Predicting the earthquake potential for Walla Walla County and the potential hazards resulting from earthquakes is an effort to aid disaster prevention. The county has been identified as an area with a substantial earthquake potential, as it contains two major fault systems, the Wallula Fault Zone and the Hite Fault. In addition to the Stateline magnitude 6 earthquake of 1936, there have been felt earthquakes in the area in the past decade.
Potential earthquake hazards include fault displacement, shaking, differential settlement, landslides, and seiches. Fault displacement most likely occurs along known faults. Shaking and differential settlement are likely to be greatest where sediments are thick, fine-grained, and/or water saturated. The landslide potential is greatest on steep slopes, particularly if the ground is saturated. Seiches could occur in Lake Wallula.
Eastern Washington and eastern Oregon are not immune to large earthquakes. The large ridges around us such as Rattlesnake Mountain provide the hard evidence for this. Earthquakes are the growing pains that accompany the ridges as they develop and they’re not yet done growing (Reidel, 1999).
There has recently been a large focus on the idea that western Washington is due for a very large earthquake, potentially with a magnitude of 8 or 9 on the Richter Scale, due to the subduction of the Juan de Fuca Plate. It is easy to think that by living on the eastern side of the mountains we do not have to worry about earthquakes since we are not close to a subduction zone. This idea is wrong, though, and could possibly be dangerous if we are not at least aware of potential earthquake hazards.
Walla Walla is located near two major faults, the Olympic-Wallowa lineament (OWL) and the Hite Fault System (HFS), as well as other minor faults. Evidence of past tectonic activity can be seen in local geomorphic features such as the Horse Heaven Hills anticlinal ridge and the faceted spurs on the hillsides near Wallula Gap. The year 1936 was the last time a major earthquake occurred in this region, which geologically speaking, was a very recent event.
There is going to be another earthquake in the Walla Walla region, and Walla Walla County wants to be prepared for it. The purpose of this thesis is to research the effects that earthquakes will have on the Walla Walla region in an effort to aid disaster prevention within Walla Walla County.
There are five main categories of hazards associated with earthquakes. These hazards include shaking, differential settlement, fault displacement, landslides, and seiches. Walla Walla County has the potential to experience all of these hazards in the event of an earthquake.
Walla Walla County contains two major fault systems, which intersect at approximately a 90-degree angle just southeast of Walla Walla in the northeastern part of Umatilla County, Oregon. These two fault systems are the Wallula Fault Zone (WFZ), which is part of a much larger feature called the Olympic-Wallowa lineament (OWL), trending west-northwest, and the Hite Fault System (HFS), trending north-northeast (Kuehn, 1995) (figure 1).
Figure 1. Map illustrating the relationships of the major faults and other major features (PowerPoint presentation, Whitman College, 1997).
Raisz first described the OWL in 1945 as a "peculiar line stretching from Cape Flattery at the entrance of Juan de Fuca Strait to the Wallowa Mts." This line essentially begins on the north side of the Olympic Peninsula, extends across the Cascade Range, across the Columbia Basin, and along the northeast side of the Wallowa Mountains, a total of 500 km in length (Raisz, 1945 in Kuehn, 1995). Raisz (1945) postulated two different hypotheses for this lineament: (1) that it is a series of independent features coincidentally aligned along the same line or that (2) it represents a transcurrent fault.
Kuehn (1995) describes the section of the OWL stretching across the Columbia Basin from near Cle Elum to the western edge of the Blue Mountains as the Cle Elum-Wallula deformed zone (CLEW). Kuehn divides the CLEW into 3 parts: the zone in the Yakima Fold Belt, the Rattlesnake-Wallula alignment (RAW), and the Wallula Fault Zone. Kuehn (1995) states that "Quaternary deformation has been documented at more than 15 locations along the CLEW," indicating that the CLEW is probably currently active.
Kuehn (1995) also points out that the OWL, determined to have a recent right-lateral strike-slip component, is parallel to other right-lateral strike-slip faults to the southwest and to the northeast, suggesting that the OWL is actually part of a more extensive deformation zone.
