THIEF RIVER WATERSHED FLUVIAL GEOMORPHOLOGY REPORT MN DNR June 2015

TABLE OF CONTENTS AUTHORS Dave Friedl Jason Vinje Lori Clark Stephanie Klamm

CONTRIBUTORS / ACKNOWLEDGEMENTS Corey Hanson, Red Lake Watershed District

Cover Photo Credit: Corey Hanson, October 2011.

FOR MORE INFORMATION CONTACT: DNR Information Center 500 Lafayette Road St. Paul, MN 55155‐4040 (651) 296‐6157 (Metro Area) 1 888 MINNDNR (1‐888‐646‐6367) TTY (651) 296‐5484 (Metro Area) 1 800 657‐3929 http://www.mndnr.gov

Thief River Watershed Fluvial Geomorphology Report  June 2015

TABLE OF CONTENTS

List of Acronyms ...... iii List of Figures ...... iv List of Tables ...... vi Executive Summary ...... vii Introduction ...... 1 Fluvial Geomorphology Study ...... 1 Healthy Watersheds ...... 2 Current Resource Concerns ...... 3 General Description ...... 4 Surface Waters ...... 4 Geology ...... 8 Land Use ...... 10 Methodology ...... 15 Field Assessments ...... 15 Data Analysis ...... 20 Hydrology...... 21 Annual Precipitation ...... 21 Stream Gages ...... 22 Mean Annual Flow ...... 23 Flood Frequency ...... 24 Magnitude & Frequency ...... 26 Streambank Erosion...... 28 BANCS Model ...... 28 Thief River ...... 29 ...... 30 Mud River ...... 31 County Ditch 20 ...... 32 Stability ...... 33

Thief River Watershed Fluvial Geomorphology Report  June 2015

Discussion ...... 34 Thief River ...... 34 Thief Lake to Agassiz National Wildlife Refuge ...... 34 Agassiz National Wildlife Refuge to Thief River Falls ...... 38 Moose River...... 40 Mud River ...... 41 County Ditch 20 ...... 44 Recommendations ...... 47 Thief River ...... 47 Thief Lake to Agassiz National Wildlife Refuge ...... 47 Agassiz National Wildlife Refuge to Thief River Falls ...... 49 Moose River...... 50 Mud River ...... 51 County Ditch 20 ...... 52 Conclusions ...... 54

Thief River Watershed Fluvial Geomorphology Report  June 2015

List of Acronyms

BANCS Bank Assessment for Nonpoint Source Consequences of Sediment BEHI Bank Erosion Hazard Index CD County Ditch CR County Road CSG Cooperative Stream Gaging HUC Hydrologic Unit Code IBI Index of Biological Integrity JD Judicial Ditch MN DNR Department of Natural Resources NBS Near Bank Stress NLCD National Land Cover Dataset NWI National Wetland Inventory NWR National Wildlife Refuge RLWD Red Lake Watershed District SPI Stream Power Index TRW Thief River Watershed USGS United States Geological Survey WMA Wildlife Management Area

Thief River Watershed Fluvial Geomorphology Report  June 2015 iii

List of Figures

Figure 1. Map of the watercourse types within the Thief River Watershed ...... 5 Figure 2. National Wetland Inventory of the Thief River Watershed ...... 6 Figure 3. Public Waterbasins and Watercourses within the Thief River Watershed ...... 7 Figure 4. Simplifed drawing of ice lobes about 14,000 years ago ...... 8 Figure 5. Geology of the Thief River Watershed ...... 9 Figure 6. 2011 Land use within the Moose River drainage area ...... 11 Figure 7. 2011 Land use within the Mud River drainage area ...... 12 Figure 8. 2011 Land use within the County Ditch 200 drainage area ...... 13 Figure 9. 2011 Land use within the County Ditch 20 drainage area ...... 14 Figure 10. 2011 Land use within the County Ditch 30/18 drainage area ...... 15 Figure 11. 2011 Land use within the Thief River main stem drainage area ...... 16 Figure 12. Map of fluvial geomorphology study sites...... 17 Figure 13. Variables determining Rosgen stream classification ...... 18 Figure 14. Annual precipitation totals of Thief River near Dakota Junction, MN ...... 21 Figure 15. Annual mean flow at USGS stream gage 0507600, near Thief River Falls ...... 23 Figure 16. Mean annual flow and precipitation for USGS stream gage 0507600 ...... 24 Figure 17. Daily mean discharge flows at USGS stream gage 05076000 near Thief River Falls, MN 25 Figure 18. Flow duration curves comparing 30‐year time periods ...... 27 Figure 19. BANCS streambank annual erosion estimate ...... 28 Figure 20. Thief River between Thief Lake and Agassiz NWR ...... 29 Figure 21. Photo along upstream Moose River reach ...... 30 Figure 22. Lower Moose River reconnaissance reach ...... 30 Figure 23. Sand bedload moving along Mud River channel bottom ...... 31 Figure 24. Lower Mud River reconnaissance reach ...... 31 Figure 25. CD 20, just downstream of Hwy 54 ...... 32 Figure 26. CD 20, approximately halfway through the downstream study reach ...... 32 Figure 27. Pfankuch Stability Ratings at the study sites ...... 33 Figure 28. Water surface profile from the outlet of Thief Lake to the Agassiz NWR boundary, using 1m DEM ...... 35 Figure 29. Cross section of a meander bend at site 3, using 1m DEM ...... 37 Figure 30. Water surface profile from Agassiz NWR to Thief River Falls, using 1m DEM ...... 39

Thief River Watershed Fluvial Geomorphology Report  June 2015 iv

Figure 31. Water surface profile from the outlet of the Moose River Impoundment to Thief Lake. using 1m DEM ...... 41 Figure 32. Water surface profile from the outlet of Thief Lake to the Agassiz NWR boundary, using 1m DEM ...... 42 Figure 33. Water surface profile of County Ditch 20, using 1m DEM ...... 44 Figure 34. Stream channel and floodplain connectivity ...... 47 Figure 35. Meander migration ...... 49 Figure 36. Rosgen B and C Stream Types ...... 50 Figure 37. Examples of poor riparian management...... 50 Figure 38. Examples of gully erosion within the Moose River Watershed ...... 51 Figure 39. Poor buffer condition along CD 20 ...... 53 Figure 40. Sand bedload deposited in Thief River near outlet of CD 20 ...... 53

Thief River Watershed Fluvial Geomorphology Report  June 2015 v

List of Tables

Table 1. Breakdown of watercourse types by subwatershed ...... 5 Table 2. Breakdown of land use by subwatershed ...... 10 Table 3. Description of stream gage locations within the Thief River Watershed ...... 22 Table 4. Streambank erosion summary by reconnaissance reach ...... 29 Table 5. Summary of study site characteristics ...... 34 Table 6. Summary of stressor sources and protection and restoration strategies ...... 48

Thief River Watershed Fluvial Geomorphology Report  June 2015 vi

EXECUTIVE SUMMARY

The MN Department of Natural Resources conducted a study of river stability within the Thief River Watershed from 2010 to 2012. The purpose of the this fluvial geomorphology study was to develop a better understanding of the impact of natural disturbances and past, present, and proposed land use practices on river channel stability. In the spring of 2010, a reconnaissance effort assessed 72.5 miles of watercourses within the Thief River Watershed. In 2011 and 2012, 13 sites were selected for more intensive field surveys to assess stability.

Reconnaissance occurred along two reaches of the Moose River and one intensive site was located adjacent to Moose River Rd. The Moose River subwatershed was dominated by emergent and woody wetlands. Upstream of CR 54/1, the river is classified as a perennial ditch. The streambank erosion estimate for this area was the highest estimate in the entire Thief River Watershed. The riparian buffer was very poor in many locations along this reach. At the intensive study site, located east of Pine Island Rd., the Moose River was inundated with fine particles and streambank erosion was evident. This stretch of the river would benefit from improvements to the riparian buffer and the reduction of upland sediment reaching the river channel.

Reconnaissance occurred along two reaches of the Mud River and three intensive sites were assessed. The upstream end of the reconnaissance had the second highest estimated erosion rate, after the upper Moose River, in the entire Thief River Watershed. This stretch of the Mud River is a channelized watercourse, immediately adjacent to Hwy 89. The Mud River for most of its course tended to have a low width‐to‐depth ratio. The river can access the floodplain at approximately the 50‐yr. flood event at some locations, but it is also confined in others. On the upstream end of the Mud River the stability would improve if the width and vegetative plant diversity of the riparian buffer increased and the floodplain accessibility was improved. Downstream of the city of Grygla, the riparian buffer condition was still poor and the inundation of excessive sediment was evident; however, the estimated streambank erosion estimates decreased to a moderate level on the downstream end. The stability of the downstream reach would improve if the existing buffers are maintained and new buffers are added where they are insufficient.

The majority of County Ditch 20 was assessed during the reconnaissance effort and three intensive sites were assessed. The most upstream site was located west of 400th Ave. NE. This area was moderately stable. The riparian vegetation was robust and diverse, and the channel bottom appeared to be very stable. The channel was efficient at moving sediment downstream with a tight width‐to‐depth ratio. The floodplain width was narrow along most of CD 20. The estimated bank erosion along the four miles on the upstream end was moderately high. The stability of this area would increase if additional capacity was built within the floodplain. The CD 20 subwatershed is also an area of flooding concern for local landowners. The downstream portion of CD 20 the streambanks were fairly stable. During the reconnaissance, a high sand bed load was observed moving in the channel. Near Elm Lake WMA the channel bottom substrates became more in flux

Thief River Watershed Fluvial Geomorphology Report  June 2015 vii

and the channel showed signs of excessive sediment. Sediment sources upstream within the channel and on the uplands should be alleviated.

The majority of the Thief River was assessed during the reconnaissance effort. Upstream of Agassiz National Wildlife Refuge (NWR) the Thief River was channelized at the outlet of Thief Lake and was moderately stable. Spoil piles were confining the floodplain capacity on the upper end. Near the center of this reach, from about 140th St. to CR 6, the Thief River was stable and in good condition. Effort should be made to maintain the river channel and riparian buffers’ current condition. By site 3, which was located just south of CR 6, fine sediment was accumulating in the river channel to a point beyond what the river could effectively move downstream. The effort to identify upland sources of sediment should continue and priority should be placed on BMPs to minimize the amount of sediment reaching the channel. Immediately north of the Agassiz NWR boundary the river again rated as only moderately stable due to excess sediment in the channel bottom and some streambank erosion. The river was previously channelized in this area as well. The old meander scrolls are still present on the landscape and the land use remains as a riparian buffer. Opportunities to remeander the Thief River north of the Agassiz NWR boundary could be explored. The restored meandering channel should be designed as a narrow channel capable of effectively moving sediment downstream.

In addition to the fluvial geomorphology work, an assessment of the historical precipitation and stream flow data was completed. Annual precipitation rates have fluctuated since 1890. Most recently, the annual rates have been consistently above average, except for handful of years since 1995. Real‐time stream flow is monitored at 3 locations in the Thief River Watershed. The USGS stream gage near 140th Ave. NE has the longest period of record, starting in 1909. Comparisons were made between the last three 30‐year time periods, 1925‐1954, 1955‐1984, and 1985‐2014. The Q10 flow, or flow that was exceeded 10% of the time, for those 30‐yr. periods was 959, 1081, and 1091 cfs, respectively. The Q90 flow, or flow that was exceeded 90% of the time, was 0, 0, and 0.2 cfs, respectively. Though there was not much variation in the extreme flood events, there has been a large increase in the frequency of small flood events. As a percent, from 1909‐1924, 1925‐ 1954, 1955‐1984, and 1985‐2014 discharge was greater than the estimated bankfull channel capacity 2.4%, 2.5%, 6.3%, and 10.2% of the days flow was recorded, respectively.

In summary, very few watercourses within the Thief River Watershed are stable. With the large amounts of glacial organic deposits and current day wetlands in the eastern portion of the Thief River Watershed, much of the headwaters area was channelized over 100 years ago. This created watercourses with narrow floodplains and limited capacity at higher flows. In addition, excessive sediment, increased frequencies of channel‐forming discharge (as estimated by the 1.5‐year return interval), and riparian buffers in poor condition are all drivers of channel instability. There are numerous opportunities for improvement to river stability within the Thief River Watershed.

Thief River Watershed Fluvial Geomorphology Report  June 2015 viii

INTRODUCTION

FLUVIAL GEOMORPHOLOGY STUDY The purpose of the this fluvial geomorphology study was to develop a better understanding of the impact of natural disturbances and past, present, and proposed land use practices on river channel stability. Stability is defined as a river or stream’s ability, in the present climate, to transport the stream flows and sediment of its watershed over time in such a manner that the channel maintains its dimension, pattern, and profile without either aggrading or degrading (Rosgen 1996).

Stability of rivers depends on maintaining a natural equilibrium between sediment supply (both size of particles and quantity), channel slope, channel length, and stream flow (Lane 1955, Leopold 1994). An alteration of the conditions of equilibrium through tectonic changes or by changes in the hydrological regimen, including changes in sediment and water yield, will result in altering the floodplain and lead to degradation and terrace formation, or to aggradation. Stream bank erosion, excess sediment in streams, and flashy stream flows were all highlighted as major concerns in the Red Lake Watershed District’s (RLWD) 10‐yr. Comprehensive Plan (2006). These issues are symptoms of a river system not currently in a state of equilibrium. For this fluvial geomorphology study, specific river reaches were selected for intensive assessment to determine which factors of stability are shifting the state from equilibrium to instability.

In May of 2010, MN Department of Natural Resources (DNR) and RLWD staff completed a reconnaissance effort to assess 72.5 miles of waterways in the Thief River Watershed (TRW). The primary goal of this assessment was to identify stream types and general locations of erosional and depositional processes occurring on the landscape. Some detailed measurements of the stream channel and bank, including vegetation characteristics, were collected. This data was used to estimate stream bank erosion rates using the Bank Assessment for Non‐point Source Consequences of Sediment (BANCS) Model (Rosgen 1996, 2001b, 2006b).

After the reconnaissance effort was complete, intensive study sites were selected to gain an overall perspective of the stability of the waterways within the TRW. Fifteen sites were selected across the watershed, including sites on the Mud River, Moose River, County Ditch 20, and the main stem of the Thief River. Thirteen of the 15 sites were then assessed using Rosgen methodology. These assessments followed the procedures outlined in the “River Stability Field Guide” levels I‐IV (Rosgen 2008).

The size, shape, and pattern of river channels are closely related to the flows that they transmit (Leopold 1994, Lane 1955). Leopold (1994) stated that the pattern of the channel itself is formed by flows which apply sufficient force to mold the channel but are also retained within the channel, rather than those which occupy the entire cross‐section of the valley during periods of a flood. This channel‐forming discharge is equal to bankfull discharge for alluvial streams in equilibrium (FISRWG 1998). In many rivers, the bankfull stage occurs on the average of once each

Thief River Watershed Fluvial Geomorphology Report  June 2015 1

year or perhaps once every two years (Leopold 1994). Other research studies have narrowed down that range by determining that channel‐forming discharge most often correlated with the 1.5‐ 1.58‐yr. recurrence flow (Dury 1973, Hey 1975).

With bankfull and higher flows naturally occurring every 1.5 years, water flowing outside of the immediate river channel is a relatively common occurrence. This makes attaining river stability a challenging aspiration when rivers are managed independently of their floodplains. A river channel and floodplain are dynamic features that function as a single hydrologic and geomorphic unit, characterized by frequent transfers of water and sediment between the two components (Kondolf 1997).

