Multiply	By	To obtain
Length
inch (in.)	2.54	centimeter (cm)
foot (ft)	0.3048	meter (m)
Area
square mile (mi²)	2.590	square kilometer (km² )
Flow rate
cubic foot per second (ft³/s)	0.02832	cubic meter per second (m³/s)

Temperature in degrees Fahrenheit (°F) may be converted to degrees Celsius (°C) as follows:

°C=(°F-32)/1.8

Vertical coordinate information is referenced to the North American Vertical Datum of 1927 (NAVD 27).

Horizontal coordinate information is referenced to the North American Datum of 1983 (NAD 83).

Altitude, as used in this report, refers to distance above the vertical datum.

Abstract

In 2003 and 2004, a geomorphic assessment of streams in 20 watersheds in the Duluth, Minn., area was conducted to identify and summarize geomorphic characteristics, processes, disturbance mechanisms, and potential responses to disturbance. Methods used to assess the streams included watershed characterization, descriptions of segment slopes and valley types, historical aerial photograph interpretation, and rapid field assessments and intensive field surveys of stream reaches. Geomorphic conditions were summarized into a segment-scale classification with 15 categories mainly based on drainage-network position and slope, and, secondarily, based on geologic setting, valley type, and dominant geomorphic processes. Main causes of geomorphic disturbance included historical logging and agriculture, and ongoing urban development, human-caused channel alterations, road and storm sewer drainage, ditching, hiking trails, and gravel pits or quarries. Geomorphic responses to these disturbances are dependent on a combination of drainage-network position, slope, and geologic setting. Geologic setting is related to drainage-network position because the geologic deposits parallel the Lake Superior shoreline. Headwater streams in large watersheds flow over glacial deposits above altitudes of about 1,200 feet (ft). Headwater tributaries and upper main stems have ditch-like channels with gentle slopes and no valleys. Urban development and road drainage cause increased runoff and flood peaks in these segments resulting in channel widening. Below about 1,200 ft, main-stem segments generally are affected by bedrock type and structure and have steep slopes and confined or entrenched valleys. Increases in flood peaks do not cause incision or widening in the bedrock-controlled valleys; instead, the flow and scour areas are expanded. Feeder tributaries to these main stems have steep, confined valleys and may be sources for sediment from urban areas, road runoff, or storm sewer outfalls. Main-stem segments near the glacial deposits/surficial bedrock contact (1,000–1,200 ft) have the most potential for response to disturbance because they tend to have narrow valleys with sandy glacial lakeshore deposits and moderate slopes. Increases in flood peaks (from upstream increases in runoff) increase the potential for landslides and mass wasting from valley sides as well as channel widening.

Introduction

Streams in the Duluth, Minn., area (fig. 1) have been affected by a variety of historical land uses but remain highly valued for their natural resources. Streams and riparian areas contain important habitat for endangered or threatened aquatic and terrestrial species. Understanding the geomorphic conditions and processes of streams within a stream network or watershed framework is important because they relate to the overall water and aquatic resource quality of the stream and to the plant and animal communities and habitat adjacent to the stream. In 2003–2004, the U.S. Geological Survey (USGS), in cooperation with the city of Duluth, studied the geomorphic characteristics and processes that affect the physical conditions of Duluth-area streams.

Figure One

Figure 1. Study area and sampling sites in the Duluth, Minn., area, 2003–2004. Airport runway location from Minnesota Department of Transportation (2001).

The Duluth Natural Areas Program (DNAP) is based on a city ordinance designed to protect and preserve the natural heritage of Duluth and the surrounding area. The program strives to protect plant and animal communities, habitat for special species, natural water features, important bird habitat areas, and landscapes. Data on stream geomorphic conditions provide a basis for evaluating the rarity and quality of natural resources nominated to DNAP for their natural water features. An assessment of geomorphic conditions within the stream channels is important for understanding the extent to which natural factors and land-use changes have affected geomorphic attributes, such as bed and channel configurations, and the likelihood that geomorphic conditions may change in the future.

A thorough stream geomorphic assessment includes a description of how channel, flood-plain, and valley landforms are spatially and temporally affected by regional and local geology, topography, climate, vegetative cover, and human modifications. Stream networks developed in homogeneous geologic deposits commonly consist of an erosion zone (where sediment and runoff are generated), a transition zone (were sediment and water are transferred through with no net gain or loss of material), and a deposition zone (where sediment from upstream erosion is dropped out of transport) (Schumm, 1977). A typical longitudinal profile of a stream illustrates how the slope of a stream typically changes from relatively steep in the erosion zone near its headwaters to gentle in the depositional zone near its mouth (Hack, 1960) (fig. 2). An understanding of sediment sources, transport, and deposition and their links to stream network position is key to understanding geomorphic responses to present and future disturbances.

Figure 2. Plan view and longitudinal profile of geomorphic zones within a stream network (modified from Schumm, 1977)

Three common spatial scales used to identify geomorphic characteristics and processes are a watershed or landscape scale, a segment scale, and a reach or channel scale (fig. 3). The segment scale is a section of stream bounded by confluences or physical or chemical discontinuities, such as major landforms, changes in slope, or point-source discharges. The reach scale is a length of stream chosen to illustrate a set of uniform channel features and represents the typical sampling unit used for field-based measurements of geomorphic characteristics. Reach-scale geomorphic conditions and processes can be used to infer conditions and processes for a given segment.

Figure 3. Spatial hierarchy of watershed, stream-segment, and stream-reach scales (modified from Fitzpatrick and others, 1998).

