Scientific Investigations Report 2007–5178
U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2007–5178
The first major turbidity event recorded in the North Santiam River basin occurred in December 1998, 3 months after the first water-quality monitors were installed in the upper basin. Heavy precipitation during December 27–29 brought 7 to 12 in. of rain to the basin (Oregon Climate Service, 2006). The heavy rain increased streamflow to greater than 2- to 5-year levels (Cooper, 2005), with 8,210 ft3/s measured at the North Santiam monitoring station, 6,940 ft3/s at Breitenbush, and 3,780 ft3/s at Blowout. Downstream of Detroit Lake, in the lower basin, streamflow reached 18,700 ft3/s at the Little North monitoring station. Turbidity increased at each of the three upper basin water-quality monitoring stations, as sediment mobilized from the landscape, with values reaching 198 FNU at North Santiam, 1,050 FNU at Breitenbush, and 874 FNU at Blowout (fig. 7). At the time, the only turbidity value available in the lower basin was 195 NTU recorded at Geren Island (Hank Wujcik, City of Salem Public Works Department, written commun., 2005). Aside from streamflow at the Little North Santiam River station, no other USGS water-quality data were available downstream of Detroit Lake because lower basin USGS water-quality monitoring stations were not established until April 2000.
During the 3-day storm, thousands of tons of sediment were transported by each upper North Santiam River basin stream (table 4). The largest suspended-sediment load transported was the calculated 17,300 T of material that passed the Breitenbush River station, a yield of 160 ton/mi2. Blowout Creek carried slightly less material at 14,100 tons, but had more than three times the yield at 540 tons/mi2. This major turbidity event generated the greatest single-event suspended-sediment load transported by Blowout Creek during the entire period of record (October 1998 to September 2004), and the second greatest single-event suspended-sediment yield of any station.
Along with suspended-sediment loads, clay-water volumes were calculated for each stream during the December 1998 storm (table 4). Responding in a similar way as it did to sediment load, the Breitenbush River also carried the largest volume of clay-water of all upper basin streams at 7,020 Mgal. Once again, Blowout Creek transported less clay-water than the larger Breitenbush River; however, similar to suspended-sediment yield, Blowout Creek produced the greatest clay-water yield of the three upper basin stations with 98 Mgal/mi2. Only the Little North Santiam River subbasin had a potentially higher clay-water yield (136 Mgal/mi2); however, the yield was computed using streamflow to estimate turbidity because water-quality data were not yet collected. Therefore, the estimate was different than if it had been calculated using continuous instream turbidity measurements, similar to the other upper basin calculations.
This 3-day event was an indicator of the influence that Blowout Creek has on the North Santiam River basin system. Not only was the Blowout Creek subbasin a major sediment contributor, but it also had the potential for supplying a large amount of clay-size material—the same type of material that can be problematic to the water-treatment facility. Although the settling effect of Detroit Lake is an impediment to the transport of clay-water to the lower basin, during extremely high flows, clay-rich, persistently turbid water can be released downstream through the dam, as in February 1996 (Bates and others, 1998; Hulse and others, 2002).
After the December storm, field reconnaissance was conducted to determine potential source areas for the increased turbidity and sediment loads in the upper North Santiam River basin. Although no specific source areas were identified for the Breitenbush River or North Santiam River loads, two potential sources were located for the Blowout Creek subbasin (fig. 8). On the basis of USFS (Douglas Shank, U.S. Forest Service, oral commun., 2005; David Klug, U.S. Forest Service, oral commun., 2005) and USGS field observations, the two sources were the Blowout Creek and Divide Creek earthflows (figs. 9 and 10). Although not the only earthflows in the subbasin, both are large and perennial sources of sediment and clay (David Klug, U.S. Forest Service, oral commun., 2005). According to soil surveys of the area (U.S. Forest Service, 1992), the Divide Creek earthflow was considered moderately stable to unstable, with moderate to severe surface soil erosion potential, a high yield potential for silt to clay sediment, and high water detention storage capacity. In contrast, the Blowout Creek earthflow was described as mostly stable but with severe soil erosion and high runoff potential, and a low silt-clay yield. Both earthflows are part of Holocene and Pleistocene-age landslide and debris flow deposits (Walker and MacLeod, 1991). Although there had been some tree harvesting in the vicinity, the earthflows were mostly forested, with some barren land and shrub coverage (U.S. Geological Survey, 2005). The bulk of the sediment, clay load, and related turbidity was likely caused by material from one or both of these locations, with some additional contribution from upstream areas.
