Landslides and Landslide Hazards in Washington State Due ... · Introduction 1 Meteorological...

Click here to load reader

  • date post

    24-Mar-2020
  • Category

    Documents

  • view

    3
  • download

    0

Embed Size (px)

Transcript of Landslides and Landslide Hazards in Washington State Due ... · Introduction 1 Meteorological...

  • U.S. Geological Survey Administrative Report

    Landslides

    and

    Landslide

    Hazards in

    Washington

    State Due

    to February

    5-9, 1996

    Storm

  • Landslides and Landslide Hazards in Washington State Due to February 5-9, 1996 Storm

    by

    Edwin L. Harp1

    Alan F. Chleborad1

    Robert L. Schuster1

    Susan H. Cannon1

    Mark E. Reid2

    Raymond C.Wilson2

    1U.S. Geological Survey, Geologic Hazards Team, 1711 Illinois St., Golden, CO 80401

    2U.S. Geological Survey, 345 Middlefield Rd., Menlo Park, CA 94025

    Publication Layout by Margo Johnson

    U.S. DEPARTMENT OF THE INTERIORU.S. GEOLOGICAL SURVEY

  • Introduction 1Meteorological setting 3Methods of reconnaissance 3Areas of high landslide concentration 4

    Blue Mountains 4Mount St. Helens area 5Ohop Creek drainage 6Willapa Bay area 6Skamokawa Creek area 8Columbia River west of Stella 8Areas of moderate landslide concentration 10Seattle area 10

    Perkins Lane landslide 10Brygger Drive landslide 10Laurelcrest Drive landslide 11California Way landslide 11

    Tacoma area 11Tacoma Narrows landslides 11Schuster Parkway and Rustin Way Landslides 11Marine View Drive Landslides 12Ollala Point landslides 12

    Carbon River area 12Carbon River bridge debris flow 12

    Enumclaw area 12Olympia area 13

    Capitol Campus landslides 13Capitol Lake landslide 13

    Mineral Lake area 13State Highway 12 area south of Mount Rainier 13

    The Glenoma landslide 14Chehalis area 15

    Kresky Avenue landslides 15North Fork Toutle River area 15

    Kid Valley debris flows 15Lewis River area 15

    Woodland landslide 15Lewis River road 16Northwest Pacific Highway landslide 17

    Vancouver area 17State Highway 14 17

    Stevenson landslide 17White Salmon area 19

    CONTENTS

  • Landslides in areas of low concentration 19Ruby Creek landslide 19Keechelus Lake debris flow 19

    Factors affecting landslide distribution 20Rainfall patterns of February storms 20Susceptibility of geologic materials 20Susceptibility affected by vegetative cover 21

    Future landslide hazards 22References cited 23

  • 1

    During the week of February 4,1996, cumulative rainfall levels ofup to 23 in. fell in areas ofOregon, Washington, and Idaho.Combined records of rainfall and waterequivalent of snowmelt were as high as39 in. for this period in the GiffordPinchot National Forest of southernWashington (National ResourcesConservation Service, Water and ClimateCenter, Portland, Oregon). Areas of high-est rainfall intensities were centered onthe Oregon-Washington border.Estimates of damage from the floods andlandslides exceed $300 million (FEMAInteragency Hazard Mitigation Team,1996) in Washington and Oregon alone.Upon receiving reports of the triggeringof landslides over a widespread area, theU.S. Geological Survey sent a team oflandslide specialists from offices inGolden, Colorado, and Menlo Park,California, to perform a regional recon-naissance of landslide distribution andthe attendant hazards. Subsequently, theFederal Emergency Management Agencycontracted with the U.S. GeologicalSurvey to provide a documentation of theoverall distribution of landslides trig-gered by the storms in the State ofWashington and an initial assessment ofspecific hazards that continue to exist as aresult of individual landslides that pose athreat to people and property (missionassignment # USGS-02, request/log # R-1100-23).

    The state-wide reconnaissance was per-formed by a combination of low-leveloverflights in fixed-wing aircraft, ground-based field inspections, and gathering oflandslide data from a variety of sources,such as federal, state, and local agencies,reports from news media, and local resi-dents. This reconnaissance identified apattern of overall landslide limits anddistributions that conform approximatelyto the February 5-9 storm precipitation

    levels as well as to various physiographicand material susceptibility factors.

    Debris flows were the most commonlandslide type triggered by the February5-9 storm. Debris flows typically origi-nate within soil overlying bedrock. As thesoil becomes saturated and pore-waterpressures increase such that the downs-lope driving forces exceed the resistingstrength of the soil, failure takes place.The initial failure may appear as a spoon-shaped rotational slump, a soil slide witha rather planar failure surface, or somecombination of these. Subsequent to fail-ure, the soil mass becomes fluidized bymixing that occurs in the failure processor by intercepting water draining fromthe failure scarp. Once fluidized, the soilmass may move rapidly downslope simi-lar to a slurry of wet concrete. Debrisflows often destroy houses or other struc-tures in their path.

    Other types of landslides observed wererotational slumps, block slides, rock falls,soil slides, and soil falls. The largest land-slide triggered by the February storm wasa block slide of 100,000 to 200,000 yd3that became fluidized by mixing andtransformed into a debris flow. This land-slide is discussed in the section on theGlenoma landslide.

    Plate 1 shows the approximate limits(pink shading) of landslides triggered bythe February storm and attendantsnowmelt. Red shading denotes areas ofhighest landslide concentration(>10/acre). Solid red circles denote loca-tions of individual landslides that werenoteworthy because of their impact onpeople and property. The red shaded areain the Blue Mountains of southeasternWashington is isolated because it is anarea of highly concentrated landslides(debris flows) with little surrounding areaof lower landslide concentrations. Within

    INTRODUCTION

  • 2

    the area of pink shading are scatteredlandslides. The limit as shown on plate 1is approximate. We are certain that somelandslides were triggered outside thisarea by the February storm. However, wejudged these limits to be the most obvi-ous as established by our reconnaissance.

    The highest concentrations of landslidesdue to the February storms occurred atthe northwest edge of the BlueMountains of southeastern Washingtonnear Walla Walla (Plate 1) which is out-side the landslide limits drawn for west-ern Washington. Here, debris flowsreached concentrations exceeding 100individual failures per square mile. Sizesranged from small flows from road cutsto long flows of several miles in length inwhich entire drainage channels failedfrom the drainage divide to the streambottom.

    Although several streambank slumpsattributable to flood flow occurred as farnorth as Washington State Highway (SH)20 in the North Cascades Range, slopefailures of all types were sparse north ofInterstate (I) 90. The northwestern limitcorresponds roughly to the KitsapPeninsula with relatively few slope fail-ures occurring here except as debris flowsalong the coastal bluffs. One shallow soilslide of about 1,500-2,500 yd3 was trig-gered by the February storm on riverbluffs of the Skokomish River about4.5 km west of U.S. Highway 101. Severalother slumps and embankment failureswere reported on the Olympic and KitsapPeninsulas by the WashingtonDepartment of Transportation (SteveLowell, oral commun.), but these wereprobably triggered by previous storms inNovember 1995.

    South of I-90, the concentration of debrisflows and other types of landslidesincreased markedly attaining high con-centrations in the area immediatelysoutheast of Mount St. Helens where

    recent deposits of volcanic ash presentedextremely susceptible material for the formation of debris flows. This area alsocorresponds to a high-precipitation cellwithin the storm of the highest estimatedrainfall levels shown in the radar stormimagery of February 7 of the NationalWeather Service (WSI Corp. EstimatedPrecipitation Report on World Wide Website, February 7, 1996.)