A hierarchical channel classification system, based on geomorphic processes, was developed for mountain streams in the Pacific Northwest to help interpret channel morphology, assess geomorphic conditions, and predict responses to natural and human-caused disturbance (Montgomery and Buffington, 1993; 1997). In the Montgomery and Buffington (1997) classification, valleys were classified into three major types based on the predominant geologic materials in the valley—colluvial, bedrock, and alluvial (confined, moderately confined, and unconfined, respectively). Colluvial deposits mainly originate from landslides and mass wasting, whereas alluvial deposits are from transport and depositional processes related to modern rivers. Channels were classified into seven types—colluvial, bedrock, cascade, step-pool, plane-bed, pool-riffle, and dune-ripple. The channel types can be related to drainage-network position, underlying geology, and slope. The Montgomery and Buffington (1993; 1997) classification is not commonly used in the Midwest because streams have relatively gentle slopes compared to mountain streams. However, tributaries along the north shore of Lake Superior drain bedrock bluff areas and have slopes that exceed 8 percent along some reaches. Bohle (2002) developed a modified version of this classification to assess geomorphic conditions for streams along the north shore of Lake Superior in Minnesota. Bohle (2002) classified stream segments into eight geomorphic map units (GMUs)—boulder/bedrock, main-stem boulder rapid, gravel-dominated pool-riffle, forced pool-riffle plane-bed, large woody debris/boulder step, tributary boulder plane-bed or rapid, beaver dam step-pool, and lacustrine/riverine wetland.

The major objectives of the present geomorphic study of Duluth-area streams are to describe geomorphic characteristics, identify major geomorphic processes and factors contributing to geomorphic characteristics, identify historical changes in channel morphology and planform (plan view of the channel pattern), and develop a segment-scale classification based on similar geomorphic characteristics, processes, and responses to disturbance. The assessment and classification will provide baseline conditions and can be used as a tool for evaluating the quality of Duluth's natural water features. Scientists, city managers, and citizens involved in the Duluth Natural Areas Program can use results from this study to help protect, preserve, and restore resources and help guide management activities. In addition to its social and conservation benefits, this study will advance the understanding of present and historical geomorphic conditions of Duluth-area streams, of other north- and south-shore tributaries to Lake Superior, and of other streams in similar hydrologic settings.

Purpose and Scope

The purpose of this report is to provide an assessment and classification of the geomorphic characteristics, processes, and responses to disturbances of streams in the vicinity of Duluth, Minnesota. The scope of this report includes: 1) description of the current geomorphic characteristics, 2) identification of major geomorphic processes, 3) identification of factors contributing to current geomorphic characteristics and expected future trends in characteristics, 4) interpretations of historical changes in channel morphology and planform, and 5) development of a segment-scale classification based on geomorphic processes. Geomorphic characterizations and the classification system were based on watershed-, segment-, and reach-scale measurements and observations collected in 2003 and 2004.

Any attempt to classify the myriad geomorphic characteristics and processes that can occur in urbanizing or urban streams with disturbed watersheds that have been altered by land clearing, channelization, and ditching into discrete categories will have limitations and problems. Therefore, the main goal of the classification portion of the study, for use with Duluth-area streams only, is to help show the importance of linking geomorphic characteristics and processes of a specific reach of stream in the context of its position within the landscape and geologic setting.

Description of Study Area

Streams in the Duluth area generally flow perpendicular to the shoreline of Lake Superior, from northwest to southeast (fig. 1). Altitudes of watersheds range from 1,400 ft (above NAVD 27) in the headwaters to 602 ft at Lake Superior. Most of the altitude change occurs along the middle and lower main stems as streams flow over bedrock bluffs and outcrops near the shore of Lake Superior.

The climate of the Duluth area is characterized by cold snowy winters, cool or moderately warm summers, and evenly distributed precipitation (Eichenlaub, 1979). Large variations from average weather can occur because of the proximity of Duluth to Lake Superior. Average annual precipitation is 29–30 in., with 18 in. of average annual evapotranspiration and 10–12 in. of runoff (Young and Skinner, 1974). Mean daily maximum air temperatures range from 22^o F in January to 78^oF in July (Eichenlaub, 1979).

Bedrock is near or at the land surface in the Duluth area and mainly consists of Proterozoic volcanic and igneous rocks, intrusions, and sedimentary rocks (Miller and others, 2002) (fig. 4). Nomenclature for geologic names of bedrock units follows that of the Minnesota Geological Survey. Bedrock type and faulting affects the pattern and density of the stream networks. Right-angle bends in the channel and stream confluences are indicative of underlying contacts between bedrock units or faulting. The Duluth Complex and miscellaneous other intrusive rocks form a prominent steep rocky bluff that parallels the shoreline of the St. Louis River Estuary and Lake Superior.

Figure 4. Bedrock geology of the Duluth, Minn., area (from Miller and others, 2002).