Since the inception of the study, Blowout Creek has become one of the most studied subbasins in the North Santiam River basin. Although the stream drains a small area, it often has high turbidity during precipitation events. In order to determine whether certain sources supply most of the sediment, temporary water-quality monitors were installed during several periods at different locations within the subbasin. For example, water-quality monitors were placed downstream of the Divide Creek earthflow, upstream of the Blowout Creek earthflow, downstream of the Blowout Creek water-quality monitor, and near the mouth of Cliff Creek (a tributary to Blowout Creek). To date, data from these additional monitors has not conclusively linked subsequent episodes of increased turbidity solely to these specific source areas. Along with monitoring water quality, attempts at monitoring both Blowout and Divide Creek earthflows also have begun. As part of the preliminary investigation, sediment samples have been collected from each earthflow. An inclinometer and calibrated stake-arrays also have been placed at the Blowout Creek earthflow to measure movement.
The next two major turbidity events occurred a year later, but in separate subbasins. In late November 1999, a large storm system in the North Santiam River basin increased streamflows to 5- to 50-year recurrence intervals (Cooper, 2005). The upper North Santiam River had a 5- to 10-year event, Breitenbush River a 10- to 25-year event, and the Little North Santiam River, the highest, with a 25- to 50-year recurrence interval. According to rainfall records at Detroit Dam, almost 9 in. of rain fell during the first 2 days of the storm—November 25–26—with an additional 0.5 in. falling on November 27 (Oregon Climate Service, 2006). Streamflow at the North Santiam River monitoring station reached 13,300 ft3/s, while streamflow at the Breitenbush and Blowout stations reached 11,200 ft3/s and 2,670 ft3/s, respectively. The highest flow recorded at Little North monitoring station was 28,900 ft3/s . During the storm, rivers became extremely turbid. For example, turbidity at North Santiam reached 739 FNU, while reaching 1,160 and 1,310 FNU at the Breitenbush and Blowout stations, respectively (fig. 11). Although discharge was recorded at the Little North monitoring station, turbidity was not recorded because the water-quality monitor was not yet installed.
As with the first major turbidity event, the heavy rainfall increased turbidity and produced high suspended-sediment loads (table 5). The November 1999 major turbidity event for the Breitenbush River was the largest single-event load ever calculated for any station during the study. The 69,800 tons of suspended sediment transported in the river during this event comprised 93 percent of the total annual sediment load for water year 2000 (September 1999–October 2000). Unlike the massive load transported by the Breitenbush River, Blowout Creek moved less than one-tenth of that at 6,060 tons (table 5). This 3-day major turbidity event transported 81 percent of the total annual suspended-sediment load for the Blowout Creek subbasin. Compared to the December 1998 Blowout Creek event, this November 1999 Blowout Creek event transported only about one-half the load, even though it had a higher maximum turbidity (1,310 FNU) than the previous event (874 FNU).
Along with a large suspended-sediment load, the Breitenbush River transported greater than 90 percent of its annual clay-water volume during the November 1999 event (table 5). The clay-water volume reached 9,110 Mgal for the 3-day period, about 98 percent of the annual total. In comparison with December 1998, more than four times the material was displaced and transported from the Breitenbush subbasin during this event. However, the clay-fraction increase was marginal, as the clay-water volume was greater by less than 25 percent. This small difference suggests that the source area supplying the sediment in 1998 was more clay-rich than the source for the 1999 major turbidity event.
