Investigation of the nature of slip surface using geochemical analyses and 2-D electrical resistivity tomography: a case study from Lapseki area, NW Turkey

9 pages
35 views
of 9
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.
Share
Description
The nature and subsurface structure of the slip surface of a landslide was studied on the basis of geochemical analyses and 2-D electrical resistivity tomography (ERT) survey. Head scarp and lateral slip surfaces of the landslide marked by clear
Tags
Transcript
  ORIGINAL ARTICLE Investigation of the nature of slip surface using geochemicalanalyses and 2-D electrical resistivity tomography: a casestudy from Lapseki area, NW Turkey Ahmet Evren Erginal   Beyhan O ¨ztu ¨rk   Yunus Levent Ekinci   Alper Demirci Received: 7 May 2008/Accepted: 3 October 2008/Published online: 25 October 2008   Springer-Verlag 2008 Abstract  The nature and subsurface structure of the slipsurface of a landslide was studied on the basis of geochemical analyses and 2-D electrical resistivitytomography (ERT) survey. Head scarp and lateral slipsurfaces of the landslide marked by clear slickensidedshear planes were composed of the average amounts of clayey silt (32.5%) and sand (67.5%). Energy dispersiveX-ray spectroscopy (EDX) data revealed the enrichment of Si (23.24%), Fe (12.2%), Al (9.51%) and C (8.34%) in theelemental composition of the disturbed slip surface. FromX-ray diffractometry (XRD) data, six main clay types weredetermined, such as Volkonskoite, Halloysite, Ferrosilite,Saponite, Illite and Nontronite. The ERT survey displayedthat the landslide developed as a reactivation on the upperpart of an old landslide body. Keywords  Landslide    Slip surface material   Geochemical analysis    ERT    Rotational slide   Lapseki Introduction Owing to the presence of favorable pre-conditional geo-environmental factors including landslide-prone geologicformations, tectonic weaknesses, and high and inclinedmorphology, active and dormant landslide areas cover asignificant area in Turkey. The occurrence of many land-slides was commonly associated with clay-rich andstructurally complex areas where precipitation amounts aresufficient for the emergence of slope failures. Most widelyknown areas exposed to slope failures in the country aresettlement areas (Ardos 1980; Uzun 1987; Ocakog ˘lu et al.2002), water reservoirs (Ertek  1999); highways (Yilmazer et al. 2003), farm lands and open-pit mines (Erginal et al.2008), and water/natural gas pipelines (Sag˘lamer 1991).The Biga Peninsula located in northwestern part of Turkey is an area where landslides are quite common. Thepeninsula has different geologic and geomorphologiccharacteristics reflecting its complex geological evolutionhistory. South-facing parts of the peninsula are morpho-logically higher and deeply incised due to the presence of Kazdag (1,774 m above sea level), an asymmetric (north-ward tilted) massive controlled by east–west trendingactive faults (Yilmaz and Karacik  2001). The northernsection of the peninsula is, however, composed of plateauareas with gently to moderately inclined slopes toward theMarmara Sea. The west side of the peninsula is delimitedby the Strait of C¸anakkale, which is the unique waterlinkage between the Aegean Sea and the Marmara Sea.In the peninsula, slope instabilities are mostly associatedwith the presence of Early Miocene aged deeply weatheredvolcanic terrains composed of andesite, tuff and agglom-erate (Tu¨rkes¸ et al. 2006) and, in particular, marinedeposits (C¸anakkale Formation) with clay intercalations of Middle-Upper Miocene age. The case study presented hereis an assessment of a small rotational landslide activitydeveloped in coastal outcroppings of clay-rich C¸anakkaleFormation in northwestern Turkey (Fig. 1a). Based on theinterviews with the villagers of Adatepe Village 700 m A. E. Erginal ( & )    B. O¨ztu¨rk Department of Geography,Canakkale Onsekiz Mart University,Canakkale, Turkeye-mail: aerginal@comu.edu.trY. L. Ekinci    A. DemirciDepartment of Geophysics,Canakkale Onsekiz Mart University,Canakkale, Turkey  1 3 Environ Geol (2009) 58:1167–1175DOI 10.1007/s00254-008-1594-4  west of the landslide area, the event took place in the rainyperiod of 2007. However, the precise date of the failure isunknown. Even though the failure was not of a notabledimension, it deserved a detailed study owing to its inter-esting characteristics in terms of the nature of its slipsurface material. Furthermore, it is of the potential of making serious damage on very active C¸anakkale-Bursahighway in possible future reactivations.In the present work, we attempt to investigate landslideinvolved by putting special emphasis on the nature andsubsurface structure of slip