Quaternary surficial deposits in the Duluth area are generally in the upper parts of most of the watersheds but thin or absent from the middle main stems where bedrock is exposed at the land surface (fig. 5). Glacial deposits in the watersheds are from the Late Wisconsinan Lake Superior Lobe (Goebel and others, 1983). Glacial deposits in the headwaters and the upper parts of the watersheds above about 1,200 ft consist of supraglacial drift composed of sandy loamy till of the Cromwell Formation (Goebel and others, 1983; University of Minnesota–Duluth Geology Department and others, 1997; Hobbs, 2004). Later glacial readvances resulted in the deposition of silty loam till of the Lakewood Member, clay loam to silty clay loam till of the Moose Lake Member, and clay till of the informally named Knife River member, all of the Barnum Formation (Hobbs, 2004). In the Duluth area, the Lakewood Member is at altitudes from about 1,150 to 1,200 ft. The Moose Lake Member is at altitudes from about 1,100 to 1,150 ft. The Knife River member is at altitudes up to about 1,050 ft. Shoreline features from wave action and beaches are present from about 1,020 to 1,100 ft and were caused by multiple phases of a glacial lake in the southwest part of Lake Superior during glacial retreats in the Late Wisconsinan and early Holocene (Leverett, 1929; Hobbs, 2004).

Figure 5. Surficial geology and geomorphology of the Duluth, Minn., area (from University of Minnesota–Duluth Geology Department and others, 1997).

Land cover in the Duluth area consists of a mix of mainly urban land (developed/barren), forest, and shrub (Minnesota Department of Natural Resources, Division of Forestry, 2002) (fig. 6). Urban land is concentrated in Duluth and along transportation corridors leading into Duluth (fig. 1). Categories of marsh, lowland shrub, and lowland forest are in wetland settings.

Figure 6. Land cover in the Duluth, Minn., area (from Minnesota Department of Natural Resources, Division of Forestry, 2002).

Population density in 2000 for the area surrounding the studied streams ranged from greater than 5,000 people/mi² near the center of Duluth to less than 250 people/mi² in rural areas (U.S. Census Bureau, 2000) (fig. 7). Comparison of population density maps for 1940, 1960, 1980, and 2000 shows how urban development has spread through the watershed of Miller Creek and along the shore of Lake Superior north of Duluth over the last 20 years (U.S. Census Bureau, 2000; Van Riper, unpublished data).

Figure 7. Population density in the Duluth, Minn., area, 1940 to 2000 (U.S. Census Bureau, 2000; Van Riper, unpublished data).

Methods

The methods used to assess the geomorphic characteristics of Duluth-area streams consisted of a variety of techniques that included data collection and interpretation at watershed, segment, and reach scales.

Watershed Characterization

Drainage boundaries were delineated for the 20 studied watersheds and also for subwatersheds of nine intensive survey sites to examine differences among watershed characteristics. Digital drainage boundaries from the USGS (Minnesota Water Science Center, written commun., 2003) were used and (or) updated based on visual interpretations from USGS 7.5-minute topographic maps. Drainage boundaries were overlaid with thematic maps of bedrock geology (Miller and others, 2002) (fig. 4), Quaternary deposits (University of Minnesota–Duluth Geology Department and others, 1997) (fig. 5), and land cover (Minnesota Department of Natural Resources, Division of Forestry, 2002) (fig. 6) by use of a geographic information system (GIS). These overlays were used to summarize differences among the watersheds for geologic setting and land cover (mainly the amount of urbanization).

Longitudinal profiles were constructed for 15 main-stem streams. Stream lengths were measured with a map measurer between contour lines on USGS 7.5-minute topographic maps. Longitudinal profiles are used to identify changes in slope that are related to geologic features or spatial position within the stream network.

Segment Characterization

Streams were first divided into segments with similar slope and valley confinement, similar to methods described in Montgomery and Buffington (1998). Slope and valley confinement characterizations initially were based on measurements and observations from USGS 7.5-minute topographic maps. Each segment was assigned a unique identification code. Segments were categorized by slope: less than 0.3 percent, 0.3 to 1 percent, greater than 1 to 2 percent, greater than 2 to 4 percent, greater than 4 to 8 percent, or greater than 8 percent. Each segment was classified as confined (valley width less than 2 times the channel width), moderately confined (valley width between 2 and 4 times the channel width), or unconfined (valley width greater than 4 times the channel width). Valley confinement was difficult to estimate from topographic maps; these initial characterizations were confirmed during the reach field surveys and changed, if necessary. Channel-segment boundaries also were confirmed in the field and changed if necessary.

After the field assessments were done, it was determined that valleys in the Duluth area fit into four groups based on modified categories assigned to valleys for Lake Superior tributaries in the western part of the Upper Peninsula of Michigan (Hack, 1965) (fig. 8). Headwater streams in the upstream parts of the watersheds have no valleys, are unconfined (N), and flow over glacial deposits. As drainage area increases and altitude decreases, the stream valleys are confined (C) V-shaped valleys as they cut through the steep, post-glacial shoreline areas, faults, or bedrock. Continuing downstream, the valleys become well developed but stream meanders follow the valley meanders (E). This type of valley and stream meander pattern was called "entrenched" by Hack (1965). The stream-meander pattern for this valley type tends to be irregular and determined by the intersection of the stream and valley with bedrock outcrops, post-glacial lake shorelines, or glacial meltwater valleys. In entrenched valleys, channels commonly intersect valley sides, which combined with enough flow and shear stress, can cause bluff and terrace erosion. Near the mouths, valleys tend to be wide and unconfined and stream-meander patterns are alluvial (A). The valleys are wider than the meander belt and alluvial stream meanders typically are a reflection of fluvial processes of runoff and sediment transport conditions.

Figure 8. Valley development and valley types for Duluth-area streams, Minn.

