Chapter 6
Types of Landslides Identified in the Study Area

The zone of extension is characterized by a large head scarp approximately 500 meters long, with a maximum 12 meters of vertical displacement showing on the surface.  This head scarp is rounded, non-continuous, and revegetated with a smaller (40 m²) active slide occurring on it on the northern side.  It is dissected by a small creek at the southern end and does not appear to have any tension cracks above the head scarp.  There is one sag pond on top of a down-dropped block located just south of the smaller active slide.  Just south of the sag pond is a flat area (top of a down-dropped block) where several structures and houses are located.

The shear zones appear as small gullies and creeks that are non-continuous, partially due to the construction of the large cutslope at the southern end of the slide.

Below Highway 213 is the toe, which is steep to very steep (70°-80°) in places and which has two smaller active slides and two inactive-young slides located on it (Figure 17).  The three northernmost younger slides are shallow-seated (depth to failure plane<4.5 m).  The southern younger slide is a deep-seated landslide.

Some of the bore-hole and well logs drilled during the construction of Highway 213 in 1979 and later, show gradual increases in sediment grain size with depth, from clays and silts to sands and gravels and other suspicious stratigraphic units.  One well-water drillers log (29), located just above the head scarp near Morton Road, shows a gradual but consistent increase in the grain size with depth from clay rich to a predominately sandy soil at a depth of approximately 50 m (Figure 18 and appendix 1).  Bore-hole log TB-104, located just north of the cross-section, displays the phreatic water surface at approximately two meters below the ground surface (Figure 18 and appendix 1).  Bore-hole log TB-108, located just below the southern head scarp, has an organic layer at a depth of 16.8 m (appendix 1).  This organic layer may represent a paleosol (old land surface) that is now restricting water movement below it causing an increased local water table.

 Table 4: Highway 213-Morton Road Landslide Stability Analysis Results, assuming undrained conditions.

To locate an approximate failure plane and a factor of safety for this deep-seated landslide, the program XSTABL was used with the Bishop analysis for circular failure planes (Table 4)(Sharma, 1994).  The current conditions used in the initial stability analysis were derived from the surficial landslide features (Figure 17), the generalized subsurface soil properties from bore holes TB-104 and TB-108 and phreatic water surfaces found in well log 29 and bore hole TB-104 (appendix 2, Figure 18).  This analysis consisted of three trials, starting with estimated material properties (cohesion and angle of internal friction) from the subsurface data, then lowering the values until a reasonable factor of safety was found (Table 4, appendix 2).  The material properties displayed in Figure 18 are most likely very close to the residual strength.  The most critical failure plane (F_s=1.2) occurs at a maximum depth of 25 meters, approximately twice the exposed scarp (12 meters), and is depicted on the cross-section (Figure 18).  To confirm this approximate depth to the failure plane, the critical thickness was calculated, resulting with a thickness of 21.5 meters, a value slightly less than that calculated by XSTABL, but within a close range of approximation considering the lack of subsurface information (appendix 2).  Further stability analysis of this landslide included the addition of a horizontal earthquake load through a coefficient (k).  Recommended values from the California Department of Conservation (1997) included k=0.10 for a 6.5 magnitude earthquake (major) and k=0.15 for a 8.25 magnitude earthquake (great).  This analysis resulted in a factor of safety ranging from 1.5 to 0.6 for a major earthquake and 1.3 to 0.5 for a great earthquake (appendix 2).

Although major development, including cutslopes and fills, occurred during the building of Highway 213, which passes across the toe of this landslide, this slide does not appear to have any deep-seated movements since this construction.  But, examination of the toe and the scarp indicate active and inactive-young slides have been modifying this deep-seated landslide continuously.  The stability analysis confirms the current conditions of this deep-seated movement with a factor of safety of 1.2 for the current conditions.  It was also discovered that a even major earthquake (magnitude=6.5) could cause a decrease in the factor of safety below one, indicating reactivation of this landslide (California Department of Conservation, 1997).  These predicted factors of safety show that there is a chance of reactivation for this inactive-mature, deep-seated landslide and others like it.