The Wallula Fault Zone (WFZ) is the section of the OWL that is most important to this study as it encompasses the region from Wallula Gap, Washington to Milton-Freewater, Oregon, a distance of about 30 km (Mann and Meyer, 1993 in McQuarrie, 1993). The latest movement along the WFZ appears to be right-lateral strike slip, based on horizontal slickenlines observed in fault zones through basalt flows in Wallula Gap (McQuarrie, 1993). A strongly linear series of faceted spurs running through Wallula Gap on the east side of the Columbia River indicate a very high angle fault component. This faulting is described by McQuarrie (1993) to be normal, as this is usually what faceted spurs indicate, but Reidel (1997) has mapped this zone as having a reverse component thinking that reverse motion is the most recent movement along this fault zone. Reidel (1993, in McQuarrie, 1993) believes this to be true based on hydrofracture tests conducted in the basalt in the region. The fact that the vertical component of the WFZ is so steeply dipping, in fact probably close to vertical, makes the difference between reverse and normal almost inconsequential.
Evidence that is present along the WFZ indicates that it has been recently active and is probably still active. Mann and Meyer (1993) discovered a fresh fault scarp about 7 kilometers south-southwest of Umapine that they refer to informally as the "Umapine fault." Due to a lack of offset drainage patterns and the presence of vertical striae, Mann and Meyer (1993) determined that this fault is dip-slip. Volcanic ash samples were collected along this fault and geochemically compared to ash samples of known ages. It was found that this ash closely correlated to 10,700 year old Mt.St.Helens’ ash. This indicates that the Umapine fault has a minimum of 5 meters of vertical offset for a maximum of 10,700 years, or about 47 centimeters/ 1000 years, based on the age of the ash and the amount of offset (Mann and Meyer, 1993).
The Hite Fault System (HFS) extends from the North Fork of the Umatilla River east of Pendleton, Oregon to Pomeroy, Washington, a length of 150 kilometers, although some geologists have extended it all the way to Colfax, Washington (Kuehn, 1995). Recent studies along the HFS indicate that it is a left-lateral strike slip system with about 80 kilometers or more of left-lateral displacement prior to the Columbia River Basalt eruptions (Kuehn, 1995). Previous to these interpretations, the Hite Fault had been thought to be normal by some scientists, and right-lateral strike-slip by other scientists, whereas others have interpreted it to have an early normal component followed by left-lateral strike slip (Kuehn, 1995). Kuehn, though, finds that the most recent interpretations discount the previous interpretations. Although Kuehn (1995) recognizes the left-lateral displacement along the Hite Fault, he later states in his report that "the most visible sign of displacement is down to the west vertical offset associated with the formation of the middle segment of the Blue Mountains Uplift." The Hite Fault was mapped as a normal fault by Schuster et al. (1997).
The total vertical displacement along the HFS associated with the Blue Mountains Uplift is 300 meters from the time of the Grande Ronde N2 portion of the Columbia River Basalt Group (as defined by Kuehn in 1995) to present time. Sixty meters of thinning occurred along a segment of the Grande Ronde N2 (a 100,000 to 200,000 year interval) and 30 meters of thinning of the post-Grande Ronde stratigraphy (a 1.1 to 2.6 million year interval). The remaining 210 meters of displacement have taken place after the eruption of the Umatilla flow of the CRBG (a period of 13 to 14 million years). From the rates of deformation presented in this information, it appears that this deformation taking place across the Hite Fault was most active during Grande Ronde N2 and has decreased since then (Kuehn, 1995).
Relationship between the OWL and the HFS
One hypothesis proposed by Kuehn (1995) regarding the OWL and the HFS is that these two fault systems are conjugate megashears. Evidence for this can be seen in that these systems are both consistent with the regional strain pattern consisting of north-northwest directed compression and east-northeast directed extension. The OWL, trending northwest-southeast displays right-lateral displacement, and the HFS trending northeast-southwest displays left-lateral displacement. Evidence for this strain pattern can be found in the east-west trending anticline/syncline sequence of the Yakima Fold Belt and Lewiston structure and the north to northwest trending CRBG dikes (McQuarrie, 1993; Kuehn, 1995). Other evidence supporting the conjugate shear hypothesis include the presence of apparently offset faults and lineaments produced by the faults of one set displacing the faults of the other set. Possible evidence of limited simultaneous movement can be seen in a crush zone at the intersection of two conjugate faults along Tiger Creek Road, which also supports this hypothesis (Kuehn 1995). Although Kuehn does present possible evidence supporting the OWL and the HFS as conjugate fault systems, he also warns about assuming that these systems are conjugate faults based solely on apparent offsets. It is possible that these offsets could have formed independently from one another, and Kuehn (1995) states that he did find "one conjugate step without a crossing fault of the opposing system" in the Lincton Mountain Faults. A contrasting viewpoint is given by McQuarrie (1993) who agrees with the megashear model as a possible scenario, but states that one problem with this model is that where the HFS intersects the OWL, there is no obvious displacement on either fault. McQuarrie (1993) also points out that the common depiction of the Hite Fault as a normal fault on geologic maps is inconsistent with the model, as it should be left-lateral strike-slip to match the model.