surface, since, as known well,slip surfaces are the most critical zone of weakness of landslides due to their various favorable chemical andphysical properties. The depth of slope failures and theamount and behavior of the failed material as well as sliprate are firmly determined by the angle and other charac-teristics of these slippery planes, which correspond mainlyto the more or less weathered pre-existing planes of faults, joints and beddings (Hutchinson 1988; Cruden and Varnes1996; Dikau et al. 1996; Skempton and Petley 1967). Defining the nature of slip surface material is therefore of great importance for both better understanding the mech-anism of a landslide and adopting efficient preventionmeasures. A very practical tool to achieve this need is tosubstantiate geochemical analyses, which provide usefuloutcomes on the srcin, rate and role of weathering agentsaccounting for development these weak pathways (Wenet al. 2004). On these grounds, we described the nature of slip surface using X-ray diffractometry, energy dispersiveX-ray spectrometry (EDX) in conjunction with scanningelectron microscopy (SEM) and fine particle content-par-ticle size distribution analyses.After geochemical analyses, electrical resistivitytomography (ERT) technique was applied to determinesubsurface structure of the landslide body. It is well knownthat the direct current (DC) resistivity method is one of themost commonly applied techniques for geophysicalinvestigation in groundwater, environmental and engi-neering applications. It is also an efficient tool because of the low cost and fast field survey procedures. In the lastdecade, depending on high technological development of computer controlled multi-electrode survey systems andthe inversion software packages, much of the investigationscarried out on DC resistivity method have mainly focusedon 2-D and 3-D data acquisition and interpretation tech-niques. This development yielded more effective surveysand reliable resistivity high resolution images (e.g., Grif-fiths and Barker 1994; Abdul Nassir et al. 2000; Beresnev et al. 2002; Drahor et al. 2006; Drahor et al. 2007; Ekinci et al. 2008). ERT technique, progressed form of the con-ventional vertical electrical sounding, is a widely used toolto study the characteristic of landslides because of itscapability to determine the lateral extension and thicknessof the landslide body, to describe the sliding surfacesbetween the units and to obtain useful information aboutthe main patterns of the underground fluid flows (e.g.,Suziki and Higashi 2001; Bateyneh and Al-Diabat 2002; Lapenna et al. 2003; Drahor et al. 2006). Methods Sampling and analytical procedureTotally six samples from slip surface material (s2–s7) andone sample from toe material (s1) were collected fordetermining sediment-size fractions (sand, silt and clayconcentrations) by a hydrometry technique (Boyoucous1963). Sampling locations are shown in Fig. 1b. A Phillips XL-30 S FEG SEM equipped with EDX detector was used Fig. 1  Location map ( a ),sampling sites ( b ) andsimplified geology map of thestudy area. ( 1 ) Alluvium(Quaternary), ( 2 ) C¸anakkaleFormation including limestone,peblestone, sandstone andmudstone (Miocene), ( 3 )Various volcanics (mainlyandesite of Eocene), ( 4 ) Gneiss(Paleozoic)1168 Environ Geol (2009) 58:1167–1175  1 3  to observe microstructure of slip surface material and toquantify its elemental composition. To identify the type of the disturbed clay minerals collected, X-ray Diffractometry(Phillips X’Pert Pro) analyses were also performed. Theanalyses were carried out in Central for Materials Researchof Izmir Institute of Technology.ERT surveyWe conducted the ERT survey along a profile which is102 m long using Wenner–Schlumberger array with anelectrode spacing of 3 m. There were 32 measuring stationsalong the survey line. A total of 252 apparent resistivitydata were gathered for 12 data levels using the Iris-SyscalR1 Plus resistivity meter. Measurements, stored in the datalogger of the resistivity meter, were transferred to thecomputer using the RS232 port. The surface topography isdetermined using optical levelling at the locations of eachelectrode. In order to improve the quality of the data, thestandard deviation of stacks of each point was taken intoaccount. The resistivity meter performed a stack of aminimum of four for each data point. When the relativestandard deviation of the stacked data was greater than 2%,the vertical stack was increased to six.The standard deviation of the measurements was mostlybelow 1%. The processing and interpretation of the mea-sured data were performed using the 2-D tomographicinversion algorithm of Loke and Barker (1996) usingRES2DINV software. This algorithm is based on