    Except in the Blue Mountains, landslideconcentrations decrease sharply east ofMount St. Helens and Mount Rainier tovirtually nonexistent west of Yakima. Theapproximate eastern boundary of land-slide occurrence in western Washingtonextends southward to intersect theColumbia River gorge about 15 mi westof the Dalles, Oregon (Plate 1). In south-western Washington, west of I-5 andsouth of Aberdeen, few landslidesoccurred except for a small pocket ofdebris flows northwest of Raymond insparsely inhabited timberland (Plate 1).South of Raymond, landslides weresparsely concentrated except along SH-4at the Washington border where twoareas of concentrated landslides occurred.One area, approximately 8 mi west ofLongview, parallels the highway forabout 6 mi and consists of numerousdebris flows that originated above basaltcliffs bordering the highway and flowedover the cliffs and across the road andinto the Columbia River. Another area ofconcentrated landslides occurred about 30mi west of Longview. This area of about 8square miles is bisected by SH-4 and con-sists of primarily debris flows of up toseveral thousand cubic yards. A few rota-tional slumps also occurred within thisarea just west of SH-4, about 4 mi north-west of Skamakowa. The largest of theseslumps is several hundred thousandcubic yards in volume and lies about 0.5mi northwest of the K M Mountain land-slide that occurred in 1990 (Steve Lowell,1990).

  • 3

    The following text describes in greaterdetail the areas of high, moderate, andlow landslide concentrations, noteworthyindividual landslide sites, factors affect-ing landslide concentrations, and estima-tions of the landslide hazards thatpotentially still exist. Both in the text andon plate 1, significant landslides are indi-cated that did not occur during the week

    of February 5-9, 1996, because of theirproximity to landslides that did formduring that time period or that may berelated to the rainfall during that period.Finally, it includes preliminary sugges-tions concerning greater documentationand analysis that may help mitigate land-slide hazards in some of these areas inthe future.

    METEOROLOGICAL SETTING

    METHODS OF RECONNAISSANCE

    The stage for major flooding andlandslide generation was set by thecombination of near-record snow-fall in January followed by warm rain ofunusually long duration. Until midJanuary, the snowpack in Washingtonand northern Oregon was below average.The last two weeks of January, however,saw large snowfall totals with daily totalsof several feet in many locations. By theend of January, the average snowpackhad risen to above 100% in many areas.An intense cold period occurred duringthe week of January 29. The hardest hitarea was Portland where icy conditionslasted for 3 days. This was then followedby a major freezing rain episode begin-ning on February 3.

    On February 6, a strong, subtropical jetstream reached the Oregon-Washington

    area which brought record rainfallamounts to the region. Although suchstorms are not rare, the duration of 3 to 4days of high intensity rainfall was unusu-al. Also, temperatures were unusuallymild during this interval, and relativelywarm rain fell upon the deep snowpackeven at elevations of up to 8,000 ft. Thisbegan to rapidly erode the snowpack,and streams rose quickly on the 6th and7th reaching flood stage in many loca-tions (George H. Taylor, Oregon StateClimatologist, Unpub. information). Theresponse of slopes to the record moisturewas not far behind as the preponderanceof landslide occurrences that were wit-nessed were reported to have occurredon the 7th and 8th.

    Approximately 4 days of aerialreconnaissance in a light fixed-wing airplane were undertaken toobserve the general landslide patternsand to locate landslides posing a poten-tial hazard to people and/or property.Approximately 3 weeks were spentexamining individual landslide sites onthe ground to field check aerial observa-tions and estimations and to gather addi-tional regional and site-specific data.Measurements and estimates relating to

    landslides and their distributions weregathered during these flights and groundsurveys. Although we made every effortto be thorough and extensive, our recon-naissance was relatively brief. Therefore,this survey cannot be considered to be acomplete inventory of all the landslidesthat were triggered by the storm, andthere are probably some significant land-slides that we failed to observe and thatmay pose hazards to people and proper-ty.

  • 4

    There were several areas within the statewhere the concentration of landslidesreached noticeably higher levels (>10landslides/acre) than elsewhere (Plate 1).

    Blue MountainsDespite the great attention given to theflooding and landsliding in westernWashington, extreme southeasternWashington had the highest landslideconcentrations from the February storms.The northeastern margins of the BlueMountains near Walla Walla contain atruly prodigious number of debris flows.

    The main areas affected by the exception-al debris-flow occurrences were the MillCreek, Blue Creek, Touchet, Tucannon,and Walla Walla drainages. Of thesestream drainages, Mill Creek is the onlyone that has been extensively developedand populated. Some of the slopes adja-cent to Mill Creek had such numerousdebris flows that more than 100 individ-ual failures could be counted within areasas small as 10 acres (Figure 1).

    Debris flows were most numerous onopen, grassy hillsides. The source areas ofthese debris flows were generally shal-low, less than 3 ft thick. They had planarfailure surfaces, and the resulting debriswas, in most cases, spread out in anapron across the unfailed slope below thefailure area. Generally, the debris fromthese flows did not reach adjacent down-slope streams. In some cases, however,large or numerous coalescing debrisflows of this type resulted in great vol-umes of sediment reaching drainage bot-toms and covering roadbeds (Figure 2). Infarmlands, debris-flow concentrationswere extremely high at the edges of fieldswhere cultivation has produced anuncompacted soil berm Powell, 1996).

    AREAS OF HIGH LANDSLIDE CONCENTRATION

    Figure 1. Grassy slopes in Blue Mountains of southeastern Washington show-ing extremely high concentrations of debris flows that occurred along thenorthwestern edges of this range.

    Figure 2. Large debrisflow in Blue Creekdrainage that, althoughshallow, has produceda large amount ofdebris. The road at thebottom of the gullywas completely cov-ered to a depth ofabout 4 ft.

  • 5

    Areas of mixed grassland and timber con-tained fewer debris flows; however, inthese areas debris flows with lengths ofup to 1 mi occurred. These debris flowsincorporated trees and brush along withrock and soil and were particularlydestructive to roads, houses, and otherstructures. The depth of failure and scouralong drainage bottoms often exceeded4 ft as it commonly removed soil andweathered rock down to the bedrock-soilinterface.

    Source material for the debris flows wasmainly colluvial and residual soils as wellas glacial loess overlying basalticbedrock. Flow deposits originating inloess contained mostly silt and clay-sizedparticles. Boulders stripped from under-lying weathered bedrock occurred indeposits where debris flows scouredchannels down to weathered basalt.

    Considerable damage to houses and vehi-cles was done by debris flows in the MillCreek area (Figure 3). Seven vehicles andfive houses were damaged or destroyedby debris flows emanating from sidecanyons to Mill Creek itself (Figure 4).Debris flows ranged in size from shallowflows less than 50 ft in length to deeplyincised flows along main side drainagesexceeding 1 mi in length.

    There have been similar occurrences offlooding and landslides in the BlueMountains in the past. Storms in 1931and 1964 produced extensive flooding.According to long-time residents of MillCreek, similar concentrations of debrisflows were triggered in 1931 but were notnearly so numerous in 1964 (Rich Klicker,personal commun., 1996).

