Proc. IVth International Conference andField Workshop on Landslides, 1985, Tokyo 71)

Geologic Zoning of Slope Movements in WesternOregon, U. S. A.Swanson, F. J.Lienkaemper, G. W.U. S. Department of Agriculture, ForestService, Oregon, U. S. A.

SYNOPSIS

The Japanese approach to geologic zoning of slope movement types has useful application in thegeologically diverse landscape of western Oregon, U.S.A. The slope movement regime of an area canbe characterized quantitatively in terms of total area affected, size distribution, and velocityand frequency of movement for slope movements of different types. Slope movements in Tertiarysedimentary rocks of the Oregon Coast Range include a high frequency (number per square kilometerper year) and low average volume of debris avalanches, high frequency of debris torrents downfirst- to fourth-order channels, and large, deep-seated slope movement features that may experiencesome of their movement in brief pulses lasting hours to days. Tertiary volcanic rocks in theCascade Range have a lower frequency of debris avalanches and torrents and the large, deep-seatedslope movements are mainly of the earthflow type that move slowly for prolonged periods (months)during the wet season. Further quantification of slope movement regimes will aid in planning landuse and erosion control programs and in studying sediment production and landscape evolution.

INTRODUCTION

Japanese researchers have classified slopemovements in Japan in a series of geologicdivisions that have characteristic'types ofslope movement features (Yamaguchi, 1980).This geological approach to zoning slopemovements and hazards is useful for groupingareas by common characteristics:(1) engineering properties of materials,including primary and secondary features ofmineralogy and fabric; (2) tectonics as theyaffect the history of uplift, potential forearthquakes, and disruption of rock materials;and (3) climate. These geologic and climaticcharacteristics of an area result intopography, and thickness and mechanicalproperties of the soil mantle distinctive toparticular rock units. These relationshipshave been recognized and incorporated into asystem for zoning prediction and mitigation ofslope movement disasters in Japan, compiled byDr. N. Oyagi (National Research Center forDisaster Prevention, Tsukuba, Japan).

Geologic zoning of slope movement regimeswould be useful in the United States also. Inwestern Oregon, for example, better zoning ofslope movement conditions could be used forevaluating risk and liability and forpreparing laws and rules to govern land usepractices, mainly logging and roadconstruction. The State of Oregon is nowconsidering such laws and rules but withoutbenefit of a systematic description of slopemovement conditions in the State.Furthermore, description and analysis ofbroad-scale slope movement conditions in the

area would be useful in analyzing sedimentproduction and the history of landscapeevolution. We are also interested in theamounts and geographic distribution of areasof the landscape experiencing various degreesand frequencies of disturbance by slopemovement. This information would be useful ininterpreting broad-scale patterns withinforest and stream ecosystems.

Gaining a general undersanding of slopemovement conditions in a region or aparticular geologic unit is difficult, becausein many cases research on slope movement hasbeen concentrated on small areas with specialproblems of slope instability. To adequatelyanalyze slope movement conditions, we need arigorous, quantitative description of theslope movement regime for an area, based on acarefully designed sampling program.Characterization of the slope movement regimeof an area should include: (1) rates andtiming of movement by active, slow movingfeatures; (2) recency of movement by inactive,large features; (3) frequency of occurrence ofrapid slope movements; and (4) total areaaffected and size distribution (volume andarea) of each type of slope movement. Thesecharacteristics can be defined in generalterms, such as "active" or "inactive" forlarge features. More rigorous description isalso possible, such as examining debrisavalanche history in relation to landmanagement history or the detailed mappingunits of Wieczorek (1982) showing dates oflast known movement for earthflow features.Because slope movement characteristicstypically vary significantly from one geologic

COAST RANGE

North ForkSiuslaw River

Mapleton

Smith River

Middle Santiam

Blue River

Lookout Creek

Alder Creek

WESTERN CASCADES

HIGH CASCADES

KLAMATHMOUNTAINS

LOCATION MAP

WILLAMETTE VALLEY

Figure 1. Map of physiographic units, slopemovement study areas, areas surveyed for lakesdammed by slope movements (outlined withdashed line), and those lakes (1. Trembly,2. Drift I, 3. Drift II, 4. Triangle,5. Wassen, 6. Loon, 7. Sitkum, 8. Moose,9. Windfall, 10. Fish).

terrane to another, geologic units can formnatural strata for sampling.