The Highway 213, Mile Post 2.1 Deep-Seated Landslide

In 1979 the Oregon Department of Transportation began investigation for the construction of Highway 213.  Included in this investigation was a series of bore holes.  Some of these bore holes were very helpful in more accurately assessing the depth to the failure plane of some of the deep-seated landslides.  The Highway 213, mile post 2.1 landslide (Plate 1), located crossing Highway 213 in the central portion of the study area is a deep-seated, inactive-mature landslide very similar to the Highway 213-Morton Road landslide described in the last section.  This slide has a double bowl shape in plan view and is approximately 300 meters wide and 300 meters long with an exposed maximum scarp of approximately nine meters.  The scarp and toe parallel Highway 213, Holly Lane and Newell Creek.  Within this landslide’s boundaries are two smaller, shallow-seated, active slides, both of which appear to have their scarps in the road prism of Highway 213.  During the building of Highway 213, the Oregon Department of Transportation recognized this deep-seated slide through surficial mapping and subsurface exploration and mitigated it with a rock buttress located upslope from the highway (ODOT, 1998).

One of the bore holes, TB-109, drilled during construction of the highway, notes the presence of slickensides at a depth of approximately 13.5 meters (ODOT, 1998) (appendix 1).  The depth of the slickensides in this bore hole (TB-109), located in the middle the Highway 213, mile post 2.1 deep seated landslide, is interpreted to display the approximate depth to the failure plane.  Although this interpreted depth to failure plane (13.5 meters) does not match the estimated depth from the exposed scarp (nine meters), their comparison confirms that this slide is deep-seated.

Deep-Seated Landslide Conclusions

Since most of these deep-seated slides are prehistoric, human caused induction can be ruled out.  The most probable initiating causes of these slides are a wetter prehistoric climate, inundation and quick lake level drop following the 40+ Missoula Floods (12,700 to 15,300 years B.P.), and earthquake shaking (CDC, 1997; Waitt, 1985).  Although most of these slides are considered inactive-mature (dormant), extreme caution should be exercised before any construction takes place on or near one of these deep-seated slides, including a complete surface/subsurface investigation and a slope stability analysis including earthquake loads by an engineering geologist and a geotechnical engineer.  This recommendation is further supported by the fact that the Dewey-Warren Street landslide reactivated in response to above average rainfall in 1996 (Figure 19).  This deep-seated slide, located in the northwest section of the study area, had a horizontal displacement of only 0.3 meters and a vertical displacement of 0.5 meters in 1996 (Figure 19).  This minor amount of movement was enough to damage the foundation of a new house to the point where repair was not economically feasible.

Shallow-Seated Landslides

There are about 65 shallow-seated landslides in the study area.  The scarps range from 0.2 to 4 meters with an average vertical displacement of two meters (± 1.0 meters)(Figure 16).  This average vertical displacement measured on the scarps of two meters was used to approximate the depth to the failure plane which corresponds to the general location of the contact between the residual soils and bedrock (Table 1)(Gerig, 1985).  These shallow-seated slides generally have a much smaller area than the deep-seated slides, ranging from 36 to 6,141 m² or 0.001% to 0.2% of the study area, in size.  This smaller size is probably due to the constraint of the depth to the failure plane.  Most of the shallow-seated slides resulted in a complete evacuation of the material, sometimes exposing all of the failure surface, unlike the deep-seated landslides.

Figure 19: Reactivation of the Dewey-Warren Street, deep-seated landslide with destroyed (red tagged) house in the background.  An 8 ½ inch by 11 inch size field notebook is on the scarp for scale.  Displacement measured on the scarp only totaled 0.3 meters of horizontal and 0.5 meters of vertical.

Many of these shallow-seated slides are active and hence younger than the deep-seated slides.  Furthermore, many of these younger shallow-seated slides have been active within the last ten to twenty years.  For example the Highway 213, mile post 2.1 shallow-seated landslide, discussed earlier, occurred during construction of the highway in 1984 (Figure 20).  Some of the active, shallow-seated landslides occur in cutslopes and fills.