    Mount St. Helens AreaA 70 mi2-area east of Mount St. Helensincluding the drainages of Bean Creek,Clearwater Creek, and to a lesser extent,Smith Creek, was virtually blanketed in

    Figure 3. Residence in Mill Creek located near the mouth of a side canyon.Debris flows emanating from this side canyon hit the garage shown in leftcenter of photo, damaged two cars, and partially covered a caterpillar tractor.Remnants of the debris flow deposit can be seen in photo center.

    Figure 4. Aerial oblique view of Mill Creek showing debris flows in almostevery tributary canyon. Debris flows extend the entire length of all of thetributary canyons to the intersection with Mill Creek.

  • 6

    places by debris flows emanating fromrecent ash deposits erupted from MountSt. Helens (Plate 1). Individual debrisflows in this area of tens of thousands ofcubic yards appeared to be common(Figure 5); however, as observed from theair, most of the flows as well as theirheadscarps appeared to be relativelysmall and shallow. These debris flowsrepresent the highest concentrations oflandslides that were triggered by theFebruary storms in western Washington.

    This area of high debris-flow concentra-tion coincides with one of the areas ofpeak estimated rainfall from WSI Corp.radar imagery for February 7, 1996.Figure 6 shows that, just southeast ofMount St. Helens, the maximum amountof estimated precipitation (5-6 in) forFebruary 7 occurs in this area of highdebris-flow concentration. Here we see agood correlation between the radarimagery of precipitation patterns and thelandslide patterns. This is an area wherethe effect of high rainfall on debris-flowgeneration were added to by rapidsnowmelt as a result of the rain on a deepsnowpack.

    Ohop Creek DrainageNortheast of Mount Rainier along theOhop Creek drainage is an area of highlandslide concentration of about 200 mi2.The area on plate 1 is elliptical in shapeparallel to Ohop Creek. Along the bluffsadjacent to Ohop Creek, abundant debrisflows occurred with volumes of up to1,000 yd3. Most of the debris flows weremuch smaller averaging about 200 yd3.Despite the small volume of most of thedebris flows, a house was destroyed by adebris flow of 200 yd3 or less thatoccurred on the east side of Ohop Lake(Figure 7). The debris flow originatednear the top of the slope about 30 ft abovethe access road to the houses. It traveledacross this road, knocked the house from

    Figure 5. Large debris flows in recent volcanic ash deposits southeast ofMount St. Helens.

    its piers, and toppled the house on itsside near the lake shoreline (#65, Table 1).

    The system of county and local roadswithin the Ohop Creek drainage wasespecially hard hit. Some stretches of theOrville E. road that connects Kapowsinand SH-161, had dozens of soil slide-debris flows issuing from roadcuts andsteep natural slopes per 100 yd of roadlength. Here, road cleanup of debris fromsoil slides and debris flows was still inprogress in April 1996.

    Willapa Bay AreaApproximately 10 mi northwest ofRaymond, Washington, an area of unde-veloped timberland about 10 mi2 in areaproduced numerous debris flows. Mostranged from several hundred to severalthousand cubic yards in volume. Loggingroads were affected by both inundationby debris-flow material and by removalfrom slumping of the road bed and inci-sion during the flow of debris acrossroads. Other than damage to loggingroads, there appeared to be no impact onpeople or property within this small area.

  • 7

    Figure 6. Radar imagery for February 7, 1996 showing high estimated-precipitation cells centered southeast of Mount St. Helens andalong the Columbia River gorge near Dodson, Oregon where several large debris flows damaged houses and closed both I-84 and theBurlington Northern rail lines (Reprinted courtesy of WSI Corporation).

  • 8

    Figure 8. Headscarparea of deep-seatedslump-earth slide westof Washington SH-4near SkamokawaCreek.

    Skamokawa Creek AreaAnother small area of about 8 mi2 alongSH-4 northwest of Stella, in the drainageof the West Fork of Skamokawa Creekwas the site of numerous debris flowsand several rotational slumps. The debrisflows ranged in volume from severalhundred cubic yards to several thousandcubic yards. These debris flows primarilydamaged local and county roads. Thelargest of the rotational slumps was sev-eral hundred thousand cubic yards involume (Figure 8). It is located about 0.5mi northwest of the K M landslide ofapproximately 2 million yd3 that disrupt-ed SH-4 in 1990 Lowell, 1990). The newlandslide removed a large section of trees,and its headscarp is about 10 ft below apaved previous detour of SH-4. This roadmay be incorporated in the landslide asretrogressive slumping takes place duringfuture rainfall events.

    Columbia River West of StellaNumerous debris flows occurred in a nar-row section, about 10 mi in length, ofshoreline of the Columbia River just westof Stella along SH-4. The highway wasclosed during the February storms inmany places by debris flows thatoccurred in marine deposits that overliethe basalt cliffs adjacent to the highway.These debris flows ranged from severalthousand cubic yards to almost 100,000yd3. The material from the debris flowscascaded over the basalt cliffs and buriedthe highway in several places. In mostcases, the basalt cliffs were not involvedin the failures except for a few pieces ofloose basalt dislodged by the flow of theoverlying material. The largest of thesedebris flows reached the Columbia River(Figure 9). By April 1996, all of the debris-flow material had been removed exceptfor debris from the largest of these fail-ures. At that time extensive excavationand slope reconstruction was still ongo-ing in the headscarp area (Figure 10) ofthis failure (#59, Plate 1).

    Figure 7. Cottage oneast shore of OhopLake knocked off itsfootings by a smalldebris flow.

  • 9

    Figure 9. Large slump-debris flow locatedabout 5 mi west ofStella, Washington onSH–4.

    Figure 10. Head-scarp area of slump-debris flow shown in figure 9.

  • 10

    Much of the landslide-affectedarea from the February stormslies outside the areas of highestlandslide concentration. Although lower(

  • 11

    Figure 12. Debris flow from steep coastal bluffs along Puget Sound in the com-munity of Salmon Beach which destroyed a residence (Reprinted courtesy ofthe Seattle Post-Intelligencer).

    the landslide was evident at the time ofobservation. Although plastic has beenplaced over these fractures, future rainsmay initiate further movement of theslump.

    Laurelcrest Drive LandslideSeveral debris flows occurred alongLaurelcrest Drive, along the shore of LakeWashington north of the University ofWashington. The largest was about300 yd3 in volume and was the mostdamaging. This debris flow issued fromsteep cliffs behind the house at 5012Laurelcrest Drive, missed the house byinches, but destroyed a large woodendeck, a patio, and brick barbeque struc-ture (#69, Table 1). Its occurrence wasnoted by the owner as just after mid-night, Thursday, February 8, 1996. As ofMarch 15, 1996, several trees were leaninginto the failure path of the debris flow assupport for their roots was partiallyremoved by the failure. Other sections ofcliff adjacent to the scarp may also failand produce small failures during futurerainfall.

    California Way Landslide

    On the northern end of the penin-sula forming West Seattle, thesteep coastal bluffs aboveCalifornia Way failed producing a soilslide in weakly cemented sandstone ofabout 600-1,000 yd3 (#70, Table 1). Theslide partially mobilized into small debrisflows near the toe but flowed only a shortdistance. Most of the failure massremained on California Way as of March15, 1996. Only the roadway was directlyaffected. Houses are located above andmore than 50 ft to the south of the failurescarp and are unlikely to be involved inany reactivation or retrogression of theslide in the near future.