The last major earthquake to shake the Walla Walla region occurred on July 15, 1936 near Stateline, Oregon on the Washington-Oregon border. This earthquake had a Richter magnitude of 6.1 and a Modified Mercalli intensity (MM) of VIII at Umapine, Oregon determined by Mann and Meyer (1993) (figure 2). Mann and Meyer (1993) determined the maximum MM to be VIII based on "apparent liquefaction" and "partial to total building collapse" between the towns of Milton-Freewater, Umapine, and Athena, Oregon. Brown (1937) described some of the damage as being broken chimneys, movement of several houses up to an inch on their foundations, the condemnation of one building, and the rotation of cemetery headstones. Brown (1937) estimated the damage to be about $100,000.
Mann and Meyer (1993) propose that the epicenter of this earthquake was near Umapine, Oregon and that the earthquake occurred along the Wallula Fault. They based this hypothesis on intensity data as well as field data. It was determined that the earthquake’s greatest intensity was between Milton-Freewater and Umapine implying a source of the earthquake coming from the WFZ. Some previously unrecognized faults and lineaments were discovered during this study, as well as the "Umapine fault" that was mentioned earlier.
Woodward-Clyde Consultants (1980), who conducted a study for the Washington Public Power Supply System in Richland, Washington, suggest that the 1936 earthquake actually occurred near Waitsburg, Washington, and propagated in a southwest direction along the Hite Fault. This conclusion was based on results obtained from seismic wave data collected from this earthquake. Seeing as how Hanford is part of the Washington Public Power Supply System for whom this study was conducted, the possibility exists that the conclusions may be slightly biased, as earthquake activity along the HFS is a much more favorable scenario for Hanford than activity on the WFZ. Reidel and Tolan (1994) are also of the opinion that this earthquake could have occurred along the HFS.
On April 8, 1979 a magnitude 4.2 earthquake occurred in the same area as the 1936 earthquake (Bogaert et al., 1983). This event was large enough to be felt. Woodward-Clyde (1980) determined that this event had an epicenter with a shallow focus (about a 5 km) depth very near the intersection of the WFZ and HFS. In their study they also found that the epicenters of these two earthquakes were coincident with one of two possible fault planes of the 1979 earthquake. This finding supports the idea that these earthquakes occurred along a northeast-trending fault (i.e. the HFS). The 1979 earthquake was not reported to not have had any foreshocks or aftershocks (Bogaert et al., 1983).
Another earthquake that occurred in this same region on March 22, 1983 was of particular interest to Bogaert et al. (1983). This 3.8 magnitude event was felt in Walla Walla, Milton-Freewater, and Helix, Oregon. This earthquake was found to be interesting as there were no reported aftershocks or foreshocks although there was a later earthquake that occurred in the same area on April 13, 1983 with a magnitude of 2.6.
Seismic shaking is one of the greatest earthquake hazards and causes the largest amount of loss to constructed works. In comparison to fault displacement, shaking is very widespread and often effects structures in areas far from the earthquake’s epicenter.
Berlin (1980) describes three parameters of earthquake shaking. These parameters include amplitude, frequency, and duration. The amplitude of ground motion is intensified in unconsolidated sediment, as is the damage done to structures. In regard to frequency, "Frequency of motion becomes paramount if it approximates or matches the frequency response characteristics of attached works." In other words, shorter buildings are subject to more damage on firm ground, whereas taller buildings are subject to more damage on unconsolidated ground. Also, high frequency earthquakes tend to cause pipelines to break. Oftentimes, broken gas lines and water lines provide for a dangerous situation where fires result from the broken gas lines, but fire-fighting efforts are severely hampered because the water lines are also broken. Duration is an important factor because for many structures the longer the shaking occurs, the weaker the structures become (Berlin, 1980).
