Autocompaction in Holocene coastal back-barrier sediments from south Devon, southwest England, UK

The present-day elevation of superficial horizons situated above a competent basal stratum is likely to be lower than the original height of deposition. This is because sediments such as minerogenic fines and peat undergo a post-depositional
of 17
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  Autocompaction in Holocene coastal back-barrier sediments fromsouth Devon, southwest England, UK  Anthony C. Massey  a, *, Michael A. Paul  b , W. Roland Gehrels  a  , Dan J. Charman  a  a  School of Geography, University of Plymouth, Drake Circus, Plymouth, Devon, PL4 8AA, England, United Kingdom  b School of Life Sciences, Heriot-Watt University, Edinburgh, EH14 4AS, Scotland, United Kingdom Received 17 December 2004; received in revised form 14 October 2005; accepted 3 November 2005 Abstract The present-day elevation of superficial horizons situated above a competent basal stratum is likely to be lower than the srcinalheight of deposition. This is because sediments such as minerogenic fines and peat undergo a post-depositional reduction in volumeas a result of the weight of overlying sediments, the downward movement being due to the cumulative compression of all thesediment below the level in question. This  b autocompaction  Q   can affect the palaeoenvironmental interpretation of lithofacies fromwhich a vertical reference is required, e.g., when quantifying the height of a sea-level index point. Geotechnical theory was used toapply a correction to Holocene coastal back-barrier sediments from North Sands and Blackpool Sands in south Devon (UK) and soto return marker horizons to a level approximating their srcinal height of deposition. In this model the total downward movement is calculated by notionally dividing the underlying soil into a number of thin (~0.1 to 0.2 m) layers and calculating the individualcompression in each one. The results were then summed to give the total compression. This approach can underestimate the fullextent of autocompaction notably due to uncertainties arising from the behaviour of organic-rich facies and from inadequateknowledge of groundwater history. The results must be considered semiquantitative and are usually minimum estimates. Results of vertical corrections from coastal sedimentary units in south Devon range from  b 0.1 m in fine-grained sediments situated above basal facies to  N 1 m at contacts between minerogenic sediments and peat, increasing to  N 2 m in more organic facies. D  2005 Elsevier B.V. All rights reserved.  Keywords:  autocompaction; geotechnical; Holocene; back-barrier sediments; compression; vertical correction 1. Introduction The post-depositional compression of Holocenesediments from thick coastal sequences alters the src-inal height of deposition of palaeo-marsh and mudflat surfaces. Autocompaction in sediments may thus be asignificant problem when attempts are made to deter-mine former sea-level heights from coastal lithofacies,especially when sea-level index points (SLIPs) arederived from non-basal fine-grained minerogenic sedi-ments or peat (Gehrels, 1999). Many index points  published by sea-level researchers in southwest Eng-land have been obtained from non-basal sediments (i.e.sediments that do not directly overlie a hard, non-compressible substrate) that are likely to have beenvertically displaced over time (e.g., Churchill,1965a,b; Morey, 1976, 1983a; Heyworth and Kidson,1982; Healy, 1995). This region is of interest to sea- 0025-3227/$ - see front matter   D  2005 Elsevier B.V. All rights reserved.doi:10.1016/j.margeo.2005.11.003* Corresponding author. Tel.: +44 7790 263094; fax: +44 1752233054.  E-mail addresses: (A.C. Massey), (M.A. Paul), (W.R. Gehrels), (D.J. Charman).Marine Geology 226 (2006) 225–  level studies because it is estimated to have the fastest relative land subsidence of any coastline in the UK (Shennan and Horton, 2002). However, estimates of  relative land-/sea-level changes are based on very lim-ited mid- to late Holocene data that have not beencorrected for autocompaction. Improvements in thisarea will therefore enhance the current sea-level dataset from southwest Britain and advance our under-standing of relative sea-level changes throughout theHolocene.From the above, it is clear that samples collectedfrom thick unconsolidated sequences of intertidal sedi-ments are likely to require a geotechnical cor rection(Allen, 1995, 2000; Shennan and Horton, 2002). On the other hand, SLIPs obtained from basal sediment arecompaction-free and, as the sediment is not