    Tacoma AreaSeveral landslides that occurred in theTacoma area caught the immediate atten-tion of the news media because of thedestruction of houses and the disruptionof a major rail line as well as the derailingof a freight train.

    Tacoma Narrows LandslidesA number of landslides originating inunconsolidated glacial deposits that formthe steep coastal bluffs on Puget Sounddestroyed two homes in Salmon Beachand damaged another (#’s 4, 5, 6; Table 1).The failures were about 1,000-5,000 yd3each and started as translational slides,subsequently mobilizing into debris flowsthat slammed into the houses, pushingthem out into the Sound (Figure 12).Utility lines coming into the area werealso damaged by the debris flows.

    Schuster Parkway and Rustin Way LandslidesNumerous small failures from the coastalbluff were observed (#12, Table 1). Mostfailures occurred below Stadium High

  • 12

    School on Stadium Way and wereapproximately 1 m wide and

  • 13

    and consists mainly of cobble- andboulder–sized debris with a minoramount of sandy matrix (Figure 14). Thepath of this flow was scoured to depthsof over 4 ft along its channel.

    Olympia AreaSeveral small slumps and debris flowsoccurred in the Olympia area around thestate capitol campus and at the southernend of Capitol Lake. Although small,these failures had considerable impact onCapitol Lake and the City sewer systemof Olympia.

    Capitol Campus LandslidesNumerous small slumps and debris flowswere triggered from the steep bluff northand west of the state capitol campus,which sits on this bluff about 100 ft aboveCapitol Lake. The largest of these is about200 yd3 in volume and failed from slopesjust west of the Governor’s mansion.Another failure of similar size originatedin the slope above a steam plant locatedjust above lake level on the shore ofCapitol Lake. This debris flow hit a con-crete diversion structure at the northeastcorner of the plant and was deflected intoCapitol Lake. Just northeast of this site,several small debris flows and slumps ofseveral tens of cubic yards volume eachcovered railroad tracks and parts of anadjacent road at the bottom of the steepbluffs (Gerstel, 1996).

    Capitol Lake LandslideImmediately below Capitol Boulevard atthe point where the Deschutes Riverenters Capitol Lake, a series of shallowcoalescing debris flows severed twosewer lines. This sewer-line breakageresulted in the discharge of 6 million yd3of sewage into Capitol Lake.

    Mineral Lake AreaClear-cut slopes west of SH-7 nearMineral Lake produced debris flows that

    Figure 14. Lower portion of debris-flow path that closed Washington SH-410east of Enumclaw. Debris flow originated in berm of logging road. Note pre-ponderance of coarse material in debris.

    closed the highway for 2 days. Severaldebris flows originating from the outerberms of logging roads in clear cuts westof SH-7 traveled up to 200 yd beforecrossing the highway. One of these debrisflows came within 50 yd of a home nearthe highway. Deposits from the flow nearthe home appeared to have been largelydeposited by fluvial action. They weredominated by wood debris and were rel-atively sparse in sediment. Thesedeposits may have been emplaced by ahyperconcentrated flow or one that ismore than 50 percent water.

    State Highway 12 Area South of Mount RainierMuch of the area adjacent to U.S.Highway 12 south of Mount Rainier ispopulated by scattered debris flows ofmoderate size (1,000-5,000 yd3). The mostnotable of the landslides in this area,however, is also the largest landslide trig-gered by the February rainstorms.

  • 14

    Glenoma LandslideAt approximately 3:00 a.m., Thursday,February 8 (Elmer Day, oral commun.,1996), a block slide of about 1.5 millionyd3 was triggered from the shoulder of apartially clear-cut ridge in volcaniclasticsediments about 1 mi north of U.S. 12, 5mi east of Glenoma and about 13 mi eastof Morton (#77, Table 1). It appears thatthe slide moved southeast into a narrowV-shaped gully where it may haveremained until overtopped by impound-ed water. Regardless of whether the slidemass mobilized immediately or sometimeafter initial failure into the gully, the slidemass eventually became extremely fluidas it traveled as a debris flow. It traveledsoutheast from this gully where it collid-ed with a house and a garage (Figure 15and 16) before stopping about 150 ydshort of U.S. 12.

    Although the house was occupied at thetime it was hit by the debris flow, nofatalities resulted, because the slide masshad nearly traveled its full extent beforehitting the house. The house was crackedbeyond repair but was not overrun orcrushed by the debris. Elmer Day, whoowns the house and lives in a trailerbeyond the slide margins, stated that heheard and felt the debris-flow movementand could see the movement of trees asthe debris flow spread out across the rela-tively flat field next to the dwellings.

    The slide headscarp, shown in figure 16 isapproximately 150 ft high, with a nearlyvertical headwall and a nearly 90˚ inter-section of this wall with the exposedbasal shear surface. The shape of thescarp suggests that the landslide’s initialmovement was mainly translational as arelatively coherent mass. After reachingthe narrow gully, its direction of move-ment changed by almost 90˚ from south-west to southeast. In the first few daysafter the landslide, numerous smallpieces of the cliff forming the headscarp

    Figure 15. House crumpled by block slide-debris flow shown in figure 16 nearU.S.-12 east of Glenoma, Washington.

    Figure 16. Block slide-debris flow thatdestroyed house shownin figure 15. House is atbottom right of photo.Near-vertical cliff inheadscarp is about 150ft high.

  • 15

    Figure 17. Numerous slumps (arrows) near Chehalis, Washington,reactivated by February storm.

    fell leaving a fairly planar scarp with fewremaining fractures above the presentheadscarp (Elmer Day, landowner, oralcommun.).

    Chehalis AreaImmediately east of the I-5 exit for northChehalis is a hillside that has been clearedof trees and prepared for residentialdevelopment. The devegetation and roademplacement for the development desta-bilized a large area of the hillside andsubsequently produced several slow-moving slump-earthflow type landslidesthat have been active in past years (#53,Table 1). The February storms reactivatedmovement on these earthflows as well astriggering several debris flows from thesame area (Figure 17).

    Kresky Avenue LandslidesTwo earth slides of about 7,000 to 10,000yd3 each blocked Kresky Avenue in northChehalis (#54, Table 1). The slides wereabout 150 ft apart and occurred in red,clay-rich sedimentary deposits. The slidescrossed the street closing all lanes of traf-fic. Although traffic was disrupted, nodamage to buildings occurred. Furtherrainfall may trigger renewed sliding fromthese slopes with similar effects. Possiblehazards are repeated road closure anddamage to passing vehicles andmotorists.

    North Fork Toutle River AreaKid Valley Debris FlowsA group of 15 to 20 debris flows, someseveral hundred yards in length, occurredin Kid Valley on the south side of theNorth Fork of the Toutle River (#19, Table1). These debris flows originated in soiland weathered bedrock overlying vol-canic bedrock. Volumes ranged from 500to 5,000 yd3. Damage from the debrisflows in the area was predominantly tologging roads and an improved roadwayto a lumber mill.

    State Highway 504 east of Kid Valley wasblocked by debris flows in two places(#20, Table 1). The first of these was about2 mi east of Kid Valley where over 200 ftof highway was buried and guard raildestroyed by a debris flow of 5,000-10,000yd3. Similar to those in Kid Valley, thisdebris flow incised a channel to theunderlying volcanic bedrock. The totallength of travel exceeded 600 ft andextended into the North Fork of theToutle River.