Aerial photographs from 1939–40, 1956, 1997, and 2002 were used to compare historical and current channel and riparian conditions for the stream segments. Channel width, canopy pattern, type and size of riparian vegetation on each stream bank, stability and position of gravel bars, location of eroding banks, wood loading, and channel position were noted. Riparian vegetation composition was described as deciduous (greater than 70 percent deciduous), coniferous (greater than 70 percent coniferous) or mixed. Riparian vegetation age was described as young (less than 10 years old), mature (between 10 and 40 years old), or old (greater than 40 years old). Riparian vegetation density was described as sparse (greater than 50 percent open ground) or dense (greater than 50 percent forested).

Drainage-network position was assigned to each segment based on the longitudinal profiles and 7.5-minute topographic maps. Main stems were divided into upper, middle, and lower categories, corresponding to main-stem length. Tributary categories were assigned based on the category assignment of the main stem at their confluence.

Predominant categories of bedrock type, glacial landforms/deposits, and land cover were assigned to each segment based on the thematic overlays (Miller and others, 2002; University of Minnesota–Duluth Geology Department and others, 1997; Minnesota Department of Natural Resources, Division of Forestry, 2002; respectively). Bedrock types included sedimentary, gabbro rocks, felsic and mafic volcanic rocks, intrusions, and bedrock buried by glacial deposits. In addition to thematic overlay of bedrock geology from Miller and others (2002), bedrock types along segments were visually checked by use of the non-digital 1:24,000-scale maps of bedrock types and outcrops in Schwartz (1949). Glacial landform/deposits categories were supraglacial drift, outwash, ice-contact, glacial lake, post-glacial shoreline/beach, bedrock surface, till plain, and undifferentiated. Dominant land cover included urban, developing urban, agriculture, forest, wetland, grass, shrubland, and roads.

Reach Characterization

Rapid Field Assessments

Rapid assessments of geomorphic characteristics and processes (mainly observation-based) were done at 48 sites from July 2003 through May 2004 (fig. 1). These sites were on segments with a range in slope and valley type. Rapid assessments were based on qualitative and quantitative field methods described in Thorne (1998) for identifying geomorphic characteristics and processes of streams for multi-disciplinary reconnaissance studies. Thorne's (1998) field methods include observations and measurements of riparian vegetation and geology, valley and flood-plain characteristics, and channel characteristics in relation to valley conditions. Channel morphology, bar formation, substrate, and controls on incision and lateral migration were measured and described. Bank conditions and interpretations also were measured and described. Descriptions also included type of road crossing, potential fish-passage problems through culverts, and any other potential local effects to the stream channel. Culverts are common in Duluth-area streams, and if inappropriately constructed, can be obstructions to fish migration (Gibson and others, 2005). Photographs of the streams were taken during the rapid assessments. Rapid assessment data were used to compare observed characteristics to predictions of dominant geomorphic channel processes based on the geomorphic segment classification.

Intensive Field Surveys

Intensive field surveys were done at nine sites in July, August, and October 2003. These were a subset of sites where rapid field assessments were done (fig. 1). The most common slope and valley types were selected. In addition, sites were concentrated in three watersheds—Mission Creek, Miller Creek, and Lester River—to describe how geomorphic characteristics and processes changed along the longitudinal profiles and also to identify upstream to downstream linkages in geomorphic processes. Assessments included qualitative and quantitative geomorphic assessment of valley, channel, and bank conditions; channel cross-section surveys, substrate measurements, cores of overbank sediment, and measurements of large woody debris (LWD) and pool area. Channel and bank measurements were based on reach-scale descriptions. Reach lengths were based on spacing of riffles, steps, or bedrock falls and ranged from 85 to 273 ft with an average reach length of 170 ft. Channel types were identified as bedrock, colluvial, cascade, step-pool, pool-riffle, plane-bed, or wetland based on definitions from Montgomery and Buffington (1998).

Two channel cross sections were surveyed at each intensive site with an automatic level (Harrelson and others, 1994). A global positioning system was used to verify the location of each cross section. Semi-permanent benchmarks (rebar stakes) were established at each cross section. The cross sections were in riffles. Water-surface and channel slope were measured along the stream distance between the two cross sections.

Quantitative measurements of coarse substrate size in riffles were made by means of Wolman pebble counts (Wolman, 1954). One riffle was chosen (out of the two with cross-section surveys) to conduct the survey. Pebble selection techniques followed Fitzpatrick and others (1998) and the diameter of the b-axis of 100 pebbles were measured and recorded in a field book. If sand or smaller-sized particles were encountered, these particles were tallied as "sand" or "fines".

Overbank sedimentation was measured through exploratory coring with a hand-held soil probe and all cores were described for texture and color in the field by use of the U.S. Department of Agriculture textural triangle and color chart (Munsell Color, 1975; Soil Survey Staff, 1951). Field grading of texture was done by rubbing soil between the fingers (Milfred and others, 1967). Photographs were taken in each reach.

The number, length, and average diameter of woody debris were measured in each reach. Geomorphic function and source for the wood were recorded. Possible functions included pool scour, bank stability, bar stabilizer, sediment trap, or step former. Possible sources for the wood included side slope, upstream, or bank. All wood within the bankfull channel was measured that was larger than 0.1 ft in diameter and 1 ft in length. There is no minimum size standard for LWD. The relatively small minimum size of wood counted in this study (as compared to other studies) was selected to maximize options for comparing LWD in Duluth streams to LWD surveys from other sources.

The area and depth of pools were measured in each reach. Pools were identified by having relatively deep, slow-moving water, and fine-grained bed material compared to the rest of the reach. Pool function also was recorded and included free, LWD, or bedrock/boulder/bank forced.