Thorough descriptions of the slope movementregime for areas in the United States arerare. We can, however, begin to compile suchdescriptions by using existing inventories ofdebris avalanches and by interpreting maps oflarge-scale slope movement features. In thispaper we characterize and contrast the slopemovement regimes of two areas in westernOregon from available data. We give examplesof the strong control of geology on slopemovement processes and of the approaches todefining and interpreting slope movementregimes. A sampling program designed toprovide a general description of the slopemovement conditions in a geologic unit was notused in collection of these data, so thedistinctions between areas can be consideredonly qualitative. Finally, we considerapproaches to compiling a quantitativedescription of the slope movement regime of anarea.

STUDY AREAS

Western Oregon can be broadly divided intofour physiographic provinces: Coast Range,Cascade Range, Klamath Mountains, andWillamette Valley (Fig. 1). A variety ofbedrock types are present in each of theseareas. In this paper we focus on sedimentary

rocks of the south-central Coast Range andvolcanic rocks of the western Cascades Range.

The central part of the southern half of theCoast Range in western Oregon is underlain byTertiary turbidite sandstone beds up to 4 mthick with thin interbeds of mudstone andsiltstone. These rocks are broadly folded andgenerally dip less than 15°. The topography -ranges from gentle hills to areas with short,steep (33-36°) slopes mantled with thinsoils. Average annual precipitation rangesfrom 1500 to more than 2000 mm, falling mainlybetween October and April. Forestry centeredon native Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) is the dominant land use.

The western Cascades are underlain by acomplex mixture of Tertiary volcanic rocks,including materials originating as lava flows,lahars, pyroclastic flows, and water-reworkedvolcanic deposits. Alteration of these rockshas been widespread, resulting in formation ofclay minerals and amorphous materials.Topography varies from gentle, hummocky slopesto steep, rock cliffs. Average annualprecipitation is about 1000 mm in the southernhalf of the range and more than 2500 mm inmuch of the northern half, falling as rain andsnow during the fall and winter. Productionof Douglas-fir and other tree species is thedominant land use.

SLOPE MOVEMENT CHARACTERISTICS IN THE WESTERNCASCADE RANGE

Debris Avalanches

Several detailed inventories of debrisavalanches have been conducted by compiling arecord of all debris avalanches in a selectedarea over a known period from aerialphotographs and field surveys (Swanson et al.,1981). These data can then be summarized interms of frequency of debris avalanches(number per square kilometer per year) andsoil transfer rate (cubic meters of soiltransported per square kilometer per year).Three studies that used similar detailed fieldinvestigations report occurrence of debrisavalanches in the central western CascadeRange (Table 1, Fig. 1). The observedfrequencies of debris avalanches, averagevolume, and soil transfer rates for forestedareas fall within moderately restricted ranges(Table 1). Here we report values for forestedareas only because frequency and size ofdebris avalanches vary among forest, clearcut,and roaded areas; and land use practices canvary between areas. Consequently, analysis ofinherent slope stability of an area andcomparison between areas are best based onslope movement under forested conditions. Theminimum size of debris avalanche sampled was75 m3.

Features of Large Slope Movements

Two studies report results of detailedmapping of landforms created by large slopemovements in the central western Cascades(Fig. 1). In the Lookout Creek basin, Swansonand James (1975) identify 3.3% of the basin inactive slumps and earthflows and 9% of thearea in inactive features. Active areas arethose with fresh scarps and tension cracks,

fable 1. Debris avalanche frequency, average volume, and soil transfer rate for forested areas

AverageArea Period

Volume TransferSite

Sampled Sampled

Frequency of Soil Rate

(km 2 ) (yr)

(events km- 2 yr-1)

(m3) (m3 km-2 yr- 1)