Realizing that this form of mass wasting predominates in Newell Creek Canyon, I decided to make a detailed study and map of a typical shallow-seated landslide, with a depth to failure plane of approximately 3.5 meters, (shallow-seated) located on Sha Spady’s property in the northwest section of the study area (Plate 3).  This particular detailed section of this study was performed in the spring of 1996.  In the spring of 1997, the Spady Landslide failed again, this time to a depth of approximately 5.5 to 6 meters (deep-seated).  This slide was mapped again at a scale of 1:100 following the second period of sliding (Plate 3).   The 1996 landslide on the Spady property is representative of shallow-seated landslides in the study area, and the following is a description of this landslide and its growth.

The Spady Landslide

The Spady property is located at the end of Alden Street in the northwest corner of the study area (Figure 21).  The property consists of a ridge trending north-northeast with slopes on either side ranging from 10°-35° (Figure 21).  The residual soils on these slopes are part of the Jory Series (45C) (Figure 4, Table 1, and Table 2) and are classified as a silty clay loam, which can contain up to 40% clay.  They are colluvial soils with a depth to the upper Troutdale Formation greater than 1.5 meters.  The vegetation on these slopes is mainly Douglas fir (up to 75 years old), maple trees, and other vegetation composed of shrubs and blackberry bushes.  At the bottom of the western slope is a creek trending northwest.  This creek begins about 140 meters south of the Spady residence as a culvert outage that funnels the runoff from the Barclay Hills apartment complex located just south of the Spady property. 

Figure 20: Oblique arial photograph of the active, shallow-seated Highway 213-mile post 2.1 landslide on the toe of the larger deep-seated slide.  This slide occurred during construction of the highway in the summer of 1984 and is indicated by arrows.  Also noted is the rock buttress located in the cutslope above the slide (after ODOT, 1998). 

During 1996 and 1997 this area had greater than normal precipitation, which resulted in a “100 year flood” event in February 1996 and another episode of flooding in January 1997 (Figure 22).

The 1996 Failure of the Spady Landslide

On March 6, 1996, Sha Spady, the property owner, observed three large cracks in the ground very close to where the scarp was located after the initial failure.  These cracks had an approximate one half to two meter spacing parallel to the slope and were approximately six to nine meters long with up to 30 centimeters of opening on each crack.

On March 7, Spady watched as the upper section of the slide failed with velocities ranging from 0.5 to 2 meters/second.  The slide stopped when it reached the small creek running northwest, located 100 meters west of her house, creating a small earth dam.  Over the next 4-5 days, a pond formed behind the landslide debris dam.  Finally, on March 12, Spady watched as the dam broke and a slurry of viscous soil and water flowed down the stream valley and stopped just before reaching the junction of the next creek about 120 meters down slope (Sha Spady, property owner in Newell Creek Canyon, personal communication, 1996).

Figure 22: Histogram of monthly precipitation for Oregon City from 1996 to 1997 (August) (after National Weather Service, 1998).

Description of the 1996 Failure of the Spady Landslide

Features of the landslide and debris flow were mapped by plane table mapping during April and May of 1996 at a scale of 1:150 (Figure 21 and Plate 3).  Soil samples were taken at two locations and several lab tests were performed including grain size analysis (sieve and hydrometer), Atterberg limits, dilatancy, toughness, dry strength, and unconfined compressive strength (Table 5- units 1 and 4 and appendix 3, sample locations are on Plate 3).  A Janbu stability analysis was performed to evaluate the conditions before failure as a benchmark for use elsewhere in the study area (appendix 3).

The 1996 evacuated area consists of an oblong-shaped bowl with an approximate volume of 2200 m³.  This area is located approximately 80 meters southwest of the Spady home (Figure 21).  Below the evacuated zone, filling the creek running northwest on the western side of the Spady property is the debris flow deposit.  This debris flow is 80 meters long and, on average, 10 meters wide.  This debris flow ends where the small creek enters another creek trending northeast.  The following section is a description of the four zones into which the 1996 failure of the Spady Landslide was divided for classification purposes: zone of extension, shear zones, zone of saturated flow, and the debris flow.

