogeneous and composed of mud . Analysis of the anisotropy of the magnetic susceptibility and the results of AMS C dating indicate that the sediments may have been deposited from turbidity currents. However, it is very difficult to identify any sedimentary cycles within the core, either visually or by measuring and analysing the grain size. This is a common problem specifically noted in other 9,10,12,13 studies . Correlation is a mathematical tool frequently used in signal processing for analysing functions or series of values, and it describes the mutual relationship between two or more random variables. Autocorrelation describes 14,15 the correlation of a set of data with itself . This is not the same as cross-correlation, which is the correlation between two different signals . A time-series data set can be separated into three components, namely the overall trend, the noise, and the cyclicity. Many researchers are interested in removing the trend and use autocorrelation to reveal important information about the temporal behaviour of the system. Autocorrelation may be also 16,17 pervasive in sedimentary profiles , and has been used between pixels in digital sediment images to measure 16,18–21 average grain-size . In the present study, we analyze Core E602, a sample collected from the shelf of the northern South China Sea (Fig. 1), develop a new mathematical method that involves the autocorrelation analysis of the grain size and discuss the mechanisms of sediments transportation involved. Our new method enables the detection of sedi- mentary cycles within any core, but within lithologically homogeneous cores in particular. Moreover, the method is quantitative and eliminates subjective interpretation that can be difficult to replicate between investigators and studies. Results Core E602 has two lithostratigraphic units determined by changes in lithology and color (Fig. 2). The lower unit (37 to 368 cm) is composed of dark-gray silty sand and is homogeneous with little variation in sand, silt and clay content. The upper unit (0 to 37 cm) changes gradually up-core from silty-sand to sandy-silt. The clay content also appears to increase gradually in the same direction. SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 1 www.nature.com/scientificreports 36 37 Figure 1 | Location of Cores E602 (A) and DD2 and PC-6 (B). Maps of sediment distribution (A) and (B) are reproduced from Liu et al. and Qin et al. , respectively. Figure 2 | Grain size, chronology, and granulometery of Core E60. SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 2 www.nature.com/scientificreports Figure 3 | Distribution of accumulation probability (a) and skewness vs. standard deviation of grain size distribution (b) for Core E602. As shown in Fig. 3(a), all of the 886 sediment samples from Core segment dominated the CM patterns for the lower unit, whereas E602 are composed of saltation and suspension fractions, the bound- the RS segment was episodic. The RS segment determined the CM ary of which is located at approximately 3.8 PHI. The suspension patterns for the upper unit, which implies that a uniform suspension fraction ranges in content from 15 to 35%. The slope of the salutation during the deposition of the unit was controlled by traction currents fraction, which reflects a sorting trend, remains almost the same (Fig. 7). Furthermore, all of the sediment samples fall within a small throughout both core units. region in the CM plot (Fig. 5a), which indicates that they are well The standard deviation of grain-size distributions of the sediment sorted. It has already been noted that the upper unit of core E602 samples obtained from the lower core unit ranges from 1.3 to 1.6, and exhibits an increase in fines sediment from top to bottom, as well as their skewness ranges from 1.4 to 2.5. Samples obtained from the decreases in grain-size and sorting. However, also evident from the upper unit exhibit little change in sediment sorting, with standard CM plot (Fig. 5b) are the relatively larger C values of sediment from deviations in the range 1.6 – 2.2. However, the skewness of grain-size 0-25 cm depth, signifying that the sediments are coarser, than those distributions of the samples decreases to 0.6 to 1.6, perhaps as a result obtained from below 25 cm. of the higher fine-sediment fraction