Clay-water volume in Blowout Creek during this event was 1,690 Mgal, which equated to about 92 percent of the annual clay-water volume for the station (table 5). This volume was less than that in December 1998, and the least percentage of annual volume of any of the upper basin stations during the November 1999 storm. In addition, although clay-water yield increased in all other subbasins as compared to the yield of the 1998 storm, the Blowout Creek subbasin yield actually decreased by 33 percent to 65 Mgal/mi2. The decrease in clay-water volume likely was caused by the short-term antecedent conditions in the subbasin as accumulated, easily erodible material was removed the previous winter.
As in 1998, the USGS and USFS made field observations to locate source areas that might have supplied sediment to streams in the upper basin. During the investigation, no specific source area was discovered for the turbidity in the upper North Santiam River. However, at least one major landslide that contributed to the sediment load was discovered in the Breitenbush subbasin. In the Blowout Creek subbasin, the Divide or Blowout Creek earthflows again were the likely sources for high turbidity values, as they were in 1998. The landslide in the Breitenbush River subbasin was a recently clear-cut area near East Humbug Creek, a small tributary to the Breitenbush River (fig. 12). The unstable area encompassed 12 acres (fig. 13) and was located 8 mi upstream of the Breitenbush River water-quality monitor. The landslide area was moderately to severely susceptible to surface erosion and had a high potential for supplying silt and clay sediment (U.S. Forest Service, 1992). This landslide was likely an example of older basaltic lava flows and breccias overlying weaker Continental sedimentary and volcaniclastic rocks (Walker and MacLeod, 1991). The landslide, though too small to supply all material that passed the Breitenbush monitoring station, likely was responsible for a portion of the estimated 69,800 tons of material that passed during the 3-day period (David Klug, U.S. Forest Service, oral commun., 2005). No other landslides or potential source areas have yet been discovered in the Breitenbush River subbasin and dated to correspond with this November 1999 major turbidity event.
A previous assessment of the known Breitenbush River subbasin landslide (U.S. Forest Service, [2001]) described the East Humbug Unit 5 Stand as a slide risk. However, trees were harvested in the mid-1990s, leaving the slope susceptible to accelerated surface erosion and landsliding. Before sliding in 1999, the first evidence of movement from this location occurred in 1996 (U.S. Forest Service, [2001]). Since the landslide of November 1999, stabilization efforts such as dewatering by overland drainage pipes, log terracing, and planting of saplings and grasses were completed to limit future mobilization (David Klug, U.S. Forest Service, oral commun., 2005; written commun., 2006). During 2004–05, a temporary water-quality monitor was installed at a decommissioned USGS streamflow-gaging station on East Humbug Creek. During its deployment, data showed instances of increased turbidity in East Humbug Creek, likely related to runoff from the landslide area; therefore, detectable sediment input from the upper East Humbug Creek watershed likely was still occurring.
In late September and early October 2000, unseasonably high temperatures in the North Santiam River basin remained in the mid- to upper 70s (Oregon Climate Service, 2006). The warm temperatures combined with a minor rainfall event (about an inch of rain) to expedite melt of the High Cascade snowpack. On October 1, streamflow at the North Santiam monitoring station increased slightly but remained less than 600 ft3/s—normal for early October. Associated with the slight increase in streamflow was an anomalously high turbidity value estimated at 3,520 FNU (Bragg and others, 2007). This particularly large spike far exceeded any other measurement in the upper basin (fig. 14). The only other monitoring station to record a change in turbidity was Breitenbush, where two elevated but short-lived turbidity measurements of 86 and 76 FNU occurred 10 hours apart. In contrast to previously mentioned events, this major turbidity event had a higher magnitude, but shorter duration. In this case, the turbidity remained greater than 100 FNU for less than 17 hours, rather than for multiple days. Besides North Santiam and Breitenbush, none of the other monitoring stations displayed noticeable changes in turbidity.