Channel widths recorded in Government Land Office (GLO) Survey notes from the Duluth area (1857, 1858, 1868, and 1870) were compared to channel and wetted widths measured in cross-section surveys in 2003 to quantify potential channel-width changes at intensive sites. In the GLO surveys, channel widths at section line crossings were measured in links, and 1 link is equivalent to 0.66 ft. It is best to compare GLO surveyed widths at multiple-section line crossings to obtain a sense of data accuracy. For this study, it was assumed that changes greater than +3 ft were appreciable.

Classification

The segment-scale classification developed for Duluth-area streams was based in part on Montgomery and Buffington's (1998) classification. Segments were grouped into 15 categories based on drainage-network position, slope, geologic setting (including a combination of bedrock and glacial landform features), valley type, dominant geomorphic processes, and potential responses to disturbance. The classification was based on results from the rapid field assessments and intensive field surveys. Channel types, geomorphic processes, and interpretations of sensitivity to disturbance summarized from the reach data were extrapolated to other segments with similar drainage-network position, slope, geologic setting, and disturbance mechanisms. Interpretations of potential for geomorphic change due to disturbance are based on observations and evidence for historical geomorphic change as well broad assumptions about the effects of slope and parent material on geomorphic processes.

Geomorphic Characteristics

Geomorphic characteristics of Duluth-area streams and the processes responsible for them are determined by topography and slope, drainage-network position, geologic setting (watershed, segment, and reach scales), valley type, base-level controls, vegetation, and human-caused alterations and disturbance. Current (2005) geomorphic processes reflect short- and long-term responses to disturbances. The source for the disturbance may be upstream or downstream of a particular segment or reach of interest.

Watershed Characteristics

Watershed characteristics help to determine the amount and timing of runoff and sediment from uplands to channels. The 20 studied watersheds are generally small and range in size from less than 1 mi² (Lenroot Creek) to 36 mi² (Lester River) (fig. 1; table 1). All of the streams are underlain by igneous rocks except Mission Creek (fig. 4). Bedrock crops out at the land surface through much of the watersheds but is buried by glacial deposits in the upper parts of the larger watersheds (fig. 5). There is little infiltration and potential for runoff is high where bedrock crops out or is thinly covered by glacial deposits. The Duluth Gabbro complex underlies many of the streams from Stewart Creek northeast to the upper watershed of Tischer Creek. Lester River and Amity Creek are underlain by felsic and mafic volcanic units interspersed with gabbro and mafic intrusions. These bedrock units have different rates of erosion. The sedimentary rocks that underlie the Mission Creek watershed are less resistant to erosion than the volcanics, gabbro complex, or the intrusions. Thus, the Mission Creek drainage pattern is denser and more dendritic than patterns for the other watersheds.

Table 1. Land-cover characteristics for Duluth-area watersheds, Minn. (based on Minnesota Gap Analysis Program Classification Level 3; Minnesota Department of Natural Resources, Division of Forestry, 2002).

[<, less than]

Watershed	Drainage area (square miles)	Developed/barren (urban) (percent)	Cropland (percent)	Grassland (percent)	Upland forest (percent)	Lowland forest (percent)	Upland shrub (percent)	Lowland shrub (percent)	Marsh (percent)	Water (percent)
Mission Creek	10.6	21	4	2	10	46	3	12	1	<1
Sargent Creek	3.1	11	1	1	31	46	1	9	<1	0
U.S. Steel Creek	2.8	34	3	3	22	16	<1	19	2	0
Morgan Park Creek	1.3	24	1	1	43	19	0	13	0	0
Stewart Creek	1.5	16	1	1	47	25	1	9	<1	0
Lenroot Creek	.4	17	<1	1	62	6	0	15	0	0
Knowlton Creek	2.2	31	3	2	24	18	5	17	<1	0
Kingsbury Creek	9.0	34	2	1	8	27	8	19	1	<1
62nd Avenue Creek	1.2	45	1	2	13	14	1	24	1	0
Keene Creek	6.0	30	1	2	11	32	6	18	<1	0
Merritt Creek	2.2	38	1	1	19	24	1	16	<1	0
Miller Creek	9.3	60	<1	1	3	21	6	8	<1	<1
Buckingham Creek	1.1	54	0	1	17	19	<1	8	<1	0
Coffee Creek	1.6	66	1	0	8	20	<1	4	0	0
Brewery and Greys Creeks	3.0	75	<1	0	6	16	<1	3	0	0
Chester Creek	6.7	33	1	1	9	36	7	12	1	<1
Tischer Creek	7.3	34	<1	3	11	44	2	5	<1	<1
34th to 50th Ave. Creeks	3.6	80	<1	<1	5	11	<1	2	<1	<1
Amity Creek	16.7	7	2	9	7	69	1	5	<1	<1
Lester River	36.0	3	2	5	7	68	2	11	<1	1

Land cover in the studied watersheds is predominantly composed of urban land (developed/barren), forest, or shrub (fig. 6, table 1). Percentages of urban land range from 3 percent in the Lester River watershed to 80 percent in the 34th to 50th Ave. Creeks watershed. Urban land is most concentrated near the mouths along the shoreline of the St. Louis River estuary and Lake Superior; however, urban land extends throughout many of the watersheds, especially those that are bisected by major transportation corridors along U.S. State Highway 53, U.S. Interstate 35, and U.S. State Highway 2 (figs. 1 and 6). Urban watersheds have higher potential for increased runoff rates and volumes and for changes in sediment inputs compared to forested watersheds. Watersheds with more than 50 percent urban land include Miller Creek, Coffee Creek, Buckingham Creek, Brewery and Greys Creeks, and 34th to 50th Ave. Creeks. Seven urban streams near downtown Duluth run underground at their mouths (table 2). Watersheds that are predominantly forest include Mission Creek, Sargent Creek, Morgan Park Creek, Stewart Creek, Lenroot Creek, Amity Creek, and Lester River (table 1). Watersheds that have had noticeable urban development from 1980 to 2000 include Mission Creek, Morgan Park Creek, Stewart Creek, Lenroot Creek, Miller Creek, and Lester River (fig. 7). Geomorphic effects from urban development are concerns for Miller Creek, Chester Creek, and Amity Creek (Camp Dresser and McKee, Inc., 2000).