    SH-504 was also buried by a debris flownear the Hofstadt Visitors’ Center nearHofstadt Creek (#21, Table 1). This debrisflow was about 5,000 yd3 in volume andburied about 90 ft of highway anddestroyed the guard rail in its path. Add-itional debris remaining in the sourceareas may mobilize during future rain-storms.

    Lewis River AreaWoodland LandslideOne of the largest impacts created by anyindividual landslide in the State ofWashington was that caused by the fail-

  • 16

    Figure 19. Small slump in northwest Vancouver that destroyed a residencelocated next to a steep slope bordering an old river terrace.

    ure of the 32,000 yd3 landslide thatblocked I-5 and the Burlington Northern-Santa Fe railroad tracks 3 mi north ofWoodland (#80, Table 1). The initial land-slide of about 12,000 yd3 occurred shortlyafter 2:00 p.m., February 8. Its initialmovement blocked only the northboundlanes of the interstate. By Friday,February 9, Washington Department ofTransportation was planning to open thenorthbound lanes at 6:30 p.m., when,shortly after 6:00 p.m., a larger landslidemass of 20,000 yd3 failed from the samesource area and covered all lanes of I-5 aswell as the rail lines adjacent to I-5(Figure 18).

    Even after reopening part of the interstateon February 10, northbound traffic wascommonly delayed for several hours,because only one lane was open in thatdirection. This situation remained untilFebruary 19. The landslide began as arotational slump in marine sediments andremained relatively undeformed untildescending a bench about 60 ft above theroadcut of I-5, which was originally about120 ft high (Blodgett, Regional HighwayEngineer, Washington Department ofTransportation, oral commun., May 1996).At this point, the toe portion of the land-slide became extremely disrupted andbehaved much as a flow as it overran I-5and the rail lines.

    Lewis River RoadAlong the Lewis River road (SH-503)from I-5 to its present passable extentwhere it is washed out, about 3 mi fromthe end of Swift Creek Reservoir south ofMount St. Helens, we noted 10 debrisflows, 2 rock falls, and 1 slump (#’s 23-35,Table 1). All but the slump covered orpartially covered SH-503 or adjacentForest Service roads. All of these failureswere 1,000 yd3 or smaller except for onedebris flow that covered Forest Road #81near the outlet of Merrill Lake which liesabout 8 mi south of Mount St. Helens.

    Figure 18. Slump north of Woodland, Washington that covered I-5 and theBurlington Northern-Santa Fe rail line.

  • 17

    This debris flow originated in the sides ofa stream channel near the outlet of MerrillLake and scoured the channel down tobedrock. It flowed across Road #81 andcontinued downslope into a campgroundwhere it spread out and deposited about5,000 yd3 of material.

    Northwest Pacific Highway LandslideAbout 0.5 mi south of Woodland, theNorthwest Pacific Highway was coveredby the toe of a large slump-earthflowwhose source area is just below a rockquarry (#36, Table 1). The volume of theslump-earthflow is about 100,000-200,000yd3, and it has overrun and partially cov-ered approximately 200 yd of the high-way. This landslide predates the Februarystorms and was moving slowly at thetime of the storms. Its movement wasaccelerated as a result of the Februaryrainfall.

    This landslide is likely to continue itsslow movement for some time unlesssources of water from the area of the rockquarry can be diverted from the slidemass and some drainage measures areundertaken.

    Vancouver AreaIn the Vancouver Lake area, tracks of theBurlington Northern-Santa Fe Railroadwere blocked by numerous slumps,debris flows, and rock debris from theshoreline bluffs adjacent to the lake. Inone case, a train had to be excavated fromlandslide debris before being able tomove (Columbian, February 7, 1996).

    In the Hazel Dell area of north Vancouverat 11612 N. E. Ninth Ct., a house was bro-ken in two when a debris flow of about1,000 yd3 carrying two large trees failedfrom a 30-foot-bluff just behind the home(Figure 19; #43, Table 1). The bluff is atthe edge of an old river terrace. Most ofthe homes in this area are built on the ter-race next to and below the low bluff that

    Figure 20. Headscarp area of deep-seated, complex rotational slump-earth-flow in subdivision in Stevenson, Washington. Landslide has resulted inremoval of three houses and has threatened six or seven additional residences.

    failed. In most cases, the houses are closeenough to the bluff that little or no outrunonto the flat terrace is necessary for alandslide mass to hit a dwelling.

    State Highway 14State Highway 14 was closed or partiallycovered by debris flows in 15 differentplaces from east of Washougal toWishram about 80 mi to the east. Thesedebris flows ranged in volume fromabout 200 to 5,000 yd3 and consisted ofcolluvial and residual soils overlyingbedrock.

    Stevenson LandslideIn the northern part of Stevenson, west ofthe junction of Loop Road and Maple HillRoad, a large complex rotational slump-earthflow was reactivated in the Februarystorms (#81, Table 1). Most of the areahere is comprised of old landslidedeposits with many visible old scarps andhummocky topography. Below Maple HillRoad, a zone of prominent scarps devel-oped following the February 1996 stormwith cumulative downslope displace-ments of 4 ft or more (Figure 20). Threehouses in this area had to be removed

  • 18

    from their foundations. The foundationshown in figure 21 illustrates the damagedone to houses by deformation acrossthis zone of fractures.

    The fractures shown in figures 20 and 21constitute the headscarp area of the land-slide. Below this, the topography steep-ens due to the presence of what appearsto be an old landslide scarp that pro-duced many coalescing debris flows fromits face during the February rainfall.These debris flows covered a section ofLoop Road just below this old scarp andflowed across the road, plugged a culvertin the road, and filled a small basin abovethe road culvert.

    The fractures forming the headscarp ofthe landslide extend for several hundredft to the east and west of the area shownin figures 20 and 21 before they becomeindiscernible. The toe of the landslidenear Loop Road is doming upwards fromthe rotation of the slide mass beneath it.This section is just below the area of theoverflowage of Loop Road by the debrisflows originating from the old scarp. Sixto seven other houses are within thedetectable slide boundaries and arebeginning to experience slight deforma-tion (David Esche, Assistant CountyEngineer, Skamania County, oral com-mun., May 1996).

    Above the headscarp shown in figures 20and 21, are other tension fractures formedby the retrogression of deformation ups-lope (Figure 22). These mark the spread-ing of displacement of the slope beyondthe initial landslide boundaries. Using acrude estimate of the average landslidedepth of about 50 ft, the landslide volumetotals about 500,000-1,000,000 yd3.According to residents of the area, move-ment continues (as of May 1996), and it isunclear how great the final extent of thelandslide mass will eventually be and

    Figure 22. Aerial oblique view of Stevenson landslide. At center right is foun-dation shown in figure 21. Arrows denote additional extension fractures up-slope from fractures shown in figures 20 and 21.

    Figure 21. Effects of headscarp fractures on foundation of residence onStevenson landslide.

  • 19

    There were a few isolated landslides out-side the pink shaded area in plate 1 thatcaused significant impact to transporta-tion routes. The two landslides discussedbelow affected major highways andcaused closure of one of WashingtonState’s two main interstate highways.

    Ruby Creek LandslideApproximately 12 mi north of BlewettPass on U.S.-97, a large rock slide of sev-eral hundred thousand cubic yards wasreactivated by both the November andFebruary rainfall. Prior to November1995, slide movement had not beenreported at the site since 1960 (SteveLowell, Washington Department ofTransportation, Chief EngineeringGeologist, oral commun., March 1996;#78, Table 1).

