Cruden_and_Varnes_1996.pdf

~Chapter 3

LANDSLIDE TYPESAND PROCESSES

1. INTRODUCTION

.T he range of landslide processes is reviewed inthis chapter, and a vocabulary is provided for

d($cribing the features of landslides relevantto their classification for avoidance, control, or re-mediation. The classification of landslides in theprevious landslide report (Varnes 1978) has beenwidely adopted, so departures from it have beenminimized and the emphasis is on the progress madesince 1978. Although this chapter i~ complete initself, particular attention is drawn to 'changes andadditions to the vocabulary used by Varnes in theprevious report and the reasons for the changes.

. The term lO.ndslide denotes "the movement ofa mass of rock, debris or earth down a slope"(Cruden 1991, 27). The phenomena described aslandslides are not limited either to the land or tosliding; the word as it is now used in NorthAmerica has a much more extensive meaning thanits component parts suggest (Cruden 1991). Thecoverage in this chapter will, however, be identi-cal to that of the previous report (Varnes 1978).Ground subsidence and collapse are excluded, andsnow avalanches and ice falls are not discussed.

This chapter also follows Varnes's expressedintention of (1978, 12) "developing and attemptingto make more precise a useful vocabulary of termsby which ... [landslides] ... may be described." Theterms Varnes recommended in 1978 are largelyretained unchanged and a few useful new termshave been added. Eliot (1963,194) noted:

36

DAVID M. CRUDEN ANDDAVID J. VARNES

... Words strain,Crack and sometimes break, under the burden,Under the tension, slip, slide, perish,Decay with imprecision, will not stay in place,Will not stay still. ...

Such displaced terms are identified in this chapter.Following Varnes (1978), the use of terms relatingto the geologic, geomorphic, geographic, or cli-matic characteristics of a landslide has been dis-couraged, and the section in the previous report inwhich these terms are discussed has been deleted.

The viewpoint of the chapter is that of theinvestigator responding to a report of a landslideon a transportation route. What can' be usefullyobserved and how should these observations besuccinctly and unambiguously described?

The technic;al literature describing landslideshas grown 'Considerably since 1978. An importantsource of landslide information is the proceedingsof the International Symposium on Landslides.The third symposium met in New Delhi, India(Swaminathan 1980), and the symposium has sincemetquadriennially in Toronto, Canada (CanadianGeotechnical Society 1984); in Lausanne, Switzer-land (Bannard 1988); and in Christchurch, NewZealand (Bell 1992); it is scheduled to meet inTrondheim, Norway, in 1996. ..

Among the other important English languagetexts and collections of descriptions of landslideshave been those by Zaruba and Mencl (1982),Brunsden and Prior (1984), Crozier (1986), and

Costa and Wieczorek (1987). Eisbacher andClague (1984) and Skermer's translation of Heim'sBergsturz und Menschenleben (1932) have madedescriptions of the classic landslides of the Euro-pean Alps more accessible to North Americans.

Important reviews of landsliding around theworld were edited by Brabb and Harrod (1989) andKozlovskii (1988). Kyunttsel (1988) reviewedexperience with classification in the USSR andnoted "considerable divergences of views betweenvarious researchers concerning the mechanismsunderlying certain types oflandslides. This appliesparticularly to lateral spreads."

A historical perspective has been added to, thediscussion of spreading to show that this type oflandslide was recognized in North America over -100 years ago and is represt:,nted here by someextremely large movements. Both the size and thegentle slopes of these movements command par-ticular attention.

Crozier commented:

The two generalized classifications most likelyto be encountered in the English speakingworld are by}.N. Hutchinson (1968; Skemptonand Hutchinson, 1969) and D.}. Varnes (1958;1978).... Both authors use type of movementto establish the principal groups.... The majordistinction between the two classifications isthe difference accorded to the status of flowmovements ... slope movements which areinitiated by shear failure on distinct, boundaryshear surfaces but which subsequently achievemost of their translational movement byflowage . . . this dilemma depends on whetherthe principal interest rests with analyzing theconditions offailure or with treating the resultsof movement. Hutchinson's classificationappears to be related more closely to this firstpurpose. . .. Both Hutchinson's and Varnes'classifications have tended to converge over

. recent years, particularly in terminology.. .' .Whereas Varnes' scheme is perhaps easierto apply and requires less expertise to use,Hutchinson's classification has particularappeal to the engineer contemplating stabilityanalysis. (Crozier 1986, Ch. 2)

The synthesis of these two classifications has con-tinued. Hutchinson (1988) included topples, andthis chapter has benefited from his comments,' InSection 4 of this chapter particularly, which deals

Landslide Types and Processes

with landslide activity, many of Hutchinson's sug-gestions from the Working Party on the WorldLandslide Inventory (WP/WLl) have beenadopted (WP/WLI 1993a,b).

Under Hutchinson's chairmanship, the Interna-tional Association of Engineering Geology (IAEG)Commission on Landslides and Other Mass Move-ments continued its work on terminology. Thedeclaration by the United Nations of the Inter-national Decade for Natural Disaster Reduction(1990-2000) prompted the IAEG Commission'sSuggested Nomenclature for Landslides (1990) andthe creation of the WP/WLI by the InternationalGeotechnical Societies and the United NationsEducational, Scientific, and Cultural Organization(UNESCO). The Working Party, formed from theIAEG Commission, the Technical Committee onLandslides of the International Society for SoilMechanics and Foundation Engineering, and nom-inees of the International Society for Rock Me-chanics, published Directory of the World LandslideInventory (Brown et al. 1992) listing many workersinterested in the description of landslides world-wide. The Working Party has also prepared theMultilinguaL Landslide Glossary, which will encour-age the use of standard terminology in describinglandslides (WP/WLI 1993b). The terminologysuggested in this chapter is consistent with the sug-gested methods and the glossary of the UNESCOWorking Party (WP/WLI 1990, 1991, 1993a,b).

2. FORMING NAMES

The criteria used in the classification of landslidespresented here follow Varnes (1978) in emphasiz-ing type of movement and type of material. Anylandslide can be classified and described by twonouns: the first describes the material and the sec-ond describes the type of movement, as shown inTable 3-1 (e.g., rock fall, debris flow).

The names for the types of materials are un-changed from Varnes's classification (1978): rock,debris, and earth. The definitions for these terms aregiven in Section 7. Movements have again beendivided into five types: falls, topples, slides, spreads,and flows, defined and described in Section 8. Thesixth type proposed by Varnes (1978, Figure 2.2),complex landslides, has been dropped from the for-mal classification, although the term complex hasbeen retained asa description of the style of activ-ity ofa landslide. Complexity can also be indicated

37

38 Landslides: Investigation and Mitigation

Table 3-1Abbreviated Classification of Slope Movements

TYPE OF MATERIALTYPE OF ENGINEERING SOILSMOVEMENT BEDROCK PREDOMINANTLY COARSE PREDOMINANTLY FINE .

Fall Rock fall Debris fall Earth fallTopple Rock topple Debris topple Earth toppleSlide Rock slide Debris slide Earth slideSpread Rock spread Debris spread Earth spreadFlow Rock flow Debris flow Earth flow

Table 3-2Glossary for Forming Names of Landslides

NOTE: Subsequent movements may be described by repeating the above descriptors asmany times as necessary.

DESCRIPTION OF FIRST MOVEMENT

RATE WATER CONTENT MATERIAL TYPE

Extremely rapid Dry Rock FallVery rapid Moist Soil ToppleRapid Wet Earth SlideModerate Very wet Debris SpreadSlow FlowVery slowExtremely slow

DESCRIPTION OF SECOND MOVEMENT

RATE WATER CoNTENT MATERIAL TYPE

Extremely rapid Dry Rock FallVery rapid Moist Soil ToppleRapid Wet Earth SlideModerate Very wet Debris SpreadSlow FlowVery slowExtremely slow

The name of a landslide can become more elab-orate as more information about the movementbecomes available. To build up the complete iden-tification of the movement, descriptors are addedin front of the two-noun classification using a pre-ferred sequence of terms. The suggested sequence .provides a progressive narrowing of the focus of thedescriptors, first by time and then by spatialloca-tion, beginning with a view of the whole landslide,continuing with parts of the movement, and finallydefining the materials involved: The recommendedsequence, as shown in Table 3-2, describes activity(including state, distribution, and style) followedby descriptions of all movements (including rate,water content, material, and type).

This sequence is followed throughout the chap-ter and all terms given in Table 3-2 are highlightedin bold type and discussed. Second or subsequentmovements in complex or composite landslidescan be described by repeating, as many times asnecessary, the descriptors used in Table 3-2.Descriptors that are the same as those for the firstmovement may then be dropped. from the name.

For instance, the very large, and rapid slopemovement that occurred near the town of Frank,Alberta, Canada, in 1903 (McConnell and Brock1904) was a complex, extremely rapid, dry rockfall-debris flow (Figure 3~1). From the full name ofthis landslide at Frank, one would know that boththe debris flow and the rock fall "Yere extremelyrapid and dry because no other descriptors are usedfor the debris flow.

As discussed in Section 4.3, the addition of thedescriptor complex to the name indicates thesequence of movement in the landslide and dis-tinguishes this landslide from a composite rockfall-debris flow, in which rock fall and debris flowmovements were occurring, sometimes siinultane-ously, on different parts of the displaced mass. The

STYLE

ComplexCompositeMultipleSuccessiveSingle

by combining the five types of landslide in the wayssuggested below. The large classification chartaccompanying the previous report (Varnes 1978,Figure 2.1) has been divided into separate figuresdisrributed throughout this chapter.