On October 1, 2000, 1,580 tons of sediment passed the North Santiam monitoring station; far exceeding the amount measured at all other stations (table 6). Because water year 2001 was a drought year, this single-day upper North Santiam River high-turbidity event was responsible for 42 percent of the annual sediment load at that station. Although the amount of material transported during this event was less than that transported during the December 1998 and November 1999 storms, the unusual seasonality, source, and magnitude of this event was noteworthy. Because this event was isolated to the upper North Santiam subbasin, it was the only stream to contain clay-water on that day. The single-day event produced a volume of 287 Mgal, hundreds of millions of gallons less than the volume during events of precipitation-driven erosion in previous examples. The North Santiam River transported 79 percent of its annual clay-water volume during this single day, because drought conditions that year eliminated large winter storms. Clay-water volume at that station would very unlikely equal such a high percentage during normal climate conditions; previous event examples for this river had volumes in the thousands of millon gallons (4,710 and 8,850 Mgal, respectively). In addition, because this event occurred during low-flow conditions, material settled before being passed through Detroit Lake.
At the time of the October 2000 major turbidity event, little was known about sediment sources in the upper North Santiam subbasin, except that the USFS observed extremely turbid water coming from Pamelia Creek as it entered the North Santiam River at Highway 22, near Marion Forks (fig. 15). Investigation of a later event (see “October 21, 2003—North Santiam River” section of this report) showed that turbidity in Pamelia Creek likely came from Milk Creek. Milk Creek flows from the Milk Creek Glacier on the western slopes of Mt. Jefferson (fig. 16). The Milk Creek watershed contains diverse environmental settings—some that are very stable with slight to moderate soil erosion potential and low silt or clay yield, and others that are only moderately stable, with severe erosion and moderate silt or clay yield (U.S. Forest Service, 1992). The watershed is mostly steep and contains basaltic and andesitic rocks with highly erodable glacial and glaciofluvial deposits of Pleistocene age (Walker and MacLeod, 1991). Land cover in the upper watershed is limited, as the landscape is mostly barren and glaciated, whereas shrubs and forests grow at lower altitudes (U.S. Geological Survey, 2005). The major turbidity event in October 2000 likely represented an episode of rapid melting in the area of the Milk Creek Glacier that carried glacial deposits downstream.
In March 2003, an automatic pumping sampler was installed at the North Santiam monitoring station. The pumping sampler was programmed to collect 1-L samples every 120 minutes when instream turbidity values exceeded a predetermined threshold (for example, greater than 20 FNU). This approach allowed for more timely samples than could be collected by field personnel. In addition, during the last 5 years, field investigations have attempted to locate potential sediment sources that supply material to Pamelia and Milk Creeks. For example, sediment samples from a variety of Cascadian sources from the upper watershed have been collected, with the aim of determining connections between specific source areas and suspended sediment collected in water-quality samples.
Starting in late November and extending into early December 2001, rain fell almost daily on the North Santiam River basin. According the USGS rain gage at Blowout Creek, a local storm cell during December 13–14 produced nearly 4 in. of rain. Two days later, a new weather front delivered an additional 1-2 in. of rain, followed by a third storm on the morning of December 17 that added a final 1-2 in. of rain. During those last few weeks of 2001, 25 in. of rain fell on the upper basin, with more than 10 in. during just 8 days (Oregon Climate Service, 2006). The heavy rains increased streamflow at the North Santiam monitoring station to 4,120 ft3/s, whereas at the Breitenbush, and Blowout, and French monitoring stations, streamflow reached 2,870 ft3/s, 1,180 ft3/s, and 565 ft3/s, respectively (fig. 17). In the lower basin, streamflow at the Little North monitoring station reached 9,930 ft3/s. Instantaneous turbidity records (fig. 17) from the water-quality network during mid-December showed turbidity spikes greater than 20 FNU at the upper-basin monitoring stations—North Santiam at 22 FNU, French at 43 FNU, Breitenbush at 37 FNU, and Blowout estimated at 3,400 FNU (Bragg and others, 2007). Turbidity levels were much lower downstream of Detroit Lake (12 FNU on the North Santiam River at Niagara) but increased downstream (at Mehama and Geren Island, 55 and 30 FNU, respectively) because of input from tributaries like the Little North Santiam River (55 FNU).