Table 2. Conditions at the mouths of Duluth-area streams, Minn.

Stream	Conditions
Mission Creek	Channel open to the St. Louis River
Sargent Creek	Channel open to the St. Louis River estuary
U.S. Steel Creek	Channel open to the St. Louis River estuary
Morgan Park Creek	Channel open to the St. Louis River estuary
Stewart Creek	Channel open to the St. Louis River estuary
Lenroot Creek	Channel open to the St. Louis River estuary
Knowlton Creek	Channel open to the St. Louis River estuary
Kingsbury Creek	Channel open to the St. Louis River estuary
62nd Ave. Creek	Runs underground by Grand Ave.
Keene Creek	Channel open to the St. Louis River estuary
Merritt Creek	Channel open to the St. Louis River estuary
Miller Creek	Runs underground through downtown Duluth
Buckingham Creek	Runs underground through downtown Duluth
Coffee Creek	Runs underground through downtown Duluth
Brewery and Greys Creeks	Runs underground through downtown Duluth
Chester Creek	Runs underground in downtown Duluth
Tischer Creek	Channel open to Lake Superior
34th to 50th Ave. Creeks	Some channels open, others run underground
Amity Creek	Channel open to Lake Superior
Lester River	Channel open to Lake Superior

Duluth-area main stems have complex longitudinal profiles (fig. 9) that are very different from the typical longitudinal profile shown in figure 2 for geologically old landscapes with homogeneous geologic deposits. Irregularities in the Duluth stream profiles are attributed to differences in the erodibility of geologic materials, extent of landforms, or changes in base level (Hack, 1965). Specifically, Duluth area longitudinal profiles reflect bedrock type and distribution of fault zones, variations in the texture of glacial deposits, glacial landforms (such as moraines and outwash plains), postglacial landforms from beach ridges, shorelines or melt-water valleys, and the present level of Lake Superior. Reaches with steep slopes are prone to long-term erosional conditions, whereas reaches with gentle slopes are prone to long-term depositional conditions (Knighton, 1998). Reaches in a transition from steep to gentle slopes may change from erosional to depositional if upstream inputs of water and sediment change.

Figure Nine

Figure 9. Longitudinal profiles of streams in the A, southwestern, and B, northeastern parts of the Duluth, Minn., area.

The upper main stems tend to have gentle slopes for headwaters in glacial deposits at altitudes higher than about 1,200 ft. Irregularities in the longitudinal profiles are common from about 1,200 to 1,000 ft because of the boundary among glacial deposits, bedrock, and glacial lake shorelines. The steep slopes are on middle and lower main stems and tributaries indicate where the streams intersect bedrock (fig. 4). Streams flowing over the Duluth Complex gabbro intrusions have the steepest slopes. Below 1,100 ft, longitudinal profiles for Mission Creek and Sargent Creek are concave upward and less steep than streams to the northeast because Mission and Sargent Creek intersect the less resistant Animikie sedimentary rocks and glacial lake plain. In the northeastern part of the city, Lester River, Amity Creek, and a portion of Tischer Creek have more gentle slopes than streams in the center of the city but steeper slopes than Mission and Sargent Creeks because the streams intersect Keweenawan volcanic rocks. Only a few streams in the southwestern part of the study area have enough length on till plain or built-out land to have a gentle slope near their mouths in the St. Louis River estuary (fig. 1).

Segment Characteristics

Streams with gentle slopes (less than 0.3 percent) generally were headwater streams flowing on glacial deposits with no valley development with wetland vegetation (lowland shrub or forest); whereas streams with moderate to steep slopes flowed on bedrock and had confined or entrenched valleys (tables 3 and 4). Alluvial valleys were rare because of the small watershed sizes, bedrock control, and geologically young age of the streams (less than 10,000 years). Main-stem segments with moderate slopes typically were in the boundary zone between glacial deposits and bedrock. Tributaries to middle or lower main stems most commonly had 2–8 percent slopes in confined valleys. The combination of steep slopes and confined or entrenched valleys are indicative of a high potential for landslides along valley sides, especially if sandy glacial lake shoreline deposits are present. If valley sides are made up of unconsolidated deposits, the potential for landslides, mass wasting, and channel erosion increases if runoff and flood peaks increase (Fitzpatrick and others, 1999).

Table 3. Number of stream segments in each slope/drainage-network position category for Duluth-area streams, Minn.

[<, less than; >, greater than]

Slope category (percent)	Upper main stem	Middle main stem	Lower main stem	Upper tributary	Middle tributary	Lower tributary	Total
<0.3	4	0	0	3	0	0	7
0.3-1	10	7	8	5	4	1	35
>1-2	4	19	6	3	9	1	42
>2-4	7	6	10	6	15	10	54
>4-8	2	4	9	3	3	10	31
>8	1	7	7	1	2	1	19
Total	28	43	40	21	33	23	188

Table 4. Number of stream segments in each slope/valley type category for Duluth-area streams, Minn.