DISTRIBUTION

AdvancingRerrogressiveWideningEnlargingConfinedDiminishingMoving

ACTIVITY

STATE

ActiveReactivatedSuspendedInactive

DormantAbandonedStabilizedRelict

Landslide Types and Processes 39

FIGURE 3-1Slide at Frank,Alberta, Canada(oblique aerialphotograph fromsouth). About 4:10a.m. on April 29,1903, about 85million tonnes ofrock moved downeast face of TurtleMountain, acrossentrance of Frankmine of CanadianAmerican CoalCompany,Crowsnest River,southern end oftown of Frank,main road fromeast, and CanadianPacific mainlinethrough CrowsnestPass. Displacedmass continued upopposite side ofvalley before comingto rest 120 m abovevalley floor. Eventlasted about100 seconds.PHOTOGRAPHNAPL T3IL213; REPRODUCED FROMCOLLECTION OF NATIONALAIR PHOTO LIBRARY WITHPERMISSION OF NATURALRESOURCES CANADA

full name of the landslide need only be given once;subsequent references should then be to the initialmaterial and type of movement, for example, "therock fall" or "the Frank rock fall" for the landslideat Frank, Alberta.

Several noun combinations may be required to. identify the multiple types of material and move-

ment involved in a complex or composite land-slide. To provide clarity in the description, a dashknown as an "en dash" is used to link these stages,as in rock fall-debris flow in the example above.(An en dash is half the length ofa regular dash andlonger than a hyphen; it is used to remove ambi-guity by indicating linkages between terms com-posed of two nouns.)

The full name of a landslide may be cumber-some and there is a natural tendency, particularly

among geologists, to establish type examples withwhich other landslides may be compared. Shreve(1968), for instance, referred to the landslide inFrank, Alberta, as belonging to the Blackhawk type.It seems clear that type examples should be historiclandslides that have been investigated in detaif'shortly after their occurrence and are of continu-ing interest to landslide specialists. In addition, fora type example to be useful, other landslides withthe same descriptors should occur in similar mate-rial. The Blackhawk landslide (Figure 3-2) was aprehistoric landslide, and thus was not subject toinvestigation at its occurrence (Shreve 1968). It istherefore not a suitable type example; neverthe-less, it may have been a Frank-type landslide.

Although Varnes (1978, 25) discussed "termsrelating to geologic, geomorphic, geographic, or

---------------------_._._._.._ .

40

FIGURE 3-2Blackhawk landslide:view upslope tosouth over lobeof dark marblebreccia spreadbeyond mouth ofBlackhawk Canyonon north slope ofSan BernardinoMountains in;outhern California.Maximum width ofobe is 2 to 3 km;1eight of scarp atlear edge is about15 m [Varnes 1978,:igure 2.28 (Shelton1966)].:OPYRIGHT JOHN s.,HELTON

:IGURE 3-3310ck diagram ofdealized complex~arth slide-earthlow (Varnes 1978,:igure 2.1 t).

Landslides: Investigation and Mitigation

climatic setting," he recommended against thepractice of using type examples because the terms"are not informative to areader who lacks knowl-edge of the locality" (1978, 26). Moreover, typeexamples are impractical because of the sheernumber required to provide a fairly complete land-slide classification. About 100,000 type exampleswould be required for all the combinations ofdescriptors and materials with all the types ofmovement defined in Table 3-2 and in the follow-ing sections, although admittedly some combina-

tions may be unlikely. The inclusion of complexand composite landslides would increase the num-ber of type examples to over a billion.

3. LANDSLIDE FEATURES AND GEOMETRY

Before landslide types are discussed, it is useful toestablish a nomenclature for the observable land-slide features and to discuss the methods ofexpress-ing the dimensions and geometry of landslides.

3.1 Landslide Features

Varnes (1978, Figure 2.It) provided an idealizeddiagram showing the features for a complex earthslide-earth flow, which has been reproduced here asFigure 3-3. More recently, the IAEG Commissiorion Landslides (1990) produced a new idealizedlandslide diagram (Figure 3-4) in which the vari-ous features are identified by numbers, which aredefined in different languages by referring to theaccompanying tables. Table 3-3 provides the defi-nitions in English.

The names of the features are unchanged fromVarnes's classification (1978). However, Table3-3 contains explicit definitions for the surface ofrupture (Figure 3-4, 10), the depletion (16), thedepleted mass (17), and the accumulation (18) andexpanded definitions for the surface of separa.tion(12) and the flank (19). The sequence of the firstnine landslide features has been rearranged to pro-ceed from the crown above the head of the dis-placed material to the toe at the foot of the

'.1"

I


FIGURE 3-4Landslide features:upper portion, planof typical landslidein which dashed lineindicates trace ofrupture surface onoriginal groundsurface; lowerportion, section inwhich hatchingindicates undisturbedground and stipplingshows extent of "displaced material.Numbers refer tofeatures defined inTable 3-3 (lAEGCommissionon Landslides 1990).

19

19

6

/---/

/I(

11..

Bfi-9

B

displaced material. This sequence may make thesefeatures easier to remember.

It may also be helpful to point out that in thezone of depletion (14) the elevation of the groundsurface decreases as a result oflandsliding, whereasin the zone of accumulation (15) the elevation ofthe ground surface increases. If topographic mapsor digital terrain models of the landslide exist forboth before and after movements, the zones ofdepletion and accumulation can be found from thedifferences between the maps or models. The vol-ume decrease over the zone of depletion is, ofcourse, the depletion, and the volume increaseover the zone of accumulation is the accumula-tion. The accumulation can be expected to belarger than the depletion because the ground gen-erally dilates during landsliding.

Table 3-3Definitions of Landslide Features

NUMBER NAME

1 Crown2 Main scarp

3 Top4 Head5 Minor scarp

6 Main body

7 Foot

8 Tip9 Toe

10 Surface of rupture

11 Toe of surface ofrupture

12 Surface of separation13 Displaced material

14 Zone of depletion

15 Zone of accumulation16 Depletion17 Depleted mass

18 Accumulation19 Flank

20 Original ground surface

DEFINITION

Practically undisplaced material adjacent 'to highest parts of main scarpSteep surface on undisturbed ground at upper edge of landslide caused by movement of

displaced material (13, stippled area) away from undisturbed ground; it is visible part ofsurface of rupture (10)

Highest point of contact between displaced material (13) and main scarp (2)Upper parts of landslide along contact between displaced material and main scarp (2)Steep surface on displaced material of landslide produced by differential movements within

displaced materialPart of displaced material of landslide that overlies surface of rupture between main scarp

(2) and toe of surface of rupture (11) .Portion of landslide that has moved beyond toe of surface of rupture (11) and overlies

original ground surface (20)Point on toe (9) farthest from top (3) of landslideLower, usually curved margin of displaced material of a landslide, most distant from main

scarp (2)Surface that forms (or that has formed) lower boundary of displaced material (13) below

original ground surface (20); mechanical idealization of surface of rupture is called slipsurface in Chapter 13

Intersection (usually buried) between lower part of surface of rupture (10) of a landslideand original ground surface (20)

Part of original ground surface (20) now overlain by foot (7) of landslideMaterial displaced from its original position on slope by movement in landslide; forms both

depleted mass (17) and accumulation (I8); it is stippled in Figure 3-4Area of landslide within which displaced material (13) lies below original ground

surface (20)Area of landslide within which displaced material lies above original ground surface (20)Volume bounded by main scarp (2), depleted mass (17), and original ground surface (20)Volume of displaced material that overlies surface of rupture (10) but underlies original

ground surface (20)Volume of displaced material (13) that lies above original ground surface (20)Undisplaced material adjacent to sides of surface of rupture; compass directions are

preferable in describing flanks, but if left and right are used, they refer to flanks as viewed'from crown

Surface of slope that existed before landslide took place

b

42

FIGURE 3-5Landslide dimensions:upper portion, planof typical landslidein which dashed lineis trace of rupturesurface on originalground surface;lower portion,section in whichhatching indicatesundisturbed ground,stippling showsextent of displacedmaterial, and brokenline is original groundsurface. Numbersrefer to dimensionsdefined in Table 3-4(IAEG Commissionon Landslides 1990).


3.2 Landslide Dimensions

The IAEG Commission on Landslides (1990)utilized the nomenclature described in Section3.1 (including Figure 3-4 and Table 3-3) to pro-vide definitions of some dimensions of a typicallandslide. The IAEG Commission diagram isreproduced here as Figure 3-5. Once again, eachdimension is identified on the diagram by a num-ber, and these numbers are linked to tables givingdefinitions in several languages. Table 3-4 gives thedefinitions in English.

The quantities Ld , Wd , Dd and Lr ,Wr ' Dr areintroduced because, with an assumption about theshape of the landslide, their products lead to esti-mates of the volume of the landslide that are use-

B

B

ful in remedial work. For instance, for many rota-tional landslides, the surface of rupture can beapproximated by half an ellipsoid with semiaxesq, W)2, LJ2. As shown in Figure 3-6(a), thevolume of an ellipsoid is (Beyer 1987, 162)

VOL =.1. 1ta . b . c.eps 3

where a, b, and care semimajor axes. Thus, thevolume of a "spoon shape" corresponding to one-half an ellipsoid is

1 4 4VOL =- -1ta . b . c =-1ta . b . CIs 2 3 6

But as shown in Figure 3-6(b), for a landslidea=D ,b=W /2, andc=L /2. Therefore, the volumer r rof ground displaced by a landsBde is approximately

4 4VOL = - 1ta . b . c =- 1tD . W /2 . L /2Is 6 .6' r r

1=-1tD W L6 r , ,

This is the volume of material before the landslidemoves. Movement usually increases the volume ofthe material being displaced because the displacedmaterial dilates. After the landslide, the volume ofdisplaced material can be estimated by 1/61tDdWd Ld(WP/WLI 1990, Equation 1).