Cascade Range

Blue River(Marion, 1981) 48.5 34 .012

3120 37

Lookout Creek(Swanson andDyrness, 1975) 50.1 25

Alder Creek(Morrison, 1975) 12.3 25

.025

.023

1460 36

1990 45

Coast Range

Mapleton(Ketcheson, 1977) 4.4 15

0.58

31 18

Mapleton(Swansonet al., 1981) 5.2 15

0.53

61 32

Smith River(Pierson, 1977) 126 15

0.060

270 16

indicating movement within the past few years,and vegetation disturbed by slope movement atany time within the life of the forest, whichis up to 500 years old in the study area.Inactive areas have broad scale landf9rmsproduced by slope movement, but they lackvegetation and microtopography indicative ofrecent movement. Using similar criteria inmapping a part of the Middle Santiam area(Fig. 1), Hicks (1982) found 15% of of thearea in active slumps and earthflows and 7%inactive. This area is noted for itsabundance of earthflows, parts of which havebeen moving at 10+ m/yr.

The timing and rates of movement of largeslope movement features in the Cascade Rangehave been documented by direct measurements onfour earthflows and informal observations ofmany other features. Active earthflows appearto move every year (our longest period ofdirect measurement is 10 yr), and movement maypersist for 6 to 12 months of the year. Atthe four sites of direct monitoring, movementduring an individual storm each year has beenonly a minor part of the total annualmovement. The maximum measured annual rate ofmovement has been 18 m/yr in the lower part ofthe Jude Creek earthflow in the Middle Santiamarea (Fig. 1). The most common rates ofmovement of active earthflows in the westernCascades appear to be a meter per year or less.

Investigation of aerial photographs since 1946shows that several large debris avalanches,each up to 1 hectare in area, have occurred inthe central western Cascades; however, we knowof no very rapid movement of areas of 5 ha ormore.

Lakes dammed by slope movements are quite rarein the Cascade Range; we have identified onlytwo in a 12000 km 2 area (bounded bylatitudes 43°N and 45°N, and longitudes 122°and 122°45'). Both are small (4-ha MooseLake, 1-ha lake on Windfall Creek) (Fig. 1)and become dry during the dry summer period.Existence of a lake dammed by slope movementindicates slope movement sufficiently fast andvoluminous that fluvial erosion could not keeppace with debris supplied by the slopemovement. In general in the western Cascadesslow encroachment of slopes into channelslocally raises the base level, but formationof a lake is prevented by rapid erosion of thetoe of the slope and deposition of sedimentupstream of channel constrictions.

There has been little dendrochronologicanalysis of large slope movement features inthe Cascades; however, we have examinedtree-ring records on many stumps where treeswere cut and removed from active earthflowareas. We have noted a consistent lack ofdevelopment of discrete episodes of eccentrictree-ring growth indicative of periods oftilting of the trees, as described by Shroder(1978). This may be a result of earthflowmovement persisting year after year, ratherthan occurring only in years of particularlyheavy precipitation.

SLOPE MOVEMENT CHARACTERISTICS IN THE COASTRANGE

Debris Avalanches

Two studies of debris avalanches have beendone in the area of sedimentary rocks in the

Coast Range by use of comparable, detailedfield investigations (Ketcheson, 1977; Swansonet al., 1981). This work centers on steep,highly dissected slopes in the Mapleton areawhich is known for its high level of slopeinstability. This area exhibits a highfrequency of debris avalanches with smallaverage volume of soil transported, based on asample of features larger than 7.5 m3sampled in forested areas (Table 1). Thisminimum size of sampled debris avalanches isonly 10% of the smallest debris avalanchessampled in the Cascade studies. These minimumvalues for sizes of debris avalanches sampledare, however, reasonable for the respectivesites, because debris avalanches between 7.5and 75 m 3 are common in the Coast Rangestudy sites, but rare in the Cascade Rangeareas studied.

