Stream cross-section profiles in the Andrews Experimental Forest and Hagan Block RNA 1978 to 2011

CREATOR: Sherri L. Johnson, Frederick J. Swanson
ORIGINATOR: Frederick J. Swanson
OTHER RESEARCHER: Julia A. Jones, Gordon E. Grant, John M. Faustini, Arianna Goodman
ABSTRACTOR: Donald L. Henshaw
FORMER INVESTIGATOR: Futoshi Nakamura, Mary Ann Madej, George W. Lienkaemper
4 Dec 1990
24 Sep 2021
Disturbance, Disturbance, Inorganic nutrients, Long-Term Ecological Research (LTER), geomorphology, disturbance, inorganic nutrients, woody debris, riparian ecosystems, streams, vegetation
This study was established to monitor long-term changes in channel cross-sectional geometry (width, depth) in response to storms and change in the disposition of large organic debris in a range of stream sizes beginning in 1978. These observations can be used to assess change in sediment storage within the sample reach and change in habitat structure. Particle counts are performed to assess bed material surface particle size and its variability at each cross section location beginning with the 1995 survey. Two Hagan Creek sites (North and South Forks) were designed to document channel changes following debris removal and streamside logging in the riparian corridor; however, logging was indefinitely postponed, and measurements were discontinued in 1986. Vegetation cover was described on the Mack Creek clearcut site along transects adjacent to the stream profiles in a 1984 survey with the intent that if major channel change occurs, it would be possible to assess effects on riparian vegetation. The Shorter Creek stream profiles were established to study the influence and downstream effects of pulling large wood (greater than 10 cm) from a stream reach on the movement of sediment, and were surveyed before and after wood pulling in 1978 and annually through 1982.
Experimental Design - GS002:

Reference cross-section profiles were first established in 1978 at the lower Lookout Creek (LOL), middle Lookout Creek (LOM) and Mack Creek old-growth (MAC) sites, and in 1980 and 1981, respectively, at the Cold Creek (COC) and Mack Creek clearcut (MCC) sites. Between 11 and 20 cross sections were established at each site. Cross sections were irregularly spaced to sample a range of channel environments (pools, riffles, etc.). Cross-sections were resurveyed annually between 1978 and 1999 except in years which did not have significant storms. The longest interval between re-surveys was five years, between 1990 and 1995, a hydrologically quiescent period during which only the 1995 peak discharge approached the mean annual flood in Lookout Creek. Currently, partial surveys may be done annually as a reconnaissance to check for channel changes, with complete surveys conducted at least every 5 years. Large winter flows or high water that may have altered cross-sectional profiles will trigger a complete resurvey. Beginning in 1995, photographs of each cross section, typically one upstream and one downstream view, were routinely taken during cross section surveys. Vegetation data was surveyed once on Mack Creek in 1984.

The five cross section sites (reaches) can be viewed as being arrayed along a hydrologic and geomorphic gradient from the headwaters (Cold Creek site) to the mouth (lower Lookout Creek site) of the Lookout Creek watershed. Along this gradient, drainage area and channel width systematically increase, while channel slope systematically decreases. Other characteristics-such as the abundance of large wood and the degree to which the channel is constrained laterally by valley walls, terraces, or other non-fluvial or relict landforms-vary in a less systematic fashion.

Cross sections in the LOL reach are quite irregularly spaced, particularly in the upper part of the reach. In the upper sub-reach above the sharp bend, the two uppermost cross-sections (XS 13 and 14) are only approximately 25 m apart, followed by a more than 60-m gap to the next cross section. This is followed by three closely spaced cross sections (XS 9 to 11) within the bend, and then by eight somewhat irregularly spaced cross sections in the next approximately 300 m downstream of the bend.

Two Hagan Creek sites (North and South Forks) were designed to document channel changes following debris removal and streamside logging in the riparian corridor; however, logging was indefinitely postponed, and measurements were discontinued in 1986. Original cross-sections were spaced at 37.5 m with additional cross-sections added for a regular 18.75 m spacing interval at the planned clearcut reaches.

On Shorter Creek, cross-section profiles were established on three adjacent reach segments: an upstream control (xsecnumbers>=200), a debris pull segment where all large wood (greater than 10 cm) was removed (xsecnumbers>=100 and less than 200), and a downstream effect segment (xsecnumbersless than 100). It is thought that the cross-sectional number represents meters upstream from the starting point. Cross-sections were measured before and after debris removal in 1978 on the middle segment, and some cross-sections were measured again later that year after the first storms. Cross-sections were generally spaced 2-4 m apart.