A term borrowed from the construction indus-try, the swell factor, may be used to describe theincrease in volume after displacement as a percen-tage of the volume before displacement. Church(1981, Appendix 1) suggested that a swell factor of67 percent "is an average figure obtained fromexisting data for solid rock" that has been mechan-

Table 3-4Definitions of Landslide Dimensions

NUMBER

12345

6

78

NAME

Width of displaced mass, WdWidth of surface of rupture, W

r

Length of displaced mass, LdLength of surface of rupture, L,Depth of displaced mass, Dd

Depth of surface of rupture, D,

Total length, LLength ofcenter line, Lei

DEFINITION

Maximum breadth of displaced mass perpendicular to length, LdMaximum width between flanks of landslide perpendicular to length, L,Minimum distance from tip to topMinimum distance from toe of surface of rupture to crownMaximum depth of displaced mass measured perpendicular to

plane containing Wd and LdMaximum depth of surface of rupture below original ground surface

measured perpendicular to plane containing W, and L,Minimum distance from tip of landslide to crownDistance from crown to tip of landslide through points on original

ground surface equidistant from lateral margins of surface of ruptureand displaced material

(a) Ellipsoid


(b) Landslide FIGURE 3-6Estimation oflandslide volumeassuming ahalf-ellipsoidshape.

43

ically excavated. His estimates may approximatethe upper bound for the swell due to landsliding.Nicoletti and Sorriso-Valvo (1991) chose an aver-age dilation of 33 percent, so 4DT~Lr '" 3Dd~Ld'More precise information is as yet unavailable.

The ground-surface dimensions of the displacedmaterial, Ld,~ , and of the surface of rupture, W:'and the total length, L, of the landslide can bemeasured with an electronic distance-measuringinstrument; a rangefinder may be sufficiently pre-cise for a one-person reconnaissance. Measure-ment of the distance L

Tmay present problems

because the toe of the surface of rupture is oftennot exposed. Its position can sometimes be esti-mated from graphical extrapolations of the mainscarp supported by measurements ofdisplacementswithin the displaced mass (Cruden 1986). Al-though Dd and DT can also be estimated by thesetechniques, site investigations provide more pre-cise methods of locating surfaces of rupture underdisplaced material (Hutchinson 1983).

The total length of the landslide, dimension L(5, Figure 3-5), is identical with length L, "themaximum length of the slide upslope," shown inFigure 3-3 (Varnes 1978); both are the straight-line distances from crown to tip. Readers are cau-tioned that several writers define the length of alandslide in terms of its horizontal extent and fre-quently use the letter L to define this horizontaldistance in tabulations of observations arid in cal-culations. This use ofL is a source of potential con-fusion and inaccuracy, and readers should makecertain that they can identify the dimension beingspecified by L in every case.

It should also be emphasized that it is unlikelythat the displaced material at the tip has traveled

','.;,

from the crown despite the frequency of this as-sumption, originally due to Heim (1932). Materialdisplaced from close to the landslide crown usuallycomes to rest close to the head of the landslide.Nicoletti and Sorriso-Valvo (1991) proposed thatan estimate of the "overall runout" ofa landslide bedetermined by measuring the length of a line con-structed along the original ground surface equidis-tant from the lateral margins of the displacedmaterial. However, such measurements may nothave immediate physical significance and are alsomore difficult and imprecise than measurements ofL. The length of the landslide measured throughthese central points is called the length ofcenter line,Lei' Note that Lei will increase w~th the number ofpoints surveyed on the center line, and the ratioL)L will increase with the curvature of the centerline in plan and section.

The difference in elevation between the crownand the tip of the landslide may be used to deter-mine H, the height of the landslide. Combiningestimates of Hand L allows computation of thetravel angle ex, as shown by Figure 3-7. If the tip isvisible from the crown, the travel angle can be mea-

FIGURE 3-7Definition oftravel angle (ex)of a landslide.

11

44

FIGURE 3-8Mobility ofsturzstroms, chalkdebris flows, andlandslides in minewastes related totravel angle (a)and debris volume

. (modified fromHutchinson 1988,Figure 12).


sured directly with a hand clinometer. The H-valuemay be conveniently estimated with an altimeterwhen tip and crown are accessible but not visiblefrom each other. Hutchinson (1988) compiled datafrom several different types of debris flows to illus-trate how debris-flow mobility appears to be relatedto the travel angle-and to the volume and lithol-ogy of the displaced material (Figure 3-8).

The measurements discussed above are ade-quate during reconnaissance for defining the basicdimensions of single-stage landslides whose dis-placement vectors parallel a common plane. Suchlandslides can be conveniently recorded on a land-slide report such as that shown in Figure 3-9.Estimates oflandslide volume determined by thesemethods are imprecise when topography divertsthe displacing material from rectilinear paths.More elaborate surveys and analyses are then nec-essary (Nicoletti and Sorriso-Valvo 1991).

4. LANDSLIDE ACTIVITY

The broad aspects of landslide activity should beinvestigated and described during initial recon-naissance of landslide movements and before moredetailed examination of displaced materials isundertaken. The terms relating to landslide age

and state of activity defined by Varnes (1978) andsome of his terms defining sequence or repetitionof movement have been regrouped under threeheadings:

1. State of Activity, which describes what isknown about the timing of movements;

2. Distribution of Activity, which describesbroadly where the landslide is moving; and

3. Style of Activity, which indicates the mannerin which different movements contribute to thelandslide.

The terms used to define these three characteristicsof landslide activity are given in the top section ofTable 3-2 and are highlighted in bold type the firsttime they are used in the following sections.

The reader is cautioned that the following dis-cussions relate to the terminology proposed by theUNESCO Working Party (WP/WLI 1990, 1991,1993a,b) and given in Table 3-2. Other reports andauthors may use classifications that apply differentmeanings to apparently identical terms. For exam-ple, in Chapter 9 of this report, a Unified LandslideClassification System is introduced that is based onlandslide classification concepts presented byMcCalpin (1984) and Wieczorek (1984). This sys-

-II)Q)

50 Q)~0)Q)'g'-"

40 w-ICIZ

30-Iw>

20 a:I-

1 0

010 8 10 9

KEY_ Chalk failures forming a taluso Chalk failures forming a flow slide Flow slide in coal mine wasteo Flow slide in kaolinized granite

7 6 Approximate value of H (m)+ More mobile Alpine and Cordilleran

sturzstroms (reported by Hsu,1 975)

lOS 10 6 10 7

DEBRIS VOLUME (m3)

0.0 L- ---J ---Ji-- ---J .....I._---'__--L ---J_....

10 3

1.6-1;:- ..........

,..., zo "-- "-

talusQ)-Z1C,

'\ format! onc: 1.2C'lI'5\ TentativeQi "- \ for chalk> '\C'lI increasing~ \... :\ degree of

'-" 0.8c: 9~~~ ,,,,,'.C'lI... \II \ '\

..... \\ \ "--I 0.4 ~}. ~8Z .............. "- 135Crt:t_:I:

'\ ' " .0 138 0 - .......... _ 85 145

granite -

r----...,..-----r------r----~----r_---___r__= 60

LANDSLIDE REPORT

Inventory Number:

Date of Report:day month year

Date of Landslide Occurrence:day month year

Landslide Locality:

Reporter's Name:

Affiliation:

Address:

Phone:

Degrees Minutes Seconds

Position: Latitude

Longitude

Elevation: crown m as.I.

Surface of rupture toe m as.I.

tip m as.I.

Geometry: Surface.of rupture Displaced Mass

Length L= LeF L=r

Width W= Wd =r

Depth D= Dd =r

Volume: V= nLdDdW./6 orV = Swell factor =

V= m3 x 10" n=

Damage: Value

Injuries Deaths

FIGURE 3-9Proposed standardlandslide reportform.

46

FIGURE 3-10Sections throughtopples in differentstates of activity:(1) active-erosion attoe of slope causesblock to topple;(2) suspended-local cracking incrown of topple;(3) reactivated-another blocktopples;(4) dormant-displaced massbegins to regain itstree cover andscarps are modifiedby weathering;(5) stabilized-fluvialdeposition stabilizestoe of slope, whichbegins to regain itstree cover; and(6) relict-uniformtree cover over slope.


tem is compared with a stability classification pro-posed by Crozier (1984). For further informationon such alternative systems, the reader should referto Tables 9-1, 9-2, and 9-6 and the associated textin Chapter 9.

4.1 State of Activity

Figure 3-10 illustrates the several states of activityby using an idealized toppling failure as an exam-ple. Active landslides are those that are currentlymoving; they include first-time movements andreactivations. A landslide that is again active afterbeing inactive may be called reactivated. Slidesthat are reactivated generally move on preexistingshear surfaces whose strength parameters approachresidual'(Skempton 1970) or ultimate (Krahn andMorgenstern 1979) values. They can be distin-guished from first-time slides on whose surfaces ofrupture initial resistance to shear will generallyapproximate peak values (Skempton and Hutch-inson 1969). Landslides that have moved withinthe last annual cycle of seasons but that are not

1~2

6

moving at present were described by Varnes(1978) as suspended.

Inactive landslides are those that last movedmore than one annual cycle of seasons ago. Thisstate can be subdivided. If the causes of movementremain apparent, the landslide is dormant. How-ever, if the river that has been eroding the toe ofthe moving slope changes course, the landslide isabandoned (Hutchinson 1973; Hutchinson andGostelow 1976). If the toe of the slope has beenprotected against erosion by bank armoring or ifother artificial remedial measures have stoppedthe movement, the landslide can be described asstabilized.

Landslides often remain visible in the landscapefor thousands of years after they have moved andthen stabilized. Such landslides were called ancientor fossil by Zaruba and Mencl (1982, 52), perhapsbecause they represent the skeletons of once-active movements. When these landslides havebeen covered by other deposits, they are referredto as buried landslides. Landslides that have clearlydeveloped under different geomorphic or climaticconditions, perhaps thousands of years ago, canbe called relict. Road construction in southernEngland reactivated relict debris flows that hadoccurred under periglacial conditions (Skemptonand Weeks 1976).