Pierson's (1977) inventory of debrisavalanches in the Smith River area in theCoast Range (Fig. 1) was based mainly onanalysis of aerial photographs and did notinvolve the same detailed field samplingmethods used in the studies of Ketcheson(1977) and Swanson et al. (1981). He reportshigher average volume of soil transport bydebris avalanches and lower frequency. Thismay reflect both an inability to see smallerdebris avalanche scars in aerial photographsand natural variation in slope movementcharacteristics within a single geologicterrane. Assuming that only 59% of debrisavalanches were visible on aerial photographs,Pierson did attempt to correct his value fordebris avalanche frequency by adjusting hiscalculations, which he estimated from hisstudies in the area.

Features of Large Slope Movements

Evidence of large (greater than one-halfhectare) slope movements is widespread in theCoast Range. Evidence includes slumps andareas of block glide on bedding planes, butthere has been little effort to map them indetail on an extensive basis. The onlyestimate of the proportion of sedimentary rocklandscape in landforms created by large slopemovements in the Coast Range comes from thework of J. D. Graham (U.S. Army Corps ofEngineers, personal communication) who hasmapped features of large slope movements for5% to 10% of a 110-km 2 area in the NorthFork Siuslaw River basin.

The timing and rates of movement of thesefeatures are very poorly documented, butseveral lines of evidence suggest that atleast some of these areas have experiencedbrief periods of rapid movement. Measurementof movement of the Wilhelm and Condon slidesin the basin of the North Fork of the SiuslawRiver (Fig. 1) reveals that pulses of movementhave occurred for hours to a few months, basedon the time between resurvey of fixedmonuments. Extensiometers at these sitesrecord movement for periods of hours to a fewdays.

large, rapid slope movements of 20 to 100 m orlore have also been reported in the area overhe past 10 years, most notably the 20 harift Creek Slide (Location 2-3, Fig. 1) andn unnamed slide of about 5 ha located about

6 km to the northwest of the Drift CreekSlide. Neither slide was observed directly,but their movement was heard and was viewedshortly afterward. Movement of up to 8 moccurred over a few years preceding the DriftCreek Slide, but catastrophic failure occurredin a peri of minutes to perhaps an hour ormore.

The Drift ;eek Slide and other large rapidslides ha s:0 dammed rivers and have formedseven lakes in a 13000 km 2 area of the CoastRange, bounded approximately by latitudes 43°Nand 45°N and longitudes 123°15' and 124°15'(Fig. 1). Four of these slope movementsformed lakes with greater than 50-ha surfacearea, the largest being 112-ha Triangle Lake.

Dendrochronologic evidence also points to ahistory of brief episodes of movement followedby periods of little movement. In samples oftilted trees growing on the Wilhelm and CondonSlides, J. D. Graham (U.S. Army Corps ofEngineers, personal communication) observedabrupt development of eccentric growth of treerings during wet periods in 1950-1958,1964-1965, and 1971. This suggests that slopemovement and tilting of trees acceleratedduring these periods, after which movementtilting of the trees was greatly reduced andtree-ring growth returned to a more concentricpattern.

CONTRASTS BETWEEN THE COAST AND THE CASCADERANGES

Based on the available data, debris avalanchesin sedimentary rocks of the south-centralCoast Range are much more frequent and ofsmaller average volume than debris avalanchesin the western Cascades. The rates of soiltransfer by debris avalanches are, however,quite similar between the two areas. Thehigher frequency and lower average volume ofdebris avalanches in the Coast Range mayresult from the effect of highly dissected,steep slopes providing many sites where debrisavalanches can occur, but debris avalanches atthese sites have very limited widths anddepths, so the volume of soil moved is small.

Although available data are inadequate todocument patterns conclusively, large slopemovement features appear to cover a higherpercentage of the landscape in the CascadeRange. We attribute this to widespreadoccurrence of hydrothermally alteredvolcaniclastic rocks in the western Cascadeswhich have resulted in development of numerouslarge-scale slope failures. The movementcharacteristics also appear to differ betweenthe two areas. Earthflows in the Cascadesexperience more prolonged periods ofwithin-year and between-year movement,probably in response to a greater abundance ofclay and amorphous materials in zones offailure. These materials deform slowly andplastically. The siltstones and mudstonesthat occupy failure planes of many of thelarge slope movements in the Coast Range failin a more catastrophic manner. Because large,slow-moving slope movement features arewidespread in the western Cascades, they formimportant sites for occurrence of debrisavalanches (Swanson and Fredriksen, 1982). Inthe Coast Range 0-order channels, following

le terminology of Tsukamoto et al. (1982),:e the dominant sites for initiation ofebris avalanches.

