South Fork Campbell Creek drains largely undeveloped land in Anchorage, Alaska, but supports heavy use near the Bureau of Land Management (BLM) Campbell Tract facility for recreation and environmental education. To help assess the impacts of human activities in the basin on biological communities, particularly aquatic and terrestrial biota, morphological changes to the channel bed and banks were monitored for 2 years. Erosion conditions and rates of change were measured and 11 transects were surveyed in three reaches of Campbell Creek near the BLM Campbell Creek Science Center in 1999. Repeat measurements at these 33 transects in 2000 documented noticeable differences between horizontal or vertical channel position at eight transects. Repeat measurements of 51 erosion pins at the survey transects provided details of bank erosion between the 2 years. Annual erosion rates at the erosion pins ranged from 0.81 foot per year of erosion to 0.16 foot per year of deposition.
South Fork Campbell Creek is the major drainage in a relatively undeveloped tract of land managed by the Bureau of Land Management (BLM) in Anchorage, Alaska. The stream and surrounding land, known as the Campbell Tract, are near urbanized Anchorage (figs. 1 and 2) and are heavily used for recreation and environmental education. Despite intensive human use, the stream supports a range of wildlife and is the spawning grounds for wild coho and chinook salmon, as well as resident Dolly Varden trout. Many entities share concerns about potential environmental degradation from increasing use of Campbell Tract and resulting impacts on aquatic and terrestrial biota. BLM established a Limits of Acceptable Change program in 1994 that addresses the balance between human use and impacts on environmental conditions at the Campbell Tract facility.
The U.S. Geological Survey (USGS) included Cook Inlet Basin (COOK) as part of its National Water-Quality Assessment (NAWQA) Program in 1997. NAWQA studies support collection of biological, chemical, and physical data to be used in an integrated assessment of water quality according to national protocols. The NAWQA effort addresses a range of sites across Cook Inlet, including South Fork Campbell Creek at the Campbell Tract, but is not intended to provide comprehensive, site-specific assessments. In order to expand the knowledge base of South Fork Campbell Creek at Campbell Tract and provide a baseline for future assessments, the USGS, in cooperation with BLM, began a study in 1999 to monitor morphological changes to the channel bed and banks.
This report describes the methods and results of channel morphology surveys and erosion monitoring at South Fork Campbell Creek. Biological, chemical, and physical habitat data collected for the NAWQA study and comparisons with data from other sites in Cook Inlet are available online at http://ak.water.usgs.gov/Projects/Nawqa/ and in reports by Glass (1999), Frenzel (2000), Huckins and others (2001), Johnson (2001), and Meyer and others ( 2001).
The purpose of this report is to provide a baseline inventory of existing channel bed and bank conditions at three reaches of South Fork Campbell Creek near BLM’s Campbell Creek Science Center, including an inventory of erosional changes between water years 1999 and 2000. A baseline monitoring study describes existing conditions in such a way that future comparisons can be made. Comparison of a baseline with future measurements allows assessment of processes that may take years to develop. Accordingly, this report describes field monumentation of erosion monitoring sites, presents monitoring data, and describes changes to channel bed and banks observed during the monitoring period. The results of related investigations on Campbell Creek for NAWQA or using NAWQA protocols are summarized in other reports and on the COOK NAWQA Website.
Three study reaches, each containing 11 transects, were established for this study and the concurrent NAWQA study. Transect surveys made in 1999 and repeated in 2000 documented lateral and vertical changes to the bed and banks of South Fork Campbell Creek. Repeat measurements of erosion pins at the transects detailed lateral changes to banks over this same study period. Aerial photographs taken in 2000 documented conditions between transects and provided detail not possible to include in transect surveys, such as the presence of large, woody debris and the extent of gravel bars.
South Fork Campbell Creek drains a mountainous area in Chugach State Park and emerges from a narrow bedrock canyon about 3 mi upstream from the study site. Downstream from the canyon, the channel flows across a broad, gently sloping lowland to the study site. Altitudes in the watershed range from over 5,000 ft above sea level in the mountainous headwaters to sea level where the stream discharges into Cook Inlet. Altitude at USGS streamflow gaging station 15274000, South Fork Campbell Creek near Anchorage, is about 260 ft, based on topographic maps.
