Annual bedload accumulation from sediment basin surveys in small gauged watersheds in the Andrews Experimental Forest, 1957 to present

CREATOR(S): Sherri L. Johnson, Jack S. Rothacher
ORIGINATOR(S): Jack S. Rothacher
OTHER RESEARCHER(S): Julia A. Jones, Gordon E. Grant, Frederick J. Swanson
FORMER INVESTIGATOR: Craig Creel, Richard L. Fredriksen, C. Ted Dyrness, Alfred B. Levno
17 Dec 2012
14 Mar 2017
Disturbance, Disturbance, Organic matter, Long-Term Ecological Research (LTER), debris flows, landslides, disturbance, erosion, sedimentation, inorganic nutrients, sediments
Forest disturbance, including timber harvest and natural disturbances, influence the rate and timing of exports from headwater basins, including dissolved, suspended, and bedload. Bedload is the material that is bounced or dragged along the base of the channel, and ranges from gravel to boulder size. The objective of this study is to quantify the effect of road construction, timber harvest and post-harvest burning on bedload production from small headwater basins. The study design (before/after, control/impacted) is intended to dsplay the effects of anthropogenic disturbance on bedload sediment production over multiple decades.
Experimental Design - HS004:

Long-term small watershed studies were established in treatment-control pairs to examine changes in bedload accumulation with respect to timber harvest and road building. Bedload accumulation is measured in a pre-treatment period, a post-road construction period (WS 3), and a post-harvesting period in 5 small watersheds. WS 2 is the control watershed paired with WS 1 (100% skyline harvesting) and WS 3 (patch-cut high-lead logging) as treatment watersheds. WS 3 has a preliminary road construction treatment 3 years before logging. WS 9 is the control watershed paired with WS 10 (100% clearcut) as the treatment watershed.

  • WS 1 - 96.8 ha, 100% clearcut by Wyssen skyline system from fall 1962 until summer 1966, debris burning Oct. 1966
  • WS 2 - 60.3 ha, undisturbed control paired with WS 1 and WS 3, no development WS 3 - 101.2 ha, 1.65 miles of road constructed 1959, high-lead logging winter 1962-63, three small clearcuts (13, 20, 28 ac), 25% of watershed, debris burning Sep 1963.
  • WS 9 - 21.2 ha, undisturbed control is paired with WS 10. WS 9 is located just south of the mouth of Lookout Creek along the Blue River Reservoir.
  • WS 10 - 25.3 ha, 100% clearcut in 1975. Clearcutting occurred during the spring and summer of 1975, and a running skyline system yarded all logs and unmerchantable material >20 cm in diameter or >2.4 meters in length uphill to a single landing. A significant debris flow in Feb 1986 destroyed the gaging station.
Field Methods - HS004:

Sediment catchment basins were constructed below the gaging station on each stream in 1955 to examine changes in bedload accumulation with respect to timber harvest and road building. WS1 and WS2 are earthen dams, not lined with concrete, and with no associated diversion structure. Neither of these basins serve as the spillway for streamflow. WS1 basin was significantly altered in basin collection area following the February, 1996 flood. The WS3 basin had a wooden containment structure. Following debris torrents in December, 1964, the catchment basin was remodeled. Further rebuilding was necessary in 1986. In February, 1996, a debris torrent completely destroyed the basin. In February, 1998, the basin was reconstructed. The newer basin is an earthen dam covered with concrete, is concrete lined and serves as the spillway. A stream diversion structure is established at the entry. The basin is configured to be cleaned on a storm by storm basin, although this has only been employed once in 1999 when the basin filled in January. The sediment catchment basins at WS 9 and 10 were constructed in 1974. The WS 9 and 10 basins are concrete lined with wooden dams which serve as the spillway. The wooden dam/spillway of the WS9 basin was reconstructed in 1996. The wooden dam/spillway of the WS10 basin was reconstructed in 2005.

A level, leveling rod, and tape are used to measure bottom elevation (.01m) at each intersection of a 3-foot grid (2-foot grid on WS 3) over the catchment basin. Monumented cross-section end points are spaced at regular intervals along a baseline at each end of the basin. A tape is stretched across the basin between surveyed end points at the baselines to locate the grid points. A permanent benchmark is established and elevation with respect to the permanent benchmark is measured at the beginning and end of the survey, and routinely checked between alternate cross-section transects. This procedure allows for comparison from year to year and reveals any changes in the elevation or level of the surveying instrument that might occur during the survey.

Sediment basin surveys are conducted annually on Watersheds 1, 2 and 3 (before 1998), and volumetric measurements made with 5-gallon bucket counts or wheelbarrow or backhoe bucket counts on the smaller Watersheds 9 and 10, and WS3 (beginning in 1998). Basin surveys are designed to build basin profiles that are used to determine a change in average bottom elevation between annual surveys. The basins are cleaned annually and resurveyed to provide a base elevation for the following year. Local contractors are employed and use a front-end loader or excavator to clean certain basins that are near or have reached capacity (typically WS 1 and WS3). Watershed 9 and 10 are usually cleaned with a crew using shovels and buckets to measure wet sediment volume. When large deposits of sediment occur in these basins they are surveyed using the same methods used at WS1 and WS2.