in the upper unit. Given that the The foraminifera samples at the base of core E602 (348–350 cm) bimodal frequency curves shown in Fig. 4 are typical of grain-size gave a C age of 10289-10584 a BP (Table 1), which suggests that the distributions for all the samples from the core, the higher content of deposition began during the early Holocene. According to a recent fine sediment fraction did not obviously affect sediment sorting in report on sea level changes in the South China Sea , sea level rose the samples, but did result in a reduced degree of the skewness. from approximately 235 m to 220 m during this period (,10 to The CM patterns in core sediments (Fig. 5(a)) plotted on the QR ,9 ka BP). Combining water depth at the coring site, and the core segment, which represents graded suspension, and the RS segment, length therefore gives a paleowater depth of ,30–45 m, placing it 22,23 which represents uniform suspension (Fig. 6(b)). The QR within the shallow continental shelf environment. The uniform content of sand, silt and clay, and granulometery in the lower unit of E602, implies reworking of late Pleistocene sedi- ments composed of medium and fine sand with high textural matur- ity. By contrast, fine-grained terrigenous sediment dominates the upper unit, with the coarsest sediment being deposited at the top of the core (Fig. 5b). The characteristics of Core E602, in terms of grain size, sediment- ary structure (Fig. 7b) and CM patterns, indicate that its sediments were, for the most part, deposited under stormy conditions, which is consistent with the modern reports of frequent tropical storms . Resuspension occurs when deep water waves enter water shallower than one-half the wave length (i.e. the wave base) . The wave base under a fair-weather is less than 10–15 m . However, the storm deposit extends about 40 m , i.e., bottom sediments below the fair-weather wave base would be reworked under storm condition . Given that each storm can gradually strengthen, reduce, or maintain its intensity, means that it could produce sedimentary units with Figure 4 | Grain size distribution of several samples obtained from Core decreasing (segment (21) in Figure 8), increasing (segment (9) in E602. Figure 8), or invariable mean sediment-size with depth respectively. SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 3 www.nature.com/scientificreports Figure 5 | CM plots of Core E602. Passega’s CM image technique, where C is the coarsest one percentile (the value at 99% on a cumulative curve) and M 23,34 is the median grain-size, has been widely used to relate the grain-size characteristics of sediments to the processes of their deposition. Passega and Passega & Byramjee made a distinction between the following types of sediment movement: (1) particles rolled on the bed, even when there is no turbulence. (2) Graded suspension (the term is preferred to ‘saltation’), composed of sand and coarseness decrease from the bottom of the river to an elevation of 2 m upwards. The size of the coarsest particles in the graded suspension (Cs) fluctuates according to the maximum turbulence when settling begins. (3) Overlying graded suspension, Passega described uniform suspension, defined by a constant concentration of particles at the top of the water column. Uniform suspension may be in direct contact with the bed when turbulence decreases below a threshold, allowing the graded suspension to settle on the bed. Slow-flowing rivers display uniform suspension, even during floods . Figure 6 | Different CM patterns within Core E602. Figure 7 | CM plot of the upper unit of Core E602 (a) and a typical sedimentary pattern resulting from a storm current (b). SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 4 www.nature.com/scientificreports Table 1 | AMS C data of Core E602 Calendar age/a BP Depth(cm) Material C age/a BP (1 sigma) (2 sigma) 48,50 Nummulites 8790 6 40 9068,9597 9019,9756 148,150 8850 6 45 9097,9742 9050,9807 248,250 8740 6 50 9038,9576 8988,9751 348,350 Nummulites, Puedorotaria, Elphidum 9690 6 50 10289,10584 10201,11119 First-order autocorrelation coefficients of mean grain-size vary periods of calm weather. Assuming sediment provenance is invariable with sample size (5–30) at different depths, as shown in Figure 8. and water depth does not change, calm hydrodynamic