Within regions, standard criteria might bedeveloped to assist in distinguishing suspendedlandslides from dormant and relict landslides.These criteria would describe the recolonizationby vegetation of surfaces exposed by slope move-ments and the dissection of the new topography bydrainage. The rate of these changes depends onboth the local climate and the local vegetation, sothese criteria must be used with extreme caution.Nevertheless, it is generally true that when themain scarp of a landslide supports new vegeta-tion, the landslide is usually dormant, and whendrainage extends across a landslide without obvi-ous discontinuities, the landslide is commonlyrelict. However, these generalizations must be con-firmed by detailed study of typical slope move-ments under local conditions; Chapter 9 providesa systematic approach for such determinations.Figure 9-7 shows some idealized stages in the evo-lution of topographic features on suspended, dor-mant, and relict landslides.

The various states of activity are also defined byan idealized graph of displacement versus time(Figure 3-11). For an actual landslide, such a graph


FIGURE 3-11(far left)Displacement oflandslide in differentstates of activity.

FIGURE 3-12 (left)Sections throughlandslides showingdifferent distributionsof activity:(1) advancing,(2) retrogressing,(3) enlarging,(4) diminishing, and(5) confined.In 1-4, Section 2shows slope aftermovement onrupture surfaceindicated by sheararrow. Stipplingindicates displacedmaterial.

5

the surface of rupture may be enlarging, continuallyadding to the volume of displacing material. If thesurface of rupture of the landslide is enlarging intwo or more directions, Varnes (1978, 23) sug-gested the term progressive for the landslide, notingthat this term had also been used for both advanc-ing and retrogressive landslides. The term is alsocurrently used to describe the process by which thesurface of rupture extends in some. landslides (pro-gressive failure). The possibility of confusion seemssufficient now to abandon the term progressive infavor of describing the landslide as enlarging.Hutchinson (1988, 9) has drawn attention to con-fined movements that have a scarp but no visiblesurface of rupture in the foot of the displaced mass.He suggested that displacements iI} the head of thedisplaced mass are taken up by compression andslight bulging in the foot of the mass.

To complete the possibilities, terms are neededfor landslides in which the volume ofmaterial beingdisplaced grows less with time and for those land-slides in which no trend is obvious. The term dimin-ishing for an active landslide in which the volumeof material being displaced is decreasing with timeseems free of undesired implications. A landslide inwhich displaced materials continue to move butwhose surface of rupture shows no visible changescan be simply described as moving. Several types of

SUSPENDED

Time in years

SUSl'ENDED-+

can be created by plotting differences in the posi-tion of a target on the displacing material withtime. Such graphs are particularly well suited toportraying the behavior of slow-moving landslidesbecause they presuppose that the target is not dis-placed significantly over the time period duringwhich measurement takes place. The velocity ofthe'target can be estimated by the average rate ofdisplacement of the target over the time periodbetween measurements.

There is some redundancy in using the descrip-tions of activity state with those for rate of move-ment (see Section 5). Clearly, if the landslide hasa measurable rate of movement, it is either activeor reactivated. The state of activity might then beused to refer to conditions before the currentmovements of the landslide. If, for instance, reme-dial measures had been undertaken on a landslidethat is now moving with moderate velocity, thelandslide might be described as a previously stabi-lized, moving, moderate-velocity landslide. Landslideswith no discernible history of previous movementwould be described as active.

4.2 Distribution of Activity

Varnes (1978) defined a number of terms that canbe used to describe the activity distribution in alandslide. Figure 3-12 shows idealized sectionsthrough landslides exhibiting various distributionsof activity.

If the surface of rupture is extending in the direc-tion of movement, the landslide is advancing,whereas if the surface of rupture is extending in thedirection opposite the movement of the displacedmaterial, the landslide is said to be retrogressive. Ifthe surface of rupture is extending at one or bothlateral margins, the landslide is widening. Move-ment may be limited to the displacing material or


I1

I2~

IiI!tf1I

tions through landslides exhibiting various stylesof activity. Varnes defined complex landslides asthose with at least two types of movement. How-ever, it is now suggested that the term complex belimited to cases in which the various movementsoccur in sequence. For instance, the toppledescribed by Giraud et al. (1990) and shown asFigure 3-13(1), in which some of the displacedmass subsequently slid, is termed a complex rocktopple-rock slide. Not all the toppled mass slid, butno significant part of the displaced mass slid with-out first toppling. Some of the displaced mass maybe still toppling while other parts are sliding.

The term composite, formerly a synonym forcomplex, is now proposed to describe landslides inwhich different types of movement occur in differ-ent areas of the displaced mass, sometimes simulta-neously. However, the different areas of thedisplaced mass show different sequences of move-ment. For example, the structures shown in Figure3-13(2), first described by Harrison and Falcon(1934, 1936), were called slide toe topples byGoodman and Bray (1976), but according to theclassification proposed in this chapter, they are com-posite rock slide-rock topples. The term composite wasintroduced by Prior et al. (1968, 65, 76) to describemudflows in which "slipping and sliding ... occur inintimate association with flowing" and "the mater-ial ... behaves as a liquid and flows rapidly betweenconfining marginal shears." In the proposed namingconvention, such movements are composite earthslides-earth flows and the convention of treating thetopographically higher of the two movements as thefirst movement and the lower of the two movementsas the second movement was adopted.

A multiple landslide shows repeated movementsof the same type, often following enlargement ofthe surface of rupture. The newly displaced massesare in contact with previously displaced massesand often share a surface of rupture with them. Ina retrogressive, multiple rotational slide, such as thatshown in Figure 3-14, two or more blocks haveeach moved on curved surfaces of rupture tangen-tial to common, generally deep surfaces of rupture(Eisbacher and Clague 1984).

A successive movement is identical in typeto an earlier movement but in contrast to a multi-ple movement does not share displaced materialor a surface of rupture with it [Figure 3-13(3)].According to Skempton and Hutchinson (1969,297), "successive rotational slips consist of an

PionA

........-.

6m

B

landslide may exhibit diminishing behavior. Move-ment may stop in parts ofboth rotational slides andtopples after substantial displacements because themovements themselves reduce the gravitationalforces on the displaced masses. Similarly, move-ments of rock masses may rapidly dilate cracks inthe masses, cause decreases in fluid pore pressureswithin these cracks, and hence decrease rates ofmovement. However, it may be premature to con-clude that the displacing material is stabilizingbecause the volume being displaced is decreasingwith time. Hutchinson (1973) pointed out that theactivity of rotational slides caused by erosion at thetoe of slopes in cohesive soils is often cyclic.

4.3 Style of Activity

The style oflandslide activity, or the way in whichdifferent movements contribute to the landslide,can be defined by terms originally established byVarnes (1978, 23). Figure 3-13 shows idealized sec-

FIGURE 3-13Sections throughlandsliaes showingdifferent styles ofactivity. (1) Complex:gneiss (A) andmigmatites (I) areforming topplescaused by valleyincision; alluvialmaterials fill valleybottom; afterweathering furtherweakens toppledmaterial, some ofdisplaced massmoves by sliding(modified fromGiraud et al. 1990).(2) Composite:limestones haveslid on underlyingshales, causingtoppling failuresbelow toe of sliderupture surface(modified fromHarrison andFalcon 1934).(3) Successive:later landslide (AB)is same type aslandslide CD butdoes not sharedisplaced materialor rupture surface.(4) Single.

(a)

Comoositehead scarps

(b)


Inactive head scarp

49

: i

assembly of individual shallow rotational slips."Hutchinson (1967,116) commented that "irregu-lar successive slips which form a mosaic ratherthan a stepped pattern in plan are also found."

Single landslides consist of a single movementof displaced material, often as an unbroken block[Figure 3-13(4)]. For instance, Hutchinson (1988)described single topples in which a single blockmoved and contrasted these with multiple topples(Figure 3-15). Single landslides differ from theother styles of movement, which require disrup-tion of the displaced mass or independent move-ments of portions of the mass.

5. RATE OF MOVEMENT

in its lower limit. These two limits now span 10orders of magnitude. Interpretation of the scalewas aided by Morgenstern's (1985) analogy to theMercalli scale of earthquake intensity. He pointedout that the effects ofa landslide can be sorted intosix classes corresponding approximately to thesix fastest movement ranges of Varnes's scale.

(a) Single Topple

FIGURE 3-14 (above)(a) Map view and(b) cross section oftypical retrogressive,multiple rotational?Iide (Eisbacherand Clague 1984,Figure 10).

The previous rate-of-movement scale provided byVarnes (1978, Figure 2.1u) is shown here as Figure3-16. This scale is unchanged from Varnes's origi-nal scale (1958) except for the addition of theequivalent SI units, which range from meters persecond to millimeters per year. Varnes--(1958) did

\

not discuss the divisions of the scale, then given inunits ranging from feet per second to feet per 5years; the scale probably represented a codificationof informal practice in the United States at thetime. Nem cok et al. (1972) suggested a fourfolddivision of a similar range of velocities.

Figure 3-17 presents a modified scale of land-slide velocity classes. The divisions of the scalehave been adjusted to increase in multiples of 100by a slight increase in its upper limit and a decrease

weak substratum

(b) Multiple Topples

HWM

LWM

FIGURE 3-15Comparison of (a) single topple (Hutchinson 1988) with (b) multipletopples [modified from Varnes 1978, Figure 2.1 d1 (de Freitas andWatters 1973)].


Velocity Description Typical(ft/sec) Velocity

102 ExtremelyRapid101 10ft/sec 3 m/sec

100 Very Rapid10-1

lft/min 0.3 m/min10-

10-3 Rapid

10-4 5ft/day = 1.5 m/day

10-5 Moderate5ft/mo = 1.5 m/mo

10-6 Slow

10-75ft/yr = 1.5 m/yr

Very Slow10-8

1ft/5yr 60 mm/yr10-9 Extremely

Slow

Velocity Description VelocityClass (mm/sec)

FIGURE 3-16Varnes landslidemovement scale(Varnes 1978,Figure 2.1 u).

FIGURE 3-17Proposed landslidevelocity scale.