Channel pattern describes how a river looks in the plan view, as seen from above, and is another measure of channel stability. Though it may first appear that rivers are haphazardly zig‐zigging across a landscape, their pattern geometry has been shown to be directly related to channel size and dominate discharge (Inglis 1949, Leopold and Wolman 1960, Williams 1986, Leopold 1994). Based on previously collected data, channel cross‐sectional dimensions, such as bankfull width, were compared against channel geometry measurements. Langbein and Leopold (1966) found meander length (wavelength) and amplitude compared to bankfull width to have a nearly linear relationship. In addition, meander length and radius of curvature also had a nearly linear relationship. Years later, in an effort to create a more robust set of data, G.P. Williams (1986) evaluated the theory of predicting channel geometry using data from 194 sites. Williams was able to derive 40 empirical equations relating channel width to channel geometry. Rosgen (1994) minimized the variability in these relationships by stratifying by Rosgen stream type.

HEALTHY WATERSHEDS Healthy watersheds with biologically diverse and connected ecosystems function to produce clean water. Healthy watersheds also produce other ecosystem services and products that contribute to the state’s economic and social vitality (e.g., habitat, fish, wildlife, timber, and recreation). The MN DNR uses a five component framework to describe watersheds as systems. These five components are biology, hydrology, geomorphology, connectivity, and water quality.

Understanding the interplay between these components will make it easier to identify the root cause of an issue that is impairing one aspect of a healthy watershed. For example, if land use changes increase the flashiness of river flows this may lead to an unstable stream condition. To adjust to these changes, the river may down cut and the banks erode as it attempts to find a new equilibrium with the changes to hydrology. As a result of this instability, variables like instream and overhead cover, substrate composition, pool quality, holding cover velocity, temperature, oxygen, spawning habitat, habitat diversity, rearing, and IBI scores would all be expected to degrade. In contrast, an evolution or restoration from the unstable form to a stable river form would result in a reversal of these negative consequences. Though the focus of this report is on only two of the five components, geomorphology and hydrology, the results are invariably tied to the condition of the other three.

Thief River Watershed Fluvial Geomorphology Report  June 2015 2

CURRENT RESOURCE CONCERNS The RLWD’s 10‐year plan covers many of the local resource concerns and highlights specific areas with flooding concerns. Due to the terrain, draining of wetlands for cropland, and/or under‐ designed structures, the TRW experiences frequent flooding (RLWD 2006). Spring flooding is almost an annual occurrence. There is little relief on the landscape once rivers overtop their banks; therefore, during even minor flooding events water spreads out across the landscape.

During the development of the RLWD’s 10‐year Comprehensive Plan, the following flooding issues in the TRW ranked as the highest concerns:  Agricultural crop damages,  Overland flooding in Goodridge,  Farmstead flooding, and  Flooding along ditches 20 and 200.

In an effort to accommodate multiple, sometimes conflicting, needs of landowners in the TRW, the RLWD is focused on resolving the conflict between the desire of landowners along the State Ditch 83 portion of the Thief River to have it repaired for flood damage reduction purposes and the interest of other natural resource‐related agencies and special interest groups in finding environmentally friendly alternatives (RLWD 2006).

The project work team’s conceptual plan included four basic features (RLWD 2006): • Diverting some of the water away from Thief River through a constructed channel from Elm Lake/Agassiz NWR, joining the Thief River four miles north of Thief River Falls • Increasing storage throughout the TRW. Specifically, off‐channel dry impoundments along CD 20 • Providing incentives and education to promote land use changes, including recommendations for changing agricultural practices in floodprone areas • Spot cleaning of Thief River (SD 83) where sedimentation or snags are major flow restrictions or deflect flow into streambanks.

In addition to flooding concerns, the causes of excessive sedimentation have also been studied. In 1910, Judicial Ditch (JD) 11 was constructed in the Mud River subwatershed to improve drainage. Judicial ditch 11 outlets into Agassiz NWR. It was originally constructed with a top width of 59 feet and bottom width of 40 feet, and a mean depth of 9.5 feet (Schottler and Engstrom 2011). Within one of the Agassiz pools, JD11 is nearly completely filled with sediment for 3.7 miles. A Soil and Water Assessment Tool (SWAT) model was developed for the Thief River Watershed. Though the model results showed that the Mud River subwatershed did contribute to the sediment yield, the Thief River main stem subwatershed, from Thief Lake to Agassiz NWR, was the highest contributor of sediment (Houston 2010). In 2012, Marshall County SWCD was awarded $178,750 in Clean Water Legacy funds to provide one‐time rental payments and cost‐share funds to help with planning, establishment, and provide education to landowners about the benefits of riparian buffers and side‐water inlets.

Thief River Watershed Fluvial Geomorphology Report  June 2015 3

GENERAL DESCRIPTION

SURFACE WATERS The TRW is located in northwestern Minnesota. It is approximately 1,068 square miles in size (RLWD 2006). There are three main river systems within the TRW: the Mud, the Moose, and Thief Rivers (Figure 1). The Moose and the Mud Rivers flow east to west and are tributaries to the Thief River. A network of drainage systems and a few natural waterways are the tributaries to these rivers. The outlet of the TRW is in the city of Thief River Falls, where the Thief River joins the .

Most rivers have been managed by humans, either to mitigate damages, such as flooding, or to enhance as a resource. These activities have contributed to changes in the dynamic equilibrium of stream systems across the country (FISRWG 1998). The rivers within the TRW are no exception. Portions of all three main rivers have been modified or channelized and are now called judicial or state ditches. Dredging and sediment clean‐outs occur in an effort to keep higher flows within the channel itself and move water through the stream channel more quickly, often times by increasing the cross‐sectional area of the channel.

Table 1 breaks the TRW into subwatersheds and shows the corresponding watercourse types. The majority of the watercourses are classified as either a perennial or intermittent ditch. Fifteen percent (15%) of the watercourses within the entire TRW are intermittent and perennial stream channels (232 miles) and 77% are intermittent and perennial drainage ditches (1,175 miles).

There are distinct differences between the subwatersheds. The Thief River main stem, Mud River, CD 20, and Moose River have the most miles of watercourses with 502, 314, 291, and 242 miles, respectively. The Mud River subwatershed has the highest amount of perennial stream miles (21.29 miles or 6.8%). The Moose River subwatershed has 15.3 miles (6.3%) of perennial streams. CD 20 has the highest percentage of intermittent streams (17.3%), followed by the Mud River and CD 30/18, 14.9% and 11.4%, respectively.

All subwatersheds are dominated by perennial and intermittent drainage ditches. The main stem subwatershed has the lowest percentage (64.7%) and CD 200 has the highest (90%). Very few watercourses are classified as undifferentiated drainage ditch, connector wetland, river centerline, or river connector.

Comparing perennial to intermittent stream and ditch watercourses, the Moose River subwatershed had the highest percentage of perennial watercourses (62.8%). CD 30/18 and the main stem subwatersheds had the lowest percentage of perennial watercourses, 23.2% and 25.6%, respectively. CD 30/18 and the main stem subwatersheds had the highest percentages of intermittent streams and ditches, 76.8% and 74.4%, respectively.

Thief River Watershed Fluvial Geomorphology Report  June 2015 4

GENERAL DESCRIPTION

Figure 1. Map of the watercourse types within the Thief River Watershed. Table 1. Breakdown of watercourse types by subwatershed. Combined Subwatersheds (based on DNR HUC 12) Full Watershed Moose Mud CD 200 CD 20 CD 30/18 Thief Mainstem Surface Water Type Miles Percent Miles Percent Miles Percent Miles Percent Miles Percent Miles Percent Miles Percent Stream (Perennial) 60.69 3.95% 15.28 6.30% 21.29 6.776% 0.21 0.23% 6.64 2.28% 0 0% 17.26 3.44% Stream (Intermittent) 172.02 11.20% 12.33 5.09% 46.74 14.88% 7.22 7.90% 50.22 17.27% 10.74 11.41% 44.78 8.92% Drainage Ditch (Perennial) 468.37 30.51% 131.93 54.42% 106.45 33.88% 34.88 38.17% 91.34 31.41% 21.88 23.24% 81.88 16.30% Drainage Ditch (Intermittent) 706.92 46.05% 74.92 30.90% 138.53 44.09% 47.37 51.84% 141.74 48.75% 61.51 65.33% 242.84 48.35% Drainage Ditch (Undifferentiated) 3.16 0.21% 0.83 0.34% 0.35 0.11% 0 0% 0 0% 0 0% 1.98 0.39% Connector (Lake) 66.75 4.35% 1.17 0.48% 0.57 0.18% 0.29 0.32% 0.27 0.093% 0.016 0.017% 64.43 12.83% Connector (Wetland) 38.22 2.49% 1.69 0.70% 0.26 0.083% 0.59 0.65% 0.55 0.189% 0 0% 35.14 7.00% Centerline (River) 18.83 1.23% 4.17 1.72% 0 0% 0.82 0.90% 0 0% 0 0% 13.84 2.76% Connector (River) 0.24 0.012% 0.11 0.045% 0 0% 0 0% 0 0% 0 0% 0.13 0.03%

Thief River Watershed Fluvial Geomorphology Report  June 2015 5

GENERALFigure 2 shows the National Wetland Inventory (NWI) for the TRW DESCRIPTION . The NWI breaks out the type of wetlands into 8 classifications (i.e. wet meadow, bog, deep marsh, etc.). The majority of the larger bodies of water/wetlands found within the watershed fall into the category of shallow marsh or shallow open water. But there are many wetlands in the east and southeast portion of the watershed that are wooded swamp, shrub swamp, and bog. Most of these wetlands are protected by either the Beltrami Island State Forest or DNR Wildlife Management Areas (WMAs).

Figure 2. National Wetland Inventory of the Thief River Watershed.

Thief River Watershed Fluvial Geomorphology Report  June 2015 6

GENERALPublic watercourses and public water basins are under the juris DESCRIPTION diction of the MN DNR. Figure 3 shows all the public water basins and public watercourses in the TRW. There are only three public watercourses within the TRW; they are the Thief River, Moose River to the north, and Mud River to the south. Both the Moose River (JD 21) and the Mud (JD 11) are perennial watercourses that are in some portions, part of a legal ditch system (noted by the dashed lines on the map). This means that they are also under the jurisdiction of the County or Judicial Ditch Authorities.

Figure 3. Public Waterbasins and Watercourses within the Thief River Watershed.

Thief River Watershed Fluvial Geomorphology Report  June 2015 7

GEOLOGY The most recent geologic period is referred to as the Quaternary Period. It is divided into the Pleistocene Epoch (2 million to 10,000 years ago) and the Holocene (10,000 years ago to present) (Lusardi 1997). Even though Minnesota’s previous geologic history is much longer, the bedrock geology of northwestern Minnesota is largely buried beneath as much as 150 m of glacial deposits, the thickest glacial cover in the state (Ojakangas and Matsch 1982).

During the Pleistocene, the Laurentide Ice Sheet covered much of northern North America. Figure 4 shows generalized coverage by specific glacial lobes of advancement. The most recent glacier to cross the state was the Des Moines Lobe. It extended from the current day Red River Basin to Des Moines, Iowa. The material deposited by this glacial advancement is gray to brown and distinctive because it contained Cretaceous shale imported from North Dakota and Canada (Lusardi 1994). Approximately 12,000 years ago, glacial Lake Agassiz Figure 4. Simplifed drawing of ice lobes about 14,000 years occupied a very large portion of the Red River ago. Source: Lusardi 1994. Basin.

The impact of glacial advancement and subsequent retreat can be seen on the landscape today. Numerous glacial advancements, or lobes, moved across the current day Red River Valley during the last ice age. Figure 5 depicts the current surface geology in the TRW. There are three distinct categories: Red River Lobe glacial till, Lake Agassiz, and organic deposits.

The majority of the surficial geology is a result of the Red River Lobe. Boulders, rocks, sand, and trees were picked up and carried with the glacier as it advanced. The area mapped as the Red River Lobe is where the glacier formed a till plain when ice melted and the till, or unsorted mix of material, was deposited in place. This material formed level terrain, which can be seen on the surface today. The deposited material is described as pebble‐clay with very few pebbles (Harris et al. 1995). It is dark gray to black with very high natural water content. The material has very low shear strength and is prone to slumping. It was deposited in or very near Lake Agassiz basin.

Glacial Lake Agassiz was the largest glacial lake to form from the melt water of receding glaciers (Lusardi 1994). At its largest extent, the lake may have covered 170,000 square miles. The surface deposits of the areas mapped as Lake Agassiz have a lacustrine, or lake, source. Lake sedimentation produces areas of level sediment and other areas have a rolling to undulating topography due to nearshore wave action and beach formation.

Thief River Watershed Fluvial Geomorphology Report  June 2015 8

Areas mapped as organic deposits are made up primarily of peat and organic‐rich clay and silt. These areas have little relief relative to adjacent topography. During a time of warmer climate during glaciation this areas probably had large amounts of organic‐rich sediments lying beneath glacial till (Ojakangas and Matsch 1982).”

Figure 5. Geology of the Thief River Watershed.

Thief River Watershed Fluvial Geomorphology Report  June 2015 9

LAND USE The stability of river channels is dependent on equilibrium between sediment supply and quantity/timing of water. The management of the land outside of the floodplain will have a direct impact on changes to these 2 factors. The quantity and size of sediment being delivered to a river channel is a product of the geologic history, present day soil types, and land use changes over time. Responses that can be measured within the river channel include changes to the width‐to‐depth ratio, channel bottom elevation, channel slope, and size of particles found within the channel. When river channels become unstable, they will go through a series of adjustments until they reach a new equilibrium. These changes not only impact channel stability, they can also have a direct impact on habitat and water quality.