conditions Autocorrelation coefficients vary considerably at some depths, for produce little variation in grain size, and thus no abrupt changes in example from 20.59 to 0.50 at ,27 cm, but change little at other the autocorrelation coefficients of mean grain-size. However, storms or depths, for example at,70 cm. Figure 9 illustrates how the points of tsunamis can alter the hydrodynamic conditions considerably, and can abrupt change in autocorrelation coefficient fit remarkably well with cause rapid accumulation of sediment with significant changes in grain 1,4,8 changes in mean grain-size of the sediment samples. size . A sedimentary cycle may be formed during a single storm event, and depending on changes in current intensity, can result in either a graded or ungraded sequence. Thus, although there are some- Discussion The accumulation of sediment can be either rapid, for example during times no discernable changes in grain size within a particular sequence storms as a result of turbidity currents, or it can be much slower during of storms, the degree of autocorrelation may be high due to the inertia Figure 8 | Variation in the first-order autocorrelation coefficients of mean grain-size with sample size for Core E602. The numbers 1-24 in parentheses refer to different sedimentary layers/cycles. SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 5 www.nature.com/scientificreports Figure 9 | Variation in the first-order autocorrelation coefficients of mean grain-size with sample size for Core DD2. A vertical abrupt change (C3, which recorded a storm event) in the gradient of the contour of autocorrelation coefficients of the core is indicated by an arrow. or internal coherence that is inherent in the rapid accumulation of base of a new cycle, which results in a high degree of autocorrelation sediment. Figure 8 shows that sharp changes in autocorrelation coef- of grain size. This may in turn result in a degree of inertia of grain size ficient tend to be found at the boundaries between different cycles that within a sedimentary profile. Thus, a cycle begins with high positive may have resulted from large differences in the energy of successive values of autocorrelation coefficient and ends with low negative ones. storm currents. It is this inertia that allows the application of the autocorrelation In addition to storm cessation, a sediment cycle could terminate analysis of grain size to the determination of sedimentary cycles in for several other reasons, including abrupt changes in water depth or cores of this type. 4,8,28 sediment supply . However, in the studies cores, relatively rapid It is relatively easy to identify sedimentary sequences or cycles sediment accumulation or more stable sedimentation occurs at the those are composed of individual layers that are either normally Figure 10 | Variation in first-order autocorrelation coefficients of mean grain-size with sample size for Core PC-6. The two black bold vertical lines indicate the boundaries of neighbouring cycles; multiple black dot-dashed lines indicate the boundaries of neighbouring subcycles; and the red vertical line, which shows an abrupt change in the gradients of the contour of the autocorrelation coefficients of the core, indicates the base of a storm layer. Note that no obvious subcycles of sedimentation can be discerned in the muddy unit above 450 cm using traditional sedimentological methods. SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 6 www.nature.com/scientificreports shows a range of sedimentary dynamics associated with foreshore, nearshore and graded (Layers 19 & 20 in Fig. 8) or reverse-graded (Layers 8 & 9 in shallow continental shelf environments . Fig. 8), because any abrupt changes in grain size at the boundaries As is known ,the correlation coefficient of a pair of random variables (X, Y)is between these layers are clear. However, in some cases there are no given by the formula: clear changes in grain size within a sequence that is composed of n ðÞ x { xðÞ y {y i i sediments deposited during a number of different events (e.g. Layers i~1 r~sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð1Þ 10,13 & 17,18 in Fig. 8). Under such circumstances, it is difficult to n n P P 2 2 ðÞ x { x ðÞ y {y determine the number of sedimentary cycles with any accuracy. In i i i~1 i~1 Core