South Fork Campbell Creek at and near the study site forms a meandering, single-thread channel with occasional islands and gravel bars. The channel contains riffles, pools, and runs. The average gradient of three study reaches is 0.011 ft/ft. Streambanks generally consist of loose, unconsolidated gravel with cobbles, sand, and silt. Streambed material contains a similar range of grain sizes. Patches of naturally consolidated gravel with sand and silt observed in the streambed form a resistant shelf near the edges of several transects.
Although located near urbanized parts of Anchorage, the South Fork Campbell Creek watershed is mostly undeveloped upstream from the study site (fig. 2). Development near the study site includes gravel-surfaced Campbell Airstrip, formerly used as a military airport and presently maintained by BLM for occasional use, and Campbell Creek Science Center, an educational facility managed by BLM. The study site is used year round by about 4,000 to 6,000 people (Van Waggoner, Campbell Creek Science Center, oral commun., 2001) in educational groups including local schools, Campbell Creek Science Center, and private programs. The stream and streambanks also are used for recreational fishing, walking, jogging, and occasional mountain biking during open-water conditions, and cross-country skiing and dogsledding in winter months. Motorized vehicle traffic is limited to maintenance vehicles along a recreational trail within several hundred feet of the stream. Foot access by humans is the most common type of use of the stream and adjacent riparian corridor. Informal trails used by humans, domestic dogs, and wildlife have developed along the top edge of streambanks and provide access to the streambed and bars. Activities observed within the stream corridor include wading, collecting water samples, releasing and capturing fish, removing and adding woody debris, and collecting minor amounts of sediment. In addition to dogs accompanying recreational users, moose and bear have been observed within the study reach.
The South Fork Campbell Creek watershed comprises mountains over 5,000 ft in altitude that are snow covered for all but summer months. The drainage area of South Fork Campbell Creek upstream from the study site is 29.4 mi2 (Meyer and others, 2001), and mean annual precipitation is 22 in. (Jones and Fahl, 1994). Other basin characteristics, used by Jones and Fahl to calculate flood frequencies for the site, also are listed in their report.
Streamflow in South Fork Campbell Creek is highest during snowmelt runoff or rainfall-induced peaks (fig. 3). The stream is mostly ice covered, with common open leads, from about late October to April. Flows generally are lowest in mid- to late March or early April (fig. 3).
The USGS measured streamflow at gaging station 15274000 from 1947 to 1971 and from 1998 to the time of preparation of this report. From 1947 to 1952, the gaging station was 0.2 mi upstream from the study site; the present station is along Reach C of this study. Mean annual discharge for the periods 1947–71 and 1998–2000 combined was 38.2 ft3/s (Meyer and others, 2001). The largest recorded instantaneous peak flow was 891 ft3/s, which is on the order of a 100-year flood (Jones and Fahl, 1994), on June 21, 1949.
Daily discharge during 1999 and 2000 was generally close to the mean of all years of record; peaks in 1999 were slightly below the mean, and peaks in 2000 were slightly above the mean (fig. 4). Daily discharge for water years 1999 and 2000 and summary statistics are published in reports by Bertrand and others (2000) and Meyer and others (2001), respectively.
Thirty-three transects (fig. 5) were established and surveyed as a repeatable measure of channel morphology from which to detect channel changes over time. Differences between the initial survey in 1999 and a repeat survey in 2000 provide a measure of the vertical changes to the streambed. Although not necessarily indicative of long-term change, these measurements provide information about the types and locations of erosional processes in the stream.
Eleven transects were established at three study reaches, Reach B (downstream from the gaging station), Reach C, and Reach D (upstream from the gaging station) (figs. 6–8). This naming scheme corresponds to NAWQA conventions. Each transect was monumented with wooden stakes labeled with black ink, metal tags, and survey tape. Wooden stakes are useful for determining transect alignment but are not considered adequate for survey control. Iron rods 1 in. in diameter and 48 in. long were installed at each reach to provide vertical and horizontal control. The rod at Reach D was removed by unknown individuals by spring 2000, but rods at Reaches B and C were present in 2000 (fig. 5).