Table 2 breaks the TRW into subwatersheds and shows the corresponding land uses. The DNR delineated level 8 HUC catchments were used to create the respective subwatersheds. Cultivated crops and wetlands, both emergent herbaceous and woody, made up the majority of the land uses. Small percentages of high intensity, developed land use were observed in the Mud and Thief main stem subwatersheds, due to the cities of Grygla and Thief River Falls. Table 2. Breakdown of land use by subwatershed. Combined Subwatersheds (based on DNR HUC 12) Full Watershed Moose Mud CD 200 CD 20 CD 30/18 Thief Mainstem Land Cover Class 2 2 2 2 2 2 2 Mi. Percent Mi. Percent Mi. Percent Mi. Percent Mi. Percent Mi. Percent Mi. Percent Cultivated Crops 378.47 36.10% 13.79 7.05% 70.49 37.91% 26.19 39.89% 77.79 37.15% 45.83 69.14% 144.34 44.34% Emergent Herbaceuous Wetlands 295.09 28.15% 74.38 38.02% 50.66 27.25% 21.70 33.05% 57.01 27.22% 7.08 10.69% 84.20 25.87% Woody Wetlands 177.49 16.93% 81.90 41.87% 36.17 19.45% 5.04 7.68% 36.86 17.60% 1.12 1.69% 16.35 5.02% Hay/Pasture 71.45 6.81% 5.28 2.70% 12.76 6.86% 4.94 7.52% 15.56 7.43% 7.60 11.47% 25.33 7.78% Deciduous Forest 62.19 5.93% 14.47 7.40% 5.02 2.70% 5.02 7.65% 11.08 5.29% 1.34 2.02% 25.27 7.76% Developed, Open Space 26.72 2.55% 2.44 1.25% 4.59 2.47% 1.57 2.39% 5.14 2.45% 2.68 4.04% 10.38 3.19% Open Water 17.59 1.68% 0.19 0.10% 0.06 0.03% 0.38 0.58% 0.13 0.06% 0.01 0.01% 16.82 5.17% Herbaceuous 7.75 0.74% 0.56 0.28% 3.13 1.68% 0.29 0.45% 3.22 1.54% 0.07 0.11% 0.48 0.15% Evergreen Forest 4.88 0.47% 0.99 0.51% 1.96 1.05% 0.18 0.28% 1.64 0.78% 0.01 0.02% 0.10 0.03% Developed, Low Intensity 2.89 0.28% 0.14 0.07% 0.27 0.15% 0.14 0.22% 0.35 0.17% 0.42 0.64% 1.57 0.48% Shrub/Scrub 2.69 0.26% 0.81 0.41% 0.52 0.28% 0.16 0.25% 0.58 0.27% 0.11 0.16% 0.52 0.16% Barren Land 0.67 0.06% 0.32 0.17% 0.24 0.13% 0.014 0.021% 0.07 0.03% 0.01 0.02% 0.01 0.00% Developed, Medium Intensity 0.56 0.05% 0.34 0.17% 0.07 0.04% 0.008 0.012% 0.00 0.00% 0.01 0.01% 0.14 0.04% Developed, High Intensity 0.018 0.002% 0.000 0.000% 0.008 0.004% 0.000 0.00% 0.000 0.00% 0.000 0.00% 0.010 0.003% Mixed Forest 0.002 0.000% 0.000 0.000% 0.002 0.001% 0.000 0.000% 0.000 0.000% 0.000 0.00% 0.000 0.000%

Thief River Watershed Fluvial Geomorphology Report  June 2015 10

The Moose River subwatershed had the highest percent of wetland coverage within the TRW (Table 2). The flat terrain, as a result of being at the bottom of Lake Agassiz, and the deposition of organic‐rich sediments beneath glacial till created the perfect landscape for wetlands to develop. Wetlands made up 79.9% of the land use in this subwatershed. Many different types of wetlands were present in this area, including emergent plant wetlands, shrub/scrub swamps, and wooded swamps (Figure 6). There was extensive coverage by state‐owned Wildlife Management Areas in this area. Cultivated crops made up only 7.1% of this subwatershed.

Figure 6. 2011 Land use within the Moose River drainage area. Source: NLCD 2011.

Thief River Watershed Fluvial Geomorphology Report  June 2015 11

The Mud River subwatershed is immediately south of the Moose River subwatershed (Figure 7). The eastern side of this subwatershed was dominated by wetlands, but the central and western portions contained more land in cultivated crops (37.9%). This subwatershed had the second highest percentage of wetland coverage with 46.7%. Most of these wetlands were adjacent to the wetland complex to the north.

Figure 7. 2011 Land use within the Mud River drainage area. Source: NLCD 2011.

Thief River Watershed Fluvial Geomorphology Report  June 2015 12

A smaller subwatershed is located between the Mud River and CD 20 subwatersheds. This drainage area ties into the Thief River through CD 200, located along the boundary between Agassiz NWR and Elm Lake WMA (Figure 8). This area had similar coverage of cultivated crops (39.9%) as the Mud River and County Ditch 20 subwatersheds. The wetland coverage was concentrated in the eastern portion within Beltrami County and within the boundaries of Agassiz NWR and Elm Lake WMA.

Figure 8. 2011 Land use within the County Ditch 200 drainage area. Source: NLCD 2011.

Thief River Watershed Fluvial Geomorphology Report  June 2015 13

The CD 20 subwatershed is located south of the CD 200 and Mud River subwatersheds (Figure 9). The land use in this area was 37.2% cultivated crops and 44.8% wetlands. The majority of wetlands were located within Beltrami County in areas that were mapped with organic deposits from glacial activity. Hay and pasture land made of 7.4% of the land use.

Figure 9. 2011 Land use within the County Ditch 20 drainage area. Source: NLCD 2011.

Thief River Watershed Fluvial Geomorphology Report  June 2015 14

South of the CD 20 subwatershed is another smaller legal drainage system that drains directly to the Thief River through CD 30 and 18 (Figure 10). Much of this subwatershed is located within Pennington County. It had the highest percentage of cultivated crops (69.1%) and lowest percentage of wetland coverage (12.4%). The wetland coverage was concentrated in the eastern portion of the subwatershed, where glacial deposits contained higher amounts of organic material.

Figure 10. 2011 Land use within the County Ditch 30/18 drainage area. Source: NLCD 2011.

Thief River Watershed Fluvial Geomorphology Report  June 2015 15

The remaining area of the TRW was grouped together as this area drains directly to the Thief River (Figure 11). All of the other subwatersheds contribute water to the main stem of the Thief River as well. The majority of the land use within this subwatershed was cultivated crops (44.3%). A large wetland complex was present within Agassiz NWR. Wetlands made up 30.9% of this entire subwatershed. Open water (5.2%) and deciduous forest (7.8%) land uses had the highest percent, compared to the other subwatersheds.

Figure 11. 2011 Land use within the Thief River main stem drainage area. Source: NLCD 2011.

Thief River Watershed Fluvial Geomorphology Report  June 2015 16

METHODOLOGY

FIELD ASSESSMENTS In May of 2010, MN DNR and RLWD staff completed a reconnaissance effort within the TRW. The goals of this field reconnaissance were to collect data on stream condition, including stream classification, bank erosion potential, stream habitat condition, riparian condition, indices of stream stability, identification of representative areas for collection of additional data, and identification of potential problem and restoration areas.

Fifteen study sites were then selected for in‐depth analysis (Figure 12). Sites 13 and 11 were ultimately not assessed. The valley type and stream classification, using Rosgen methodology, were determined at each location (Rosgen 2008). Figure 13, depicts the dominant slope range, cross‐ section view, and plan view of each broad‐level stream classification. Each study reach was also assessed to determine its morphological condition, including dimension, pattern, and profile.

Figure 12. Map of fluvial geomorphology study sites.

Thief River Watershed Fluvial Geomorphology Report  June 2015 17

Figure 13. Variables determining Rosgen stream classification.

The information collected at each site included:  Longitudinal profiles  Cross‐section data at a minimum of one pool and one riffle  Determinations of bankfull elevation  Pebble counts  Riparian vegetation assessment  Pfankuch stability assessment  Bank Erosion Hazard Index estimates  Near Bank Stress estimates  Sediment samples

Using global navigation satellite system (GNSS) equipment, elevation data was collected. With data being corrected in real time, using MN Department of Transportation’s base station network, the vertical accuracy was within 20 mm (Frontier Precision, 2013). A site calibration was completed on the GNSS data using nearby MN Department of Transportation geodetic monuments. A longitudinal profile was developed by taking measurements of the channel bed along the thalweg, or lowest points of the channel, for at least two meander lengths where feasible or 20‐30 bankfull widths.

During the first field season, cross‐sectional data were collected at a minimum of one pool and one riffle location along the study reach. Often several pools and riffles were assessed to best represent the natural variability within a site. Pool cross‐sections were reassessed the following year (2012) to validate the estimated erosional rate data. At least one riffle cross‐section per site was also

Thief River Watershed Fluvial Geomorphology Report  June 2015 18

reassessed in 2012. Cross‐section start and/or end survey points were monument with a 2’ rebar stake, capped with a yellow, plastic cover labelled “MN DNR CWL”.

Bankfull elevation was determined at each cross‐section using field indicators, such as floodplains, slope breaks or change in particle size distribution, inundation features, and/or staining on rocks. These bankfull elevations were then validated during analysis of the data, comparing cross‐ sectional areas and bankfull slope along the longitudinal profile. The USGS StreamStats tool was used to compare predicted flows with confidence estimates (Lorenz et al. 2009).

Based on the bankfull elevation, the cross‐sectional area, stream width, mean depth, maximum depth, and width‐to‐depth ratio were calculated. The entrenchment ratio, or the channel width at two times the maximum bankfull depth divided by the bankfull width, was also calculated.

At least two Modified Wolman pebble counts were conducted at each study site (Wolman 1954, Rosgen 2012), one along the logitudinal profile and one at a riffle cross‐section. The intermediate axis, defined as the limiting axis for a pebble to move through a gravelometer hole, was measured on at least 100 particles. The 50th percentile size of the representative pebble count along the longitundinal profile was used as a factor for stream classification. The 84th percentile size of the pebble count from the active bed of the riffle cross‐section was used in the hydraulic calculations. The 84th percentile may seem arbitrary, but it is used because experience of many investigators has shown that the particles larger than the median size play an important role in flow resistance, and therefore a single parameter to describe bed particle size should be some size larger than the median (Leopold and Wolman 1960). The 84th percentile is two standard deviations larger than the mean size, assuming a normal distribution. If possible, an additional pebble count was completed on a pool cross‐section as a measure of channel stability and stream habitat condition.

To determine the current condition (stability) of each study reach, a series of assessments were completed. A basic assessment of the riparian vegetation included species names and percentage coverage, broken into canopy cover, shrub layer, herbaceous, leaf or needle cover, and bare ground. Information was also gathered on flow regime, stream order, meander patterns, depositional patterns, channel blockages, degree of channel incision, and channel confinement. A channel stability rating was created by evaluating the upper bank, lower bank, and channel bed using Pfankuch Stability Assessment, modified by stream type (Pfankuch 1975, Rosgen 1996).

Stream bank erosion rates were estimated using the “Bank Assessment for Non‐point source Consequences of Sediment” (BANCS) Model (Rosgen, 1996, 2001b, 2006b). This empirical model uses the Bank Erosion Hazard Index (BEHI) and Near Bank Stress (NBS) erosion estimation tools. During the reconnaissance, features of the streambanks were measured to create a BEHI rating. The BEHI variables included full study bank height, bankfull height, root depth, root density, bank angle, and surface protection (Appendix B). Waypoints and photographs were collected and the length of similar streambanks were measured using ArcGIS. NBS was estimated through analysis of aerial photos using the ratio of the radius of curvature of the meanders to the bankfull width of the

Thief River Watershed Fluvial Geomorphology Report  June 2015 19

channel. This ratio is a measure of the tightness of the bends in the river and the degree of boundary shear stress acting on those banks.

The annual streambank erosion rate can then be estimated using the BEHI and NBS ratings. The final annual erosion rate is adjusted by using a regional curve, created from local data collection. Erosion rate curves have been developed for North Carolina, Colorado, and Yellowstone National Park. With the collection of additional BANCS data in NW Minnesota, relationsips will be developed for this specific area, which will greatly strengthen the estimates. The estimates in this report used the most appropriate of the three currently developed relationships, based on aerial photo validations.

DATA ANALYSIS Field worksheets and GNSS data were entered into the software program RIVERMorph. This program is a software package, developed by Stantec, designed to analyze, store, sort, and share fluvial geomorphology data. Shapefiles were created to store the handheld GPS waypoints. Pictures taken with the handheld waypoints were also saved to the project folder. The geometry of channel pattern along the study reaches was measured using ArcGIS 10.1. Sinuosity, belt width, meander wavelength, and radius of curvature were all measured using a series of aerial photos, from 1992‐2012.

Precipitation and annual flow statistics and summary data were analyzed and presented using Microsoft Excel. Flow data were retrieved from the USGS stream gaging website. The site nearest to the pour point in the Thief River watershed was used (Thief River USGS gage 05076000). The period of record used for flow data was from 1909 to 2014. Precipitation data was also analyzed for the location near the pour point or closest to the USGS gaging station. The precipitation data was retrieved from the Wetland Delineation Precipitation Data website for a site named Dakota Junction (T154N, R43W, Section 18). Precipitation data were analyzed for the same time period as the USGS stream flows.

Thief River Watershed Fluvial Geomorphology Report  June 2015 20

HYDROLOGY

Precipitation

ANNUAL PRECIPITATION

Since the early 1900s the annual precipitation has fluctuated between dry and wet periods. Early in the 20th century, the periods were about a decade long with the 1910s, 1930s, and mid‐1950s to early 1960s having below average annual rates (Figure 14). In general, the rates were near or slightly above average in the 1900s and late 1940s to early 1950s. From approximately 1964 to 1975 there was a more pronounced wet period, with the highest annual rates on record occurring in the mid‐1960s. Since 1995, the annual rates have been consistently above average, except for a handful of years.

Annual Precipitation Trend Analysis USGS 05076000 Thief River near Dakota Junction, MN 35 33 31 29 27 (inches) 25 23 21 19 Precipitation 17 15 13 Annual 11 9 7 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 Ann. Precip (in) Long‐term Average 90th Percentile 10th Percentile 7 per. Mov. Avg. (Ann. Precip (in))

Figure 14. Annual precipitation totals of Thief River near Dakota Junction, MN.

Thief River Watershed Fluvial Geomorphology Report  June 2015 21

Stream Flow

STREAM GAGES Stream channel size, shape, and pattern are formed by the frequency and magnitude of flow events moving through the channel. Increases in flow can be caused by changes to vegetative cover, increased agriculture or urban drainage, increased precipitation, or a combination of these.

The gage with the longest period of record is located on the Thief River, approximately 4 miles north of Thief River Falls. Historical daily average discharge readings were first collected in July of 1909 (Table 3).

Table 3. Description of stream gage locations within the Thief River Watershed.

River ID Location Record of Period Description Thief River USGS: 05076000 Nr. Thief River 07/01/1909 – Daily discharge MPCA: S002‐079 Falls, MN present Thief River RLWD: S002‐088 Nr. Holt, MN 07/05/2007‐ Water level and MPCA: S002‐088 (CSAH 7) present flow Mud River USGS: 05075700 Nr. Grygla, MN 05/29/2007‐ Water level and RLWD: S002‐078 (89) present flow1 MPCA: S002‐078 Moose River RLWD: S004‐211 Hwy 54, 10.5 06/14/2007 – Water level and miles N of Grygla present flow Marshall RLWD: S004‐494 180TH Ave NE, 05/25/2007 ‐ Water level and County Ditch 8.25 miles N of Present flow 20 Thief River Falls Thief River RLWD: S004‐055 Northern 6/9/2012 ‐ Water level and NWS: TRGM5, FWS boundary of Present flow ID: 482537095581201 Agassiz NWR, USGS: 05075500 near Gatzke Branch A of RLWD: S006‐540 CSAH 48, 7 miles 03/30/2010 ‐ Water level and Judicial Ditch SW of Casperson Present flow 21 Branch 200 of RLWD: S004‐493 190th Ave NE, 6 05/09/2007 ‐ Water level and Judicial Ditch miles SE of Holt Present flow 11 Judicial Ditch NWS: AGDM5, FWS Nr. Holt, MN on 10/13/2010 – Water level and 112 ID: 481840096033001 Agassiz NWR Present flow Judicial Ditch RLWD: S004‐966 149th Ave, 3 03/31/2010‐ Water level and 30 miles N of Thief Present flow River Falls Agassiz Pool NWS: AGPM5, FWS Nr. Middle River, 2/16/2011 ‐ Pool level ID:481918095583201 MN on Agassiz Present NWR 1USGS annual peak flow data has also been collected since 1979 2Location of monitoring has changed over time since initial install in 2010

Thief River Watershed Fluvial Geomorphology Report  June 2015 22

On the Mud River the USGS gage is located near Grygla, where the river flows under State Highway 89. Annual flow peak data readings began in 1979. In June 2013, MN CSG Program monitoring equipment was installed to collect flow data every 15 minutes.

MEAN ANNUAL FLOW Annual mean streamflow is the average of all daily stream flows for each year. Figure 15 shows the annual means since 1929. The higher annual mean flows occurred during years that annual precipitation rates were also high, specifically in 1950, 1966, 1999, and 2011. In general, annual flow rates appear to be increasing since 1929. Fluctuations in annual mean stream flows may be a response to a variety of factors including precipitation rates, land use, land management techniques, hydrological soil group, evapotranspiration (ET), and other physiographic watershed variables.