Citation: Faustini 2000 (pub no. 2781)
Field Methods - GS002 Overview:

Between 11 and 22 cross-sections are established at 5 long-term sites. Distance and elevation are measured along a cross-sectional transect at 0.5-m intervals along the tape and at slope break points, geomorphic feature changes, or substrate changes. Substrate, channel boundaries and other information is noted for each measurement. Raw elevation differences between end points and total cross-section distance can vary from year to year. Early data were adjusted with additive and multiplicative adjustments to align cross-sections based on known features for comparability across years. Beginning in the late 1980's the zero-distance post (z-post) elevation has always been adjusted to zero elevation, but minor adjustments for differing measured lengths or elevation differences between end posts may require adjustment to better align a cross-section for multi-year comparison. Beginning with the 1995 survey, modified Wolman (1954) particle or "pebble" counts are performed in conjunction with the cross-section surveys to characterize bed material surface particle size and its spatial and temporal variability, and note the presence of bedrock.

In 1984 riparian vegetation is surveyed alongside the Mack Creek clearcut cross-sections. Cover of vegetation and substrate type are measured in 1x1 meter plots along a transect across the stream channel. Biomass measurements are also made, but biomass data was never run through the allometric equations to calculate biomass and this raw data is only available on request. Biomass measurements are made based on species and might include basal diameter, leaf length, height, frond length, blade length, or the number of leaves, stems, or fronds. In some cases percent cover is sufficient to estimate biomass. Contact for more information.

Citation: Faustini, John M. 2000. Stream channel response to peak flows in a fifth-order mountain watershed. Corvallis, OR: Oregon State University. 339 p. Ph.D. dissertation.
Field Methods - GS002 Cross-section protocol:

At each cross-section a tape is strung between two endpoints and pulled taut to avoid sag. The tape attachment points on the endposts are marked so the tape is positioned the same way in subsequent measurements of the cross-section. The attachment points were established by leveling a stringline between the endposts. By convention the zero end of the tape is attached to the "Z-post", the left-hand post looking upstream. The "X-post" is the right hand post looking upstream. The survey instrument (auto-level on tripod) is placed directly above or adjacent to the Z-post. If placement at the Z-post is not possible, use the X-post with the zero end of tape at the Z-post. The stadia rod operator starts at the opposite endpost and places rod on substrate vertically and adjacent to the tape for elevation readings. Readings are taken at 1.0-m intervals outside the active channel, and 0.5-m intervals inside the active channel along the tape. Readings are also taken at changes in slope, geomorphic feature, or substrate. The automatic level operator uses the scope and records the elevation reading at the stadia rod cross-hair. The distance along the tape is recorded. The substrate at each rod position is recorded as well as channel geomorphology and other information in standard codes. The chief substrate categories (within the active channel) are boulder (less than 25 cm diameter), cobble (5 to 25 cm), gravel (2 mm to 5 cm), fine sediment (less than 2 mm; typically sand), log, suspended log, and organic debris (woody debris less than 10 cm in diameter). Channel geomorphology codes include bankfull, edge of active channel, and edge of water. Other codes have been established to capture special notes regarding undercut banks and other features.

Spatial resolution of the cross-section survey data is variable, particularly for the early surveys. Prior to 1986, elevations were measured at irregular intervals along the line of section at slope breaks and other points selected to capture important topographic and geomorphic features and changes in substrate type. The frequency of surveyed points along each cross section varied considerably from year to year, resulting in a variable degree of resolution in characterizing the channel cross-sections. In 1986, surveying protocols were standardized to improve data consistency, and the horizontal spacing of elevation data points was set at 0.5 m and 1.0 m, with additional points taken as needed to capture abrupt changes in slope. Up until the mid-1980s, the cross-sections were surveyed using a Dumpy level, an instrument which can be used for either level or inclined sightings. This instrument required careful attention to maintaining a level line of sight to achieve accurate results. Inclined sightings (in which the angle between the telescope line-of-sight and the horizontal plane was recorded and a trigonometric correction applied to the elevation data) were occasionally used during this period when vertical relief within the cross section exceeded the height of the stadia rod. An automatic-level was first used for surveying the cross sections in either 1986 or 1988. The categories used to classify the substrate material also varied from year to year until 1982, when substrate codes were standardized and data from previous surveys were reclassified to conform to the new scheme.

In addition to standardizing the sampling interval, several additional steps were taken to improve data quality in 1986. Cross-section end posts were better anchored (where necessary) and labeled; attachment points on each cross section end post were identified and marked to ensure that the tape would be at least approximately level (so that horizontal distance measurements would be accurate and consistent between survey dates); and a check of the cross section length was made prior to surveying to ensure that it was consistent with previously measured values. Previously, no effort was made to ensure that the tape was approximately level (or to correct the horizontal distance for the effects of an inclined tape) or that the measured section length matched previous survey data. At several cross-section locations at the lower Lookout Creek site, where one end post was located on a high terrace, intermediate posts were established lower on the bank in 1986 at an elevation closer to that of the post on the opposite bank; the intermediate posts subsequently served as the new cross section end points.