verticallydisplaced over time, are likely to improve the reliabilityof relative sea-level histories (K aye and Barghoorn,1964; Kiden, 1995; Gehrels, 1999; Donnelly et al.,2004). This issue is often ignored by sea-level research-ers or, at best, only approximate estimates of compac-tion are made. This may be a reflection of the current lack of applied theory and the apparent shortage of geotechnical models currently available to quantifythe autocompaction of sediments (e.g., Paul et al.,1995; Pizzuto and Schwendt, 1997; Paul and Barras,1998; Rybczyk et al., 1998; Allen, 1999; Tovey andPaul, 2002; Williams, 2003; Bird et al., 2004).Earlier attempts to quantify the vertical displacement of a dated sample in the sedimentary sequence and itsrelationship to contemporary sea level (e.g., Godwin et al., 1958; Heyworth and Kidson, 1982) underestimatedthe effects of sediment compaction (Haslett et al.,1998). This often resulted in samples plotting lower on an age–depth curve than their original level of deposition. Paul and Barras (1998) observed that  b elevation errors are especially significant when study-ing Flandrian (Holocene) sea-level change, because themagnitude of the change can be similar to the magni-tude of the errors  Q  . Allen (1999) also noted that vertical errors are likely to be much greater in Holocene peat sequences as organic material is generally highly com- pressible. K idson and Heyworth (1973) and Devoy (1982) attempted a decompaction of Holocene peat sequences from sites in southern Britain but encoun-tered similar problems relating to the generally unpre-dictable nature of organic lithofacies.The numerical quantification of autocompaction inHolocene sediments is also complicated by spatial var-iability of coastal stratigraphy. For example, K aye andBarghoorn’s (1964) geometrical measure of peat auto-compaction, using the deformation of logs, was limited by  d slip T  between timber in the sequence and the sur-rounding finer peat material. Similar geometricalapproaches (e.g., Bloom, 1964; Belknap and Kraft ,1977) have been hampered by limited borehole andchronological data and assum ptions that  some peat  bed tops may be isochronous (Allen, 1999). In this paper we present new data from coastallithofacies from south Devon and use them to determinethe height of intertidal surfaces in sediments depositedduring the early to mid-Holocene. The paper also pre-sents a correction of these heights for post-depositionalautocompaction, using the model of  Paul and Barras(1998), and discusses the implications of this correctionfor the interpretation of our results. In this way, we aimto illustrate both the practical application of this rela-tively simple model to a Holocene sequence and alsothe likely magnitude of autocompaction and its effect on the interpretation of our data. 2. The model Fine-grained sediments contain large amounts of  pore water that is expelled, over time, following depo-sition. This reduces sediment porosity and increasesinterparticle effective stress until sediment volumeequilibrates with this stress. The volume reduction process is known geologically as compaction and thevertical correction applied by the model as decompac-tion. In geotechnical usage, compression arises from thestress-dependent reduction in equilibrium volumewhereas consolidation is the time-dependent process by which this reduction is achieved.A simple model of sediment autocompaction has been described by Paul and Barras (1998). It relies on the so-called Terzaghi law of compression, which statesthat the volume reduction in sediment is directly pro- portional to the logarithm of the increase in effectivestress (Fig. 1). The Terzaghi law is known to be fol- lowed closely by fine-grained minerogenic sedimentssuch as clays and some evidence (Hobbs, 1986) sug- gests that it is broadly followed by peat, although with astrong time-dependent secondary component. In themodel, the sedimentary sequence is divided into a seriesof notional increments (in this case of ~0.1 to 0.2 mthickness, corresponding to the sampling interval for water content) and the compression in each layer iscalculated according to the stress increase that it hasexperienced. The total compression at any depth is therunning sum of these increments (from the base up-wards) and hence the sequence can be decompacted togive the srcinal elevation at any required level. Thismodel is derived from a standard geotechnical proce-  A.C. Massey et al. / Marine Geology 226 (2006) 225–241 226  dure for estimating foundation settlement: further details are given in Paul and Barras (1998). The geotechnical parameters