Transects were surveyed on June 17, June 18, July 28, and August 17, 1999, with a total station (theodolite and electronic distance-measurement instrument) (figs. 9–11, back of report; appendix 1). The survey was based on an arbitrary horizontal and vertical datum assigned to the left bank monument of Transect 11C and an arbitrary azimuth. Transect 01B was not surveyed in 1999 because vegetation obscured substantial parts of the transect.
All transects were resurveyed in 2000 (figs. 9–11, appendix 2) using a total station. Parts of Reach B were surveyed on May 24 and May 26, and the remaining transects were surveyed on September 26 and October 2, 6, 10, and 11, 2000. The 2000 survey was based on an arbitrary vertical datum and a horizontal datum close to actual latitude and longitude that was acquired for the iron rod monument on the left bank near Transect 10B using a survey-grade Global Positioning System (GPS) receiver. This coordinate system is estimated to be horizontally within about 30 ft of actual position (Bob McCool, System Dividends, oral commun., 2000) and was used for the presentation of all transects in this report. All 1999 data were converted to this coordinate system first by translating the entire 1999 survey by an amount equal to the difference between 1999 and 2000 coordinates for the left bank monument at Transect 11C, then rotating the entire survey about that monument until a best fit was obtained on the iron rod near Transect 10B. The remaining relatively permanent monument, the iron rod at Reach C, provided a check on the rotation. The 2000 northing, easting, and altitude for the iron rod at Reach C are – 0.11, 0.10, and 0.07 ft from the 1999 coordinates, respectively. These values include errors in surveying as well as errors in adjusting the 1999 and 2000 coordinate systems. This level of accuracy was considered acceptable for the purposes of this study, in which the magnitude of channel change detectable from transect profiles was expected to be about 0.5 to 1.0 ft.
Original monuments that could not be located were reestablished by surveying to the original location and installing a new monument as close to that location as possible. The iron bar at Reach D was missing but was not replaced. A wooden stake was installed at this location to serve as a monument for the transects.
In addition to the transects, the 2000 survey included the locations of the iron rods at Reaches B and C, some erosion pins, some additional stakes placed to mark instrument locations, and the four outermost corners of the concrete bridge abutment for the dogsled trail footbridge upstream from Reach D (appendix 2).
Despite efforts to survey along a straight line between transect monuments, individual survey points were usually offset a slight distance upstream or downstream. If the actual coordinates for these points were used to plot the transect profile, horizontal distances shown on the transect would be longer than actual distances. Survey points were projected onto a straight transect to eliminate this problem. Altitudes at the projected location may vary from the altitude at the surveyed location, but there was no basis for corrections to altitude. Horizontal distances between points were calculated from projected locations and plotted with the surveyed altitude to generate transect profiles (figs. 9–11). The original surveyed location, horizontal distance along the transect as calculated from the projected points, and the offset distance upstream (negative values) or downstream (positive values), are presented in appendices 1 and 2.
South Fork Campbell Creek contains many features common to small streams (fig. 12), including meanders, pools, riffles, bars, and islands. At the study reach, the channel is incised about 5 to 10 ft into the surrounding lowland. Short, gently sloping point bars are present on the inside bank of meanders within each of the three study reaches and are opposite tall, near-vertical banks. Wide gravel bars are present along some straighter reaches. Gravel islands vegetated with shrubs are present at Reach B and between the reaches. Average bankfull width is 36 ft, and average wetted width at the time of the 2000 surveys was 26 ft. The channel transects include shallow riffles with bankfull depths of about 1.5 ft and pools as much as 4.8 ft deep at bankfull flow. Mean bankfull depth at the transects is 2.4 ft. Bankfull altitude was determined for the purposes of this study as the altitude of the transition between vegetation species tolerant of exposure to air and species tolerant of inundation by water.
Large, woody debris is present in the channel in all three study reaches. A large log spans the channel between Transects 08D and 09D, and other logs partly span the channel in several locations (fig. 13). Multiple logs form a debris jam at Transects 06B and 07B.