    The reactivated movement in November1995 caused extensive erosion of thehighway due to the displacement ofPeshastin Creek against the road fill. The

    LANDSLIDES IN AREAS OF LOW CONCENTRATIONslide movement almost dammedPeshastin Creek. Erosion repair costabout 2 million dollars.

    Renewed activity in February resulted inalmost 30 ft of additional movement onthe slide and considerable renewed ero-sion of the newly repaired highway.Highway repair was still underway whenwe visited the site in April 1996. Potentialfor future movement remains high duringstorms such as those of November 1995and February 1996. Renewed movementof the present slide mass serves to desta-bilize slopes that are adjacent and ups-lope of this slide.

    Keechelus Lake Debris FlowDuring the February 5-9 storm, a moder-ate-sized debris flow of about 5,000 yd3emerged from about 20 yd above I-90 andflowed across all lanes, closing the inter-state for about 1 day. This debris flowoccurred just east of Keechelus Dam (#79,Table 1).

    how many additional homes it mayinclude.

    White Salmon AreaIn the city of White Salmon and along theslopes adjacent to the White Salmon Rivernear its mouth, numerous small, shallowdebris flows occurred along road cuts andgrassy natural slopes. The depths of theseflows as well as the vacated scarps aregenerally 1 ft or less, and flow distancesare generally less than 20 ft. North of themouth of the river, few debris flows areseen until reaching Phelps Creek, a dis-tance of about 12 mi.

    Unlike the surrounding terrain, the chan-nel of Phelps Creek has been scoured bymixed debris flow and flood activity to awidth of about 20 yd and for a length ofat least 1 mi (#82, Table 1). Scattered treesand boulders are evident along thestreambanks. State Highway 141 was cov-ered when the box culvert was plugged,and debris-flow material flowed over thehighway, burying it to a depth of about20 ft. The debris flow mass continuedover the highway and down PhelpsCreek for perhaps another 200 ft (KenKirkland, Washington Department ofTransportation, Regional Engineer, oralcommun., May 1996).

  • 20

    FACTORS AFFECTING LANDSLIDE DISTRIBUTION

    Rainfall Patterns of February StormAs previously mentioned, the radarimagery for February 7, 1996, from theNational Weather Service (Figure 6)shows cells of high intensity rainfall thatcorrelate well with high concentrations oflandslides in western Washington. Thehighest rainfall levels shown in figure 6are 5-6 in. One of these cells is roughlycircular and located just southeast ofMount St. Helens where the highest con-centrations of landslides in westernWashington occurred immediately south-east of Mount St. Helens. Although notpreviously discussed in this report,another area of similar rainfall level isalong the Columbia Gorge near Dodson,Oregon. This area is one of elongatedcoalescing cells that corresponds wellwith the occurrence of numerous destruc-tive debris flows that originated from thesteep cliffs of the gorge that destroyedone house, damaged several others, andclosed both I-84 and the Union PacificRailroad tracks.

    Although accounting for only part of themoisture delivered to slopes, the relative-ly warm rainfall was directly responsiblefor the rapid snowmelt. Thus, the spatialvariation in rapid snowmelt should close-ly match the spatial variation in rainfallintensities from the storms.

    Susceptibility of GeologicMaterialsAlthough no thorough evaluation of theeffect of the spatial variation in geologyon the distribution of landslides has beenattempted, a few first-order patterns areobvious. To the west of Vancouver alongthe Columbia River, soils and deposits ofmarine sediments overlying basalt cliffsproved more susceptible than the basaltor its residual soils in producing debris

    flows. Virtually no failures occurred with-in the basalts or within soils developedon them. However, the relatively soft,marine sediments that overlie the basaltshave produced numerous soil slumpsand debris flows.

    In the Blue Mountains of southeasternWashington, the effects of glacial loesscan be seen in the debris flow concentra-tion. Because the loess is extremely sus-ceptible to failure, debris flows in thesedeposits on farmlands closely spaced,and both scarps and deposits are relative-ly thin, generally less than 2 ft thick(Figure 23). Nowhere else in Washingtondid debris-flow concentrations achievethis level from the February storm. Theconcentrations of debris flows just south-east of Mount St. Helens were nearly ashigh but, as previously discussed, theywere also located at one of the maxima ofrainfall intensities—no accurate radarimagery was available for the BlueMountains.

    High debris-flow concentrations occurredin both loess and in the recent ashdeposits of Mount St. Helens; both havelow densities. Loess is formed by the set-tling of wind-blown silt and clay particlesderived from glacial rock flour. Its lowdensity makes it especially erodable andits interparticle structure easily collapsi-ble upon saturation and water flowthrough its pore spaces.

    Man-made fill in general, and road-fillembankments in particular, was especial-ly susceptible to landsliding. The outerportions of roads on steep slopes are usu-ally poorly compacted and, because thismaterial has relatively low density, it isextremely susceptible to rainfall-inducedinstability. Many of the obvious debrisflows in clear-cut areas were initiated atthe outer portion of logging roads. Even

  • 21

    Figure 23. High concentration of debris flows in glacial-loess deposits onnorthwestern flank of Blue Mountains.

    improved roads and highways were sus-ceptible to debris flows and slumps.Improved highways and roads seemed tohave as numerous associated failures asthe unimproved roads. Where steep roadcuts existed on the uphill side of roads,materials were susceptible to landslidesbecause of the oversteepening andremoval of downslope support.Numerous small debris flows and slumpsformed at and just above cut slopeboundaries.

    Undoubtedly, other examples of high-susceptibility deposits exist; however,more detailed investigations will berequired to identify all such materials.

    Susceptibility Affected byVegetative CoverThe most noticeable differences in vegeta-tive cover on slopes in western Washing-ton are those provided by the two endmembers of timber harvesting: forestedslopes and clear-cut slopes. There haslong been controversy regarding theeffects of logging and clear cutting on thestability of slopes.

    Our regional aerial reconnaissance gaveus the overall impression that clear cutscontained more debris flows and associat-ed other types of landslides than similarslopes in forested areas. However, wequalify this statement with the observa-tion that landslides were often more diffi-cult to see on forested slopes. Wecertainly found numerous landslideswithin forested areas but had trouble see-ing them clearly from the air even whenwe had previous knowledge of their pres-ence. Ground-based field inspections gen-erally allowed us to discover many morelandslides in forested slopes than wewere aware of from the air.

    In the Blue Mountains of southeasternWashington where clear cutting is notpresently done, we saw debris flows bothin areas of timber and on slopes that werealmost free of timber. The only major dif-ference we saw in this area was thatslopes completely free of timber tendedto have high concentrations of smaller,very shallow debris flows, while slopeswith mixed forest and grassland hadsomewhat lower concentrations of debrisflows in which longer debris flows withdeeper channels occurred.

    From our observations, we do not feelthat any conclusive statement can bemade regarding the effects of logging orclear-cutting on the distribution of land-slides. To accomplish this, several similarareas, both forested and deforested, willhave to be studied carefully to determineif there are indeed differences in the dis-tribution of landslides that are inducedby the removal of trees.