Sediment type contributes largely to the extent of damage to buildings that results from shaking. Shaking is amplified greatly on unconsolidated surficial deposits or nonengineered fill, especially when moist. An example of this is seen in an M=7 earthquake that occurred in Varto, Turkey on August 16, 1966. In one area, twelve out of fourteen houses built on old channel deposits collapsed, while buildings that had been built on the adjacent terraces suffered only minor damage. A water table close to the ground surface also contributes to instability (Berlin, 1980).
Because of the many variables associated with earthquakes, it is usually difficult to determine how much damage will result from shaking. One opinion, though, is that earthquakes with a local magnitude of 5 or larger will result in damage to weak structures (Housner, 1970, in Berlin, 1980). An earthquake with a magnitude of less than 5 has a lower damage potential because of its short duration and only moderate ground motion acceleration (Berlin, 1980). From this standpoint, Walla Walla could have a fairly high damage potential as it has been estimated that an earthquake with a magnitude of about 6.5 could occur in the region (Miklancic, 1989 in VanAuken, 1997).
The magnitude of shaking in an area is also dependent on the type of fault involved in the earthquake, and the location of the area in question in relation to the fault. In the case of a strike-slip fault, the shaking intensity is likely to be similar on either side of the fault. In the case of a thrust fault the ground motion will probably be more intense on the upper side of the fault. According to Nason’s elastic strain release mechanism, the upper side of a thrust fault is unconfined, resulting in more elastic strain (Berlin, 1980).
On the geologic map of the Walla Walla quadrangle, Schuster (1994) mapped the Wallula Fault as a thrust fault with the southern side of the fault being the upthrown side. Thus, it is reasonable to think that the southern side will experience a larger degree of shaking in an earthquake that occurs on this fault. Schuster also mapped some normal faults in the southeastern part of the quadrangle, one of which is the Hite Fault and two of which strike north-south. The east side of all of these faults is the upthrown side, and may thus experience the most shaking if an earthquake were to occur on any of these. There are no strike-slip faults present on this map (Schuster, 1994), although horizontal slickenlines can be seen along the Wallula Fault in Wallula Gap suggesting that the latest movement was strike-slip.
Differential settlement occurs when a structure settles at different rates so that one part may be settling into the ground at a faster rate than another section of the same structure. This phenomena often occurs in sediments of varying thickness or consolidation (Berlin, 1980), and has the largest impact on rigid structures (Mathewson, 1981).
Factors that influence differential settling include water content, sediment thickness, and the consolidation of the sediments.
Water content has a large influence on the settling capacity of soils as water has the ability to completely change the phase of a soil from a solid to a liquid. The amount of water that is required to change the phase of a sediment is measured in terms of a liquid limit (LL) and a plastic limit (PL) for the sediment.
The liquid limit of a soil is defined as "the moisture content at which a soil passes from a plastic to a liquid state (Rahn, 1986)." At the liquid limit, all cohesion is lost. The liquid limit is a measurement that is in terms of the moisture content as a percentage of the weight of the dry soil. Soils with a high clay content generally have high liquid limits, indicating that the soil demonstrates a high degree of cohesion. A high liquid limit is indicative of high cohesion. A high liquid limit also often indicates a low load-carrying capacity.
The plastic limit of a soil is defined as "the moisture content at which a soil changes from a semi-solid to a plastic state (Rahn, 1986)." An example of a plastic state in clayey soils is when the soil can be rolled into threads 3 millimeters in diameter. Soils that cannot roll into these threads without breaking at any moisture content are described as being nonplastic. This may occur in silt or sand soils containing little or no clay. A high plastic limit in a soil is often a good indication that the soil contains large amounts of clay and silt-sized particles, although clay is the component that enables a soil to have a plastic phase (Rahn, 1986).