Endpoints of each cross section are marked by steel fence posts or in some cases (on bedrock or resistant cohesive banks) by an eyebolt or a steel rebar stake. Cross-section endpoints are surveyed with tape and compass to establish lengths and bearings to other endpoints. When bank failure compromises an endpoint stake, the end point is re-established and the new endpoint position is established with similar bearing to the previous position if possible. Generally, cross-section numbers followed by a letter "A" or letter "B" indicate a change in the cross-section endpoints. For example, Lower Lookout cross-section 009 is listed as 009A after one endpoint was reestablished. While these modified cross-sections can still be compared over time, they may actually represent a different bearing or length from their earlier namesakes. On Hagan Creek, the 900 series of cross-sections were general new sections placed between existing sections in 1982. For example cross-section 912 was placed between cross-section 012 and 013. For Shorter Creek cross-sections where large wood was removed in 1978, cross-section 100 is listed as 100P after the woody debris is removed and 100Q following the first storm after debris removal. In these cases cross-section 100 is exactly sampled (no changes in end points), and the letters (P,Q) only differentiate among the multiple surveys within that year. See the enumerated domain for the attribute "xsecnum" in entity 1 or 2 for a complete description of each cross-section.

Instrumentation: Automatic level on tripod (several models have been used) and stadia rod
Field Methods - GS002 Particle count protocol:

Modified Wolman (1954) pebble counts are performed in conjunction with the cross-section surveys to characterize bed material surface particle size (mode and median particle size) and its spatial and temporal variability. Wolman suggested measuring the intermediate axis of 100 particles by averting ones eyes and sampling the first particle touched at the toe of one's boot. Modified Wolman (1954) particle or "pebble" counts of 100 particles selected in a "random walk" from an ~ 2-m wide band along each cross-section transect are performed beginning with the 1995 survey. Pebble counts are performed at each cross-section location at the Lookout and Mack Creek sites and within 3-5 sub-reaches at the Cold Creek site, where the small channel size makes individual cross-section based counts impractical. In the data (Entity 3, 4) Cold Creek cross-section xsecnum=COC001 includes collection over xsecnum 001-003, COC004 include 004-005, COC006 includes 006-009, COC010 includes 010-013, and COC014 includes 014-017.

The modified protocol: Run a tape between the X and Z end posts of the cross-section. Randomly sample within a swath approximately one meter upstream and downstream of the tape line and within the active channel boundaries. Sample after each step taken across the swath until 100 sample diameters are taken. If bedrock is encountered, note the count of bedrock encounters and do not count as a particle sample. The swath may be traversed several times. At each step sampling point, reach down with one finger while looking away from the streambed to avoid biasing selection and sample the particle touched. Measure the diameter of the B-axis (intermediate axis) of the particle using a steel tape measure to the nearest millimeter for particles less than 5 cm, and to the nearest 5 mm for larger particles. Particles less than 2 mm in diameter (i.e., sand) are recorded simply as 2 mm. For large partially buried particles, the length of the apparent intermediate axis of the exposed portion of the particle is measured.

Citation: Wolman, M.G., 1954, A method of sampling coarse river-bed material: Transactions of the American Geophysical Union (EOS), v. 35, p. 951-956.

Starting in the 1970s, several long-term studies of stream channel change were established in the Pacific Northwest of the US and British Columbia, Canada. These include studies at the HJ Andrews Experimental Forest (Faustini & Jones, 2003; Nakamura & Swanson, 1993), Redwood Creek in Northern California, USA (Madej, 1999; Madej & Ozaki, 1996, and data link below), the Toutle River drainage at Mt. St. Helens, USA (Major et al., 2019; Meyer et al., 1985), and Carnation Creek on Vancouver Island, BC (Hartman & Scrivener, 1993; Reid et al., 2019).

Related Publications:

Hartman, G. F., & Scrivener, J. C. (1993). Impacts of forestry practices on a coastal stream ecosystem, Carnation Creek, British Columbia (Canadian bulletin of fisheries and aquatic sciences No. 223). Reviews in Fish Biology and Fisheries. doi:10.1007/BF00043304

Major, J. J., Zheng, S., Mosbrucker, A. R., Spicer, K. R., Christianson, T., & Thorne, C. R. (2019). Multidecadal Geomorphic Evolution of a Profoundly Disturbed Gravel Bed River System—A Complex, Nonlinear Response and Its Impact on Sediment Delivery. Journal of Geophysical Research: Earth Surface, 124(5), 1281-1309. doi:10.1029/2018JF004843

Meyer, D. F., Nolan, K. M., & Dodge, J. E. (1985). Post-Eruption Changes in Channel Geometry of Streams in the Toutle River Drainage Basin, 1980-82, Mount St. Helens, Washington (85-412). Retrieved from