  • 22

    Our regional survey highlights sev-eral areas where additional workmay help to avoid future hazardsto people and property. Patterns of land-slide distribution discerned in our studyhave indicated that attendant hazards arecontrolled by material properties and thepresence of high intensity/duration rain-fall. These hazards can be predicted ifdetailed inventories of landslides andknown geological and geotechnical dataare compiled and analyzed together.Three areas where additional analysis ofthe February 1996 landslide event canserve to help avoid similar occurrences inthe future are as follows:

    (1) Large scale (1:18,000 or larger) aerialphotography of the area affected by land-slides in the Blue Mountains should betaken to document the distribution ofdebris flows in detail so that an inventorymap of the debris flows could be made.Such a detailed inventory could be com-bined with existing geologic mapping andgeotechnical data to prepare a hazardmap for the area. Only one of the canyonsin this area is densely developed. A haz-ard map could assist decisionmakers inavoiding landslide hazards in the futureor to develop areas using appropriate mit-igative measures to protect people andproperty in areas highly susceptible tolandslide or debris-flow hazard.

    FUTURE LANDSLIDE HAZARDS

    (2) Steep coastal bluffs in the PugetSound area were the sites of numerousdebris flows and other types of landslidesin the February storms, yet most of thelandslides occurring there could not beseen from aerial reconnaissance. Thesefailures were only clearly visible fromfairly close quarters on the ground. Wesuggest that the coastal cliffs of the PugetSound area be inspected and accuratelocations be documented to the extentpossible. This is an area of intense devel-opment pressure, and an accurate pictureof where landslides were triggered dur-ing this storm is vital in making intelli-gent land use planning decisions. Aconsideration of existent landslide sus-ceptibilities and potential hazards willreduce the risk to people and propertyboth now and with future development.

    (3) The Stevenson landslide may continueas an ongoing problem. Once in motion,deep-seated complex landslides such asthis may continue to move for extendedperiods of time. Also, landslide bound-aries may enlarge to include significantportions of surrounding terrain. Withthree houses already removed and six toseven more affected, detailed geotechni-cal studies and mitigative engineeringmeasures may have to be undertaken ifthe area presently affected is to be stabi-lized and the surrounding area protectedagainst becoming unstable.

  • 23

    FEMA Interagency Hazard MitigationTeam, 1996, February 1996 Flooding,Landslides, and Stream Erosion in theState of Oregon: Interagency HazardMitigation Team Report.

    _____1996, State of Washington WinterStorm of 1995-1996: InteragencyHazard Mitigation Team Report.

    Gerstel, W.J., 1996, Slope stability analysisof bluffs along the Washington StateCapitol Campus, Olympia,Washington: Washington Division ofGeology and Earth Resources, Open-File Report 96-3, 1 v.

    Lowell, Steve M., 1990, The K MMountain landslide near Skamokawa:Washington Geologic Newsletter, v. 18,no. 4, p. 3-7.

    Powell, Jack, 1996, Appendix A—Masswasting assessment, in South ForkTouchet watershed, in McKinney,Charles, ed., Touchet WatershedAnalysis, Washington Department ofNatural Resources, Olympia,Washington, 21 p.

    Tacoma News Tribune, 9 February 1996.

    The Columbian, 8 February 1996.

    ———, 7 February 1996.

    Walsh, Timothy J., Korosec, Michael A.,Phillips, William M., Logan, Robert L.,and Schasse, Henry W., 1987, Geologicmap of Washington-Southwest quad-rant: Washington Division of Geologyand Earth Resources Geologic MapGM-34, 1:250,000.

    REFERENCES CITED

  • 24

    Table 1. Specific Landslide Sites(Landslide numbers plotted on Plate 1)

    EstimatedLandslide Type Volume (yd3) Source Material Damage Continuing Hazard Comments

    1,2 Soil slide-debris 1,200 Granular sediments Pushed one home Other homes Shallow failures onflow. (Whidbey Fm (?) & into Colvos Passage threatened by bluff.

    ovelying surficial & damaged fronts of adjacent unstablesediments. three others. slopes.

    3 ? ? ? ? Threatens town- Reported in Tacomahouses on Lake News Tribune 2/9/96Fenwick Rd., Kent, (site not visited).Wash.

    4,5,6 Soil slide-debris 1,000-5,000 Unconsol. glacial Two Salmon Beach Other homes along Several shallowflow. sand and silt. homes pushed into Salmon Beach failures.

    Puget Sound at threatened byTacoma Narrows. possibility of moreDamaged a third landslides duringhouse and utility intense storms.lines.

    7 Debris flow. 3,000 Mostly sand & silt; Collapse of about Potential for cont. Inadequate culvert orprobably fill under 65 ft of roadway failure could be broken drainpipepaved road on steep where road intercepts diminished by probable cause.bluff. drainage. improving drainage.

    8,9,10,11 Soil slides-debris 500-2,000 Glacial sand, gravel, Two homes severely Future storms may Shallow failure;flows. & silt. damaged; blocked cause similar some channelized.

    road & damaged failures along thisguard rail. stretch of Marine

    View Drive.

    12 Debris flow. 500-1,000 Colluvium and Covered part of May recur during Shallow failures onunderlying gravel, Schuster Parkway future storms. steep bluffs.sand, silt, and clay. and overtopped

    median barrier.

    13 Slump(?)

  • 25

    Table 1. Specific Landslide Sites—continued(Landslide numbers plotted on Plate 1)

    EstimatedLandslide Type Volume (yd3) Source Material Damage Continuing Hazard Comments

    19 Debris flow. 500-5,000 Soil & weathered Blocked road to Some debris flows 15 debris flow—rock over volcanic lumber mill on south likely to recur in some severalbedrock. side of North Fork intense storms. hundred yards in

    of Toutle River in length.several places. Also some damaged logging roads higheron mountain.

    20 Debris flow. 5,000-10,000 Soil & weathered Covered about 75 yd Possible recurrence Partly channelizedrock over volcanic of SH-504 & flowed during periods of shallow flow.

    into N. Fork of intense rainfall.Toutle River. Destroyedguard rail.

    21 Debris flow. 5,000 Soil & weathered Blocked SH-504 & Remaining material Shallow failure.rock over volcanic destroyed 30 yd of in source area maybedrock. guard rail. mobilize during

    intense rainfall.

    22 Undetermined. ? ? Derailed Amtrak train ? Reported inin Ostrander area Columbian 2/7/96—north of Kelso. site not visited.

    23,24 Debris flows.

  • 26

    Table 1. Specific Landslide Sites—continued(Landslide numbers plotted on Plate 1)

    EstimatedLandslide Type Volume (yd3) Source Material Damage Continuing Hazard Comments

    39 Debris flow. 5,000-10,000 Fill & underlying Removed 50-yd Likely to recur Small culvert unableTroutdale Fm. section of Timmons Rd. unless drainage to accommodate

    problem is solved. water flow, resulting in saturation of fill & underlying sediment.

    40,41,42 Soil slides-debris ? ? Covered railroad ? Reported in flows. tracks & part of a Columbian, 2/7/96—

    train. site not visited.

    43 Slump-debris flow. 1,000-2,000 Colluvium & House broken in two. Adjacent homes Susceptible bluff alluvium. might be threatened above old river

    by similar failures terrace.from bluffs duringfuture rainstorms.

    44 Coalescing soil 500-2,000 Troutdale Fm. & Covered & damaged Future intense rainfallslides-debris flows. overlying soil. NW Forest Home Rd. would likely generate

    more shallow slides &debris flows.

    45,46 Debris flows.