A method of determining the moisture content range in which a soil can remain in a plastic state is with the plasticity index (PI). The plasticity index is defined as the difference between the liquid limit and the plastic limit (PI= LL-PL) (Rahn, 1986). A small plasticity index indicates a small plastic phase, which means that the soil is able to quickly change from a semisolid to a liquid with only a small amount of added moisture. Rahn (1986) states that "this is an undesirable condition for a foundational material". Rahn (1986) uses 5% as an example of a small plasticity index. From an engineering standpoint, siltlike soils tend to be poorer than claylike soils based, in general, on the moderate to high liquid limits and moderate to low plasticity index of siltlike soils (Rahn, 1986). Walla Walla County has many soils with a PI of 5% and less (U.S. Dept. of Agriculture, 1964), indicating that many of the soils within the county do not act as a good foundation material for structures, especially during times of high precipitation, such as in the winter months.
Additionally, the load-carrying capacity of soils is often directly related to the moisture content. The plastic limit is a critical point, because the load-carrying capacity quickly decreases above the plastic limit, and quickly increases below the plastic limit (Rahn, 1986) further suggesting that Walla Walla soils are poor engineering soils during times of high precipitation.
Compaction refers to the degree that the density of a soil can increase to its maximum density in connection with a decrease in volume and settlement of sediment. The most important factors that control the maximum density that result from soil compaction are: 1. moisture content, 2. nature of the soil (i.e. texture, gradation, and other physical properties), 3. type and amount of the compactive effort (Rahn, 1986).
Compaction most commonly occurs in loosely consolidated or unconsolidated sediments that are subjected to seismic vibrations. Compaction of unconsolidated sediments occurred in many areas affected by the 1964 earthquake in Alaska. In particular, the Homer Spit on the Kenai Peninsula dropped a total of 1.4 meters. Tectonic subsidence was also a factor in this case and contributed to 0.6 meters of the drop, but the remaining 0.8 meters was due to the compaction of loosely consolidated gravel (Berlin, 1980).
Another example of compaction occurred in Peru’s 1970 earthquake. Alluvial deposits in the town, Chimote, were compacted 1 meter.
Collapsing soils are another contributing factor to differential settlement. The volume of an accumulation of collapsing soils will abruptly decrease when saturated with water. These soils are normally composed of clay, silt, and sand deposited at the base of mountains and are often found in semiarid to arid environments where they are dry or only slightly saturated. Under these conditions they do not present a problem, but as soon as the moisture content increases to the point where they are saturated, they "collapse" (Mathewson, 1981).
Sediment thickness is another factor of differential settlement. Differential settlement is most likely to occur in a thick accumulation of soils, containing a variety of soils each having different thickness. The thicker an accumulation of soils is, the more it can be compacted, and the more a structure will settle. For instance, if a 5 inch layer of unconsolidated material were to be put in a bucket and then compacted, the thickness would only be able to decrease a small amount because the original thickness was small. On the other hand, if the bucket is filled to the top with the same material and then compacted, the thickness could decrease to a greater extent because the original layer was of a greater thickness.
Soil compaction in Walla Walla County
The main types of sediment for Walla Walla County that have been studied in this report are alluvium, loess, and Touchet Beds. The only exposed bedrock in the county is Columbia River basalt.
Based on studies done by Bob Carson and Mackey Smith in Washington’s Puget Lowland in the 1970’s, the most compressible soils in Walla Walla County are alluvium, especially in the Walla Walla River delta, soils of slight compressibility include loess and Touchet beds, and basalt is noncompressible.
In this study, compressibility maps have been constructed based on the above classifications (figures 3 and 4). These maps were constructed using Schuster’s (1994) geologic map for the Walla Walla quadrangle as a base and color coding the geologic units based on their compressibility characteristics. This classification for compressibility is adapted from the studies conducted by Bob Carson and Mackey Smith in the Puget Lowland. Green indicates that bedrock is close to, or at the surface and that these areas are noncompressible. The yellow areas consist of slightly compressible loess and some Touchet beds. The red areas contain alluvium. The compressibility of alluvium is highly variable, but has the highest compressibility potential. Alluvium that is fine-grained and/or contains organic matter is more compressible than alluvium consisting mainly of gravel. Compressibility also increases with increasing water content.
Figure 3. Compressibility map for the cities of Walla Walla and College Place and the surrounding area (geologic base map from Schuster, 1994).
Figure 4. Compressibility map for Walla Walla River delta and surrounding area (geologic base map from Schuster, 1994).