7 ExtremelyRapid

6 Very Rapid

5 Rapid

4 Moderate

3 Slow

2 Very Slow

1 ExtremelySlow

5xl03

5xl0 l

5x 10-1

5xl0-3

5xl0-5

5x 10-7

TypicalVelocity

5 m/sec

3 m/min

1.8 m/hr

13 m/month

1.6 m/year

16 mm/year

An added seventh class brings these effect classesinto correspondence with the divisions of thevelocity scale.

The Mercalli scale is based on descriptions oflocal effects of an earthquake; degrees of damagecan be evaluated by investigating a house or a sec-tion of a street. Yet the intensity value can be cor-related with the total energy release of the eventbecause both local damage and the area affected arerelated to the magnitude of the earthquake. The sit-uation is different for landslides. Small, rapid debrisavalanches are known to have caused total destruc-tion and loss of lives. In contrast, a large slopemovement ofmoderate velocity can have much lessserious effects because it can be avoided or thestructures affected can be evacuated or rebuilt. It issuggested that a measure of landslide risk shouldinclude both the area affected and the velocity; theproduct of these two parameters is approximatelyproportional to the power release of the landslide.

Varnes (1984) drew attention to the UnitedNations Disaster Relief Organization terminologyin which the specific risk, R" or the expected degreeof loss due to landsliding or any other natural phe-nomenon, can be estimated as the product of thehazard (H) and the vulnerability (V). The hazard isthe probability of occurrence of the phenomenonwithin a given area; the vulnerability is the degreeof loss in the given area of elements at risk: popula-tion, properties, and economic activities. The vul-nerability ranges from 0 to 1. In this terminologythe vulnerability of the landslide might wellincrease with velocity because it can be expectedthat extremely rapid landslides would cause greater

Iloss of life and property than slow landslides.

A parameter that is difficult to quantify is theinternal distortion of the displaced mass. Struc-tures on a moving mass generally are damagedin proportion to the internal distortion of theirfoundations. For example, the Lugnez slope inSwitzerland (Huder 1976) is a 2S-km2 area mov-ing steadily downward at a IS-degree angle at avelocity as high as 0.37 m/year. The movementshave been observed by surveying since 1887. Yetin the six villages on the slope with 300-year-oldstone houses and churches with bell towers, noneof these structures have suffered damage when dis-placed because the block is moving withoutdistortion. Damage will also depend on the typeof landslide, and each type may require separateconsideration.


Table 3-5Definition of Probable Destructive Significance of Landslides of DifferentVelocity Classes

Landslide velocity is a parameter whose destruc-tive significance requires independent definition.Table 3-5 defines the probable destructive signifi-cance of the seven velocity classes on the newlandslide velocity scale (Figure 3-17). Several casehistories in which the effects of landslides onhumans and their activities have been welldescribed and for which the landslide velocities arealso known are given in Table 3-6, which suggestsa correlation between vulnerability and landslidevelocity. An important limit appears to liebetween very rapid and extremely rapid move-ment, which approximates the speed of a personrunning (5 m/sec). Another important boundaryis between the slow and very slow classes (1.6m/year), below which some structures on the land-slide are undamaged. Terzaghi (1950, 84) identi-fied as creep those slope movements that were"proceeding at an imperceptible rate.... Typicalcreep is a continuous movement which proceedsat an average rate of less than a foot per decade.

LANDSLIDEVELOCITY

CLASS

7

6

5

4

3

21

PROBABLE DESTRUCfIVE SIGNIFICANCE

Catastrophe of major violence; buildings destroyed byimpact of displaced material; many deaths; escape unlikely

Some lives lost; velocity too great to permit all persons toescape

Escape evacuation possible; structures, possessions, andequipment destroyed

Some temporary and insensitive structures can betemporarily maintained

Remedial construction can be undertaken duringmovement; insensitive structures can be maintained withfrequent maintenance work if total movement is not largeduring a particular acceleration phase

Some permanent structures undamaged by movementImperceptible without instruments; construction possible

with precautions

Table 3-6Examples of Landslide Velocity and Damage

LANDSLIDE LANDSLIDE EsTIMATEDVELOCITY NAME OR LANDSLIDE

CLASS LOCATION REFERENCE VELOCITY DAMAGE

7 Elm Heim (1932) 70 m/sec 115 deaths7 Goldau Heim (1932) 70 m/sec 457 deaths7 Jupille Bishop (1973) 31 m/sec 11 deaths, houses destroyed7 Frank McConnell and Brock (1904) 28 m/sec 70 deaths7 Vaiont Mueller (1964) 25 m/sec 1,900 deaths by indirect damage7 Ikuta Engineering News Record (1971) 18 m/sec 15 deaths, equipment destroyed7 St. Jean Vianney Tavenas et al. (1971) 7 m/sec 14 d~aths, structures destroyed6 Aberfan Bishop (1973) 4.5 m/sec 144 deaths, some buildings

damaged5 Panama Canal Cross (l924) 1 m/min Equipment trapped, people

escaped4 Handlova Zaruba and Mend (1969) 6 m/day 150 houses destroyed, complete

evacuation3 Schuders Huder (l976) 10 m/year Road maintained with difficulty3 Wind Mountain Palmer (1977) 10 m/year Road and railway requir~ frequent

maintenance, buildings adjustedperiodically

2 Lugnez Huder (l976) 0.37 m/year Six villages on slope undisturbed2 Little Smoky Thomson and Hayley (1975) 0.25 m/year Bridge protected by slip joint2 Klosters Haefeli (l965) 0.02 m/year Tunnel maintained, bridge

protected by slip joint2 Ft. Peck Spillway Wilson (l970) 0.02 m/year Movements unacceptable,

slope flattened

52

FIGURE 3-18Mudslide fabric andassociated variabilityin water content(modified fromHutchinson 1988,Figure 9).

,

etJ1


Higher rates of creep movement are uncommon."Terzaghi's rate is about 10-6 mm/sec. The limit ofperceptible movements on the scale given inFigure 3-17 and Table 3-5 is conservatively lowerthan Terzaghi's. Still lower rates of movement canbe detected with appropriate instrumentation(Kostak and Cruden 1990).

Varnes (1978, 17) pointed out that "creep hascome to mean different things to different persons,and it seems best to avoid the term or to use it ina well-defined manner. As used here, creep is con-sidered to have a meaning similar to that used inthe mechanics of materials; that is, creep is simplydeformation that continues under constant stress."The term creep should be replaced by the appro-priate descriptors from Figure 3-17, either veryslow or extremely slow, to describe the rate ofmovement of landslides.

Estimates of landslide velocities can be made byrepeated surveys of the positions of displacedobjects (Thomson and Hayley 1975; Huder 1976),by reconstruction of the trajectories of portions ofthe displaced mass (Heim 1932; McConnell andBrock 1904; Ter-Stepanian 1980), by eyewitnessobservations (Tavenas et al. 1971), by instrumen-tation (Wilson 1970; Wilson and Mikkelsen 1978),and by other means. The Colorado Department ofTransportation experimented with the use of time-lapse photography to document the movement ofarelatively slow-moving but very large landslide.The estimates reported were usually the peak veloc-ities of substantial portions of the displaced masses;these estimates are suitable for damage assessments.Rates of movement will differ within the displacedmass of the landslide with position, time, and theperiod over which the velocity is estimated. Such

~ Location of Typical water-contentwater-content values for London Claysample

Site A Site B(Beltinge) (Sheppey)

OverallSample 43% 48%

Lump 41% 34%

GeneralMatrix 46% 52%

True Matrix 1(>46%) 1(>52%)

differences argue against very precise reports oflandslide velocity in reconnaissance surveys.

6. WATER CONTENT

Varnes (1978) suggested the following modifica-tions to terms first proposed by Radbruch-Hall(1978) to describe the water content of landslidematerials by simple observations of the displacedmaterial:

1. Dry: no moisture visible;2. Moist: contains some water but no free water;

the material may behave as a plastic solid butdoes not flow;

3. Wet: contains enough water to behave in partas a liquid, has water flowing from it, or supportssignificant bodies of standing water; and

4. Very wet: contains enough water to flow as aliquid under low gradients.

These terms may also provide guidance in esti-mating the water content of the displaced materi-als while they were moving. However, soil or rockmasses may drain quickly during and after dis-placement, so this guidance may be qualitativerather than quantitative. Individual rock or soilmasses may have water contents that differ con-siderably from the average water content of thedisplacing material. For example, Hutchinson(1988) noted that debris slides (which Hutchinsontermed mudslides) generally were composed ofmaterials that exhibited a fabric or texture con-sisting of lumps of various sizes in a softened claymatrix. Samples taken from different portions ofthis fabric had considerably different water con-tents, with lumps having much lower water con-tents than that of the matrix (Figure 3-18).

7. MATERIAL

According to Shroder (1971) and Varnes (1978),the material contained in a landslide may bedescribed as either rock, a hard or firm mass thatwas intact and in its natural place before the initi-ation of movement, or soil, an aggregate of solidparticles, generally of minerals and rocks, thateither was transported or was formed by the weath-ering of rock in place. Gases or liquids filling thepores of the soil form part of the soil.

Soil is divided into earth and debris (see Table3-1). Earth describes material in which 80 percent


o

or more of the particles are smaller than 2 mm, theupper limit of sand-size particles recognized bymost geologists (Bates and Jackson 1987). Debriscontains a significant proportion of coarse mater-ial; 20 to 80 percent of the particles are larger than2 mm, and the remainder are less than 2 mm. Thisdivision of soils is crude, but it allows the materialto be named by a swift and even remote visualinspection.

The terms used should describe the displacedmaterial in the landslide before it was displaced.The term rock fall, for instance, implies that thedisplacing mass was a rock mass at the initiation ofthe landslide. The displaced mass may be debrisafter the landslide. If the landslide is complex andthe type of movement changes as it progresses, thematerial should be described at the beginning ofeach successive movement. For instance, a rockfall that was followed by the flow of the displacedmaterial can be described as a rock fall-debris flow.