When comparing the highest 25% of annual maximum peaks since 1909, 12 of those 26 years were tallied since 1995. Though that 20‐yr. time frame (1995‐2014) accounts for almost 20% of the period of record, it contains 48% of the highest annual peaks. The 20‐yr. time frame that contained the next highest number of annual peaks was from 1955‐1974. Seven of the top 25% occurred then and accounted for 28% of the highest annual peaks.

Annual Flow Thief River near Thief River Falls 800 R² = 0.178 700 Annual Flow Linear (Annual Flow)

600 (cfs) 500 Flow

400 Annual 300

200

100

0

Figure 15. Annual mean flow at USGS stream gage 0507600, near Thief River Falls.

Thief River Watershed Fluvial Geomorphology Report  June 2015 23

These years with higher annual mean flows and peaks do correlate with higher precipitation rates. Though the mid‐1950s to the early 1960s was drier on average, the next decade had a more pronounced wet period; and since 1995 the precipitation has been greater than average as well. However, when comparing the increases of precipitation and streamflow, the rate at which streamflow is increasing appears to be greater (Figure 16). Precipitation rates undoubtedly impact the amount of water that reaches the Thief River each year; however, the additional variables that impact streamflow may be influencing the higher mean and peak flows observed at USGS gage 05076000 on the Thief River.

900 40

800 y = 0.0346x + 19.481 35 700 R² = 0.4903 30 600 (cfs) 25

500 (in) Flow

400 20 Average 300

15 Precipitation 200 Annual 10 100 y = 2.6923x + 30.778 R² = 0.8593 5 0

‐100 1910 1914 1921 1930 1934 1938 1942 1946 1950 1954 1958 1962 1966 1970 1974 1978 1983 1987 1991 1995 1999 2003 2007 2011 0

Annual Average Flow (cfs) Running 30 Yr Ave Flow (cfs) Annual Precipitation (in) Running 30 Yr Ave Precip (in) Linear (Running 30 Yr Ave Flow (cfs)) Linear (Running 30 Yr Ave Precip (in))

Figure 16. Mean annual flow and precipitation for USGS stream gage 0507600.

FLOOD FREQUENCY The effective discharge of a river at a specific location is defined as the discharge which forms or maintains the channel (Leopold et al. 1964). It is the discharge that over a long period of time transports the most sediment (Benson and Thomas 1966). Though flows just over bankfull elevation don’t transport as much sediment as larger floods, they do occur more frequently. Leopold et al (2010) stated that this effective discharge can often be approximated by the bankfull discharge, and in many rivers this discharge has a recurrence interval of about 1.5 years.

Thief River Watershed Fluvial Geomorphology Report  June 2015 24

At USGS gage 0507600, the bankfull discharge was estimated to be 1006 cfs. Figure 17 is a graph of the mean daily discharge since 1909. The red line is the estimated bankfull discharge. In the last 30 years (1985‐2014) the mean daily discharge has been equal to or greater than 1006 cfs 1113 times. From 1955‐1984 bankfull discharge or greater occurred 686 times. Three years (1925‐1927) are missing from the next 30‐yr. time period of 1925‐1954. During those 27 years of record, the mean daily flow reached 1006 cfs 242 times. Prior to 1925, mean daily flows were measured for 14 years. As a percent, from 1909‐1924, 1925‐1954, 1955‐1984, and 1985‐2014 discharge at USGS gage 0506700 exceeded 1006 cfs 2.4%, 2.5%, 6.3%, and 10.2% of the days flows were recorded, respectively.

Figure 17. Daily mean discharge flows at USGS stream gage 05076000 near Thief River Falls, MN.

Thief River Watershed Fluvial Geomorphology Report  June 2015 25

MAGNITUDE & FREQUENCY A flow duration curve is an estimate of the percentage of time a given streamflow was equaled or exceeded over a specific period of time. Figure 18 was created using the mean daily flows at USGS gage 0507600 near Thief River Falls. Time periods were broken up into three 30‐yr. periods. Each curve does not represent the probability of specific discharges within a given year; rather it is an average curve for each 30‐yr. period. The 1925‐1954 time period had a few years of no data, specifically early on in this 30‐yr. period.

The shape of an individual flow duration curve is determined by the hydrologic and geologic characteristics of the watershed (USGS 1959). A stream with flows largely from direct runoff would have a steep slope. A watershed with surface‐ or groundwater storage would have a flatter slope. The end of the flow duration curve shows the characteristics of the perennial storage within a watershed. A flat slope near the end indicates a large amount of storage and a steep slope indicates a negligible amount.

The highest mean daily flow for 1925‐1954, 1955‐1984, and 1985‐2014 was 5580, 3840, and 4080 cfs, respectively. The 1925‐1954 time period had 18 of the highest mean daily flows on record. All of these 18 days were in May of 1950. The top 45 mean daily flows during that earliest time period were in April, May, or June of 1950. After those recorded flows the 1925‐1954 flow duration curves quickly dropped below the other two more recent periods. The Q10 flow, or flow that was exceeded 10% of the time, for the 3 time periods, was 959, 1081, and 1091 cfs (oldest to newest).

At lower flows the flow duration curve dropped steeply. This indicates limited availability of perennial storage. This is also reflected in the higher percent of intermittent streams and ditches within the TRW. Perennial streams and ditches make up 34.46% of the watercourses, while 57.25% are intermittent. The CD 20 subwatershed enters the main stem just a few miles north of the USGS gage, and 66.02% of the watercourses within this subwatershed are intermittent.

Comparing between the 3 time periods, discharge over 1000 cfs was most frequent during the recent 30‐yr. period and least frequent from 1925‐1954. Except for the extreme flood event in 1950, the frequency of any discharge was highest during the 1985‐2014 time period. No discharge (0 cfs) was recorded 35.5%, 13.4%, and 6.8% of the time (oldest to newest). Mean daily flows were collected prior to 1925. No discharge (0 cfs) was recorded 8.2% of the time (1909‐1917, 1920‐ 1924). The Q90, or flow that was exceeded 90% of the time, for each time period was 0, 0, 0.2 cfs (oldest to newest). The Q90 before 1925 was 0.2 cfs.

Thief River Watershed Fluvial Geomorphology Report  June 2015 26

Flow Duration Curve for Thief River 10000

1925‐1954 discharge 1000 1955‐1984 discharge 1985‐2014 discharge 100

10

1

0.1

0.01 0 102030405060708090100 Figure 18. Flow duration curves comparing 30‐year time periods.

Thief River Watershed Fluvial Geomorphology Report  June 2015 27

STREAMBANK EROSION

BANCS Model

INTRODUCTION The stability of a river is determined by its ability to move water and sediment without changing its dimension, pattern, and profile. That is not to say that river channels should never erode. Stable rivers will have slowly eroding banks as the channel is reshaped by flood flows. It will develop point bars and the meanders will migrate down the valley fall line. The concern lies with accelerated rates of streambank erosion. According to the 2010 Minnesota Assessment Data report, turbidity was the leading cause of impairment for Minnesota’s rivers and streams (EPA 2015). The BANCS Model provides a quantitative prediction of streambank erosion, contributing to the sediment load within a river channel (Rosgen 1996). BEHI and NBS data were collected on the Thief, Moose, and Mud Rivers between 4/14/10 and 5/12/10. County ditch 20 data was collected on 6/14/11. Figure 19 shows the overview of the total rates (tons/yr./mi).

241.3

239.2

54.4 210.1

Figure 19. BANCS streambank annual erosion estimate.

Thief River Watershed Fluvial Geomorphology Report  June 2015 28

Streambank erosion was estimated along a total of 72.55 miles of watercourses in 2010 and 2011 (Table 4). The total annual tons of erosion per mile ranged from 1.056 to 241.296.

Table 4. Streambank erosion summary by reconnaissance reach. Length Length Total Erosion Total Erosion Total Erosion Total Erosion Total Erosion River Name Reconnaissance Reach (mi) (ft) (ft3/yr) (yd3/yr) (tons/yr) (tons/yr/ft) (tons/mile) Thief River Thief Lake to Agassiz NWR (Upper) 4.54 23971.2 6699.36 248.12 322.56 0.0134 70.752 Thief River Thief Lake to Agassiz NWR (Lower) 1.87 9873.6 2784.56 103.13 134.07 0.0136 71.808 Thief River Co Rd 7 to Co Rd 12 4.38 23126.4 2891.78 107.1 139.23 0.006 31.68 Thief River Co Rd 12 to Gage 9.28 48998.4 13758.66 509.58 662.45 0.0135 71.28 Thief River Gage to Thief River Falls 5.07 26769.6 16988.67 629.21 817.97 0.0305 161.04 Moose River NFR 55 to Co Rd 54 (Upper) 6.58 34742.4 32976.37 1221.37 1588 0.0457 241.296 Moose River Hwy 89 to Co Rd 127 (Lower) 3.91 20644.8 82.13 3.04 3.95 0.0002 1.056 Mud River Co Rd 703 to Waypoint 152 1.3 6864 6467.85 239.55 311.51 0.0453 239.184 Mud River T395 to T407 (Upper) 1.91 10084.8 5405.61 200.21 260.27 0.0259 136.752 Mud River T395 to T407 (Lower) 2.44 12883.2 6842.7 253.43 329.46 0.0256 135.168 County Ditch 20 Co Rd 707 to Co Rd 53 (Upper) 12.87 67953.6 56050.65 2075.95 2699.55 0.0398 210.144 County Ditch 20 Co Rd 53 to T291 (Lower) 18.4 97152 20856.4335 772.4605 1004.5 0.0103 54.384

THIEF RIVER BANCS data was collected along the Thief River from County Road 49, just downstream of Thief Lake, to the boundary with Agassiz NWF. This 6.4 mile reach was broken into 2 segments. The study reaches were separated just north of the farmstead on Hwy 6. The upstream and downstream reaches produced very similar estimates of 71 and 72 tons of sediment/yr./mi., respectively (Table 4).

A second reconnaissance reach was completed just south of the Agassiz NWR boundary. The study reach began at CR 7 and ended at CR 12. The BANCS model predicted approximately 31.68 tons of sediment/yr./mi. (Figure 20). The soil Figure 20. Thief River between Thief Lake and ANWF. Photo taken on 5/3/10.

Thief River Watershed Fluvial Geomorphology Report  June 2015 29

mapping through this area is less detailed. It is mostly mapped as Foxlake loam (Soil Survey Staff 2015). This type has loam from 0 to 19 in. and then silty clay from 19 to 80 in. The silty clay horizons contain signs of frequent saturation.

A third reconnaissance reach was completed from County Road 12 to the location of the USGS gage, just north of Thief River Falls. The BANCS model predicted approximately 71.28 tons/yr./mi. A fourth reconnaissance reach was completed from the USGS gage down to the north edge of Thief River Falls. The study reach began downstream of 140th Ave NE and ended upstream of the bridge on CR 63. The BANCS model predicted approximately 161 tons of sediment/yr./mi.

MOOSE RIVER The upstream reconnaissance reach was from the north‐south stretch of Moose River Rd NW to County Highway 54. This reach was a confined ditch, but it had good vegetative cover. The BANCS model predicted approximately 241 tons of sediment/yr./mi. (Figure 21). During the reconnaissance assessment, numerous upland sources of sediment were also identified. The majority of the riparian area soil type is mapped as Pengilly very fine sandy loam (Soil Survey Staff 2015). This has a thin, 4 in. layer of organics mixed with mineral matter classified Figure 21. Photo along upstream Moose River reach. Photo as a very fine sandy loam, followed by 50+ in. taken on 5/11/10. of stratified loamy very fine sand to silt loam. Areas of bank erosion in this area would likely be contributing very fine sand and silt to the bed load furthering downstream.

The downstream reconnaissance reach was from State Highway 89 to 310th Ave NE. This reach was very stable and had the lowest erosion estimate within the TRW. The BANCS model predicted approximately 1.056 tons of sediment/yr./mi. (Figure 22). The riparian area had a very diverse and dense vegetation cover. The water slope was also very low (0.000104). The majority of the riparian area soil type was mapped as Huag muck (Soil Survey Staff 2015). This soil type has a 9 in. organic layer of muck over organic and mineral layers of loam. Starting Figure 22. Lower Moose River reconnaissance reach. at about 15 in. the soil profile shows signs of Photo taken on 5/11/10. frequent saturation.

Thief River Watershed Fluvial Geomorphology Report  June 2015 30

This entire study reach was located in the Thief Lake Wildlife Management Area. The land adjacent to the river channel is delineated as shrub wetlands (type 6). Type 6 wetlands usually have water logged soil and the vegetation can include alders, willows, dogwoods and leatherleaf. Wet meadow wetlands (type 2) are also present in this complex. A solid wetland complex is at least a mile wide along this stretch of the Moose River. In this area, the river has been ditched and old spoil piles are present along the channel. To improve the lateral connection to the floodplain, holes in the spoil piles would allow flood waters to flow out onto the wetland complexes as surface water and reduce velocities within the stream channel during flooding.

MUD RIVER The upstream reconnaissance reach was from Flintlock Rd. NW to the west about 1.25 miles where the Mud River starts flowing northwest. This reach had high erosion rates. The left bank was adjacent to State Highway 89 and neither side of the river had a good riparian buffer. The BANCS model predicted approximately 239 tons of sediment/yr./mi. The soils present within the riparian zone are mapped as Kratka fine sandy loam and Eckvoll loamy fine sand (Soil Survey Staff 2015). The soil profiles for these 2 soil types contain horizons of fine sandy loam and Figure 23. Sand bedload moving along Mud River channel loam. Figure 23 shows the large amount of sand bottom. Photo taken on 9/11/2011. bedload moving downstream within the river.

The downstream reach was from 390th Ave NE to 360th Ave NE, just west of Grygla. This reach was fairly stable. The presence of a diverse riparian corridor was helping maintain the stability of this reach (Figure 24). During the reconnaissance it was noted that a high sand bedload was moving through the channel in this reach. The assessment of the BANCS data was broken into 2 separate segments. The break was placed about 0.5 miles east of CR 53, where another watercourse flows in from the south. Both reaches produced similar rates of erosion; upstream was 137 tons/yr./mi. and downstream was 135 tons/yr./mi.

Figure 24. Lower Mud River reconnaissance reach. Photos taken on 5/4/10.

Thief River Watershed Fluvial Geomorphology Report  June 2015 31

COUNTY DITCH 20 The upstream reach of CD 20 went from Jelle Rd (CR 707) to Hwy 53. The BANCS model predicted approximately 210 tons of sediment/yr./mi. on this 12.87 mile reach (Figure 25). The soil types shifted just upstream of Hwy 54 from an assortment of muck types to fine sandy loams (Soil Survey Staff, 2015). The riparian zone on the upper end of this study reach is delineated as a mix of scrub‐shrub, emergent vegetation and forested wetlands. The presence of wetlands thinned as the ditch moved into the sandier soils.

Where the soils began to get sandier the riparian Figure 25. CD 20, just downstream of Hwy 54. Photo area got narrower in some areas. Maintaining a taken on 5/10/10. diverse vegetated cover within the riparian zone will maintain and/or improve the stability of this reach.

The downstream reconnaissance reach of CD 20 was from Hwy 53 to the confluence with the angle ditch, coming from Elm Lake WMA. This 18.4 mile reach was fairly stable. The BANCS model predicted approximately 54 tons of sediment/yr./mi. (Figure 26). The soil map in the riparian zone was mainly Kratka fine sandy loam and Smiley loam (Soil Survey Staff 2015). Both soil types contain fine sand and loam. The Smiley loam also has a soil horizon of clay loam from 12 to 19 in.