that are required for the calculation are the compression index (slope of compression line) of the sediment, the bulk densityand the groundwater level. From these are derived theeffective stress and hence the volume reduction. Themeasurement of these parameters is routine, expect for the compression index which is normally mea-sured on an undisturbed sample in the oedometer test (e.g., Smith and Smith, 1988). We may note in  passing that it is also possible to calculate the prog-ress of the volume reduction over time during thesedimentation process (Gibson, 1958) and that, for  most natural sediments the volume reduction keeps pace with loading (Skempton, 1970). We calculate this to be the case in the sequence considered hereand thus do not need to consider effects due todelayed consolidation.In the case of sediments examined for geological purposes, it is often the case that the requisite geotech-nical parameters, especially the compression index,may not have been obtained at the time of collection.This is so for the cores considered here, on which thecompression index was not measured directly. Howev-er, for fine-grained sediments the compression index iscorrelated with the liquid limit (Atterberg, 1911), which can be measured retrospectively. This correlation wasinitially empirical but has later been supported theoret-ically by the principles of critical state soil mechanics(e.g., Schofield and Wroth, 1968). The correlation used in our model is that described in Terzaghi et al. (1996), from the work of  Skempton (1944), which was srci- nally derived on results from mineralogenic sediments.Thus work showed that the compression index ( C  C )can be estimated from the liquid limit ( W  L ) using theexpression C  C  ¼  0 : 009  W  L    10 ð Þ We note that, in the case of very soft materials, thiscorrelation usually underestimates the compressionindex and thus our estimates of autocompaction will be minimum values. We also stress that the use of thiscorrelation is not an implicit feature of the model: it ismerely a convenient way of estimating the compressionindex in the absence of direct measurements of this parameter.A question arises over the presence of peat, which isusually a very compressible material of high initialwater content. It has, however, clearly been shown byHobbs (1986) that peat behaves geotechnically as afine-grained sediment and obeys the above correlation between compressibility and liquid limit up to values of  W  L  in excess of 800%. This is considerably above thevalues that we have encountered and thus we have usedthe same correlation throughout our model. Fig. 1. Theoretical basis of the geotechnical model (after  Paul and Barras, 1998). In the assumed palaeo-setting of the sedimentary sequence, a sample  S   deposited at a height   H  0  above a thickness of fine-grained minerogenic sediment   Z  a   has an effective stress of   P  0 . The sample  S   wassubsequently overlain by a unit   Z   b  raising the effective stress to  P  1  and compressing the unit   Z  a   to a thickness  Z  c . This reduced the srcinal height of the sample to  H  1  and the model must calculate the vertical compression  D  H   to decompact the sequence. The sample moves down a compressionline from  S  0  to  S  1  in response to the sediment loading and  D  H   is dependent on the ratio of effective stresses (  P  0  to  P  1 ) and the slope of thecompression line. See Section 2 for determination of the compression index ( C  C ).  A.C. Massey et al. / Marine Geology 226 (2006) 225–241  227  3. Site description and rationale The current configuration of back-barrier sites alongthe Channel coast of southwest England was formedwhen back-barrier intertidal lagoons were disconnectedfrom the open sea in the late Holocene (Morey, 1976;Healy, 1995, 1996; Massey, 2004). This occurred pre-sumably as the result of a reduction in the rate of sea-levelriseandtheinfillingoftidalinlets.LowerDevonian bedrock and Pleistocene deposits were overlain by basal peat facies of high intertidal srcin which in turn were buried by thick Holocene minerogenic sediments depos-ited in a shallow lagoonal setting (Clarke, 1970; Hails,1975a,b; Lees, 1975; Morey, 1983a,b). Non-basal inter-tidal facies preserved in present-day back-barrier sedi-ments are suitable for sea-level reconstruction but require a correction for autocompaction. Compaction-free index points were obtained from some basal peat faciesandallowforindependentverificationofcorrectedmarker horizons. Sites from south Devon (Fig. 2) t hat   provided these sequences but required a geotechnicalcorrection are North