Table 5: Material properties for residual soils and bedrock units (beds) within the extent of the Spady Landslide.

Zone of Extension - The head scarp was a concave downslope break, with a maximum vertical displacement of approximately 3.5 meters (Figure 9 and Plate 3).  This part of the landslide was approximately 21 meters wide, with no vegetation left standing.

There were a few small (1-2 cm opening) tension cracks located above the head scarp.  The interception of the flanks and the head scarp had many tension cracks extending out from the main body as far as 15 meters, characteristics of a zone of extension (Fleming and Johnson, 1989).  There was one main spring line that parallels the head scarp and was located at the base of the scarp.

Shear Zones - Both flanks were littered with tension cracks, some of which were in an en-echelon style.  These tension cracks were usually left-stepping along the right flank and right-stepping along the left flank.  They were orientated between approximately parallel along the main shear zones and closer to perpendicular where the shear zones were interrupted by a minor scarp, usually extending out from the main shear zone.  Some of the tension cracks had openings as wide as 30 centimeters.  Slickensides were located on the lower right flank where the slide entered the channel and became a debris flow.

Zone of Saturated Flow - The main body of the slide was marked by the saturated soil flows (Plate 3).  Below the head scarp, several spring lines mark the beginning of this saturated flow zone, which extended to the small creek running northwest (Plate 3).  This zone contained features that formed after the debris dam failed on March 12, as a result of smaller sloughing off the head scarp and flanks, as they were still intact after March of 1996.  These saturated flow features were generally at least twice as long as they were wide and were composed of non-blocky saturated residual soil that resemble lobes.

The point where the slide becomes a debris flow (bottom of cross section A-A’ on Plate 3 was the only place where the Upper Troutdale Formation was exposed.  The rest of the slide occurred in the residual soils that covers the parent material on the slope.

Debris Flow - The channelized section of the slide, from elevation 73 to 90 meters, was 80 meters long and has debris levees on both flanks of the main body (Figure 11).  These levees mark the highest elevation the saturated soil and debris reached during flow down the channel.  They were located along most of the debris flow and range from 0.3 to 1.2 meters in height and 3 to 28 meters in length.  The tops and sides of some of these debris levees had scour marks resembling slickensides and trended parallel to the general direction of flow.  This section of the 1996 failure of the Spady Landslide also exhibited saturated flow lobe features, like the zone of saturated flow, that contained toe-like thrusts of blocky debris at the ends of each lobe.  When this debris flow reached the intersection of the next creek, it formed a snout with a vertical thrust of about 1.5 meters.

Pre-Failure Stability Analysis - A pre-failure stability analysis was performed along the slope where the 1996 failure took place (appendix 3).  This analysis was performed using Janbu’s method of slices, which calculates the factor of safety of a known failure plane.  The cross section line A-A’ was used as the line of analysis.  The analysis variables were derived from Table 5 and a water table was assumed to be at a depth of 0.5 meters below the ground surface.  The material properties were lowered until a factor of safety at one (F_s=1.0) was reached for the higher than normal precipitation year of 1996 (Figure 22 and appendix 3).

Growth of the Spady Landslide: The 1997 Failure

The second failure occurred early in January 1997 (Figure 23), following the heavy rains of December 1996.  The first and most obvious difference was the change in shape and increase in size of the evacuated area from one to two oblong shaped bowls with a volume increase from 2,200 m³ to 14,600 m³ (Figure 21 and Figure 23).  The direction along the slope where this increase took place was not upslope, as seen in the 1996 failure, but along the slope (north, and next to the 1996 failure).  The debris flow section of this second landslide also increased to about twice the length it was after the 1996 failure and has ponded the next channel at the intersection of the two creeks (Figure 12).

Figure 23: Map of the entire extent of the 1997 Spady Landslide failure. Includes the debris flow section of the landslide and the ponded channel (Wilson et al., 1997).