[<, less than; >, greater than]

Slope category (percent)	No valley	Confined (origin not determined)	Entrenched	Alluvial	Bedrock (confined)	Total
<0.3	7	0	0	0	0	7
0.3-1	26	1	4	4	0	35
>1-2	17	10	12	0	3	42
>2-4	1	37	10	1	5	54
>4-8	0	15	8	0	8	31
>8	0	0	9	0	10	19
Total	51	63	43	5	26	188

Characteristics for all the segments are listed in table 5, with a list of abbreviations and definitions in table 6. Drainage-network position, slope, valley type, bedrock type, glacial landform, and dominant land cover are described in table 5. The table has site identification numbers for segments with rapid assessment or intensive survey sites. Based on field reach-scale data and air photos, characteristics of channel type, dominant substrate, mode of sediment transport, flood attenuation, observed/expected causes for disturbance, presence of storm drains, and potential for geomorphic change are described for segments with field sites and extrapolated for other segments. Almost all of the segments have one or more observed or expected causes for disturbance, usually land clearing, urban development, road drainage, and inputs from upstream segments. The only exception is six tributary segments near the middle/lower main stem of Mission Creek that have complete forest cover and no roads.

Table 5. Geomorphic characteristics of stream segments for Duluth-area streams, Minn., 2003–2004.

[See table 6 for explanation of abbreviations. Stream segments are shown on figure 13. Data in italics are extrapolated from reach assessments of sites with the same segment category. Data in bold are rapid assessment or intensive survey sites. USGS, U.S. Geological Survey; ID, identification number; --, no detailed measurements taken; %, percent; >, greater than; <, less than]

Table is large; scroll down and to the right to view entire table.