Figure 26. CD 20, approximately halfway through the Though the stream banks in this study reach downstream study reach. Photo taken on 5/10/10. were contributing a low amount of sediment to the channel, during the reconnaissance a high sand bed load was observed moving in the channel. Some of this sediment may be coming from the streambanks higher up in the watershed, but upland sources of this sand load should be investigated. Existing stream power index (SPI) adjacent to the stream channel would be a tool that could be used to analyze the upland terrain.

Thief River Watershed Fluvial Geomorphology Report  June 2015 32

STABILITY

Pfankuch Stability Rating

The Pfankuch Stability Rating was developed by Dale Pfankuch (US Forest Service) with methods created in 1975. Years later Dave Rosgen, modified the final rating scale to be tiered by Rosgen stream type (1996). The rating is created by answering 15 questions, related to the upper banks, lower banks, and channel bottom (Appendix B). The individual ratings of each question provide information about the capacity of streams to adjust and recover from potential changes in flow and/or increases in sediment product. Figure 27 shows the Pfankuch ratings at each of the study locations.

Figure 27. Pfankuch Stability Ratings at the study sites.

Thief River Watershed Fluvial Geomorphology Report  June 2015 33

DISCUSSION

Thirteen of the original 15 sites were selected to complete detailed assessments on. Table 5 highlights some of the outcomes of these assessments. All of the sites were a B, C, or E stream type with gravel or sand substrate as the dominant particle size. The Pfankuch stability ratings ranged from unstable to stable.

Table 5. Summary of study site characteristics. Substrate

Predicted Actual Pfankuch Stream Drainage Site River Slope Erosion Erosion Stability Type Area (ft/yr) (ft/yr) Rating Bedrock Boulders Cobble Gravel Sand Silt/Clay 12 Moose River E5  174 0.000410.253 0.2373 Unstable 6Mud River E4  94.4 0.00062 0.153 0.6273 Stable 5Mud River B5c / E5  139 0.0003 0.25 0.3617 Unstable 4Mud River C5c‐ / E5  153 0.00024 0.036 0.0837 Moderately Stable 10 County Ditch 20 B4c  86.4 0.00065 0.153 0.37 Moderately Stable 9County Ditch 20 B5c  174 0.0003 0.153 0.2982 Moderately Stable 8County Ditch 20 B5c  210 0.00018 n/a n/a Unstable 1Thief River B4c  249 0.00025 0.016 0.0145 Moderately Stable 2Thief River C4c‐  251 0.00022 n/a n/a Stable 3Thief River C4c‐  257 0.00011 0.06 0.0888 Stable 15 Thief River E5  258 0.0001 0.03 0.0046 Moderately Stable 7Thief River B5c  641 0.00022 0.06 0.0768 Unstable 14 Thief River C4c‐  967 0.00035 0.03 0.4542 Stable

Thief River

THIEF LAKE TO AGASSIZ NATIONAL WILDLIFE REFUGE The Thief River starts at the outlet of Thief Lake and flows west for about 0.5 miles before it crosses Hwy 49 and turns in a southerly direction. There is a dam about 750 feet upstream of Hwy 49 that creates about a 3.5 feet vertical drop in the water surface. Sections of the river have been channelized. The center section of this reach appears to still be in a natural channel, primarily in Whiteford Township, section 5, and the northern half of section 8.

The water surface gradient through this reach is relatively flat. Over the 7.3 miles the river travels to reach the Agassiz NWR boundary, the vertical drop is approximately 17.5 feet, which is a slope of 0.00045. Some of the field measured slopes within the study reaches were even less than this: 0.00025, 0.00022, and 0.00011. Figure 28 shows the water surface longitudinal profile, using the

Thief River Watershed Fluvial Geomorphology Report  June 2015 34

elevation data from LiDAR. The control structure downstream of Thief Lake shows a vertical drop near the top of the profile. After site 2 the slope increases slightly, but then decreases again near site 15.

Drop Structure

Site 2

Site 3 Site 1

Site 15

Figure 28. Water surface profile from the outlet of Thief Lake to the Agassiz NWR boundary, using 1m DEM.

The land use through this section was mainly cultivated crops and some pasture or hay land (NLCD 2011). According to the National Agricultural Statistics Service (2012) the dominant crop types were soybeans and spring wheat. The soil type at all four study sites within this reach was mapped as Fluvaquents, frequently flooded‐Hapludolls complex. The typical soil profile for this map unit was 0‐16 inches of fine sandy loam, then 16‐80 inches of stratified loamy sand to silt loam. The adjacent soil types transitioned within this reach from sandy and fine sandy loams upstream to clay loams downstream, near Agassiz NWR.

The SWAT model for this watershed identified this stretch of the Thief River as a major contributor of sediment (Houston 2010). Follow up fingerprinting analysis on sediment entering the Agassiz pool within Agassiz NWR showed that greater than 65% of the sediment enters through a river or ditch system and the dominate source, based on its Cesium fingerprint, is an upland source (Engstrom and Schottler 2011).

The first study site downstream of the Thief Lake outlet was site 1. This reach of the Thief River has been straightened, and piles of the excavated stream channel line the upper banks. The approximate elevation of the 50‐yr. flood recurrence is 1165 feet. The flood‐prone area at this

Thief River Watershed Fluvial Geomorphology Report  June 2015 35

elevation is confined in areas, due to the spoil piles, to a width of 65 feet. Due to the confinement within the floodplain, this reach was classified as a B4c stream type. The number 4 denotes the channel bottom particles, which were dominated by gravels, though silt, clay, and very fine sands were also present. The lowercase c signifies a low gradient (less than 0.02).

The Pfankuch stability rating was 84, which is moderately stable for a B4c stream type. The factors ranking high (more unstable) in this assessment included steep upper banks, lower density of riparian vegetation and rooting mass, and significant lower bank cutting. The width/depth ratio for a B stream type was very low at this site. The river would be more stable at this location if the floodplain was wider. B4 stream types are only moderately sensitive to disturbance and are moderately dependent on robust vegetation to remain stable.

Two riffle cross‐sections were measured at site 1. The upstream riffle (0+29) was inundated with fine particles. Silt and clay made up 33.64% of the particles measured on the active channel bed. The full study reach count was 13.64% silt and clay. The lower riffle (13+20) had larger particle sizes, but combined, silt, clay, and very fine sand made up 24.56% of the active channel bed count as well. Gravel dominated the riffle and full reach particle counts, ranging from 48.18‐63.16%. The one pool particle count (9+95) did show signs of scour, as no larger particle sizes (greater than 180 mm) were tallied. The mean pool depth‐to‐mean riffle depth was 1.212, which means the pools were deeper than the riffles.

The next study site downstream was site 2, located just upstream of 400th St. NE (Appendix A). Based on aerial photos, this location also appeared to have been modified by straightening of the channel, though it either re‐meandered naturally or was not as extensively modified as site 1. Though there was evidence of spoil piles adjacent to the stream channel, they were not as high and pronounced as they were at site 1. The approximate elevation of the 50‐yr. flood recurrence is 1163 feet. The left bank berm has an elevation of approximately 1160 feet. The width of the floodprone area was at least 350 feet at this site. Based on its slight entrenchment and slightly larger width to depth ratio, this study location was classified as a C4c‐. The number 4 denotes the channel bottom particles, which were dominated by gravels, though silt, clay, and very fine sands were also present. The lowercase c signifies a low gradient (less than 0.001).

A Pfankuch stability rating was not completed at this site, because ratings were evaluated during the second year of assessments and site 2 was only visited in 2011. Though a formal rating was not gathered, DNR staff present in 2011 estimated that site 2 was very stable. The riparian vegetation was dense and more robust. In addition, bank cutting was less evident than at site 1. The wider floodplain allowed the stream channel to form a C stream type, which is a very stable form. C stream types are low gradient, meandering stream channels with riffle‐to‐pool sequences. They naturally form point bars on the inside bends. C stream types also have well‐defined floodplains.

One riffle cross‐section (6+99) was measured at site 2. The 50th percentile particle size in this active channel bed was 52.78 mm. This riffle was clearly dominated by gravel particle sizes, though silt, clay, and fine sand were also tallied. The pool particle count (4+16) did show signs of scour, as

Thief River Watershed Fluvial Geomorphology Report  June 2015 36

no larger particle sizes (greater than 128 mm) were tallied. The riffle clearly had larger particle sizes than the pool and full study reach counts. In addition, the pool showed signs of scour as the particles are much smaller than in the riffle and full study reach counts. The mean pool depth‐to‐ mean riffle depth was 1.3, which means the pools were deeper than the riffles.

Site 3 was just over 1 mile downstream from site 2. This stretch of the river appeared to be flowing in its natural meander pattern; however, spoil piles existed on the inside of the river bends. Figure 29 is a cross‐section of the ground elevation through one of these meanders. The higher elevation point on the right (1155.2 feet) is the valley wall. The middle point (1152.8 feet) appears to the old berm and the left point (1151.7 Figure 29. Cross section of a meander bend at site 3, using 1m DEM. feet) is the left valley wall.

The approximate elevation of the 50‐yr. flood recurrence is 1153.9 feet. The valley walls (Figure 29) and the adjacent agricultural land was not this high. The nearest land that reaches this elevation is the adjacent roads at the section lines. With this high entrenchment ratio, the Thief River at site 3 was classified as a C4c‐. It was a low width/depth ratio C stream type, and could be considered an E4.

The Pfankuch stability rating was 89, which is stable for a C4c‐ stream type. The substrate within the channel appeared dull and dark, and there was abundant green perennial growth of aquatic plants within the stream channel. The factors that rated more unstable within the final Pfankuch rating were the presence of frequent or large mass erosion of the upper banks and significant cutting of the lower banks.

One riffle cross‐section was measured at site 3. This riffle appeared to have an excessive amount of fine particles. Silt and clay made up 27.78% of the particles measured on the active channel bed. At the pool cross‐section (3+12) silt and clay made up 16.82% of the pool particle count. The pool particle analysis had more gravel particles than the riffle count, 66.36% and 50%, respectively. The pool was deeper than the riffle with a ratio of 1.24, but it may be struggling to continually move the amount of sediment being supplied to the channel from upstream.

Site 15 was the last study location along this reach. It was located between Thief Lake and Agassiz NWR, just upstream of the Agassiz NWR boundary (Appendix A). The study reach started in a natural meander and transitioned into a straightened ditch. It was classified as an E5 stream type,

Thief River Watershed Fluvial Geomorphology Report  June 2015 37

with a width‐to‐depth ratio of 7.1. E5 stream types have a lower width‐to‐depth ratio (less than 12.0) and a wide floodplain.

The Pfankuch stability rating was 92, which is moderately stable for an E5 stream type. The factors ranking high (more unstable) in this assessment included moderate to heavy amount of debris in the upper banks, moderate level of new deposition occurring on the lower banks, and little consolidation of the channel bottom substrate. Scour and/or deposition were occurring in 30‐50% of the study reach.

One riffle and one pool particle analysis was completed at site 15. Both locations had similar amounts of fine sediment. Silt, clay, and very fine sand made up 29.52% of the riffle and 29.91% of the pool counts. The pool did have more sand‐sized particles (54.21%) and the riffle had more gravel (60.0%). Similar to site 3, the pool was deeper than the riffle with a ratio of 1.2, but it may be struggling to continually move the amount of sediment being supplied to the channel from upstream.

AGASSIZ NATIONAL WILDLIFE REFUGE TO THIEF RIVER FALLS The Thief River enters Agassiz NWR in a channelized form. The river flows through Agassiz Pool within the Refuge then follows the western boundary until it crosses CR 7. It takes a southwestern flow path towards Thief River Falls, where it flows into the Red Lake River.

The slope appears to be fairly consistent from its starting point at CR 121 (west of Agassiz Pool) to just downstream of CR 2 (Figure 30). The energy slope steepens just upstream of the USGS gage, east of 140th Ave. The Thief River is classified as a perennial drainage ditch south of Westgate Rd, within Agassiz NWR. The classification changes to a perennial stream after the confluence with CD 20.

The land use adjacent to the Thief River through this lower reach was dominated by herbaceous wetlands, cultivated crops, and pasture/hay land. According to the National Agricultural Statistics Service (2012) the dominant crop types were soybeans, spring wheat, and alfalfa. Some fields were planted to corn, dry beans, and barley.

The soil type within the riparian zone is a poorly drained, fine sandy loam and transitions into the Fluvaquents, frequently flooded‐Hapludolls complex, similar to sites 1, 2, 3, and 15. The typical soil profile for this map unit is 0‐16 inches of fine sandy loam, then 16‐80 inches of stratified loamy sand to silt loam. The adjacent soil types are loamy fine sands and fine sandy loams. The profiles contain fine sands and silt.

Thief River Watershed Fluvial Geomorphology Report  June 2015 38

CR 7 Site 7 CR 12 CR 2

Site 14

Golf Course

Figure 30. Water surface profile from Agassiz NWR to Thief River Falls, using 1m DEM.

Two study sites were located on the Thief River between Agassiz NWR and Thief River Falls. Site 7 was located one mile south of County Road 7, near 280th Ave. The Thief River was delineated as a perennial drainage ditch at this location. The Rosgen stream type was a B5c. The width‐to‐depth ratio was 13.39 and the entrenchment ratio was 1.56.

The Pfankuch stability rating was 111, which is unstable for a B5c stream type. The factors ranking high (more unstable) in this assessment included almost continuous cuts (some over 24 inches high) on the lower banks and no packing and a marked distribution change of channel bottom substrates. Scouring and deposition were also occurring in at least 30‐50% of the channel.

One riffle and one pool cross‐section was completed at site 7. Both the riffle and pool cross‐sections had a high percentage of silt/clay particles, 35.66% and 33.64%, respectively. The riffle cross‐ section had more gravel‐sized particles than the pool, but the pool had more sand‐sized particles. The mean pool depth‐to‐mean riffle depth was 0.996, but the pool had a larger cross‐sectional area. There were a few additional locations on the longitudinal profile with thalweg depths greater than these two cross‐sections.

The downstream study location was site 14. This site was located near the USGS gage, just upstream of 140th Ave. The Thief River was delineated as a perennial stream at this location. The Rosgen stream type was a C4c‐. The width‐to‐depth ratio was 21.72 and the entrenchment ratio was 2.22.

The Pfankuch stability rating was 86, which is stable for a C4c‐ stream type. The factor ranking high (more unstable) in this assessment was almost continuous cuts (some over 24 inches high) on the

Thief River Watershed Fluvial Geomorphology Report  June 2015 39

lower banks. The factors ranking low (more stable) included little to no obstructions to flow on the lower banks and channel bottom substrates appeared dull, dark, or stained.

One riffle and one pool cross‐section was completed at site 14. The riffle cross‐ section was dominated by larger particle sizes and the pool s had finer sizes. Silt, clay, and sand‐sized particles made up 25.41% of the riffle count and 47.22% of the pool count. In contrast, gravel‐sized particles made up 65.57% of the riffle count and 35.19% of the pool count. The mean pool depth‐ to‐mean riffle depth was 1.117. The pool also had a slightly larger cross‐sectional area.

Moose River

The Moose River begins at the outlet of the North Pool of the Moose River Impoundment. It is delineated as a perennial stream downstream of the impoundment outlet and forms a sinuous channel. Approximately 1 mile south of the outlet, the river enters a wetland and the delineation changes to a perennial drainage ditch. From this point, the Moose River is a channelized form adjacent to the Moose River Rd. At CR 54, the Moose River Rd ends and the river continues to flow west through a wetland complex. It remains in the historically dug channel until it outlets into Thief Lake.