LPPROACHES TO CHARACTERIZING SLOPE MOVEMENT1EGIMES

Description of the slope movement regime of anarea depends on the objectives of the work.We envision a basic description including thearea in active and inactive large slopemovement features, their rates and timing ofmovement, and the frequency and sizedistribution of small, rapid slope failures.These data can be collected by mapping thelarge features and compiling an inventory ofsmall, rapid failures occurring over specificareas and periods of time (Swanson et al.,1981). This sampling program should bestratified by geologic units, and possibly byvariation within these units, such asconsidering degree of hydrothermal alterationin volcanic terranes like the western CascadeRange. On Federal and industry lands in manyareas of western Oregon, soils and landformshave been mapped with units down to 2 ha insize, providing a ready basis for fine-scalestratification for sampling slope movements.

REFERENCES

Hicks, B. A. (1982). Geology, geomorphology,and dynamics of mass movement in parts ofthe Middle Santiam River drainage, westernCascades, Oregon. Oregon StateUniversity, Corvallis, Oregon. 170 p.M.S. thesis.

Ketcheson, G. L. (1977). Hydrologic factorsand environmental impacts of mass soilmovements in the Oregon Coast Fange.Oregon State University, Corval14,Oregon. 93 p. M.S. thesis.

Marion, D. A. (1981). Landslide occurrencein the Blue River drainage, Oregon.Oregon State University, Corvallis,Oregon. 114 p. M.S. thesis.

Morrison, P. H. (1975). Ecological andgeomorphological consequences of massmovements in the Alder Creek watershed andimplications for forest land management.Oregon State University, Corvallis,Oregon. 102 p. B.A. thesis.

Pierson, T. C. (1977). Factors controllingdebris-flow initiation on forestedhillslopes in the Oregon Coast Range.University of Washington, Seattle,Washington. 166 p. Ph.D. thesis.

Shroder, J. F., Jr. (1978).Dendrogeomorphological analysis of massmovement on Table Cliffs Plateau, Utah.Quaternary Research 9:168-185.

Swanson, F. J., Dyrness, C. T. (1975).Impact of clearcutting and roadconstruction on soil erosion by landslidesin the western Cascade Range, Oregon.Geology 3(7): 393-396.

Swanson, F. J., Fredriksen, R. L. (1982).Sediment routing and budgets:Implications for judging impacts offorestry practices. In: Swanson, F. J.,Janda, R. J., Dunne, T., Swanston, D. N.,eds. Sediment budgets and routing inforest drainage basins. U.S. Departmentof Agriculture, Forest Service ResearchPaper PNW-141. Pacific Northwest Forestand Range Experiment Station, Portland,Oregon. 129-137.

Swanson, F. J., James, M. E. (1975). Geologyand geomorphology of the H. J. AndrewsExperimenal Forest, western Cascades,Oregon. U.S. Department of Agriculture,Forest Service Research Paper PNW-188,14 p. Pacific Northwest Forest and RangeExperiment Station, Portland, Oregon.

Swanson, F. J., Swanson, M. M., Woods, C.(1981). Analysis of debris-avalancheerosion in steep forest lands: An examplefrom Mapleton, Oregon, U.S.A. In:Erosion and sediment transport in PacificRim Steeplands. IAHS Publ. No. 132, 67-75.

Tsukamoto, Y., Ohta, T., Noguchi, H. (1982).Hydrological and geomorphological studiesof debris slides on forested hillslopes inJapan. IAHS Publ. No. 137, 89-98.

Wieczorek, G. F. (1982). Map showingrecently active and dormant landslidesnear La Honda, Central Santa CruzMountains, California. U.S. Department ofthe Interior, Geological Survey MapMF-1422.

Yamaguchi, S. (ed). (1980). Landslides inJapan. Japan Society of Landslides, 44 p.