8. TYPE OF MOVEMENT

(a)

(b)

(c) ~lo 3m-=-

8.1.1 Modes of Falling

Observations show that the forward motion ofmasses of soil or rock is often sufficient for free fallif the slopes below the masses exceed 76 degrees

FIGURE 3-19Types of landslides:(a) fall, (b) topple,(c) slide, (d) spread,(e) flow. Brokenlines indicate originalground surfaces;arrows show portionsof trajectories ofindividual particlesof displaced mass[modified fromVarnes 1978, Figure2.1 (Zaruba andMencl 1969)].

5(X)m

(d)

(e)

extremely rapid. Except when the displaced masshas been undercut, falling will be preceded bysmall sliding or toppling movements that separatethe displacing material from the undisturbed mass.Undercutting typically occurs in cohesive soils orrocks at the toe of a cliff undergoing wave attackor in eroding riverbanks.

8.1 Fall

A fall starts with the detachment of soil or rockfrom a steep slope along a surface on which littleor no shear displacement takes place. The mater-ial then descends mainly through the air by falling,bouncing, or rolling. Movement is very rapid to

The kinematics of a landslide-how movement isdistributed through the displaced mass-is one ofthe principal criteria for classifying landslides.However, of equally great importance is its use as amajor criterion for defining the appropriate responseto a landslide. For instance, occasional falls from arock cut adjacent to a highway may be contained bya rock fence or similar barrier; in contrast, topplingfrom the face of the excavation may indicate ad-versely oriented discontinuities in the rock massthat require anchoring or bolting for stabilization.

In this section the five kinematically distincttypes of landslide movement are described in thesequence fall, topple, slide, spread, and flow(Figure 3-19). Each type oflandslide has a numberofcommon modes that are frequently encounteredin practice and that are described briefly, oftenwith examples of some complex landslides whosefirst or initial movements were of that particulartype. These descriptions show how landslides ofthat type may evolve.


(0.25:1). The falling mass usually strikes a slopeinclined at less than this angle (Ritchie 1963),which causes bouncing. Rebound from the impactwill depend on material properties, particularlyrestitution coefficients, and the angle between theslope and the trajectory of the falling mass (Hungrand Evans 1988). The falling mass may also breakup on impact.

On long slopes with angles at or below 45degrees (1:1), particles will have movement pathsdominated by rolling. There is a gradual transitionto rolling from bouncing as bounces shorten andincidence angles decrease. Local steepening of theslope may again project rolling particles into theair, restarting the sequence of free fall, bouncing,and rolling (Hungr and Evans 1988).

8.1.2 Complex Falls

Sturzstroms are extremely rapid flows of dry debriscreated by large falls and slides (Hsu 1975). Theseflows may reach velocities over 50 m/sec. Sturzs-troms have also been called rock-fall avalanches(Varnes 1958) and rock avalanches (Evans et al.1989; Nicoletti and Sorriso-Valvo 1991). Twoexamples of historic sturzstroms in Switzerland,the Rossberg landslide of 1806 and the Elm land-slide of 1881, are discussed in Chapter 1. Hsu(1975) suggested that 5 million m3 is the lowerlimit of the volume of significant sturzstroms, butHutchinson (1988) demonstrated that falls inhigh-porosity, weak European chalk rocks withvolumes two orders of magnitude smaller have thesame exceptional mobility because of collapse ofthe pores on impact and consequent high pore-water pressures. Some of Hutchinson's data arereproduced in Figure 3-19.

The motion ofsturzstroms probably depends onturbulent grain flow with dispersive stresses arisingfrom momentum transfer between colliding grains.Such a mechanism does not require the presenceof a liquid or gaseous pore fluid and can thereforeexplain lunar and Martian sturzstroms. Van Gassenand Cruden (1989) showed that the motion of thecomplex, extremely rapid, dry rock fall-debns flowthat occurred near the town of Frank, Alberta,Canada, in 1903 (see Figure 3-1) could be explainedreasonably well by momentum exchange betweenthe moving particles and measured coefficients offriction.

8.2 Topple

A topple [Figure 3-19(b)] is the forward-rotationout of the slope of a mass of soil or rock about apoint or axis below the center ofgravity of the dis-placed mass. Toppling is sometimes driven by grav-ity exerted by material upslope of the displacedmass and sometimes by water or ice in cracks in themass. Topples may lead to falls or slides of the dis-placed mass, depending on the geometry of themoving mass, the geometry of the surface of sepa-ration, and the orientation and extent of the kine-matically active discontinuities. Topples rangefrom extremely slow to extremely rapid, sometimesaccelerating throughout the movement.

8.2.1 Modes of TopplingFlexural toppling was described by Goodman andBray as

occurring in rocks with one preferred disconti-nuity system, oriented to present a rock slopewith semi-continuous cantilever beams....Continuous columns break in flexure as theybend forward.... Sliding, undermining or ero-sion of the toe (of the displaced mass) lets thefailure begin and it retrogresses backwards withdeep, wide tension cracks. The lower portionof the slope is covered with disoriented and dis-ordered blocks.... The outward movement ofeach cantilever produces interlayer sliding(flexural slip) and ... back-facin{scarps (obse-quent scarps).... It is hard to say where thebase of the disturbance lies for the change isgradual. . . . Flexural toppling occurs mostnotably in slates, phyllites and schists. (Good-man and Bray 1976, 203)

A flexural topple in the proposed classification is aretrogressive, complex rock topple-rock fall. Typicalexamples are shown in Figures 3-20(a) and 15-16.

In contrast, block toppling occurs

where the individual columns are divided bywidely-spaced joints. The toe of the slope withshort columns, receives load from overturning,longer columns above. This thrusts the toecolumns forward, permitting further toppling.The base of the disturbed mass is better definedthan in the case of flexural toppling; it consistsof a stairway generally rising from one layer to

I!

I


result of deformation of the massif. (Giraudet al. 1990, 250)

Goodman and Bray (1976) identified four com-posite landslide modes in which toppling wascaused by earlier sliding. These landslide condi-

(c)

(a)

FIGURE 3-20Three modes of toppling. (a) Flexural: cracks indicate tension-crack topple;fallen blocks below topple show movement is complex rock topple-rockfall. (b) Chevron: multiple block topple; hinge surface of chevron maydevelop into surface of rupture of slide forming complex multiple rocktopple-rock slide. (c) Block-flexure.

the next. The steps of this stairway are formedby cross-joints.... New rock breakage ...occurs much less markedly than in flexuraltopples .... Thick-bedded sedimentary rockssuch as limestones and sandstones, as wellas columnar-jointed volcanics exhibit block-toppling. (Goodman and Bray 1976, 203)

Typical examples of block toppling are shown inFigure 15-16.

Chevron topples are block topples in which thedips of the toppled beds are constant and thechange ofdip is concentrated at the surface of rup-ture (Cruden et al. 1993). This mode was namedfrom its resemblance to chevron folds (Ramsay1967, 436). Chevron topples occur on steeperslopes than other block topples. The surface ofrupture is often a sliding surface [Figure 3-20(b)]forming a complex rock topple-rock slide.

Block-flexure toppling is characterized by"

Several distinct movements may be identifiedas the phenomenon progresses, along with amodification of water flows within the rockmass, due to considerable changes in perme-ability over time and probably in space as a

pseudo-continuous flexure of long columnsthrough accumulated motions along numerouscross joints. Sliding is distributed along severaljoint surfaces in the toe [of the displaced mass]while sliding and overturning occur in closeassociation through the rest of the mass.(Goodman and Bray 1976, 204)

8.2.2 Complex and Composite T()pplesA complex rock topple-rock slide is shown in Figure3-13(1). This cross section of the La Clapiere land-slide was described by Giraud et al. as follows:

Sliding occurs because accumulated overturningsteepens the cross joints. There are fewer edge-to-face contacts than in block toppling but stillenough to form "a loosened, highly open ... dis-turbed zone [displaced mass].... Interbedded sand-stone and shale, interbedded chert and shale andthin-bedded limestone exhibit block flexure top-pling" (Goodman and Bray 1976, 204). Typicalblock-flexure topple examples are shown in Figures3-20(c) and 15-16.

56

FIGURE 3-21Debris topple(Varnes .1978,Figure 2.1 e).


tions are discussed in detail in Chapter 15 and areillustrated in Figure 15-17. A slide head toppleoccurs as blocks from the crown of the slide toppleonto the head of the displaced mass. According tothe naming convention, such a landslide is a com-posite rock slide-retrogressive rock topple.

Rotational sliding of earth or debris above asteeply dipping sedimentary rock mass can causeslide base toppling as the sliding induces shear forcesin the top of the rock mass (Goodman and Bray1976). The resulting landslide is, according tothe proposed naming convention, a composite earthslide-advancing rock topple.

Toppling below the toe of the surface of ruptureof a rock slide may be caused by load transmittedfrom the slide. Such a failure is call~d a slide toe top-ple (Goodman and Bray 1976). According to theproposed naming convention, it is a compositerock slide-rock topple.

The formation ofextension cracks in the crownof a landslide may create blocks capable of top-pling, or a tension crack topple (Goodman and Bray1976). According to the proposed naming con-vention, this is a retrogressive multiple topple, per-haps forming part of a compositefall or slide. Suchfailures may also occur in cohesive soils beingundercut along streambanks (Figure 3-21).

8.3 Slide

A slide is a downslope movement of a soil or rockmass occurring dominantly on surfaces of ruptureor on relatively thin zones of intense shear strain.Movement does not initially occur simultaneouslyover the whole of what eventually becomes thesurface of rupture; the volume of displacing mate-rial enlarges from an area oflocal failure. Often thefirst signs of ground movement are cracks in theoriginal ground surface along which the main scarpof the slide will form. The displaced mass may slide

beyond the toe of the surface of rupture coveringthe original ground surface of the slope, whichthen becomes a surface of separation.

8.3.1 Modes ofSliding

Varnes (1978) emphasized the distinction betweenrotational and translational slides as significant forstability analyses and control methods. That dis-tinction is retained in this discussion. Figure 3-22shows two rotational slides and three translationalslides. Translational slides frequently grade intoflows or spreads.