The slope of the river from the outlet of the Moose River Impoundment to Moose River Road was approximately 0.0014 (Figure 31). The slope decreased to 0.00064 from where the river joins up with Moose River Rd. to just upstream of CR 54. After the CR 54 road crossing the slope continued to decrease. It was approximately 0.00017 until the Hwy 89 crossing.

Emergent and woody wetlands make up 79.89% of the Moose River’s drainage area. Cultivated crops and hay/pasture land make up 7.05% and 2.7%, respectively (NLCD 2011). According to GAP Stewardship, 58% of the area is in private ownership (2008). The MN DNR and Red Lake Tribe own 17.44% and 12.89%, respectively. Much of the MN DNR land is designated Wildlife Management Areas (WMA) including the Thief Lake and Wapiti WMAs.

Only one intensive study reach was located on the Moose River. Site 12 was located adjacent to the Moose River Rd, immediately upstream of Pine Island Rd. within Beltrami County. The river was delineated as a perennial drainage ditch at site 12. The Rosgen stream type was an E5. The width‐to‐depth ratios of the riffles were 7.80 and 6.71. The Moose River at site 12 had a wide flood prone width, approximately 315 feet.

Thief River Watershed Fluvial Geomorphology Report  June 2015 40

Moose River Rd.

Morel Rd.

Powder Horn Rd.

CR 54

Hwy 89

Site 12

Figure 31. Water surface profile from the outlet of the Moose River Impoundment to Thief Lake. using 1m DEM.

The Pfankuch stability rating was 111, which is unstable for an E5 stream type. The factors ranking high (more unstable) in this assessment included a steeper upper bank slope (40‐60%), significant bank cutting (12‐24” high) in the lower banks, a mostly loose assortment, with stable materials 0‐ 20%, of channel bottom substrate, and a channel bottom that was in more than 50% of a state of flux nearly yearlong.

Two riffle and one pool cross‐sections were completed at site 12. One riffle, one pool, and the full reach had pebble counts completed. At the riffle, 80.2% of the particle sizes were very coarse sand or smaller. At the pool, 86.36% of the particle sizes were very coarse sand or smaller. The largest particle size tallied was on the riffle count and was within the 32‐45 mm range. The mean pool depth‐to‐mean riffle depth was 1.08, so the pool was slightly deeper than the riffle. The pool area to riffle area ratio was 1.09.

Mud River

Based on the historic J.WM Trygg map (1967), the Mud River area was surveyed in 1892. The river started about 4 miles west of Grgyla, MN, just NW of the current Flintlock Rd. The Mud River is currently delineated as a drainage ditch, starting at Gunpowder Rd, 6 miles east of Grygla. The classification changes to a perennial stream once the watercourse flows NW from Hwy 89, about 1.5 miles east of Grygla.

Thief River Watershed Fluvial Geomorphology Report  June 2015 41

The slope appears to be fairly consistent from its starting point at Gunpowder Rd to CR 54 (Figure 32). The slope decreased slightly from CR 54 to the Agassiz NWR boundary. The watercourses within this subwatershed are classified as 59.0% intermittent and 40.66% perennial streams and ditches. Ditches make up the majority of the watercourse type (77.97%).

Site 6

CR 54

Site 5

Site 4 Channelization Hwy 89 Begins

CR 53

Figure 32. Water surface profile from the outlet of Thief Lake to the Agassiz NWR boundary, using 1m DEM.

The land use in the Mud River subwatershed was dominated by wetlands (46.7%) and cultivated crops (37.91%) (NLCD 2011). According to the National Agricultural Statistics Service (2012) the dominant crop types were soybeans and spring wheat. Some fields were planted to alfalfa, oats, sunflower, or corn. The soil type within the riparian zone is a poorly drained fine sandy loam and transitions into the Fluvaquents, frequently flooded‐Hapludolls complex, similar to sites 1, 2, 3, and 15. The typical soil profile for this map unit is 0‐16 inches of fine sandy loam, then 16‐80 inches of stratified loamy sand to silt loam. The adjacent soil types are loamy fine sands and fine sandy loams. The profiles contain fine sands and silt.

Site 6 was the most upstream study location on the Mud River (Appendix A). It was located 3 miles east of Grygla, north of Hwy 89. The Mud River is along the gravel road cut‐off between County Hwy 89 and Flintlock Rd. The river was delineated as a perennial drainage ditch at site 6. The Rosgen stream type was an E4. The width‐to‐depth ratios of the riffles were 8.89 and 9.47. The river was slightly entrenched on the upstream end of the longitudinal profile and moderately entrenched towards the bottom.

Thief River Watershed Fluvial Geomorphology Report  June 2015 42

The Pfankuch stability rating was 63, which is stable for an E4 stream type. The factors ranking high (more unstable) in this assessment included significant bank cutting (12‐24” high) in the lower banks. The more stable factors included no obstructions to flow, little to no scour or deposition, no particle size shift within the substrate, and abundant aquatic plants were present.

Two riffle and two pool particle counts were completed at site 6. Both riffles had a high percentage of silt/clay and very fine sand sizes. Gravel‐sized particles accounted for about a third of the samples at the riffles. The pools also had a high amount silt/clay and very fine sands; however, the pools contained more gravel than the riffles. The mean pool depth‐to‐mean riffle depth was 1.307, so the pools were deeper than the riffles; however, the longitudinal profile did not show much variation in depths.

The next study location was 4 miles to the west, just downstream of 390th Avenue. The river was delineated as a perennial stream at site 5. The Rosgen stream type was a borderline B5c/E5. The width‐to‐depth ratio at the only riffle cross section was 9.38. The entrenchment ratio was 2.21. Depending on where exactly the 1m DEM was measured, the flood prone width was slightly wider.

The Pfankuch stability rating was 101, which is unstable for both B5c and E5 stream types. The factors ranking high (more unstable) in this assessment included frequent or large mass erosion in the upper banks, significant bank cutting (12‐24” high) in the lower banks, new excessive deposition, only 0‐20% stable particles on the channel bottom, 30‐50% of the channel was experiencing scour or deposition, and perennial aquatic plants were scarce or absent. One riffle and one pool particle counts were completed at site 5. Both the pool and riffle had a high percentage of very fine sand sizes, 33.03% and 38.53%, respectively. Gravel‐sized particles accounted for 24.77% of the pool sample and 55.05% of the riffle sample. The mean pool depth‐to‐ mean riffle depth was 1.13, so the pool was deeper than the riffle; however, the longitudinal profile did not show much variation in depths.

The last study location along the Mud River was site 4. This site was 2.5 miles from site 5. The longitudinal profile was located in a straightened stretch. The Mud River was classified as a perennial stream at this location. It was a borderline C5c‐/E5 stream type. The width‐to‐depth ratios at the riffles were 11.31 and 12.39. The flood prone width was greater than 500 feet.

The Pfankuch stability rating was 91, which is moderately stable for both C5c‐ and E5 stream types. The factors ranking high (more unstable) in this assessment included moderate level of new deposition occurring on the lower banks and the channel bottom substrates had loose assortment with little overlap and moderate change in particle sizes. Scour and/or deposition were occurring in 30‐50% of the study reach. Perennial aquatic plants were scarce to absent.

One riffle and one pool particle analyses were completed at site 4. The riffle particle analysis was dominated by gravel‐sized particles (68.63%) and the pool was dominated by sand‐sized particles (75.68%). The riffle did seem to have excessive amounts of fine‐sized particles (<0.125 mm). The pool and riffle essentially had the same mean depth, with a pool‐to‐riffle ratio of 1.03.

Thief River Watershed Fluvial Geomorphology Report  June 2015 43

County Ditch 20

SUBHEADING County Ditch 20 (CD20) is a perennial drainage ditch that flows east to west. It is adjacent to County Rd. 2 on the eastern portion and 240th St. NE further downstream. Historical survey maps depict this area as a mix of marsh and prairie in the late 1800s (Trygg 1967). The water surface profile in figure x begins near Korstad Road in Hamre Township and ends at its confluence with the Thief River.

The slope appeared to be fairly consistent from its starting point to just upstream of Sharon Rd. in Lee Township (Figure 33). The energy slope steepened and maintained a relatively constant slope until a few miles upstream of its confluence with the Thief River.

Sharon Road

Site 10

Hwy 219 Site 8

Site 9

Figure 33. Water surface profile of County Ditch 20, using 1m DEM.

According to the 2011 National Land Class Database (NLCD) 37.15% of the CD 20 drainage area was cultivated crops. Emergent and woody wetlands made up 44.82% of the area. Hay/pasture land made up 7.43% and only 0.17% was developed (low intensity). The dominant crop types were soybeans and spring wheat (NLCD 2012). Some fields were planted to corn, alfalfa, and barley.

Thief River Watershed Fluvial Geomorphology Report  June 2015 44

The soil type within the riparian zone starts on the eastern side as fine sandy loams and Smiley loam, which contains a layer of clay loam from 12 to 19 inches within the soil profile (Soil Survey Staff 2015). As the watercourse flows west, the riparian soil area changes between the fine sandy loams, Smiley loam, and patches of muck. Near the Elm Lake WMA the riparian zone soil type is dominated by Smiley loam, with areas of Berner and Cathro ponded soils and various loams. From the confluence with the angle ditch to the outlet at the Thief River, the soil types are an assortment of various loams and Clearwater clay map units.

Three intensive sites were completed on CD 20. The most upstream site was site 10. This site was located just downstream of County Rd. 54, south of Grygla. CD 20 was delineated as a perennial drainage ditch at site 10. The Rosgen stream type was a B4c. The width‐to‐depth ratio on the upstream end was 9.08 and downstream end was 10.59. The entrenchment ratios were 1.81 and 1.53.

The Pfankuch stability rating was 73, which is moderately stable for a B4c stream type. The factors ranking high (more unstable) in this assessment included upper bank slope gradients from 40‐60%, significant cutting on the lower banks (12‐24” high), and a moderate shift in particle sizes on the channel bottom. Factors ranking low (more stable) included no evidence of past or future mass erosion and no debris jam potential within the upper banks, riparian vegetation was robust and diverse, there were no obstructions to flow in the lower banks, and there was abundant aquatic plant growth on the channel bottom.

Two riffles and one pool cross‐section was completed at site 10. Pebble counts were completed on the upstream riffle and pool cross‐sections. Both the riffle and pool cross‐sections had a high percentage of gravel‐sized particles, 55.96% and 53.33%, respectively. The riffle cross‐section contained 33.94% silt/clay and 9.17% sand‐sized particles. The pool cross‐section contained 19.05% silt/clay and 26.67% sand‐sized particles. The mean pool depth‐to‐mean riffle depth was 1.26 and the pool area‐to‐riffle area was 1.2.

The next site was 19 miles downstream of site 10. Site 9 was located south of Elm Lake Wildlife Management Area on 240th St., 1.5 miles east of the outlet of the angle ditch. CD 20 was delineated as a perennial drainage ditch at this location. The Rosgen stream type was a B5c. The width to depth was 9.14 and the entrenchment ratio was 1.81.

The Pfankuch stability rating was 74, which is moderately stable for a B5c stream type. The factors ranking high (more unstable) in this assessment included new deposition on the lower banks, mostly loose assortment of channel bottom substrate, and present but spotty aquatic vegetation within the stream channel. Factors ranking low (more stable) included no evidence of past or future mass erosion and no debris jam potential within the upper banks, riparian vegetation was robust and diverse, and there were no obstructions to flow in the lower banks.

Thief River Watershed Fluvial Geomorphology Report  June 2015 45

Three riffles and one pool cross‐section were completed at site 9. Pebble counts were completed at one riffle and the pool cross‐sections. Silt/clay made up 28.04% of the riffle pebble count and 19.27% of the pool pebble count. Except for the fines presence at the riffle cross‐ section, the pool pebble count had a higher abundance of small particle sizes. At the pool cross section, 90.83% of the particles were a very fine gravel or smaller. Gravel (all sizes) made up 50.47% of the riffle count. The mean pool depth‐to‐mean riffle depth was 1.04 and the pool area to riffle area was 0.98.

The most downstream study location on CD 20 was site 8. It was located immediately upstream of the confluence with the Thief River. CD 20 was delineated as a perennial stream at this location. The Rosgen stream type was a B5c. The width‐to‐depth ratio was 12.63 and the entrenchment ratio was 1.71.

The Pfankuch stability rating was 120, which is unstable for a B5c stream type. The factors ranking high (more unstable) in this assessment included extensive deposition of predominantly fine particles within the lower banks, a marked distribution of particle sizes with the channel bottom, more than 50% of the channel bottom was within a state of flux, and aquatic perennial plants were scarce or absent.

Three riffles and one pool cross‐section were completed at site 8, though grade control structures were installed within this study reach in 2011. The location of the pool cross‐section and one of the riffle cross‐sections were replaced by these structures. These cross‐sections were not surveyed in 2012. Pebble counts were completed at one riffle and the pool cross‐sections. The riffle, pool, and full study reach counts were dominated by fine particle sizes. The 50th percentile particle size of these three pebble counts was 0.61 mm, 0.28 mm, and 0.12 mm, respectively. The riffle cross‐ section did have slightly larger particles than the pool. The mean pool depth‐to‐mean riffle depth was 1.06 and the pool area‐to‐riffle area was 1.12.

Thief River Watershed Fluvial Geomorphology Report  June 2015 46

RECOMMENDATIONS

Thief River

THIEF LAKE TO AGASSIZ NATIONAL WILDLIFE REFUGE Reconnaissance on the upper reach of the Thief River, from Thief Lake to Agassiz NWR, averaged about 71.5 tons/yr./mi. of stream bank erosion. On the upstream end of this reach, the Thief River had access to its floodplain, but the floodplain width was narrow. Site 1 was classified as a relatively stable B stream type, but the upper banks did have steep, vertical banks. If the floodplain had more capacity, the flood flows could dissipate over a larger area. The 50‐yr. flood elevation was confined to about 70 feet, but the riparian zone was approximately 300 feet wide. It is recommended that the spoil piles from historical channelization be pulled back to the outside edge of the riparian zone (Figure 34). In the example of figure 34, the spoil piles are closest to the location represented by the letter “A”. It would be beneficial for stream stability to pull the spoil pile back to the location of letter “B”. In addition, site 1 rated low for riparian zone vigor and rooting depth. This area would benefit from a rejuvenation of the plant vigor.

B A

Figure 34. Stream channel and floodplain connectivity. Image Source: FISRWG (10/1998).

Further downstream, sites 2 and 3 were located near the middle of this river reach. Both of these locations had a Pfankuch rating of stable. Both sites were typed out as C4c‐, which is a very stable stream type in northwestern MN. Site 2 had a robust and diverse riparian buffer. Site 3 also had an adequate buffer. One recommendation for this area would be to maintain the floodplain access and connectivity, as well as the current riparian buffer condition (Table 6). The Pfankuch rating at site 3 did show evidence of potential mass erosion on the upper banks,

Thief River Watershed Fluvial Geomorphology Report  June 2015 47

Table 6. Summary of stressor sources and protection and restoration strategies.