thesmoothness–constrained least-squares and produces ageoelectrical section for a profile data. The optimizationadjusts the 2-D resistivity model by reducing the differencebetween the calculated and measured apparent resistivityvalues in the iterations. The smoothness–constraint leadsthe algorithm to yield a solution with smooth resistivitychanges (Drahor et al. 2006). This is the case for our studyarea in which clayish unit overlays sandy units and resis-tivity of both units are in close range. Due to thetopographical changes in the survey area finite-elementscheme was performed in the inversion process. There wasnot high random noise in the investigation area, so themaximum number of iteration was set to be four in order toavoid overfitting of the data. The inversion process pro-duced a fit with an RMS error of 4.5%. It was thought thatthe obtained results provided a realistic true resistivitydistribution of the subsurface depending on the RMS error.Location and environmental settingsThe landslide area is located in about 4 km east of thenortheast exit of the Strait of C¸anakkale, which circum-scribes the Biga Peninsula in the west (Fig. 1a). The arealies between the latitudes 40  23 0 51 00 –40  23 0 54 00 andlongitudes 26  48 0 16 00 –26  48 0 19 00 . The sliding event, locatedin about 13 km east of the Lapseki district of C¸anakkalecity has developed as an individual partial reactivationwithin a small valley cut in north-facing (seaward) down-slope of an old landslide mass, which is divided in itscentral part into two portions by the recently enlargedC¸anakkale-Bursa highway. Coastal landslides are commonalong this coastal area and are mostly spoon-shaped.Morphologically, the coastal region is backedpredominantly by old landslide areas that developed onsteep slopes of a moderately dissected plateau cut both inC¸anakkale Formation and Eocene volcanics composedprimelyofweathered andesite(Fig. 1c).Elevations increasegradually southwards and are not more than 150 m. Shallowchannels incised by small creeks are common on clay-richunitsthroughoutthecoastalarea;severalofthesevalleys arepresently exposed to landslide occurrences. The old land-slide area is dominated by a hummocky topography, reverseinclinations, stepped slopes and back tilted threes, implyingcontinual activity of rotational failures. Its relict scarp isdefined by an escarpment of 60 m (Fig. 2a, b).The geological material involved in the landslide isGu¨zelyal ı  member of the C¸anakkale Formation (Atabeyet al. 2004). It is a 60–90 m thick sequence composedpredominantly of pebblestone, coarse grained sandstone, Fig. 2  Contour map ( a ) and 3-D view ( b ) of the old landslide areaand studied landslideEnviron Geol (2009) 58:1167–1175 1169  1 3  and to lesser extent, siltstone and mudstone of Miocene. Inthe study area, coastal outcroppings of the landslide-proneGu¨zelyal ı  Formation are composed mainly of claystoneand mudstone with brown reddish color.The study area is placed at the ‘‘semi-humid forest sub-region’’ of the Marmara transitional region. Based on thedata of C¸anakkale Meteorological Station placed at 6 mabove sea level with geographical coordinates of 40  09 0 Nand 26  25 0 , the area receives maximum and minimumprecipitations in winter and summer seasons, respectively(Tu¨rkes¸ 1996). Long-term averages of monthly and annual precipitation totals show that the wettest and driest monthsof the year are December (112.3 mm) and August(6.6 mm), respectively. The annual precipitation amount is619.7 mm. In spite of obvious decreasing trend of precip-itation amounts (Fig. 3a) since 1992, the landslide areareceives an average annual precipitation amount of 715.91 mm. Based on the interviews with villagers of theAdatepe village located in 700 m west of the landslidearea, landslide activity commenced in rainy period of 2005and showed several recurrences in every rainy period. Therelative increase in total precipitation amounts in 2005,2006 and 2007 (Fig. 3b) was likely responsible for therecurrence of these failures. In addition, three reactivationsoccurred in March (151.5 mm) and November (126.6 mm)of 2007 and March (26.6 mm) of 2008 (Fig. 3c).Although the landslide area is not drained by any majorstream, it takes a significant runoff from the old landslidearea with an area of 0.55 km 2 . In terms of drainage density(total stream length per unit area; Horton 1945), drainagedensity of the old landslide area estimated by MapInfo GISsoftware is 0.19 km km - 2 . The annual water input byprecipitation into the landslide area and its close environswas also computed regarding the surface drainage area of the old landslide area and annual mean value of precipi-tation as follows: P  ¼  AP m where  A  (55,490 