Rotational slides (Figure 3-23) move along a sur-face of rupture that is curved and concave. If thesurface of rupture is circular or cyc10idal in profile,kinematics dictates that the displaced mass maymove along the surface with little internal defor-.mation. The head of the displaced material maymove almost vertically downward, whereas theupper surface of the displaced material tilts back-ward toward the scarp. If the slide extends for aconsiderable distance along the slope perpendicu-lar to the direction of motion, the surface of rup-ture may be roughly cylindrical. The axis of thecylindrical surface is parallel to the axis aboutwhich the slide rotates. Rotational slides in soilsgenerally exhibit a ratio of depth of the surface ofrupture to length of the surface of rupture, .q. /L

r

(see Table 3-4 and Figure 3-5 for definitionsof these dimensions), between 0.15 ar;,d 033(Skempton and Hutchinson 1969).

Because rotational slides occur most frequentlyin homogeneous materials, their incidence in fillshas been higher than that of other types of move-ment. Natural materials are seldom uniform, how-ever, and slope movements in these materialscommonly follow inhomogeneities and disconti-nuities (Figure 3-24). Cuts may cause movements

. that cannot be analyzed by methods used for rota-tional slides, and other more appropriate methodshave been developed. Engineers have concen-trated their studies on rotational slides.

The scarp below the crown of a rotational slidemay be almost vertical and unsupported. Furthermovements may cause retrogression of the slideinto the crown. Occasionally, the lateral marginsof the surface of rupture may be suffiCiently highand steep to cause the flanks to move down andinto the depletion zone of the slide. Water findingits way into the head of a rotational slide may con-

I!III

I


(b)

(a)FIGURE 3-22Examples ofrotational andtranslational slides:(a) rotational rockslide; (b) rotationalearth slide;(c) translational rockslide (upper portionis rock block slide);(d) debris slide;(e) translational earthblock slide[Varnes1978, Figures 2.1 g,2.1i, 2.1j2, 2.1k, 2.11(Hansen 1965)].

(d)

(c)

tribute to a sag pond in the backward-tilted, dis-placed mass. This disruption of drainage may keepthe displaced material wet and perpetuate theslope movements until a slope of sufficiently lowgradient is formed.

In translational slides (Figures 3-22, 3-25, and3-26) the mass displaces along a planar or undu-lating surface of rupture, sliding out over the orig-inal ground surface. Translational slides generallyare relatively shallower than rotational slides.Therefore, ratios ofDJLr for translational slides insoils are typically less than 0.1 (Skempton andHutchinson 1969). The surfaces of rupture oftranslational slides are often broadly channel-shaped in cross section (Hutchinson 1988).Whereas the rotation of a rotational slide tends torestore the displaced mass to equilibrium, transla-tion may continue unchecked if the surface ofseparation is sufficiently inclined.

As translational sliding continues, the displacedmass may break up, particularly if its velocity or

FIGURE 3-23Cut through rotational slide of fine-grained, thin-bedded lake deposits,Columbia River valley; beds above surface of rupture have been rotated byslide to dip into slope (Varnes 1978, Figure 2.7).

58

FIGURE 3-24Rotational slides(Varnes 1978,Figure 2.5).

FIGURE 3-25(below)Translational slideof colluviumon inclinedmetasiltstonestrata along 1-40,Cocke County,Tennessee (Varnes1978, Figure 2.11).


1.1 SLOPE FAILURE IN HOMOGENEOUS MATERIAL.CI RCULAR ARC. (1) SLIDE WHOLLY ON SLOPE AND121 SURFACE OF RUPTURE INTERSECTS TOE OFSLOPE.

(ul FAILURE OF EMBANKMENT. GRAVEL COUNTERWEIGHT ON LEFT SIDE PREVENTS SLIDE.

(bl SLOPE FAILURE IN NONHOMOGENEOUS MATERIAL.SURFACE OF RUPTURE FOLLOWS DIPPING WEAKBED.

water content increases. The disrupted mass maythen flow, becoming a debris flow rather than aslide. Translational sliding often follows disconti-nuities such as faults, joints, or bedding surfaces orthe contact between rock and residual or trans-ported soils.

Translational slides on single discontinuities inrock masses have been called block slides (Panet1969) or planar slides (Hoek and Bray 1981). AsHutchinson (1988) pointed out, there is a transi-tion from rock slides of moderate displacement thatremain as blocks on the surface of rupture to slideson steeper and longer surfaces that break up intodebris or transform into sturzstroms. Where the sur-

face of rupture follows a discontinuity that is paral-lel to the slope, the toe of the displaced mass mayform a wedge that overrides or ploughs into undis-placed material causing folding beyond the toe ofthe surface of rupture (Walton and Atkinson1978). Composite rock slide-rock topples [Figure313(2)] or buckles and confined slides may result.

The surface of rupture may be formed by twodiscontinuities that cause the contained rock massto displace down the line of intersection of the dis-continuities, forming a wedge slide [Figure 3-27(a)and Chapter 15, Figures 15-8 through 1514].Similarly shaped displaced masses may be boundedby one discontinuity that forms the main scarp ofthe slide and another that forms the surface of rup-ture. The mode of movement depends on the ori-entations of the free surfaces relative to thediscontinuities in the rock masses (Hocking 1976;Cruden 1978, 1984). Stepped rupture surfaces mayresult if two or more sets of discontinuities, such asbedding surfaces and some joint sets, penetrate therock masses. As shown in Figure 3-27(b), one setof surfaces forms the risers of the steps and theother forms the treads, creating a stepped slide(Kovari and Fritz 1984).

Compound slides are intermediate between rota-tional and translational slides and their D, IL, ratiosreflect this position (Skempton and Hutchinson1969). Surfaces of ruptUre have steep main scarpsthat may flatten with depth [Figure 3-22(e)]. Thetoes of the surfaces of rupture may dip upslope.Displacement along complexly curved surfaces ofrupture usually requires internal deformation andshear along surfaces within the displaced materialand results in the formation of intermediate scarps.Abrupt decreases in downslope dips of surfaces ofrupture may be marked by uphill.facing scarps indisplaced masses and the subsidence of blocks ofdisplaced material to form depressed areas, grabens[Figure 322(e)]. A compound slide often indicatesthe presence of a weak layer or the boundary be-tween weathered and unweathered material. Suchzones control the location of the surface of rupture(Hutchinson 1988). In single compound slides, thewidth of the graben may be proportional to thedepth to the surface ofrupture (Crudenet al. 1991).

8.3.2 Complex and Composite Slides

Complex and composite slide movements arecommon and the'litenlture contains numerous ref-


(a)

(b)

erences to a variety of specialized names. Two ofthese names, mudslide andfIowslide, are believed tobe imprecise and ambiguous, and therefore theiruse is not recommended.

According to Hutchinson, mudslides are

slow-moving, commonly lobate or elongatemasses of accumulated debris in a softenedclayey matrix which advance chiefly by slidingon discrete, bounding shear surfaces. '.' . In

59

FIGURE 3-26(above)Shallow translationalslide that developedon shaly rock slopein Puente Hills ofsouthern California;slide has low 0, IL,ratio; note wrinkles0.0 surface [Varnes1978, Figure 2.33;(Shelton 1966)].COPYRIGHTJOHN S.SHELTON

FIGURE 3-27(left)(a) Wedge and(b) steppedtranslationalslides.


long profile, mudslides are generally bilinearwith a steeper ... slope down which debris isfed (by falls, shallow slides and mudslides) to amore gently-inclined slope. . .. Mudslides areespecially well-developed on slopes containingstiff, fissured clays, doubtless because of the easewith which such materials break down to pro-vide a good debris supply. (Hutchinson 1988,12-13)

Evidently these movements begin in eitherweak rock or earth and retrogress by falls and slideswhile advancing by sliding. Hutchinson and Bhan-dari (1971) suggested that the displaced material,fed from above onto the mudslide, acted as an un-drained load. Clearly mudslides are composite orperhaps complex both in style ofactivity and in thebreakdown of displaced material into earth ordebris. The displaced material is generally moist,though locally it may be wet. A mudslide is there-fore often a retrogressive, composite rock slide-advancing, slow, moist earth slide. Other mudslidemodes may include single earth slides (Brunsden1984). The current use of mudslide for several dif-ferent landslide modes suggests that more preciseterminology should be used where possible.

.The term /lowslide has been used to describe sud-den collapses of material that then move consid-erable distances very rapidly to extremely rapidly.As Hutchinson (1988) and Eckersley (1990)pointed out, at least three phenomena can causethis behavior: (a) impact collapse, (b) dynamicliquefaction, and (c) static liquefaction. Impact-collapse flowslides occur when highly porous,often-saturated weak rocks fall from cliffs, resultingin destruction of the cohesion of the material andthe generation of excess fluid pressures within the'flowing displaced mass. Clearly these are complexfalls in which the second mode of movementis adebris flow; if the displaced material is dry debriS,the movement may be a' sturzstrom, previouslydefined in Section 8.1.2. This mode of movementcan be recognized by both the materials involvedand the topography of the flow.

If the structure of the material is destroyed byshocks such as earthquakes, saturated material mayliquefy and then flow, sometimes carrying masses ofoverlying drier material with it. Such a movementis a flow or liquefaction spread, which is furtherdefined in Section 8.4.1. Such movements arecharacteristic of loess, a lightly cemented aeolian

silt. Landslides occurring in loosely dumped an-thropogenic materials, both stockpiles and wastedumps, have also been termed flowslides. Theseloose, cohesionless materials contract on shearingand so may generate high pore-water pressures aftersome sliding (Eckersley 1990). Similar landslidesmay also occur in rapidly deposited natural silts andfine sands (Hutchinson 1988, 14). Since thesemovements involve both sliding and then flowage,they may be better described as complex slide flows.