From these maps, it can be seen that the cities of Walla Walla and College Place mainly exist on alluvial soils of variable compressibilities and to a lesser degree on the slightly compressible Touchet Beds. The fact that these cities were built on alluvial soils of variable compressibilities suggests that in the event of a large earthquake, structures in both Walla Walla and College Place may experience some degree of settling or differential settlement depending on the thickness, organic content, grain size, and water content of these sediments. If the sediment thickness is thin, shaking and compaction will be less amplified, and the foundations of structures may be anchored in the underlying basalt therefore preventing settlement. According to the Walla Walla Soil Survey (1964), though, the soils in these areas that have been studied generally range in thickness from 6 to more than 20 feet. Thus, it is unlikely that many structures are firmly anchored in the bedrock and could potentially undergo settlement or differential settlement in a large earthquake.
The area that may experience the most soil compaction is the Walla Walla River delta, as mentioned above. This is mainly due to the fact that not only does the delta consist of alluvial soils, but also that it is saturated with water from the river. The high water content makes it especially susceptible to compaction. Additionally, this area is very near the Wallula Fault System, so shaking would be intense in this area if an earthquake occurred on this fault. Fortunately this area has been designated as a wildlife refuge, so no one is allowed to build here.
Fault displacement simply refers to the amount that the ground is displaced along a fault in the event of an earthquake. In the case of a strike-slip fault, fault displacement may be easily seen and measured if a road or stream that was thought to be originally straight and running perpendicular to the fault is displaced. In the case of a dip-slip fault, faceted spurs may develop on hillsides, such as is the case near Wallula Gap along the WFZ.
Fault displacement becomes hazardous when a structure lies across the fault, and displacement along the fault causes the structure to become deformed or break. The damage that results from fault displacement depends on: 1. the type (strike-slip, reverse, normal) and amount of displacement and 2. the type of structure involved and the condition it is in. In general, the amount of damage increases as the amount of displacement increases, and engineered buildings tend to sustain less damage then nonengineered buildings (Berlin, 1980). Buildings engineered to sustain earthquake damage often have concrete walls reinforced with rebar and/or well-braced wood frames with shear walls.
Often when faults rupture, other cracks in the ground form besides the main fault. Smaller branch and secondary units often accompany the main fault and can also show displacement. Nontectonic (or auxiliary) ground cracks or fissures also sometimes form by tensional or compressional forces that are approximately perpendicular to the primary fault. Damage can result if structures cross any of these ground cracks (Berlin, 1980).
An example of the difference in damage done to buildings within and outside of a fault zone was seen in the earthquake that occurred in San Fernando, California in 1971. In that earthquake, 80% of the houses within the fault zone were moderately to greatly damaged whereas only 30% of the houses nearby on either side of the fault zone were damaged to the same extent (Berlin, 1980).
Berlin describes the way in which building type effects the intensity of the damage that can result from fault displacement with the 1906 San Francisco earthquake and the 1972 Managua, Nicaragua earthquakes. He describes a wood-frame house in the San Francisco earthquake that straddled the San Andreas Fault. The fault moved about 2 meters and the house was ripped in half, but neither half collapsed and none of the inhabitants inside were hurt. On the other hand, in the Managua earthquake, the fault displacement was smaller than that in the San Francisco earthquake, but the unreinforced concrete block and old tarquezal (wood and frame adobe) houses that were on the fault completely shattered and collapsed (Berlin, 1980). Berlin did not mention the casualties, but it is safe to assume that there were many.
Other damage that occurred in the earthquakes above due to fault displacement included cracking, buckling, and displacement of roads and sidewalks, broken pipelines and disrupted underground utilities, and blocked streets. In some cases, the broken water pipelines prevented effective fire fighting.
In Walla Walla County, a couple of structures above the ground that may be severely affected if a large earthquake were to occur along the WFZ would be highway 12, as it crosses the WFZ at Wallula Gap, and the Boise Cascade Paper Mill, as it is very near the WFZ. The paper mill is a highly engineered structure which may be able to sustain a fairly high amount of shaking, but if the fault was to be displaced to a large degree as it could be if a magnitude 6.5 earthquake were to occur, the paper mill may undergo significant damage.