E602, for example, the small variation in mean grain-size values from ,183 to ,100 cm (Layers 10,13 in Fig. 8) formed a cycle of Where x and y are the average values of X and Y, respectively; Xi and Yi are the i-th specific variable values of {(x , y ): i 5 1,.., n; i i sedimentation, similar to an aggradational stack. It was difficult to i 5 a particular case i; identify the sedimentary cycles or sequences of the core in terms of n 5 the number of cases. changes in mean grain-size alone. Fortunately, autocorrelation ana- 16 The autocorrelation coefficient of a series of data {x :i 5 1,.., n}, is also calculated lysis can reveal the hidden relationships that affect sediments at by the same formula above except y is replaced with x in a form lagged by one (i.e., i i11 first order autocorrelation) or with x in a form lagged by k (i.e., k order different depths. Abrupt changes in autocorrelation coefficient of i1k autocorrelation). grain size can provide a clear indication of the boundaries between Because the number of cases adopted has direct influence on the values of both the neighbouring sedimentary layers. For example, at least 24 deposi- autocorrelation coefficient of the variable X and correlation coefficient of a pair tional cycles have been identified in Core E602 (Fig. 8). variables (X, Y), varying number of cases (range: 3,20) from a successive grain size data are taken here to calculate the autocorrelation coefficients and disclose their In a further example, the muddy unit of Core DD2 was formed in changes with depth. a relatively stable sedimentary environment and during a period of calm weather . However, the sediments at ,107 cm in the core (C3 on Fig. 9) may be the result of the combined influence of a winter 1. Kortekaas, S. & Dawson, A. Distinguishing tsunami and storm deposits: an example from Martinhal, SW Portugal. Sedimentary Geology 200, 208–221 coastal current and a storm current, because there is also an abrupt (2007). change in the autocorrelation coefficients of the core at this point. 2. Horikawa, K. & Ito, M. Non-uniform across-shelf variations in thickness, grain Core PC-6 provides a third example (Fig. 10). In this core, sedi- size, and frequency of turbidites in a transgressive outer-shelf, the Middle ments deposited during different environmental conditions appear Pleistocene Kakinokidai Formation, Boso Peninsula, Japan. Sedimentary Geology to have different characteristics of grain size and stacked patterns. 220, 105–115 (2009). 3. Gonza´ lez-Alvarez, R. et al. Paleoclimatic evolution of the Galician continental Abrupt changes in the autocorrelation coefficients of the core shelf (NW of Spain) during the last 3000 years: from a storm regime to present occurred not only at these boundaries, but also at the boundaries conditions. Journal of Marine Systems 54, 245–260 (2005). of the subcycles of sedimentation. The latter are sometimes difficult 4. Morton, R. A., Gelfenbaum, G. & Jaffe, B. E. Physical criteria for distinguishing to identify within relatively homogeneous sedimentary profiles. At sandy tsunami and storm deposits using modern examples. Sedimentary Geology 200, 184–207 (2007). the same time, the storm deposit event reported at ,100 cm is 5. Sun, Y., Gao, S. & Li, J. Preliminary analysis of grain-size populations with clearly revealed by the abrupt change in the autocorrelation coeffi- environmentally sensitive terrigenous components in marginal sea setting. cients of grain size within the core (Fig. 10). Chinese Science Bulletin 48, 184–187 (2003). A degree of autocorrelation in sedimentary profiles is a widespread 6. Anthony, E. J. & He´ quette, A. The grain-size characterisation of coastal sand from the Somme estuary to Belgium: sediment sorting processes and mixing in a tide- phenomenon, and is a result of the similarity of the hydrodynamic and storm-dominated setting. Sedimentary Geology 202, 369–382 (2007). conditions present in the stable sedimentary environment that char- 7. Sedgwick, P. E. & Davis, R. A. Stratigraphy of washover deposits in Florida: acterized the sedimentation . Sediments derived from storms or implications for recognition in the stratigraphic record. Marine geology 200, other events have a higher degree of autocorrelation as a result of 31–48 (2003). 