Reid, D. A., Hassan, M. A., Bird, S., & Hogan, D. (2019). Spatial and temporal patterns of sediment storage over 45 years in Carnation Creek, BC, a previously glaciated mountain catchment. Earth Surface Processes and Landforms, 44(8), 1584-1601. doi:10.1002/esp.4595


These sites are lower Lookout Creek (LOL; 61.5 km2 drainage area), middle Lookout Creek (LOM; 31.7 km2), Mack Creek clearcut (MCC; 5.9 km2, Unit HJA L110) and Mack Creek old growth (MAC; 5.6 km2), Cold Creek, Shorter Creek, and two sites along Hagan Creek outside of the Andrews Forest in the Hagan Block RNA. Drainage areas of surveyed streams range from 0.6 to 61.5 km2.

The Mack Creek old-growth (MAC) and clearcut (MCC) sites are nearly adjacent, contrasting 320-m long reaches at the downstream end of the undisturbed portion of the Mack Creek watershed. The Mack Creek study site lies at an elevation of ~ 750 m within the H.J. Andrews Experimental Forest (Andrews Forest). Mack Creek is a third-order watershed with a drainage area of ~ 5.8 km2 at the gaging station located about 1 km upstream of the junction with Lookout Creek. The upper reach (old-growth site) lies just upstream of the gaging station and a road crossing; this reach lies within an old-growth forest dominated by Douglas-fir (Pseudotsuga) with lesser amounts of western hemlock (Tsuga) and western redcedar (Thuja). The lower reach (clear-cut site) lies immediately downstream of the gaging station. The old-growth reach contains 12 cross sections established in 1978 which are numbered XS 1 through XS 12 in the upstream direction; the lowermost (XS 1) is located approximately 30 m upstream from the bridge. The clearcut reference reach contains 20 cross sections established in 1981, assigned 3-digit numbers (XS 101 through XS 120) that, contrary to the old-growth reach, increase in the downstream direction. The uppermost MCC cross section is located approximately 75 m downstream of the road crossing.The hillslopes adjacent to the reach were clear-cut in 1964-65, and most of the LWD remaining in the channel was flushed downstream by a major flood in December 1964. (Faustini 2000)

The middle Lookout Creek (LOM) site consists of 11 irregularly spaced cross sections in a 300-m long reach. It is located roughly in the middle of the Lookout Creek watershed on 4th-order Lookout Creek at an elevation of approximately 575 m. The upper of two low elevation mainstem Lookout Creek sites, its upstream drainage area, estimated from 30-m DEM, is 31.7 km2, approximately half that of the entire Lookout Creek watershed and a little more than five times that of the Mack Creek sites. The average channel gradient is 3.2%, or about one third of the gradient at the Mack Creek sites and less than a fifth of the gradient of the Cold Creek site. The average channel width is approximately 24.7 m, roughly double that of the Mack Creek sites and nearly 5 times that of the Cold Creek site. The lower Lookout Creek (LOL) site consists of 14 cross sections along a 470 m long reach of 5th-order Lookout Creek located approximately 450 m upstream of the Lookout Creek gaging station. The upstream drainage area is approximately 61.5 km2, very nearly double that of the LOM site, while the average channel gradient is slightly less than half that of the LOM site at 1.5%. With an average channel width of approximately 27.3 m, however, it is only slightly wider than the LOM site (24.7 m), although the lower two-thirds of the reach is slightly wider (29.3 m). (Faustini 2000)

The Cold Creek (COC) site is a high elevation (1000 m) headwaters channel with a drainage area of approximately 71 ha (0.71 km2). While Cold Creek likely experienced some degree of past glaciation, the study reach is narrow (5.3-m average channel width), highly constrained by valley sidewalls, and quite steep (average gradient of 17%). The channel contains abundant woody debris, but most of the larger pieces and the majority of the volume of LWD is suspended well above the channel bed. The Cold Creek reference reach contains 17 cross sections, labeled XS 1 to XS 17, with numbers increasing upstream (Figure 9.3). The lowermost cross section is located approximately 50 m upstream of Forest Road 1506, the main access road into the Andrews Forest, and the uppermost cross section is located approximately 170 m farther upstream. (Faustini 2000)

Garrison et al., 1976
Five sites (Mack Creek clearcut and old growth, lower and middle Lookout, and Cold Creek) are located on the H.J. Andrews Experimental Forest; the other two sites are on North and South Forks of Hagan Creek located approximately 15 km west of the Andrews.
1-5 years
Ground condition
Beginning in 1995, photographs of each cross section, typically one upstream and one downstream view, were routinely taken during cross section surveys. Photos are stored with Craig Creel in the Forestry Sciences Lab (FSL 331W). Establishment surveys and stream reach maps are also stored with the raw collection sheets.