Landslides have the potential to be a widespread hazard as they are more commonly induced by shaking rather than by faulting. Depending on other factors involved in landslides, just a small amount of shaking could induce damaging landslides. This occurred in Hollister, California in 1972. A series of small earthquakes in this region, the largest having a magnitude of 5, induced damaging landslides (Berlin, 1980). Landslides have the potential to occur outside of an earthquake’s epicentral region and may occur months after an earthquake increases slope instability (Berlin, 1980). Factors include slope steepness, water content, and soil type.
Common types of landslides include debris slides, earthflows, slumps, and rockfalls. Slumps and earthflows generally occur on gentler slopes whereas rockfalls and debris slides generally occur in mountainous terrains where slopes are steeper. Slumps and earthflows are the more common type of landslides in urbanized areas (Berlin, 1980).
Slumps and earthflows are also often induced by liquefaction where the soil is clay free, granular, and water saturated and has undergone seismic shaking (Berlin, 1980).
Loess landslides may be of particular concern to the Walla Walla region as loess is abundant in the region, and as there is historical documentation of devastating, earthquake induced, loess landslides in China.
An earthquake with a magnitude of 8.5 occurred near Ganyan Chi, Haiyualn Count of the Ningxia Hui Autonomous Region, China on December 16, 1920. The shaking lasted for ten minutes and resulted in the collapse of 500,000 houses and cave dwellings, and 234,117 deaths. Many more houses and people were damaged and wounded. Dense areas of loess landslides affected more than 4,000 km2 of area resulting in blocked roads and buried farmlands and villages. This further resulted in quite an economic sting for the region (Zhang and Lanmin, 1994).
The occurrence and degree of damage of loess landslides is not only dependent on the degree of shaking, but also is largely dependent on the thickness of the loess. In a mountainous area affected by this earthquake, the magnitude of shaking was very high, but due to relatively thin loess coverage, the density of the landslides was low and individual slides were small. On the other hand, another area that experienced a lower magnitude of shaking, but had a hilly topography with layers of loess ranging from 20 to 50 meters in thickness experienced about 650 landslides (Zhang and Lanmin, 1994).
In their 1994 study, Zhang and Lanmin concluded that low water content, low slope angles, and high sliding speeds are main characteristics of seismically induced loess landslides. Zhang and Lanmin also found that for one of the larger landslides that occurred, the Huihui Chuan Landslide, the original slope was stable before the earthquake occurred (Zhang and Lanmin, 1994). Thus slopes that are stable under normal conditions and appear to be safe may be very dangerous in an earthquake.
Zhang and Lanmin (1994) also state that loess is susceptible to subsidence from seismic conditions due to its macroporous structure.
Potential landslide sites in Walla Walla County
Steep slopes can be seen on topographic maps where the topographic lines are close together, indicating a change in elevation over a relatively short distance. Steep slopes are present in the southeast corner of the county in the Kooskooskie area in the foothills of the Blue Mountains. The southwest corner of the county also contains steep slopes along the Columbia River, near where Spring Gulch intersects the Columbia just south of Wallula Junction. Steep slopes are also seen to a lesser degree in Vansycle Canyon east of Spring Gulch. The valley walls of the Snake River are also very steep in the northwest corner of the county. These areas mainly have basalt as the bedrock with overlying thin patches of Missoula Flood deposits (Touchet beds). The steep slopes are due to the walls of rigid Columbia River basalt beds. If landslides occur in these areas due to seismic shaking, rock falls may be the prevalent type of landslide. Along the Snake River where Touchet beds are abundant, though, slumps and flows may be more common, especially if the beds are water saturated.
A geologic map for the Walla Walla quadrangle (Schuster, 1994) illustrates areas where faults occur, which could result in landslides due to strong seismic shaking. The Wallula Fault Zone strikes northwest-southeast across Wallula Gap. Secondary normal faulting extending from the Wallula Fault Zone occurs within Vansycle Canyon and near Spring Gulch where steep slopes exist. A normal fault striking directly north-south passes through Kooskooskie to just south of Wilson Creek. This fault extends across loess beds that occur from Blue Creek north to Wilson Creek. A short, concealed normal fault striking NNW-SSE is present across Mill Creek a few miles east of McNary National Wildlife Refuge. This fault mainly extends across loess. The Hite Fault is in the very southeast corner of the county in the Umatilla National Forest where many steep slopes occur due to the presence of basalt. Depending on whether or not the Hite Fault is still active, there may or may not be a fairly high potential for rock falls in this area. An extensive amount of loess occurs in the central part of the county. This area may be particularly susceptible to loess failures blocking roads and damaging homes and farms.