By mid-December, consistent rain eroded surfaces and mobilized landslides, causing turbidity to increase at each station. Associated with the elevated turbidity was an increased sediment load for each location (table 7). The largest suspended-sediment load during the storm was an estimated 4,900 tons transported past the Blowout monitoring station, which was the third largest event load for Blowout Creek during the study period. The Blowout Creek subbasin was the only subbasin to transport a large suspended-sediment load, about 36 percent of the annual load, a yield of 190 tons/mi2. Similarly, Blowout Creek was the only stream to transport clay-water. Over the 3-day period, 522 Mgal of clay-water issued from Blowout Creek into Detroit Lake, a yield of 20 Mgal/mi2. That was about 17 percent of the total annual clay-water volume for the subbasin that year. Clay yield in the December 2001 event was considerably less than clay yield during previous Blowout Creek events because the source this time was a landslide rather than the earthflows.
Sampling during the storm revealed that the high turbidity in Blowout Creek originated from Ivy Creek, an upstream tributary. The muddy water was traced back to a landslide and associated road failure along Forest Road 1012, which released material down an unnamed channel into Ivy Creek (fig. 18). The damage to Forest Road 1012, including the exposed scarp and road-grade deformation, is shown in figure 19. Based on calculations for a half-ellipsoid-shaped landslide (Cruden and Varnes, 1996), an estimated 178 yd3 of material slipped from the roadbed. On the basis of fluvial-sediment discharge computations (Porterfield, 1972), roughly 377 tons of saturated soil and debris was mobilized and lost from the road failure. As the saturated material traveled downhill, it created a debris flow, gaining considerably more mass as material was eroded from the 1,300-ft channel before entering the stream (fig. 20). After the sediment and debris entered Ivy Creek, it continued downstream into Blowout Creek, eventually passing the USGS water-quality monitor. Because the water in Blowout Creek was clear upstream of the turbid Ivy Creek confluence, most of the turbidity registered at the Blowout monitoring station came from this landslide and associated debris flow.
In general, slopes in the Ivy Creek drainage basin are stable, with moderate soil erosion and runoff potential (U.S. Forest Service, 1992). The sedimentary and volcaniclastic geology (Walker and MacLeod, 1991) and dense evergreen land cover (U.S. Geological Survey, 2005) also help stabilize the area. However, intense and prolonged rainfall in 2001 weakened the surface and caused sliding. Since 2001, the road-failure location has been surveyed multiple times to estimate feature size and sediment displacement. Following the original failure, subsequent movement has widened the scarp (fig. 21). Recently, this section of Forest Road 1012 was decommissioned and then obliterated, and the fill material was removed (David Klug, U.S. Forest Service, written commun., 2006).
From January 29 to February 2, 2003, as much as 3 to 5 in. of rain fell on the North Santiam River basin, depending on locality, according to measurements at Detroit Dam (Oregon Climate Service, 2006) and the USGS Blowout Creek rain gage. The heavy rainfall increased streamflow at the North Santiam monitoring station to 6,670 ft3/s, Breitenbush to 5,270 ft3/s, French to 1,470 ft3/s, and Blowout to 1,240 ft3/s. In the lower basin, streamflow increased at the Little North monitoring station to 9,900 ft3/s. In correspondence with the increased streamflow, turbidity at the North Santiam monitoring station reached 86 FNU, Breitenbush reached 162 FNU, and Blowout reached 74 FNU (fig. 22). On January 31 at 06:30 a.m., the water-quality monitor at French Creek recorded the highest turbidity value of any station, a spike of 1,530 FNU. The turbidity probably rose higher, but because of the sudden onset and short duration of the event, along with the instrument reaching its maximum measurement capacity, the actual peak is unknown. Downstream of Detroit Dam, turbidity was high along the lower North Santiam River, with values reaching 16 FNU at Niagara, 98 FNU at Mehama, and 93 FNU at Geren Island. The Little North Santiam River also was turbid, reaching 133 FNU.