8. Bentley, S. J., Sheremet, A. & Jaeger, J. M. Event sedimentation, bioturbation, and their rapid deposition , and abrupt changes in autocorrelation coef- preserved sedimentary fabric: Field and model comparisons in three contrasting ficient usually indicate the boundaries between successive sediment- marine settings. Continental Shelf Research 26, 2108–2124 (2006). ary cycles. The analysis of autocorrelation coefficients is therefore a 9. Tuttle, M. P., Ruffman, A., Anderson, T. & Jeter, H. Distinguishing tsunami from useful tool in the identification of sedimentary sequences or cycles, storm deposits in eastern North America: the 1929 Grand Banks tsunami versus and our results confirm its potential of for detecting sedimentary the 1991 Halloween storm. Seismological Research Letters 75, 117–131 (2004). 10. Dawson, A. G. & Shi, S. Tsunami deposits. Pure and applied geophysics 157, cycles in homogeneous sediment sequences. Combined with the 875–897 (2000). 16,18–21 use of rapid techniques to determine grain-size from digital 11. Tang, X., Chen, M., Liu, J., Zhang, L. & Chen, Z. The anisotropy of magnetic sediment images, autocorrelation analysis should therefore improve susceptibility of Core NS97-13 sediments from the Nansha Islands sea area in the the accuracy of sedimentary environmental determination. southern South China Sea. Acta Oceanologica Sinica 31, 69–76 (in Chiness with English Abstract) (2009). 12. Allison, M. A., Sheremet, A., Gon˜ i, M. A. & Stone, G. W. Storm layer deposition on Methods the Mississippi–Atchafalaya subaqueous delta generated by Hurricane Lili in Core E602, with a length of 3.68 m, was collected with gravity sampling pipe in 2002. Continental Shelf Research 25, 2213–2232 (2005). October 2004 from a sandy area on the northern shelf of the South China Sea 13. Goff, J., McFadgen, B. & Chague´ -Goff, C. Sedimentary differences between the (112u14.889E, 20u44.939N), in a water depth of 65 m, and distant from any estuarine 2002 Easter storm and the 15th-century Okoropunga tsunami, southeastern influence. In the laboratory, the core was described in detail and then split into a total North Island, New Zealand. Marine geology 204, 235–250 (2004). of 886 samples (300 samples at 0.25 cm intervals from 0 to 75 cm and 586 samples at 14. Phillips, C., Parr, J. & Riskin, E. Signals, Systems, and Transforms. (Prentice Hall 0.5 cm intervals from 75 to 368 cm). The samples were initially pre-processed by the 1999). addition of excess H O (Q 5 30%), followed by the addition of HCl (3N). The 15. Herman, E. K., Toran, L. & White, W. B. Quantifying the place of karst aquifers in 2 2 distribution of grain size was classified following Wentworth , and measured using a the groundwater to surface water continuum: A time series analysis study of storm Mastersizer 2000 from Malvern Instruments Ltd., with a range of measurement of behavior in Pennsylvania water resources. Journal of Hydrology 376, 307–317 0.02 to 2000 mm and a resolution of 0.01 PHI. The errors in mean grain size for the (2009). same sample obtained following repeated measurements were less than 3%. The mean 16. Rubin, D. M. A simple autocorrelation algorithm for determining grain size from grain size (unit: PHI) was used to calculate the autocorrelation coefficients of the grain digital images of sediment. Journal of Sedimentary Research 74, 160–165 (2004). size data. 17. Xiao, S., Liu, W., Li, A., Yang, S. & Lai, Z. Pervasive autocorrelation of the chemical Radiocarbon dating of the core sediments was carried out at the National Ocean index of alteration in sedimentary profiles and its palaeoenvironmental Sciences Accelerator Mass Spectrometry Facility (NOSAMS), Woods Hole implications. Sedimentology 57, 670–676 (2009). Oceanographic Institution. All the C ages (Table 1) were calibrated to give calendar 18. Barnard, P., Rubin, D., Harney, J. & Mustain, N. Field test of an autocorrelation ages using CALIB4.3 . technique for determining grain size using a digital camera. Sedimentary Geology The significance of the autocorrelation of grain size is herein discussed in the 201, 180–195 (2007). context of two other sedimentary cores obtained from the inner shelf of the East 19. Warrick, J. A. et al. Cobble Cam: Grain-size measurements of sand to boulder 29 30 China Sea that have previously been described, namely DD2 and PC-6. Core DD2 from digital photographs and autocorrelation analyses. Earth Surface Processes is composed of muddy sediments formed in a shallow continental shelf. Core PC-6 and Landforms 34, 1811–1821 (2009). SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 7 www.nature.com/scientificreports 20. Buscombe, D. & Masselink, G. Grain-size information from the statistical 33. Hughen, B. K., McCormac, G., van der Plicht, J. & Spurk, M. INTCAL98 properties of digital images of sediment. Sedimentology 56, 421–438 (2008). radiocarbon age calibration, 24,000-0 cal BP. Radiocarbon 40, 1041–1083 (1998). 