Because these separate and distinguishable phe-nomena are comparatively distinct types of land-slides that may be more accurately described bystandard descriptors, the use of the term f/.owslidefor all these types ofmovements is redundant, con-fusing, and potentially ambiguous.

In contrast, one form of compound sliding fail-ure appears to warrant a special term. Sags (or sack-ungen) are deformations of the crests and steepslopes of mountain ridges that form scarps andgrabens and result in some ridges with doublecrests and small summit lakes. Material can be dis-placed tens of meters at individual scarps. Thestate ofactivity, however, is generally dormant andmay be relict. The term sag may be useful to indi-cate uncertainty about the type of movement vis-ible on a mountain ridge.

Movement is often confined, and small bulgesin local slopes are the only evidence of the toes ofthe displaced material. Detailed subsurface inves-tigations of these features are rare, and classifica-tion should awa'it this more detailed exploration.As Hutchinson (1988, 8) demonstrated, the geom-etry of the scarps (which often face uphill) may beused to suggest types of movement, which mayinclude slides, spreads, and topples. The modes ofsagging depend on the lithology of the displacedmaterial and the orientation and strength of thediscontinuities in the displacing rock mass. Varneset al. (1989, 1) distinguished

1. "Massive, strong (although jointed) rocks lyingon weak rocks,"

2. "Ridges composed generally of metamorphosedrocks with pronounced foliation, schistosity orcleavage," and

3. "Ridges composed of hard, but fractured, crys-talline igneous rocks."

Sags of the first type are usually spreads(Radbruch-Hall1978; Radbruch-Hall et al. 1976),

which are discussed in Section 8.4. Sags in foliatedmetamorphic rocks are often topples, and thus arediscussed in Section 8.2, although bedrock flowmay also occur (see Section 8.5). Sags in crys-talline igneous rocks (Varnes et al. 1989,22) weremodeled by a plasticity solution for "gravity-induced deformation of a slope yielding under theCoulomb criterion." Sags may thus be slides,spreads, or flows, depending on the extent anddistribution of plastic flow within the deformingrock mass.

Sags are often associated with glacial features.The absence of Pleistocene snow and ice covers,and the resulting loss of the permafrost and highpore-water pressures these induced, may accountfor the present inactivity of many sags. Varnes etal. (1990) described an active sag in the mountainsof Colorado. Earthquakes, however, can also pro-duce uphill-facing scarps along reactivated normalfaults. The surface features of sags require carefulinvestigation before any conclusions can be drawnabout the cause .and timing of slope movement.Such investigations may be sufficient to allow theidentification of sags as other types of landslides.

In many landslides, the displaced material, ini-tially broken by slide movements, subsequentlybegins to flow (Figure 3-28). This behavior is espe-cially common when fine-grained or weak materi-als are involved. These landslides have beentermed slump-earth flows. Slump has been used as asynonym for a rotational slide, but it is also used todescribe any movement in a fill. It is therefore rec-ommended that this mode ofmovement be termeda complex earth slide-earth flow and that the use ofthe term slump be discontinued.

In permafrost regions, distinctive retrogressive,complex earth slide-earth flows, known as thaw-slumps (Hutchinson 1988, 21) and bimodal flows(McRoberts and Morgenstern 1974), develop onsteep earth slopes when icy permafrost thaws andforms flows of very wet mud from a steep mainscarp. These special landslide conditions are dis-cussed in more detail in Chapter 25.

8.4 Spread

The term spread was introduced to geotechnicalengineering by Terzaghi and Peck (1948) todescribe sudden movements on water-bearingseams of sand or silt overlain by homogeneousclays or loaded by fills:


ZONE A~chieflyby""'"-=ate slumping .'ont tnp....-

:;i~i~;~ump unitsB,S'Narrow slump unit withaxes perpendicular toaxes of main slump unitsand parallel with lengthof main slideC"'$1800" remaining afterdownward mcwement ofunit D 1rom IIfea E.

ZONE 8Zone of earth flow;movement chiefly byflowage.

ZONECToe of slide area; originalform aitered bV railroadr8l;onstruetion work.

100m

328ft

It is characteristic ... that a gentle clay slopewhich may have been stable for decades or cen-turies, moves out suddenly along a broad front.At the same time the terrain in front ... heavesfor a considerable distance from the toe. Oninvestigation, it has invariably been found thatthe spreading occurred at a considerable dis-tance beneath the toe along the boundarybetween the clay and an underlying water-bearing stratum or seam of sand or silt. (Ter-zaghi and Peck 1948, 366)

Recognition of the phenomenon is considerablyolder. One of the three types of landslides distin-guished by Dana (1877, 74) occurs "when a layerof clay or wet sand becomes wet and softened bypercolating water and then is pressed out laterallyby the weight of the superincumbent layers." Anearly use of spread to describe this phenomenon isby Barlow:

In a landslip [British term for some types oflandslide], the spreading ofsome underlying bedwhich has become plastic through the percola-tion ofwater or for some other cause drags apartthe more solid, intractable beds above and pro-

61

FIGURE 3-28Plan of Ames slidenear Telluride,'Colorado. Thisenlarging complexearth slide-earthflow occurred in tilloverlying Mancosshale. Crown ofslide retrogressed bymultiple rotationalslides after mainbody of displacedmaterial moved.Surface of rupturealso widened onleft lateral margin[Varnes 1978, Figure2.10 (modified fromVarnes 1949)].

,500m -.l

, .- - - ~ .".

may also fill with broken, displaced material[Figure 3-29(a)]. Typical rates of movement areextremely slow.

Such movements may extend many kilometersback from the edges of plateaus and escarpments.The Needles District of Canyonlands NationalPark, Utah, is an example ofa block spread (McGilland Stromquist 1979; Baars 1989). Grabens up to600 m wide and 100 m deep stretch 20 km alongthe east side of Cataract Canyon'on the ColoradoRiver (Figure 3-30). The grabens extend up to 11km back from the river. A 450-m-thick sequenceof Paleozoic sedimentary rocks has been spreaddown a regional slope with a 4-degree dip by theflow of an underlying evaporite that is exposed invalley anticlines in the Colorado River and its trib-utaries (Potter and McGill 1978; Baars 1989). Thisapproximately 60 km3 of displaced material con~stitutes one of North America's largest landslides.

Liquefaction spreads form in sensitive clays andsilts that have lost strength with disturbances thatdamaged their structure [Figpres 3-29(c) and3-31]. These types of landslides are discussed fur-ther in Chapter 24. Movement is translational andoften retrogressive, starting at a stream bank or ashoreline and extending away from it. However, ifthe underlying flowing layer is thick, blocks may

(b)

(a)


8.4.1 Modes of Spreading

duces fissures and fractures transverse to thedirection of movement. (Barlow 1888, 786)'Spread is defined here as an extension ofa cohe-

sive soil or rock mass combined with a general sub-sidence of the fractured mass of cohesive materialinto softer underlying material. The surface of rup-ture is not a surface of intense shear. Spreads mayresult fr~m liquefaction or flow (and extrusion) ofthe softer material. Varnes (1978) distinguishedspreads typical of rock, which extended withoutforming an identifiable surface of rupture frommovements in cohesive soils overlying liquefiedmaterials or materials flowing plastically (Figure3-29). The cohesive materials may also subside,translate, rotate, disintegrate, or liquefy and flow.Clearly these movements are complex, but they aresufficiently common in certain materials and geo-logical situations that the concept of a spread isworth recognizing as a separate type of movement.

. In block spreads, a thick layer of rock overlies softermaterials; the strong upper layer may fracture andseparate into strips. The soft underlying material issqueezed into the cracks between the strips, which

62

FIGURE 3-29Rock and earthspreads: (a), (b) rockspreads that haveexperienced lateralextension withoutwell-defined basalshear surface orzone of plastic flow[Varnes 1978, Figure2.1 m2 (Zaruba and

, Mend 1969); Figure2.1 m3 (Ostaficzuk1973)]; (c) earthspread resultingfrom liquefactionor plastic flow ofsubjacent material(Varnes 1978,Figure 2.10).

FIGURE 3-30Geological cross section showing formation of The Grabens in NeedlesDistrict of Canyonlands National Park, Utah. Colorado River has carvedCataract Canyon to within a few meters of top of Paradox Salt beds (IPps),undercutting inclined layers of overlying Honaker Trail (IPht), ElephantCanyon (Pee), and Cedar Mesa (Pern) formations. These formations havebroken up and are moving toward canyon by flowage within salt. ColoradoRiver follows crest of Meander anticline, a salt-intruded fold located abovea deep-seated fault zone (Baars 1989).

sink into it, forming grabens, and upward flow cantake place at the toe of the displaced mass. Move-ment can begin suddenly and reach very rapidvelocities:

Spreads are the most common ground failureduring earthquakes.... [They] occur in gentlerterrains (commonly between 0.5 per cent [0.3degrees] and 5 percent [3 degrees)) with lateralmovement of a few meters or so.... [S]preadsinvolve fracturing and extension of coherentmaterial owing to liquefaction or plastic flow ofsubjacent material. ... [S]preads are primarilytranslational although some associated rota-tion and subsidence commonly occurs. (Andrusand Youd 1987, 16)

Van Hom (1975) described two movements ofmore th~ 8 km2 in area on very gentle slopes inflat-lying beds ofSilty clay, clayey silt, and very finesand deposited in prehistoric Lake Bonneville,Utah: The displaced material forms ridges parallelto the main scarp of the landslide; distinctiveinternal structures include gentle folds, shears, andintrusions of liquefied sediment.

8.4.2 Complex Spreads

w


The Grabens

CataractCanyon

ENeedles

63

Major deformations in rock strata were foundalong many valleys in north-central England dur-ing the construction of dams in the late 19th cen-tury. These deformations occurred where a nearlyhorizontal, rigid-jointed cap rock overlaid a thicklayer of stiff-fissured clay or clay shale that in tumoverlaid a more competent stratum. A bending, orcambering, of the rigid strata caused blocks of thiss

Documents

Cruden_and_Varnes_1996.pdf