During the January–February storm, suspended-sediment loads increased at all locations in the basin (table 8). The most noticeable increased load was in French Creek, where values typically had been low. Now, for the first time since its installation in 2001, the water-quality monitor recorded a major turbidity event, with a corresponding event sediment load of 400 tons (40 tons/mi2). In comparison with other upper basin streams, French Creek transported the smallest total load, but the second largest percentage of annual load (60 percent) behind Breitenbush (63 percent). This event marked the largest suspended-sediment load and percentage of annual load for French Creek during the study. In addition, the January–February 2003 major turbidity event was the only event during the study to produce clay-water in French Creek (67 Mgal). Because the French Creek subbasin is small (10 mi2) with coarse, immature soils (U.S. Forest Service, 1992), clay-water volume for most of the other stations was expectedly higher, although absent at Blowout Creek (table 8).
With most monitors reading turbidity at equal to or less than 100 FNU, effort for field verification was spent looking for the source of the major turbidity event in French Creek. On January 31, USGS scientists visited French Creek and observed that a debris flow had crossed French Creek and French Creek Road less than 0.5 mi upstream of the monitoring station (fig. 23). Large cobbles, boulders, and debris littered the road (fig. 24). The area around the debris flow was naturally stable, although susceptible to severe soil erosion and runoff (U.S. Forest Service, 1992). No calculations were made to estimate the size or volume of material displaced by the debris flow; however, most of the suspended-sediment load in French Creek originated from this source.
Because of the rarity of such high-turbidity events, and the difficulty in maintaining active telemetry in the heavily forested French Creek subbasin, the station was decommissioned in 2005.
A second major turbidity event, similar to that of October 2000, occurred in the North Santiam subbasin in October 2003. Once again, in late September and early October 2003, record-setting high temperatures were recorded in the basin. Temperatures rose to 88°F on September 29, and daily high temperatures remained in the upper 70s for most of October (Oregon Climate Service, 2006). The warm temperatures increased snow and glacial melt and slightly increased runoff in upper-basin streams. On October 21, an extraordinarily large spike in turbidity estimated at 5,550 FNU exceeded sensor measurements at the North Santiam monitoring station (Bragg and others, 2007). The high turbidity was relatively short-lived, remaining greater than 100 NTU for just more than 14 hours before decreasing to less than 10 FNU. The North Santiam monitoring station was the only one of the upper-basin monitor stations to record high turbidity (fig. 25). The high turbidity estimate was verified by a series of water samples collected by the newly installed automatic pumping sampler, which was triggered during the event (fig. 26). Although no rain fell during the actual turbidity event, nearly 3 in. of rain had fallen in the High Cascades during the weeks leading up to the October 21 event (Oregon Climate Service, 2006).
The total suspended-sediment load transported during this major turbidity event was 2,030 tons (table 9). This event load was about 25 percent larger than the previous glacial-outwash event in October 2000, yet accounted for less of the annual load at 27 percent because of more winter rainfall. The clay-water volume for the event was 222 Mgal, representing 12 percent of the annual clay-water volume. This was much less significant than the 79 percent of annual clay-water volume for the October 2000 event.
Sediment responsible for the elevated turbidity in the upper North Santiam River came from upstream in the Milk Creek by way of Pamelia Creek (fig. 27), as in October 2000.
As mentioned earlier (see section, “October 1, 2000—North Santiam River: Follow-Up Investigations”), field investigations during the last 5 years have attempted to locate potential sources for future events from Milk Creek and the Milk Creek Glacier. As a result of high temperatures, snowmelt, and glacial outwash likely eroded exposed glacial sediments (fig. 28) or landslide material (fig. 29) and transported it downstream.