Segment ID	Stream name	Topographic map	USGS reach ID	Drainage-network position	Slope %	Valley type	Bedrock type	Glacial landform	Dominant land cover	Inter- mittent	Segment ID	Channel type	Dominant substrate	Mode of sediment transport	Flood attenuation	Dominant geomorphic processes	Observed/expected causes for disturbance	Storm-drain outfalls	Potential geomorphic changes	Sensitivity to disturbance	Geomorphic segment category	Segment ID
b1	Buckingham Creek	Duluth	--	LM	>8	E	GAB	BED	U	N	b1	BE	BE	T, SIp	NO	EX	UP, HI, UR	Y	Y	LO	B	b1
b2	Buckingham Creek	Duluth Heights	8	MM	>1–2	N	GAB	BED	U	N	b2	PB	GV	SIa	FP	A	UP, CA, UR, GV, CL	N	Y	MOD	M1	b2
b3	Buckingham Creek	Duluth Heights	--	UT	>2–4	C	GAB	BED	F	Y	b3	CO	CO/BO	SOi, SOw	NO	I, W	CL, CA, RO, GV	N	Y	MOD	T2	b3
c1	Chester Creek	Duluth	9	LM	>4–8	E	VOL	BED	U	N	c1	BE	BO/BE	T, SOb	NO	EX, EBL	UP, HI, UR	Y	Y	LO	B	c1
c2	Chester Creek	Duluth	43	MM	>1–2	N	GAB	OW/BED	U/F	N	c2	PR/SP	GV	SOi, SOw, T, SIc, SIp	FP	I, EBA, W, LM, BF	UP, UR, RO, CL	Y	Y	HI	M1	c2
c3	Chester Creek	Duluth, Duluth Heights	--	UM	0.3–1	N	GAB	SD/BED	S/W	N	c3	WE	GV	SOw, SOi, T, SIa	WE	I, W, EBA, A	CL, UR, RO, CA	N	Y	MOD	W.3	c3
c4	Chester Creek	Duluth	--	MM	<0.3	N	BUR/GAB	OW	U	N	c4	AR	FINES	T, SOw	NO	EBA, LM, W, A	UR, CA, RO, CL	Y	Y	MOD	W	c4
c5	Chester Creek	Duluth, Duluth Heights	--	UM	>1–2	N	GAB	SD	S/U	N	c5	WE	GV	SOw, SOi	FP	I, W, EBA	CL, UR, RO	Y	Y	HI	U1	c5
c6	Chester Creek	Duluth, Duluth Heights	--	UM	>1–2	N	GAB	BED	U	N	c6	AR/PB	CO	SOb, SOi, SOw, T	NO	I, W, EBA, LM	UR, RO, CA, CL	Y	Y	HI	U1	c6
c7	Chester Creek	Duluth	--	LT	>2–4	C	VOL	BED	F/U	Y	c7	CO	BO	SOi, SOb, SOw	NO	I, W, EBA, EBL, EX	UR, RO, CL	Y	Y	MOD	LT	c7
k1	Keene Creek	West Duluth	--	LM	0.3–1	N	GAB	TP	U	N	k1	AR	BO	T, SIa	NO	A	UP, UR, RO, CA	N	Y	LO	A	k1
k2	Keene Creek	West Duluth	--	LM	>2–4	C	GAB	BED	U	N	k2	BED/CA	BO	SOb, SOw, T	NO	W, EBA, EBL, EX	UP, RO, UR	Y	Y	MOD	L2	k2
k2a	Keene Creek	West Duluth	42	LM	>1–2	N	GAB	TP	U	N	k2a	PB/AR	BO	T	NO	ST	UP, UR, RO, CA	Y	Y	LO	L1	k2a
Segment ID	Stream name	Topographic map	USGS reach ID	Drainage-network position	Slope %	Valley type	Bedrock type	Glacial landform	Dominant land cover	Inter- mittent	Segment ID	Channel type	Dominant substrate	Mode of sediment transport	Flood attenuation	Dominant geomorphic processes	Observed/expected causes for disturbance	Storm-drain outfalls	Potential geomorphic changes	Sensitivity to disturbance	Geomorphic segment category	Segment ID
k3	Keene Creek	West Duluth, Duluth Heights	--	MM	>4–8	E	GAB	BED	U/F	N	k3	CA/BE	BO	T, SOw, SOb	NO	W, EX, EBL, EBA	UR, RO, UP, GV, CL	Y	Y	MOD	B	k3
k4	Keene Creek	Duluth Heights	6	MM	>1–2	E	GAB	BED/SH	F/U	N	k4	PR/SP	BO/CO	T, SOw, SOi	NO	W, EBA, I, LM	UP, GV, UR, RO, CL	N	Y	HI	M1	k4
k5	Keene Creek	Duluth Heights	--	UM	0.3–1	N	GAB	SD/BED	F/U/W	N	k5	WE	SA/FINES	SOw, SOi, SIa	WE	W, EBA, I, A	UR, RO, GV	N	Y	MOD	W.3	k5
k6	Keene Creek trib	Duluth Heights	5	MT	>4–8	B	GAB	BED/SH	U/F	N	k6	CA/CO	BO	T	NO	EX	UR, RO, GV	N?	Y	MOD	BT	k6
kg1	Kingsbury Creek	West Duluth	--	LM	0.3–1	A	BUR	TP	U	N	kg1	AR	BO	T, SIa	NO	ST, A	UP, RO, CL	Y	Y	LO	A	kg1
kg2	Kingsbury Creek	West Duluth	--	LM	>4–8	C	GAB	BED/SH	U	N	kg2	BE	BE	SOw, SOi, SIc, T	NO	W, EBA, EX, BF	UR, RO, UP	Y	Y	MOD	B	kg2
kg2a	Kingsbury Creek	West Duluth	--	LM	>2–4	C	GAB	BED	F/U	N	kg2a	BED/CA	BO/BE	SOw, T	NO	W, EBA, EX, BF	UP, RO, UR	N	Y	LO	L2	kg2a
kg2b	Kingsbury Creek	West Duluth	--	LM	>4–8	C	GAB	BED	U	N	kg2b	BE	BE	SOw, T	NO	W, EBA, EX, BF	UR, RO, UP	N	Y	LO	B	kg2b
kg2c	Kingsbury Creek	West Duluth	--	LM	>8	C	GAB	BED	U	N	kg2c	BE	BE	SOb , SOw, T	NO	EX	UP, UR, RO	Y	Y	LO	B	kg2c
kg2d	Kingsbury Creek	West Duluth	--	LM	>2–4	C	GAB	BED	U	N	kg2d	BED/CA	BO	SOb , T, SOw, SIc	NO	EX, EBA, W, BF	UP, UR, RO	Y	Y	MOD	L2	kg2d
kg3	Kingsbury Creek	West Duluth	41	MM	>2–4	B	GAB	SH/SD/BED	U/F	N	kg3	SP/CA	BO	SOb, SOw, T, SIc	NO	EBL, EBA, LM, BF, W	UP, RO, UR	N	Y	HI	M2	kg3
kg4	Kingsbury Creek	West Duluth	--	MM	0.3–1	N	GAB	SD	W/U	N	kg4	AR	SA/FINES	SOb, T, SOw, SIc	WE	EBA, I, W, A	UP, CA, UR, RO, CL	N	Y	MOD	M.3	kg4
kg5	Kingsbury Creek	Duluth Heights	--	UM	<0.3	N	GAB	SD/BED	W	N	kg5	AR/WE	FINES	SOw, T, SIa	WE	EBA, LM, I, A	CA, RO, CL	N	Y	MOD	W	kg5
Segment ID	Stream name	Topographic map	USGS reach ID	Drainage-network position	Slope %	Valley type	Bedrock type	Glacial landform	Dominant land cover	Inter- mittent	Segment ID	Channel type	Dominant substrate	Mode of sediment transport	Flood attenuation	Dominant geomorphic processes	Observed/expected causes for disturbance	Storm-drain outfalls	Potential geomorphic changes	Sensitivity to disturbance	Geomorphic segment category	Segment ID
kg6	Kingsbury Creek	Duluth Heights	--	UM	0.3–1	N	GAB	SD/BED	W	N	kg6	AR/DW	SA/FINES	SOw, SIa, T	WE	A, I, W	UP, RO, CA, CL	N	Y	LO	W.3	kg6
kg7	Kingsbury Creek	Duluth Heights	--	UT	>1–2	N	GAB	SD/BED	W	Y	kg7	PR	CO	SOw, SOi	NO	I	CL, UR, RO	N	Y	MOD/HI	T1	kg7

Geomorphic Characteristics and Classification of Duluth-Area Streams, Minnesota

By Faith A. Fitzpatrick, Marie C. Peppler, Michele M. DePhilip1, and Kathy E. Lee

The PDF for the report is 7,385 kb

Table of Contents

Figures

Tables

Conversion Factors, Vertical Datum, and Abbreviated Units of Measurement