Substrate Stressor Sources Protection & Restoration Strategies

to

Stream sources Existing

Site River buffers upland

water

Erosion buffers Type

Sediment

channel Change

Buffer

meander Agricultural BMPs Bedrock Boulders Cobble Gravel Sand Silt/Clay None Incised Aggraded Entrenched Channelized Slope Excess Upland Poor Maintain Condition Protect Grazing Management Restore storage Improve Increase Floodplain Identify sediment Allow 1Thief River B4c  2Thief River C4c‐    3Thief River C4c‐  4Mud River C5c‐ / E5   5Mud River B5c / E5  6Mud River E4  7Thief River B5c       8 County Ditch 20 B5c  9 County Ditch 20 B5c  10 County Ditch 20 B4c  12 Moose River E5  14 Thief River C4c‐  15 Thief River E5 

Thief River Watershed Fluvial Geomorphology Report  June 2015 48

significant bank cutting on the lower banks, and some new deposition. The riffle pebble count also appeared to be inundated with fine sediments. Based on the results of this assessment, along with the results of the SWAT model and sediment fingerprinting assessment, it would be recommended to continue locating upland sources of sediment. The condition of the Thief River dropped to moderately stable by the time it reached site 15, just upstream of Agassiz NWR. The Pfankuch rating at this location indicated that new deposition was occurring and that the channel bottom was being inundated with excessive sediment. It would be important to identify upland sediment sources and improve the riparian buffers at this location. Opportunities to remeander the Thief River north of the Agassiz NWR boundary could also be explored.

AGASSIZ NATIONAL WILDLIFE REFUGE TO THIEF RIVER FALLS At the outlet of Agassiz NWR, the Thief River is in a channelized form. From CR 7 to 12 the BANCS model estimated a relatively low rate of streambank erosion. Site 7 was located within this reconnaissance reach. It was a B5c stream type with an unstable Pfankuch stability rating. Most of the features of the upper and lower banks rated in the fair to good stability range, but the channel bottom scores dramatically increased the rating at this location. The channel bottom in this reach (CR 7 to 12) was in flux.

From CR 12 to 140th Ave. the BANCS model predicted a slightly larger amount of erosion, a similar rate to the prediction upstream of Agassiz NWR. Site 14 was located within this reconnaissance reach. It was a C4c‐ stream type with a stable Pfankuch stability rating. At this location there was adequate pool habitat, and at the riffles, the stream had enough power to move the sediment supplied at this location downstream. The recommendation would be to maintain the existing condition of the channel and riparian buffers. This recommendation includes the consideration of long‐range planning for riparian zones. Stable rivers are in a dynamic equilibrium, meaning they are stable in their current pattern and profile, not necessarily location. A natural channel migrates laterally by erosion of one bank, maintaining on the average a constant channel cross‐section by deposition on the opposite bank. In other words, there is equilibrium between erosion and deposition (Leopold 1994). Figure 35 demonstrates how a river channel migrates down the valley over time. Protecting the buffers will maintain the existing boundary condition and maintain a naturally slow rate of channel migration.

Figure 35. Meander migration. Source: DNR Resource Sheet – Streambank Erosion and Restoration.

Thief River Watershed Fluvial Geomorphology Report  June 2015 49

The Thief River in this area is also designated as State Ditch 83. It is periodically cleaned out to remove sediment. Prior to initiating a new clean‐out it would be beneficial to assess whether the river is currently in a stable form. The stable stream type for the Thief River downstream of Agassiz NWR appears to a C4/5. This stable stream type has a higher width‐to‐depth ratio and moderate sinuosity. Figure 36 shows the comparison between Rosgen B and C stream types. Note that the dotted line above the water surface represents the floodplain width. The C stream type is unconfined on one bank, while the B stream type remains confined in both directions. B stream types also tend to have a higher slope.

While some stable stream types do not have obvious sediment deposition, C stream types do exhibit deposition along the inside of the bends (Figure 36). The outside bank at a bend normally has a slightly higher bank and is where bank erosion naturally occurs. If it is determined that State Ditch Figure 36. Rosgen B and C 83 is unstable due to excessive sedimentation, the cause should be Stream Types. identified and alleviated. The Thief River in its stable form should be able to efficiently move both the water and sediment supplied from upstream, while at the same time maintaining those depositional point bars characteristic of C stream types.

On the most downstream portion of the Thief River (140th Ave to CR 63) the BANCS model predicted an annual erosion rate of 161 tons/yr./mi. This was the highest estimate along the main stem of the Thief River. Creating setbacks within the riparian zone will minimize future problems with the Thief River encroaching on development, and it will allow for easier management of the river channel and floodplain as one hydrologic resource. Figure 37 is a series of photos of poor riparian buffer management, taken on the lower portion of the Thief River.

Figure 37. Examples of poor riparian management. Photos taken during reconnaissance in 2010.

Thief River Watershed Fluvial Geomorphology Report  June 2015 50

Moose River

Two segments of the Moose River had BANCS erosion estimates. The upper segment predicted a higher erosion rate (241 tons/yr./mi.), while the lower segment predicted an extremely low erosion rate (1.06 tons/yr./mi). During the data collection for the BANCS model numerous upland erosion sources were identified along the upper reach (Figure 38).

Figure 38. Examples of gully erosion within the Moose River Watershed. Photos taken during 2010 reconnaissance.

Site 12 was located on the downstream end of the upper reconnaissance segment. The Pfankuch stability rating at this location produced an unstable rating. Many of the factors that rated high, or unstable, were related to bank cutting and channel bottom substrates that were not stable. One recommendation for the upstream portion of the Moose River would be to identify upland sources of sediment and lessen the contribution of excessive sediment to the river channel. A stream power index was created for the entire Thief River Watershed and would aid in identifying potential locations. Appendix C is an example of how to identify near‐channel erosion using this index.

In addition to removing the sources of excessive sediment, the upstream portion of the Moose River would benefit from riparian buffer improvements. As an E stream type, stability of the river is very highly dependent on robust riparian vegetation. E stream types have a low width‐to‐ depth ratio, which naturally creates steeper, more vertical banks. Without robust vegetation, streambank erosion is likely to occur. At site 12 the Moose River does have adequate access to the floodplain. This allows the river dissipate energy during flood events. Ensuring that the Moose River has access to its adjacent floodplain in additional areas will increase river stability as well.

Downstream of Hwy 89, the BANCS model estimated an extremely low rate of erosion. The Moose River is well‐buffered within the Thief Lake WMA and the river appears to be in a stable condition, based on the reconnaissance effort. The recommendation for the downstream segment of the Moose River is to maintain its current condition. If stability issues do arise, this segment would benefit from flattening of the spoil piles to increase floodplain capacity.

Thief River Watershed Fluvial Geomorphology Report  June 2015 51

Mud River

Three reconnaissance reaches were assessed for streambank erosion estimates. The most upstream reach, starting at Flintlock Rd, had a high erosion rate of 239 tons/yr./mi. Site 6 was located at the start of this reconnaissance reach. The Pfankuch rating did highlight that significant streambank cutting was occurring on the lower banks. Based on pictures and aerial photos, the riparian zone adjacent to the Mud River at site 6 appears to mimic a drainage ditch buffer. Even with the presence of bank cutting and lack of riparian buffer, site 6 still produced a stable Pfankuch rating. The channel bottom substrate appeared to be in good condition with little to no excessive sediment sources or the stream channel had enough stream power to move any sediment inputs downstream. The recommendations for this upper reach of the Mud River would be to maintain the current form of the channel, maintain access to the floodplain, and improve the riparian buffer.

The middle reconnaissance reach, located just downstream of Grygla, produced a streambank erosion rate of 137 tons/yr./mi. Site 5 was located near the top of this study reach. The Pfankuch Stability Rating at this location was unstable. Numerous factors played a role in this poor rating including evidence of mass erosion on the upper banks, significant lower bank cutting, only 0‐20% of the channel substrate was stable, and perennial aquatic plants were scarce or absent. This location, just downstream of Grgyla, was inundated with excessive fine sediments. The left bank had a decent riparian buffer, but on the right bank it was nearly absent. The recommendations for this middle reach of the Mud River would be to identify upland sources of sediment and lessen the contribution of excessive sediment to the river channel. The stream power index would be helpful in identifying priority areas. In addition, the left riparian buffer should be maintained, while the right buffer should be improved.

The downstream reconnaissance reach appeared to have relatively stable stream banks, with a streambank erosion estimate of 136 tons/yr./mi. Site 4, located near the middle of this reach, was moderately stable. The factors pushing this site towards instability included new deposition on the lower banks, a loose assortment of channel bottom substrate, a moderate change is particle sizes, and perennial plants were absent from the channel bottom. The pebble counts also highlighted the presence of excessive sediment within the channel bottom. The recommendations for this lower reach would be the same as the middle reconnaissance reach: identify upland/upstream sources of sediment and maintain the existing riparian buffer.

County Ditch 20

During the reconnaissance in 2011, BANCS erosion estimates were collected on 2 long segments of CD 20. The upstream segment, starting at Jelle Rd., had a bank erosion estimate of 210 tons/yr./mi. The lower segment, starting at Hwy 53, had a much lower estimate of erosion (54 tons/yr./mi.).

Thief River Watershed Fluvial Geomorphology Report  June 2015 52

Site 10 was located within the upper reconnaissance segment. It was moderately stable with evidence of bank cutting and steep, upper banks. One of the main stressors at this location was the narrow floodplain width. With the low width‐to‐ depth ratio, the channel was very effective at moving water and sediment, but during higher flows the floodplain capacity was lacking. Flooding along CD 20 and 200 were noted as concerns in the local RLWD 10‐yr. management plan. CD 20 at this location had access to its floodplain, but it would benefit from a wider floodplain. Creating a Figure 39. Poor buffer condition along CD 20. wider floodplain would alleviate some of the flooding concerns expressed by landowners and allow this waterway to be more stable. The channel should be maintained at its current dimensions because it has enough shear stress to move the water and sediment supplied to it. When a clean‐out project is proposed on this section of CD 20 it would be recommended to maintain the narrow width‐to‐depth ratio, but create additional capacity within the floodplain. Establishing a diverse riparian buffer will increase the roughness of the streambank, thereby reducing near bank velocities and likelihood of bank cutting (Figure 39).

Sites 8 and 9 were located within the downstream reconnaissance reach. CD 20 upstream and downstream of site 9 was in the best condition. There was evidence of bank cutting, but it was more minimal than at other locations. The main concern in the area south of Elm Lake WMA was excessive sediment within the stream channel. During the reconnaissance, large quantities of sand bedload were observed moving downstream. Some areas were experiencing new deposition by some of this load. Figure 40 shows the sand bedload at the confluence with the Thief Figure 40. Sand bedload deposited in Thief River near outlet of River. Aquatic plant life was scarce, as well, CD 20. which would indicate an unstable channel bottom. One recommendation would be to identify upstream sources of sediment and lessen the contribution of excessive sediment to the river channel. These sources could be from the bank cutting that is occurring upstream or from upland sources. A stream power index was created for the entire Thief River Watershed and would aid in identifying potential locations.

Thief River Watershed Fluvial Geomorphology Report  June 2015 53

One additional concern is the limited floodplain capacity. With the low width‐to‐depth ratio (9.14), this channel is functioning like an E stream type, but with the narrow floodplain width it is typed as a B channel. It is recommended to look for additional flood storage within the floodplain, and not directly within the bankfull channel. Conclusions

In conclusion, the stability of a river depends on maintaining a natural equilibrium between sediment supply (both size of particles and quantity), channel slope, channel length, and stream flow. All four of these factors have changed over time within the TRW, whether due to natural variability or human influences. Comparing historical stream flow records with precipitation rates, both are increasing and stream flow rates are increasing faster than the precipitation rates. Straightening of the meandering river channels and natural meander cut‐offs have changed slope and length; thereby possibly changing the sediment supply. Inadequate buffers allow upland sediment to enter the channels. Confined stream channels keep higher velocity stream flows within the channel, creating excessive bank erosion.

Maintaining the natural equilibrium of the Thief, Moose, Mud, and other altered natural watercourses will not be successful by creating a patchwork of fixes on outside eroding banks or digging slugs of deposited sediment from the river channel. The question must be asked – why is the outside bank eroding at an accelerated rate? Why is the river unable to effectively move the sediment being supplied from upstream? Once answered, those driving forces should be corrected or at least mitigated for. Once stable, a healthy watershed has natural buffering ability against shifts to more unstable conditions.

If the factors pushing these watercourses towards instability cannot be alleviated, at a minimum, the conditions near the watercourses must be improved. The rate at which instability worsens can be slowed by ensuring the following:  Perennial vegetation with robust and dense root mass is present adjacent to the stream  The vegetative planting is of an appropriate width to function as a buffer  Watercourses must have access to a floodplain at the bankfull elevation (1.5‐2 year flood recurrence interval)  Artificial and altered natural watercourses should be designed and maintained with an appropriate base flow channel with floodplain access

This report does not go as far as laying out a prioritization strategy for restoration and protection projects. The identified concerns (stressor pathways) and suggested recommendations (restoration and protection strategies) (Table 6) can be combined with other natural resource management goals, as well as land management concerns (e.g. bank sloughing in specific areas) to create a watershed‐wide vision for management. Further developing a collaborative management plan for the Thief River Watershed will guide these watercourses towards a dynamic stable condition.

Thief River Watershed Fluvial Geomorphology Report  June 2015 54

BIBLIOGRAPHY

Benson, M.A. and D.M. Thomas. 1966. A definition of dominant discharge. Hydrological Sciences Journal. 11(2), pp. 76‐80.

Dury, G.H. 1973. Magnitude‐frequency analysis and channel morphology. Fluvial Geomorphology, Binghamton Symp. Geomorphol. vol. 4, pp. 91‐121, State Univ. of N.Y. at Binghamton, Binghamton.

Environmental Protection Agency (EPA). 2010. Minnesota Causes for Reporting Year 2010. Online at http://ofmpub.epa.gov/waters10/attains_state.control?p_state=MN&p_cycle=2010&p_re port_type=A#causes. Retrieved May 2, 2015.

FISRWG. 10/1998. Stream Corridor Restoration: Principles, Processes, and Practices. By the Federal Interagency Stream Restoration Working Group (FISRWG)(15 Federal agencies of the US gov't). GPO Item No. 0120‐A; SuDocs No. A 57.6/2:EN 3/PT.653. ISBN‐0‐934213‐59‐3.

Frontier Precision. 2013. Four Questions about Site Calibration. Message posted to http://fpisurvey.com/2013/04/29/four‐questions‐about‐gnss‐site‐calibrations/. Retrieved May 29, 2015.

GAP Stewardship. 2008. Public Ownership by Agency. Raster GIS layer. Assessed from DNR Data Deli on January 15, 2015.

Harris, K.L., West, S.A., Lusardi, B.A., and Tipping, R.G. 1995. Quaternary Stratigraphy. RHA‐3, PART A. Plate 2. Minnesota Geological Survey.

Hey, R.D. 1975. Design discharge for natural channels. Science, Technology and Environmental Management. pp. 73‐88. Saxon House, Farnborough, U.K.

Houston Engineering, Inc. 2010. Thief River SWAT Modeling Thief River Watershed, Minnesota. Report prepared for Red Lake Watershed District.

Inglis, C.C. 1949. The behavior and control of rivers and canals (with the aid of models). No. 13. Yeravda Prison Press.

Kondolf, G.M. 1997. Profile: Hungry Water: Effects of Dams and Gravel Mining on River Channels. Environmental Management. Vol 21: Issue 4, pp. 533‐551.

Lane, E.W. 1955. Design of stable channels. Transactions of the American Society of Civil Engineers, Vol. 120, pp. 1234‐1279.

Thief River Watershed Fluvial Geomorphology Report  June 2015 55

Langbein, W.B. and L.B. Leopold. 1966. River Meanders – Theory of Minimum Variance. US Government Printing Office.

Leopold, L.B. and M.G. Wolman. 1960. River Meanders. Bulletin of the Geological Society of America, vol. 71, pp. 769‐794.