Seiches are not a major hazard as they are only rarely associated with earthquakes. However, as the Walla Walla region does contain reservoirs upon which seiches could occur, they do deserve some attention.
Berlin (1980) defines seiches as "oscillating waves set up on surface water bodies such as lakes, bays, fjords, and rivers (p. 193)." Their heights are usually less than 2 meters. Factors controlling the distribution of seiches include the thickness of low-rigidity sediments (sediments that do not resist deformation when stress is applied), varied thickness of these sediments, and surrounding major tectonic features (thrust faults, basins, domes, etc.) that have been found to control seismic waves (Berlin, 1980).
Damage that can result from seiches includes flooding, sudden rises in water level, and recessions followed by surges that can damage boats, docks, and other structures that may be present on the shore.
One area in particular that may show an occurrence of seiches during an earthquake is Lake Wallula. Lake Wallula is the reservoir that formed upon the building of McNary Dam near Umatilla, Oregon. The pooled water reaches 103 km upstream to the vicinity of the Tri-Cities, Washington. Lake Wallula has also affected the Snake River as it extends almost 16 kilometers up the Snake River. The total amount of shoreline that surrounds Lake Wallula is about 400 km (USACE, 1974 in VanAuken, 1998). Thus, Lake Wallula is quite extensive, affecting a large area where seismic seiche activity could potentially damage structures along the shoreline in a major earthquake.
Walla Walla County is an area that is prone to earthquakes. Evidence for this exists in the geomorphic features such as faceted spurs, as well as documentation on historic earthquakes. The location of the magnitude 6 earthquake that occurred in 1936 in Stateline, Oregon, is one area in particular that has experienced numerous earthquakes.
As Walla Walla County may be prone to earthquakes, earthquake hazards do exist and it is important that the county be aware of these hazards. Even though these hazards are present, they probably do not pose as great a danger as in other earthquake prone areas because of Walla Walla County’s low population, probable low magnitude earthquakes, and a probable long recurrence interval.
I would like to thank my advisor, Bob Carson, for his help and guidance in researching this subject and writing this paper. I also thank Walla Walla County’s Project Impact Coordinator, Don Marlatt, and Connie Kruegar with Walla Walla County Regional Planning, for presenting me with the idea for this project.
Figure 5. Peak ground acceleration for the Pacific Northwest
Figure 6. Peak ground acceleration for the continental United States.
Reidel, S.P., and T.L. Tolan, 1994, Late Cenozoic structure and correlation to seismicity along the Olympic-Wallowa lineament, northwestern United States: Discussion and reply: Geological Society of America Bulletin, v. 106, p. 1634-1638.
Schuster, J.E., 1994, Geologic map of the Walla Walla 1:100,000 quadrangle, Washington: Washington DGER, open file report 94-3.
Schuster, J.E., C.W. Gulick, S.P. Reidel, K.R. Fecht, and Stephanie Zurenko, 1997, Geologic map of Washington- southeast quadrant: Washington DGER.
United States Army Corps of Engineers, Walla Walla District, 1974, McNary: Draft environmental impact statement: Walla Walla, Washington, 275 p.
Woodward-Clyde Consultants, 1980, Seismological review of the July 16, 1936 Milton-Freewater earthquake source region: San Francisco, California, prepared for Washington Public Power Supply System, 44 p.
Van Auken, Heidi, 1998, The changing Walla Walla river: a 200 year perspective, with emphasis on inundation due to the construction of McNary Dam on the Columbia River: unpublished thesis, Walla Walla, Washington, Whitman College, 49 p.
Zhang, Zhenzhong, Wang Lanmin, 1995, Geological disasters in loess areas during the 1920 Haiyuan earthquake, China: GeoJournal, v. 36, p. 269-274.
**Ground acceleration maps found at the USGS National Seismic Hazard Mapping Project webpage, 1997: www:http//geohazards.cr.usgs.gov/eq/ in a PowerPoint presentation given by John Winter’s Environmental Geology class, Whitman College, 1997.
Personal communication: Bob Carson, 2000