For over a week, beginning on January 23, a series of storms passed over the North Santiam River basin, depositing more than 10 in. of rain (Oregon Climate Service, 2006). During January 28–30, more than 5 in. of rain fell, raising streamflow to 4,960 ft3/s at the North Santiam monitoring station, 4,690 ft3/s at Breitenbush, 2,350 ft3/s at French, and 1,090 ft3/s at Blowout. The highest flows were recorded at the Little North monitoring station at 12,700 ft3/s. The continuous rainfall mobilized sediment, increasing turbidity to 25 FNU at the North Santiam monitoring station, 71 FNU at Breitenbush, 27 FNU at Blowout, and 251 FNU at Little North (fig. 30). Along the lower mainstem North Santiam River, turbidity reached as high as 290 and 154 FNU at Mehama and Geren Island, respectively. High turbidity during this storm prompted the closure of the water-treatment facility (Hank Wujcik, City of Salem Public Works Department, written commun., 2005). Values were missing at the North Santiam River at Niagara station because the water-quality monitor was malfunctioning.
During the 3-day storm period, the Little North Santiam River carried 11,394 tons of suspended-sediment, a yield of 102 tons/mi2 (table 10). Even though turbidity values were relatively low compared to those measured during previous major turbidity events, the long duration and high streamflow made this a major event. This was the third highest event yield for the Little North Santiam subbasin during the study duration and also was the largest Little North Santiam high-turbidity event as calculated using instream turbidity rather than using streamflow as a surrogate for turbidity, as with previous Little North Santiam River estimates. For comparison, using streamflow to estimate loads would have resulted in an estimate of 7,060 tons, 40 percent less than what was predicted using instream turbidity. That would decrease the percentage of annual load for this event from 64 to 39 percent. Clay-water volumes would have been underestimated as well. For example, there would be 2,384 Mgal less if discharge were used rather than instream turbidity. Therefore, previous estimates for 1998 and 1999 could have been much greater for the Little North Santiam River had monitoring stations been in place.
Field investigations pointed to a landslide in the Evans Creek watershed, near the community of Elkhorn, as the source of turbidity to the Little North Santiam River during this event (fig. 31).The landslide on Evans Creek covers more than 20 acres on the western slope of Evans Mountain (fig. 32) and displays all the characteristics of a large, natural, complex landslide: multiple scarps, debris flow channels, and transverse cracks (fig. 33). Its earliest documentation dates to the late 1970s (Hank Wujcik, City of Salem Public Works Department, written commun., 2005), although storms from the flood of 1964 probably had mobilized this area earlier. The failure of Evans Mountain Road occurred during the flood of 1996, with additional movement in 1997 (Jerry Pierce, Upward Bound Camp, oral commun., 2006). Because the landslide intersects the creek, material is constantly being eroded whenever higher flows occur. In addition, erosion of exposed surfaces of the landslide transports turbid water into the creek. Because of the proximity of the Little North Santiam River to the City of Salem water-treatment facility (about 10 mi upstream), turbid water emanating from the subbasin is a major concern.
To assess sediment input from Evans Creek, temporary water-quality monitors were installed on Evans Creek and on the Little North Santiam River at Elkhorn, just upstream from Evans Creek. Turbidity data from these monitors confirms earlier assumptions that Evans Creek flows often are considerably more turbid than those of the mainstem of the Little North Santiam River during runoff events. For example, in late December 2005, turbidity in Evans Creek was, on average, more than 30 times greater than that in the Little North Santiam River at Elkhorn, yet only 3.5 times greater than turbidity measured downstream at the Little North monitoring station (fig. 34). Although Evans Creek may not be the only sediment source, it appears to be one of the uppermost subbasin sources. Further investigation into other potential source areas between Evans Creek and the permanent water-quality monitor at the mouth of the Little North Santiam River is currently underway.
U.S. Department of the Interior | U.S. Geological Survey
Persistent URL: https://pubs.water.usgs.gov/sir20075178
Page Contact Information: Publications Team
Page Last Modified: Thursday, 01-Dec-2016 19:51:36 EST