Sands (50 8 13.8  V  N, 03 8 47.0  V W)and Blackpool Sands (50 8 19.2  V  N, 03 8 36.8  V W). Auto-compaction was calculated in two cores from each site. North Sands is located in the lower reaches of theKingsbridge estuary, south Devon (Fig. 2). The back-  barrier system comprises a freshwater marsh coveringabout 5 ha. The local geology is dominated by schistsderived from altered Devonian rocks (Ussher, 1904;Durrance and Laming, 1982). The two cores are located150 m inland from the limit of mean high water of spring tides (MHWST), at +2.5 to +3 m above the UK Ordnance Datum at Newlyn (OD).Blackpool Sands is a 700 m long gravel barrier insoutheast Devon (Fig. 2) and a continuation of the more extensive Slapton Sands barrier (Hails, 1975a,b). A lake was dammed behind the gravel barrier ca. 4400cal. yr ago and gradually filled in with organic silts andmud (Massey, 2004). The back-barrier system covers roughly 2 ha and is grassland, low scrub and woodedwetland valley. The local geology comprises slates,some hard grit and quartzite beds, and sedimentaryrocks of the Meadfoot Group (Ussher, 1890; Orme,1960; Dineley, 1961). Cores are located about 150 minland from MHWST, at +4 m OD. 4. Methods 4.1. Fieldwork  Coring was carried out in 1997 using a percussiondrilling rig to extract sediments in 1 m long by 94 mmdiameter hard plastic sleeves. The coring method inev-itably creates some disturbance in soft sediments and isan additional reason why the Terzaghi et al. (1996)correlation was used to obtain the compression indexrather than direct measurement in an oedometer. Partialrecovery of some sections occurred upon extraction andis due to a number of factors related to the coringmethod. For example, sandy sediment settled in somesleeves after removal above the water table. The verti-cal datum was recorded at the top of each (1 m) sectionand core loss at the base. Compaction of clay, silt and peat is a problem with percussion rigs and other non-rotary techniques (Rees, 1999) and was measured in the field in each 1 m section. A sediment catcher attachedto the base of each sleeve prevented sediment from being lost from the bottom of the barrel during extrac-tion and the  d captured T  sediment was later retained in a bag or smaller tube. Cores were collected as part of awider project and the uppermost sections in some loca-tions were not retrieved. These sections are predomi-nantly freshwater peat facies or infill and weredescribed in the field. Core sections were placed incold storage and opened horizontally to avoid anyvertical (upward or downward) movement of sediment in the liner. Sediments were described using the Troels-Smith (1955) classification scheme.Core sites were surveyed into OD using an Electron-ic Distance Measurer (EDM) and Ordnance Survey(OS) benchmarks within 300 m of the boreholes (Fig.2). Altitude and horizontal distance were recorded tothe nearest millimetre. Angular measurements wererecorded to the nearest 1/1000 of a degree. The hori-zontal and angular measurements were used to locatethe borehole positions on a large-scale OS map. Loca-tions were confirmed using a handheld Global Position-ing System (GPS). 4.2. Laboratory methods The methods and classifications for geotechnical parameters are based on standard techniques (e.g., Ter -zaghi, 1925; Terzaghi et al., 1996). To determine thecompressibility of the sediments, cores were sampled at 1 m intervals, for plastic and liquid (Atterberg, 1911)limits. Liquid limits (using a drop-cone penetrometer complying with British Standard 4691, 1974) and plas- tic limits tests were carried out according to BritishStandard 1377 (1975). The plasticity indices were cal-culated using the equation PI=LL  PL; wherePI=Plasticity Index (%), LL=Liquid Limits (%) andPL=Plastic Limits (%). Cores were sampled every 20cm for particle size analysis (PSA) of the  b 2 mm  A.C. Massey et al. / Marine Geology 226 (2006) 225–241 228  fraction using a Laser Mastersizer Autosampling Sys-tem and the  N 2 mm fraction using a Camsizer (ParticleSize and Shape Analyser). Classifications and particlesize distributions are according to Folk (1954) andShepard (1954). The percentage of sand, silt and clayin each sample was calculated by laser diffraction from Fig. 2. Location of core sites at North Sands and Blackpool Sands, south Devon, UK. Coordinates are Ordnance Survey of Great Britain NationalGrid System and Latitude and Longitude.  n Crown Copyright/Database right 2005. An Ordnance Survey/EDINA supplied service.  A.C. Massey et al. / Marine Geology 226 (2006) 225–241  229
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!