  • 27

    Table 1. Specific Landslide Sites—continued(Landslide numbers plotted on Plate 1)

    EstimatedLandslide Type Volume (yd3) Source Material Damage Continuing Hazard Comments

    54 Complex earth 15,000-20,000 Red, clayey sediments Both slides blocked Could happen again Two slides about 50slide-earth flow. (total for two probably of Lincoln Kresky Ave. in in intense rainfall; if yd apart crossed

    slides) Creek Fm. north Chehalis. so, would block Kresky Ave.Kresky Ave., butnot a hazard to nearbybuildings.

    55 Earth slide-debris 500-1,000 Soft sedimentary Reached road at Probably not likely High on loggedflow. rocks, probably of bottom of slope. to trigger additional hillside above Coal

    Lincoln Creek Fm. flows from source Creek.area that wouldreach road.

    56 Debris slide. 1,000 Colluvium & None to logged slope. None. Occurred on sloperesidual soil. above Bear Creek,

    tributary of TiltonRiver.

    57 Rock slides & 2,000 total. Road cuts in Rock had to be Probably will recur Several small rockfalls. andesite. removed from SH-508. in heavy rain. falls & rock slides

    accumulated in borrow pit & on pave-ment. Just east of SH-508 crossing ofBear Creek.

    58 Rock falls, rock 2,000 total. Highway cuts in Rock had to be Likely to recur in At least 10 small slides, & debris andesite. removed from U.S.-12. heavy rain. landslides (mainlyflows. rock falls, one debris

    flow) from shortlength of highwaycuts on north side of U.S. 12.

    59 Debris flow. 100 Red sediments above Blocked SH-4, 0.3 mi Could be reactivated Occurred April 26,Columbia River basalt. west of Germany Cr. by future rain. 1996. May be a

    reactivation of a failure triggeredin February.

    60 Earth slide 200 Glacial sediments in Knocked house from All 12 houses along Occurred on April 23,coastal bluff. foundation. this section of 1996, due to heavy

    coastal bluff may be rain. Head of slide in danger in intense about 10 yd aboverainfall. house.

    61 Earth slide-earth 500-1,000 Glacial sediments in No damage because Continuing problem Occurred on Feb. 8,flow. coastal bluff. of absence of houses. with slides along 1996 from heavy rain.

    Slide ran out onto beach. bluffs in heavy rainfall. About 0.5 mi north ofslide #60.

    62 Debris flows. 200 total Glacial sediments in Blocked Westlake Could easily block Two debris flows onbluff along west Ave. in 2300 block. Westlake Ave. again. 45° slope occurredshore of Lake Union. Head of debris flow Apartment building on April 23, 1996.

    close to foundation does not appear to Both 25 yd long.of large apartment be in serious danger.

  • 28

    Table 1. Specific Landslide Sites—continued(Landslide numbers plotted on Plate 1)

    EstimatedLandslide Type Volume (yd3) Source Material Damage Continuing Hazard Comments

    63 Earth slide. 500 Glacial sediments in Destroyed large house All houses along Occurred on April 23,coastal bluff. on Perkins Lane. Perkins Lane south 1996, 0.4 mi south of

    of intersection with Rayle St. Perkins Rayle St. are in on- Lane intersection.going danger during Head of slide justfuture periods of below Magnolia Blvd.heavy rainfall.

    64 Debris slide. 1,000 Glacial sediments in Blocked Perkins Entire section of Occurred on April 23,coastal bluff. Lane in section with bluff above Perkins 1996. About 75 yd

    no houses. Lane susceptible to north of south end offailure during Perkins Lane. Sepa-intense rainfall. rate from but related

    to March 8 slide atsouth end of Perkins Lane.

    65 Debris flow. 300 Fluvial sediments in Destroyed cottage Other small debrisslopes above east on lakeshore. flows from slopesshore of Ohop Lake. above houses suggest

    that other damagingflows could occur inperiods of intenserainfall.

    66 Rotational slump- 1,500-rock Glacial sandy con- Rock fall piled up Further movement Occurred March 8,rock fall. fall. Slump glomerate in steep against house on of deep-seated rota- 1996. Head scarp

    unknown but coastal bluff. Perkins Lane. tional slump beneath below 1734much larger. house will continue Magnolia Dr.

    to deform it andpossibly destroy it.

    67 Rotational slump. Not sure. ? Four houses are Recurring movementProbably at being deformed on is likely during intenseleast 5,000. Brygger Dr. rainfall unless exten-

    sive mitigation is done to slide.

    68 Debris flow. 200 Glacial sediments Blocked Portage Bay May have additionalin coastal bluff. Dr. and piled up against debris flow from head-

    a garage on north side scarp area during of street. Head scarp intense rainfall. Retro-is below a house. gression of head scarp

    may jeopardize houseabove scarp.

    69 Debris flow. 300 Sediments in coastal Destroyed patio next Future intense rainfallbluff. to house on may trigger other failures

    Laurelcrest Dr. from slopes that couldendanger houses.

    70 Slump-debris flow. 600-1,000 Coastal bluff Blocked California Way. May recur in futuresediments. intense storms.

    71 Debris flow. 200 Colluvium overlying SH-165 blocked. Some added flow mayEocene continental occur from same sourcesediments, probably during high-intensityCarbonado Fm. rainfall.

  • 29

    Table 1. Specific Landslide Sites—continued(Landslide numbers plotted on Plate 1)

    EstimatedLandslide Type Volume (yd3) Source Material Damage Continuing Hazard Comments

    72 Debris flow. 2,000 Colluvium overlying Removed downhill Road is exposed to Carbonado Fm. lane of SH-165 and erosion during intense

    covered highway. rainfall.

    73 Soil slide-debris 3,000 Soil overlying Head scarp undercuts Intense rainfall in futureflow. andesite. bridge abutment. will cause head scarp

    to cut farther into bridgeabutment.

    74 Debris flow. 2,000 Soil overlying Blocked SH-165. Intense rainfall could Source about 100 ftandesite. possibly initiate above road.

    additional debris flowsfrom source.

    75 Debris flow. 500,000 Colluvium overlying Blocked SH-410. Additional flow may Debris flow track isrhyolite. result from future about 1.5 mi long

    intense rains, but in clear-cut area.small volumes willprobably not make itall the way to the highway.

    76 Soil slump- 100 Glacial sediments Knocked house off Intense rainfall may 7970 SH-106 mile-debris flow. from coastal bluff. wooden footings. initiate further slope post 8.

    failure above housesin this area.

    77 Block slide- 1,500,000 Volcaniclastic Knocked house off Little remaining hazard Largest failure debris flow. sediments. foundation & hit garage. to residents at distal triggered by

    margin. February storm.

    78 Rockslide. 200,000 Renewed erosion of Renewed movement Slide reactivatedU.S. 97. possible in future in Nov. 1995, then

    storms with attendant in Feb. 1996.deflection of creektoward highway andrenewed erosion.

    79 Debris flow. 5,000 Closed I-90 for about Possible additional 1 day. debris flows from

    retrogressive failureof soil in scarp duringfuture storms.

    80 Rotational slump. 32,000 Marine sediments. Blocked I-5 & Burlington May pose no furtherNorthern-Santa Fe Rail- hazard due to excava-road tracks. tion of landslide mass.

    81 Complex rotational 500,000- Reactivated land- Three houses removed Will probably continueslump/earthflow 1,000,000 slide deposits. and threatens six to movement in future

    seven others. storms.

    82 Debris flow. 20,000- Covered SH-141 to a Small failures may100,000 ? depth of 20 ft. occur from margins

    of scoured riverbank during future storms.