m 2 ) is the old landslide area and  P m  is theannual average amount of precipitation (715.91 mm) in thestudy area. Thus, annual water input was calculated as4  9  10 4 m 3 , which is of importance with the potential of increasing water content and pore water pressure in thelandslide area. Results Geomorphological characteristics of the landslideBased on the interpretations from black and white aerialphotograph with scale of 1:10,000 taken in 1958, thelandslide area had been a shallow valley cut into theGu¨zelyal ı  member of the C¸anakkale formation in the northof old landslide body. The landslide displays typical mor-phological features of rotational slides. It has a total lengthand width of 100 and 30 m, respectively. The slideoccurred in the direction of N30  E and has a head arearepresented by a spoon-shaped slip surface that dipstowards the sea with angles varying between 44   and 63  (Fig. 4a). The slide body was brown reddish owing to itsgeochemical composition and in particular to the abundantpresence of iron oxide precipitates. Head scarp is 3–5 m inheight and has evident vertical striations (Fig. 4b).En echelon cracks are present as indicators of longitu-dinal and diagonal shear stresses along two sides of themain body. The western flank of the slide body is obviouslymarked by slip surfaces with slide-parallel striations and Fig. 3  Changes in annualprecipitation amounts between1975 and 2007 ( a ), Start of landslide activity period from2005 to March 2008, ( b ) andlast reactivations in 2007 and2008 ( c )1170 Environ Geol (2009) 58:1167–1175  1 3  polishing (Fig. 4c). These striations possess fresh appear-ance and dip angles ranging between 16   and 31  , implyingthe existence of a concavely upward-curving failure sur-face according to definition of Varnes (1978).The morphological details determined between headscarp and accumulation zone showed the presence of morethan one sliding surface marked by minor scarps, whichintersect the main slip surface at high angles. These sec-ondary steep slip surfaces have dips greater than 45  . Opentransverse and radial tension cracks are very common onthe toe areas at downslide end. The toe material shows afan-like spreading on beach materials (Fig. 4d).The nature of slip surface materialThe nature of slip surface material was studied on the basisof grain-size distribution analyses, EDX-SEM and XRD.Before clay mineral determination, grain-size distributionanalyses on the samples extracted from slip surfaces werefirst analyzed using hydrometry technique.According to the results of hydrometer tests, three slipsurface materials taken from main head scarp are, on theaverage,composedof37%sand,34%siltand29%clay.Thematerial does not contain particles (gravel or larger parti-cles). Three representative samples collected from differentparts of the lateral shear surface, however, consist of 30%sand and 30% silt and 40% clay. One sample taken from thezone of accumulation (toe material covering beach sands)also showed the total amounts of sand (45%), silt (48%) andclay (7%). These data indicate that slip surface materialcontains more abundant mixture of silt and clay. In otherwords, slide-prone nature of the sliding body is associatedwith the excessive amount of clayey silt. Such a mixedcomposition of slip surfaces of landslides has been reportedby several authors previously (Thevanayagam 1998; Youdand Gilstrap 1999; Vallejo and Mawby 2000; Wang and Sassa 2003). In addition, the abundance in sand fractionmight be related to the adjacent sandstone escarpment lim-iting the landslide area in the west side, since friction duringthe slide with sandstone might have caused increase in sand-size materials composed mainly of quartz.In the light of the abundance of fine particles, EDX/ SEM and XRD analyses were carried out to explain ele-mental composition of the slip surface material. EDX datawere given in Table 1 to reveal the nature of slip surfacematerial.Three samples extracted from the main head scarp with5 m vertical displacement showed the enrichment of Si(20%) and Fe (13%). Three samples of lateral slip surfacewere also found to be similar dominated by total amountsof 27 and 12% in Si and Fe fractions, respectively. Here,the Si content is relatively higher due likely to the addi-tional Si induced by friction with sandstone along thelateral slip surface as indicated above. Fig. 4  Pictures from the landslide; head scarp area ( a ), vertical striations on the main head scarp ( b ) and lateral shear surface ( c ), landslide pondand toe in the accumulation area covering beach materials ( d )Environ Geol (2009) 58:1167–1175 1171  1 3
Related Search
Similar documents
View more...
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