Permanent benchmarks are established near all gaging stations. In 1977 permanent auxiliary benchmarks (1/4" bolts set in concrete) were established near each of the catchment basins and are routinely measured in annual surveys. These new auxiliary benchmarks replaced nails or spikes driven into stumps or trees as reference points. Annual checks, which monitored the elevational difference between benchmarks and reference points, showed unexpected changes. The reference point on the Watershed 3 dam was actually sinking, while at Watershed 1 and 2 stumps containing the reference spikes were deteriorating and reliable measurements became increasingly difficult. The elevational difference between the new auxiliary benchmarks and permanent benchmarks were monitored until 1983 when surveys indicated the new auxiliary benchmarks were stable.

In an effort to identify or capture the timing and volume of material delivered to the Watershed 1 sediment basin during individual storms, a device for measuring the profile of the sediment deposit was installed in September 1981. Pulleys were fixed to two trees growing in appropriate positions at either end of the long axis of the sediment delta. A plastic-coated cable was run between. the two pulleys - washline style. A hook was attached to the cable and a fabric measuring tape attached to a lead clock weight was strung through the hook. A nail driven into one of the trees was used as a measurement point. The cable was advanced at one meter intervals, where the weight was lowered to the surface of the water and the tape distance recorded. Then the weight was lowered to the bottom of the basin and a second tape distance was recorded. This procedure was repeated at each meter along the axis of the delta. A meter stick (partially submerged) attached to another basin-side tree served as a staff gage and was used to normalize differences in water surface level between surveys, and determine the amount of cable stretch or contraction due to temperature variations. Changes in the water surface tape distances not accounted for by the change in the basin water level were assumed to be due to cable stretch/contraction, and were used to adjust the bottom tape distances to make more accurate comparisons between periodic measurements.

Instrumentation: Level, leveling rod, and tape
Processing Procedures - HS004:

Determination of sediment basin accumulation is based on the change in average bottom elevation by comparing the same survey points in two annual surveys. While all points along the cross-section line are recorded, not all points are used in calculating mean basin elevation and the number of points included in any comparative calculation is variable. Initially all points between cross-section end posts were included in the calculations, but in years of little or no bedload accumulation small errors began to compound and led to negative accumulation calculations. Errors in rod placement or instrument readings were hard to quantify, but some potential errors were identified and eliminated. Points on steep slopes of the sediment basins, which accumulate virtually no sediment, may have produced erroneous rod readings and slope points are typically eliminated from the calculations. These points occupy less than ten percent of the basin area at all three watersheds. When a basin is filled to near capacity, points on the bottom that may have been on a steep slope in a previous survey are included in the calculations.

Rod measurements for survey points included in the calculations are totaled and averaged, yielding an average rod reading. A line of sight is determined by adding the mean of the benchmark readings to an arbitrary elevation of the auxiliary benchmark (100.00 meters) and further adjusting by any change in elevational difference between the permanent and auxiliary benchmarks. The average rod reading subtracted from the line of sight provides an average bottom elevation. By subtracting the previous bottom elevation from the current value and multiplying by sediment basin area, the volume of accumulated sediment is determined. This volume divided by watershed area determines the yield of bedload per unit area of watershed.

Since a small difference in bottom elevation multiplied by sediment basin area results in a substantial value for accumulated material; and since adjustments for errors in rod readings and instrument level cannot be made, a difference in bottom elevation of plus or minus .01 meters is regarded as “no change”.

  • Example:
  • Average rod reading = sum of rod readings / number of points
  • Line of sight = Arbitrary elevation auxiliary benchmark + mean auxiliary benchmark reading+ correction value
  • Change in bottom elevation = line of sight – average rod reading
  • Accumulation = Change in bottom elevation * sediment basin area
  • Accumulation per unit area = accumulation / watershed area

Annual calculations were originally hand calculated, 1957-1978; calculated with hp 97 desk calculator,1979-1985, and calculated with desktop application programs, 1986-present. After 1990, field recorders are used to record the measurements (CMT recorders (1990) and Husky Hunter recorders (1994), waterproof laptop computers (2005).

WS 3 debris basin was reconstructed after the original installation was destroyed by a slide in December, 1964, and subsequently reconstructed after February 1996 debris flow avalanche. WS 1 debris basin is not believed to have captured all bedload material in the immediate years following burning in fall, 1966.

WS 2 (149 ac., 60 ha), control, undisturbed; WS 1 (239 ac., 96 ha), 100% clearcut by Wyssen skyline system, fall, 1962 to summer,1966, debris burning Oct., 1966, planting; WS 3 (250 acres, 101 ha), 1.65 miles of roads constructed in 1959, high-lead logging winter 1962-63, three small clearcuts (13, 20, 28 ac), 25% of watershed, debris burning Sept., 1963; WS 9 (9 ha), control, undisturbed; WS 10 (10 ha), clearcut, no burning 1975.
HJ Andrews Experimental Forest, western Cascades of Oregon, Watersheds 1, 2, 3, 9, and 10
annual surveys
Ground condition
Andrews WS 1, 2, 3 bedload accumulation data have been summarized by annual reports on file at Corvallis, Oregon, Forestry Sciences Lab, and now scanned and online in association with this study data.