Repeated surveys of a large number of points at the same transect can be used to determine the amount of vertical change (erosion or deposition) at a transect. However, changes evident in transect surveys also can be the result of differences in resolution (the number of points surveyed) or of surveying slightly upstream or downstream from the transect. For example, apparent vertical changes reflect actual physical changes when two consecutive measurements are made in the same location but cannot necessarily be confirmed when consecutive measurements are several feet apart (fig. 14). For the Campbell Creek transect surveys, vertical changes within the wetted channel are generally actual changes, whereas changes to banks that result in obvious differences in form often are related to differences in survey resolution. Plots of Campbell Creek transects over 2 years (figs. 9–11) are generally similar, but noticeable differences are apparent on 14 transects (table 1). Of these, at least 8 transects have apparently changed. Changes in consecutive transects were observed only at the upstream end of Reach D, where the thalwegs of Transects 09D, 10D, and 11D each moved up or down about 1 ft.
The limit of detection associated with repeat transect surveys for horizontal changes is greater than for vertical changes because the delineation of features such as banks with abrupt changes in slope is sensitive to the choice of survey point location. Little definitive horizontal change was detected from the Campbell Creek surveys between 1999 and 2000 (table 1). These baseline surveys will be more effective for detecting future major changes in channel planform than for detecting minor amounts of lateral bank erosion or bar growth.
Some changes occurred between transects that were not detectable from transect surveys. For example, the downstream edge of the point bar near Transect 05B eroded shoreward about 1 ft, but no change in the bar occurred at the transect. Likewise, some events that occur upstream or downstream from the study reaches could have an impact within the reach. For example, a scour hole is developing upstream from Transect 11C behind the root ball of a tilted, dead tree as the bank is being actively eroded.
Although repeat transect surveys can be used to detect channel changes of about 1 ft or more, erosion pins can be used to monitor bank changes directly with a detection limit of about 0.1 ft. Erosion pin monitoring is an established method of assessing bank erosion in which a permanent marker (a “pin”) is installed and measurements of the amount of pin exposed are made over time. Applications vary from a dense array of pins at a single bank location to a sparse array at many locations. The latter method was chosen for this study to provide a reconnaissance-level assessment of bank erosion over a large area. An assumption inherent in the method is that a single pin can represent the change across the entire vertical bank profile, that is, that every point on the bank moves at the same rate. This assumption is generally true at near-vertical banks and very gently sloping point bars. Because these conditions are not always met at South Fork Campbell Creek and because erosion rates varied at the few study transects where two erosion pins were installed, it is possible that the erosion rates calculated from erosion pin measurements do not reflect the true migration rate of the bank as a whole. However, they do provide a general assessment of the amount and spatial distribution of bank change.
Erosion pins consisting of a smooth, 1/4- or 3/8-in.-diameter, 24-in.-long (except for erosion pin 09DL1, which is 20 in. long) metal rod were driven horizontally into the streambank (fig. 12) at or near the established transects on July 28, 1999. Fifty-one pins were installed such that about half of the transects had an erosion pin on each bank (table 2) and two transects had more than one pin on the same bank. Pin exposure was measured with a metal tape at the time of installation and again on May 12, 2000, and October 5, 2000. Change in pin exposure then was divided by the time between measurements to provide erosion rates for two periods (table 2). The erosion pins were left in place at the end of the study to be available for future use.
Bank change (determined from erosion pin measurements in May and October 2000, subtracted from the initial measurement in July 1999) included undetectable change, progressive erosion or deposition, and reversals between erosion and deposition (fig. 15). The erosion rate at each study transect was calculated from erosion pin measurements over the entire monitoring period, July 1999 to October 2000. Rates for this period ranged from deposition of 0.16 ft/yr to erosion of 0.81 ft/yr (table 2). The average erosion rate was 0.06 ft/yr.
Plots of erosion rates by transect (fig. 16) show changes on one side of the channel relative to the other side. Pin measurements identified erosion on one bank and deposition on the opposite bank (lateral migration) at three transects, erosion on both banks (channel widening) at five transects, a change on only one of the banks at two transects, and both erosion and deposition on the same bank at one transect. None of the measurements identified deposition on both banks (channel narrowing). Pins at the remaining transects either had not been installed on both banks, or had been installed but were not found again during part of, or during the rest of, the monitoring period.
Continue to References cited, or return to Table of Contents
| AccessibilityFOIAPrivacyPolicies and Notices | |
  |
|