21. Buscombe, D. Estimation of grain-size distributions and associated parameters 34. Passega, R. Texture as characteristic of clastic deposition. AAPG Bulletin 41, from digital images of sediment. Sedimentary Geology 210, 1–10 (2008). 1952–1984 (1957). 22. Bravard, J.-P. & Peiry, J.-L. The CM pattern as a tool for the classification of 35. Passega, R. & Byramjee, R. GRAIN-SIZE IMAGE OF CLASTIC DEPOSITS. alluvial suites and floodplains along the river continuum. Geological Society, Sedimentology 13, 233–252 (1969). London, Special Publications 163, 259–268 (1999). 36. Liu, Z. S., Zhao, H. T., Fan, S. Q. & Chen, S. Q. Geology of the East China Sea (in 23. Passega, R. Grain size representation by CM patterns as a geologic tool. Journal of Chinese). Beijing: Science Press (2002). Sedimentary Research 34, 830–847 (1964). 37. Qin, Y. S., Zhao, Y. Y., Chen, L. R. & Zhao, S. L. Geology of the East China Sea (in 24. Hori, K. et al. Delta initiation and Holocene sea-level change: example from the Chinese). Science Press, Beijing (1987). Song Hong (Red River) delta, Vietnam. Sedimentary Geology 164, 237–249 (2004). 25. Davies, J. Geographical variation in coastal development. (Longman Publishing Group, 1972). Acknowledgements 26. Evans, R. D. Empirical evidence of the importance of sediment resuspension in We thank Dr. Paul Blanchon, who is at Institute of Marine Sciences and Limnology, lakes. Hydrobiologia 284, 5–12 (1994). National University of Me´ xico, for giving highly constructive suggestions for the first 27. Wright, M. E. & Walker, R. G. Cardium Formation (U. Cretaceous) at Seebe, manuscript. This work was sponsored by National Science Foundation of China Alberta-storm-transported sandstones and conglomerates in shallow marine (No. 41273110) and the China Postdoctoral Special Science Foundation (No. 200801440). depositional environments below fair-weather wave base. Canadian Journal of Earth Sciences 18, 795–809 (1981). 28. Kakinoki, T., Tsujjimoto, G., Yamada, F., Sakai, D. & Uno, K. Beach profile and Author contributions sediment characteristics of a mixed sand beach under diurnal sea level variations. Xiao, S. B. carried out grain-size analysis and prepared the primary manuscript. Li, R. Journal of Coastal Research SI 64, 765–770 (2011). prepared the figures. All authors reviewed and discussed the manuscript. 29. Xiao, S. et al. Recent 2000-year geological records of mud in the inner shelf of the East China Sea and their climatic implications. Chinese Science Bulletin 50, 466–471 (2005). Additional information 30. Xiao, S. et al. Coherence between solar activity and the East Asian winter monsoon Competing financial interests: The authors declare no competing financial interests. variability in the past 8000 years from Yangtze River-derived mud in the East License: This work is licensed under a Creative Commons China Sea. Palaeogeography, Palaeoclimatology, Palaeoecology 237, 293–304 Attribution-NonCommercial-NoDerivs 3.0 Unported License. To view a copy of this (2006). license, visit http://creativecommons.org/licenses/by-nc-nd/3.0/ 31. Martino, R. L. & Sanderson, D. D. Fourier and autocorrelation analysis of estuarine tidal rhythmites, lower Breathitt Formation (Pennsylvanian), eastern How to cite this article: Xiao, S., Li, R. & Chen, M. Detecting Sedimentary Cycles using Kentucky, USA. Journal of Sedimentary Research 63, 105–119 (1993). Autocorrelation of Grain size. Sci. Rep. 3, 1653; DOI:10.1038/srep01653 (2013). 32. Wentworth, C. K. A scale of grade and class terms for clastic sediments. The Journal of Geology, 377–392 (1922). SCIENTIFIC REPORTS | 3 : 1653 | DOI: 10.1038/srep01653 8

Journal

Scientific ReportsSpringer Journals

Published: Apr 11, 2013

There are no references for this article.