Leopold L.B., M.G. Wolman, and J.G. Miller. 1964. Fluvial Processes in Geomorphology. W.H. Freeman, San Francisco.

Leopold, L.B. 1994. A View of the River. Harvard University Press, Cambridge, Massachusetts.

Lorenz, D. L., C. A. Sanocki, and M. J. Kocian. 2009. Techniques for estimating the magnitude and frequency of peak flows on small streams in Minnesota based on data through water year 2005. U.S. Geological Survey Scientific Investigations Report: 2009‐5250. 54pp.

Lusardi, B.A. 1997. Quaternary Glacial Geology. Minnesota at a Glance. Minnesota Geological Survey, University of Minnesota. Retrieved 2014‐03‐13.

National Agriculural Statistics Service (NASS). 2012. Online at www.nass.usda.gov. Retrieved April 14, 2015.

National Land Cover Dataset (NLCD). 2011. Online at www.mrlc.gov. Retrieved April 14, 2015.

Ojakangas, R.W. and Matsch, C.L. 1982. Minnesota’s Geology. Minneapolis, University of Minnesota Press.

Pfankuch, D.J. 1975. Stream Reach Inventory and Channel Stability Evaluation. USDA Forest Service, R1‐75‐002. Govt. Printing Office, #696‐260/200, Wash. D.C. 26 pp.

Red Lake Watershed District. 2006. 10‐Year Management Plan.

Rosgen, D.L. 1994. A classification of natural rivers. Catena. 22: 169‐199.

Rosgen, D.L. 1996. Applied River Morphology. Pagosa Springs, CO: Wildland Hydrology.

Rosgen, D.L. 2001(a). A practical method of computing streambank erosion rate. In Proceedings of the seventh Federal Interagency Sedimentation Conference: Vol. 1, pp. II9‐II‐15. Reno, N.V.: Subcommittee on Sedimentation.

Rosgen, D.L. 2006(b). Watershed Assessment of River Stability and Sediment Supply (WARSSS). 2nd edition 2009. Fort Collins, CO: Wildland Hydrology.

Thief River Watershed Fluvial Geomorphology Report  June 2015 56

Rosgen, D.L. 2008. River Stability Field Guide. Fort Collins, CO: Wildland Hydrology.

Rosgen, D. L. 2012. Applied River Morphology: Training Manual. Wildland Hydrology, Fort Collins, CO. 268 pp.

Schottler S.P. and D.R. Engstrom. 2011. Sediment loading and sources to Agassiz National Wildlife Refuge. St. Croix Watershed Research Station Final Report, 22 p

Soil Survey Staff, Natural Resources Conservation Service, United States Department of Agriculture. Web Soil Survey. Available online at http://websoilsurvey.nrcs.usda.gov/. Retrieved November 15, 2015.

Trygg, J.W. 1967. Composite Map of United States Land Surveyors’ Orginal Plats and Field Notes. Sheet 22. Minnesota Series, Ely, Minnesota.

Williams, G.P. 1986. River Meanders and Channel Size. J. Hydrol., 88:147‐164.

Wolman, M. G. 1954. A method of sampling coarse riverbed material. Transactions of the American Geophysical Union 35: 951‐956.

Thief River Watershed Fluvial Geomorphology Report  June 2015 57

APPENDIX A DETAILED SITE ASSESSMENTS

MORPHOLOGICAL DESCRIPTION

Site 1

INTRODUCTION Site 1 was located on the Thief River, downstream of the outlet from Thief Lake, in section 29 of Thief Lake Township (Figure 1). According to StreamStats, the drainage area is 249 mi2. The subwatershed that drains to site 1 is outlined pink in Figure 1.

Figure 1. Watershed that drains to site 1.

This area of the Thief River was ditched and is fairly straight (Figure 2). The valley walls are confined by spoil piles from those clean‐outs. Areas of State Ditch 83 (Thief River) are still periodically spot cleaned where sedimentation or snags are major flow restrictions or deflect flow into streambanks.

Figure 2. Aerial view of site 1. Source: 2013 DNR aerial photo, 3m DEM, and ArcGIS Image Service best available resolution hillshade.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 1

LONGITUDINAL PROFILE The thalweg, or channel bottom, through site 1 appeared to follow a normal pattern (Figure 3). The river became shallower through the glides and riffles and deeper in the pools. The thalweg towards the end of the longitudinal profile had an extended glide as it came out of the pool at 9+95. The bankfull slope was 0.00018 using just two bankfull calls, one at riffle cross section 0+29 and the other at pool cross section 9+95. The water surface slope was 0.00025. Since there were only two bankfull calls to create a bankfull slope, the water surface slope was used for the velocity and discharge calculations.

Thief River: Site 1 Longitudinal Profile 2011

1165

1163 Thalweg

1161 Bankfull Slope = 0.00018 GM1 Pool XS 9+95, 2011 GM1 Riffle XS 0+29, 2011

GM1 Riffle XS 13+20, 2011 Water Surface

1159 GM1 Pool XS 9+95 SB, 2011

1157 Water Surface Slope = 0.00025 Bankfull

1155 PINS Elevation (ft)

1153 P2 1151

1149 P3

1147 P4 1145 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 1000 1050 1100 1150 1200 1250 1300 1350

Distance along stream (ft)

Figure 3. Longitudinal profile at site 1 in 2011.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 2

CROSS‐SECTIONS Two riffle cross‐sections were completed at site 1. Riffle 0+29 was used for the velocity and discharge calculations (Figure 4). The bankfull area was 161.73 ft2 and the bankfull width was 41.55 ft. The second riffle was located 1,320 ft. from the start of the longitudinal profile. The bankfull area was 190.67 ft2 and the bankfull width was 49.91 ft.

The entrenchment ratio, comparing flood‐prone width to bankfull width, came out as moderately entrenched for both riffle cross‐sections. Riffle 0+29 had a flood‐prone width of 64 ft., giving it an entrenchment ratio of 1.54. Riffle 13+20 had a flood‐prone width of 107 ft., giving it a ratio of 2.15.

Thief River: Site 1 Riffle 0+29, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 41.5 Dbkf = 3.89 Abkf = 161.7 1185

1180 Bankfull Width = 41.55 ft. Mean Depth = 3.89 ft. 1175 Bankfull Area = 161.73 ft2

1170 Flood‐prone Elevation 1165

1160 Bankfull Elevation Elevation (ft)

1155

1150

1145 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Horizontal Distance (ft)

Figure 4. Riffle cross section 0+29 at site 1.

The D50, or 50th percentile, particle size was 4.24 mm. Silt and clay made up 33.64% of the particles measured at this cross‐section. Gravel, cobble and boulder‐sized particles were also tallied. The D84 used for the velocity and discharge calculations was 15.06 mm (Figure 5). A pebble count was also completed at riffle cross‐section 13+20. The D50 and D84 were 10.12 mm and 35.31 mm, respectively. Silt and clay only made up 11.4% at this cross‐section.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 3

40.00% Active Riffle Particle Analysis 0+29 110.00% 35.00% 90.00% (%) 30.00%

(%) 70.00% 25.00% Volume 20.00% 50.00%

Volume 15.00% 30.00% 10.00% 10.00% 5.00% Cumulative 0.00% ‐10.00% 4 2 8 ‐ ‐ ‐ 16 32 45 64 90 1.0 5.7 .50 .25 2 ‐ ‐ ‐ ‐ ‐ 180 256 362 512 128 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 1.0 4 5.7 ‐ <.062 32 45 64 ‐ 8 .50 .25 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 5. Active riffle particle analysis at cross section 0+29.

One pool cross‐section was completed at site 1 (9+95). The bankfull area was 202.23 ft2 and the bankfull width was 43.24 ft. (Figure 6). Compared to riffle cross‐section 0+29, the pool was slightly wider and had a greater cross‐sectional area.

Thief River: Site 1 Pool 9+95, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 43.2 Dbkf = 4.68 Abkf = 202.2 1180

1175 Bankfull Width = 43.24 ft. Mean Depth = 4.68 ft. 1170 Bankfull Area = 202.23 ft2 1165

1160 Bankfull Elevation

Elevation (ft) 1155

1150

1145

1140 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 Horizontal Distance (ft)

Figure 6. 2011 Pool cross section 9+95 at site 1.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 4

The study bank overlay at pool cross‐section 9+95 indicated an actual erosion rate of .0145 ft/yr from 2011 to 2012 (Figure 7). This rate closely matched the predicted erosion rate of .016 ft/yr derived from the BANCS model.

Thief River: Site 1 Pool Study Bank Overlay 2011-2012 Site 1 Pool 9+95 Study Bank, 2011 Site 1 Pool 9+95 Study Bank, 2012

1162

1161

1160

1159

Elevation (ft) 1158

1157

1156

Toe Pin 1155

1154 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Horizontal Distance (ft)

Figure 7. Site 1 pool cross‐section 9+95 study bank overlay showing the study bank profile in 2011 and 2012.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 5

DISCHARGE TABLE The estimated velocities and discharges at riffle cross‐section 0+29 were very similar between different methods (Table 1). Generally the velocities were approximately 2.3 ft/sec and the estimated discharge was 380 cfs. StreamStats estimated the 1.5, 2, 5, 10, 25 and 50 year recurrence interval discharges at 388, 510, 837, 1060, 1330 and 1530 cfs, respectively.

Table 1. Velocity and discharge estimates for riffle cross‐section 0+29.

BANKFULL VELOCITY & DISCHARGE ESTIMATES Stream: Thief River Location: Site 1 Observers: Dave Friedl Date: 8/19/2011

Riffle Cross‐section: Riffle 0+29 Bankfull Slope: 0.00025 D84 of Active Bed: 15.06 Dia. (mm) Drainage Area: 249 Mi2

Estimation Methods Bankfull Velocity Bankfull Discharge Manning Limerinos n 2.368 ft / sec 382.977 cfs Darcy‐Weisbach 2.427 ft / sec 392.574 cfs U/U* 2.28 ft / sec 369.53 cfs Stream Type (medium to large) 1.498 ft / sec 242.272 cfs

PARTICLE DISTRIBUTION The full reach particle analysis was dominated by fine sand‐sized particles (23.9%). Silt and clay, as well as an assortment of gravel sizes, was also present (Figure 8). The Web Soil Survey (2015) maps the riparian zone as a Fluvaquents, frequently flooded‐Hapludolls complex (map unit I16F).

The typical soil profile is: A ‐ 0 to 16 in.: fine sandy loam

Cg ‐ 16 to 80 in.: stratified loamy sand to silt loam

A C soil horizon contains glacial or post‐ glacial material with little affected by soil forming processes. A lowercase g denotes the presents of gray colors, indicative of permanent or periodic intensive reduction due to saturation. Though there is a dominant presence of clay, silt and fine sands the D50 for the full reach count was 5.7 mm (gravel). Cobble and Figure 8. Bank along site 1. Photo taken 8/18/2011. boulder‐sized particles were also tallied, but made up a small percentage of the overall count (Figure 9).

Thief River Watershed Fluvial Geomorphology Report  Appendix A 6

30.00% Reach Particle Analysis ‐ Site 1 110.00%

25.00% 90.00% (%)

20.00% 70.00% (%)

Volume 15.00% 50.00%

Volume 10.00% 30.00%

5.00% 10.00% Cumulative

0.00% ‐10.00% 4 2 8 ‐ ‐ ‐ 16 32 45 64 90 .25 5.7 .50 1.0 2 ‐ ‐ ‐ ‐ ‐ 180 256 362 512 128 ‐ ‐ ‐ ‐ 11.3 .125 22.6 ‐ 1024 2048 ‐ ‐ 1.0 4 5.7 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 9. Particle analysis for the full study reach at site 1. RIPARIAN ZONE The land adjacent to Thief River at site 1 was dominated by grass and forbs, making up approximately 50% of the riparian zone cover (Figure 10). The dominant species present were 50% Bromus inermis (smooth brome), 20% Phalaris arundinacea (reed canarygrass), 5% Solidago spp. (goldenrod), 5% Melilotus officinalis (sweet clover), 5% Asclepias spp. (milkweed), 5% Cirsium arvense (Canada thistle) and 10% other forbs. Understory shrubs covered approximately 10% of the riparian zone. Prunus virginiana (Chokecherry) made up 40% of the shrub Figure 10. Site 1 vegetation cover. Photo taken on 8/18/2011. layer. Other shrubs present included 25% Cornus sericea (redosier dogwood), 20% Amelanchier spp. (Juneberry), 5% Lonicera spp. (honeysuckle), 5% Crataegus spp. (Hawthorne) and 5% Corylus spp. (hazelnut).

The tree canopy covered approximately 55% of the aerial cover and 20% of the basal cover. Tree species included 35% Fraxinus pennsylvanica (green ash), 30% Ulmus americana (American elm), 20% Quercus macrocarpa (bur oak), 8% Acer neguno (box elder), 5% Populus balsamifera (balsam poplar) and 2% Populus deltoides (cottonwood).

Thief River Watershed Fluvial Geomorphology Report  Appendix A 7

CLASSIFICATION Table 2. Stream classification input factors at The Rosgen stream type was classified as a B4c (Table 2). site 1. The number 4 denotes the 50th percentile size of particles. With a D50 of 5.7 mm, site 1 was considered a gravel STREAM CLASSIFICATION ‐ Site 1 stream. The lowercase c signifies that the energy slope at STREAM TYPE B4c this site was less than 0.02. The entrenchment ratio of both VALLEY TYPE X riffle cross‐sections fell within the moderately entrenched category. B stream types usually have a width‐to‐depth Slope 0.00025 (w/d) ratio of greater than 12.0. Riffle 13+20 came in just above this at 12.4, but riffle 0+29 was slightly below with a w/d ratio of 10.68. Sinuosity was 1.16. Entrenchment Ratio 1.55 The modified (by stream type) Pfankuch stability rating on 10/23/2012 was an 84. A good, or stable, rating range for a Width/Depth Ratio 10.68 B4 is 40‐64. A fair, or moderately stable, rating is 65‐84. B stream types are moderately entrenched with a moderate Sinuosity 1.16 gradient. They usually are riffle dominated with infrequently spaced pools. B4 stream types are moderately sensitive to disturbance, but have an excellent recovery 5.7 mm Channel Material (D ) potential when the disturbance is removed. This stream 50 (GRAVEL) type is moderately dependent on robust stream bank vegetation to remain stable and the potential for streambank erosion is low.

DIMENSIONLESS RATIOS The upstream and downstream riffle had similar morphology. The downstream riffle (13+20) was slightly larger than the upstream riffle (Table 3). Only 1 pool cross‐section was completed at this site. The mean pool depth‐to‐mean riffle depth ratio was 1.21, so the riffles were shallower than the pools. The entrenchment ratio of a B stream type ranges from 1.4‐2.2.

Table 3. Channel morphology and dimensionless ratios at site 1.

CHANNEL MORPHOLOGY & DIMENSIONLESS RATIOS Stream: Thief River Site: 1 Observers: Dave Friedl Date: 8/19/2011

Riffle Dimensions Mean Min Max Dimensionless Ratios Mean Min Max

Riffle Width (Wbkf) 45.7 41.6 49.9 Width of Flood‐Prone Area (Wpfa) 85.8 64.6 107

Mean Riffle Depth (dbfk) 3.86 3.82 3.89 Riffle Width/Depth Ratio (Wbkf / dbkf) 11.87 10.68 13.07

Maximum Riffle Depth (dmax) 5.37 5.19 5.54 Max Riffle Depth to Mean Riffle Depth (dmax / dbkf) 1.392 1.359 1.424 2 Riffle Cross‐Sectional Area (Abkf) (ft ) 176.2 161.73 190.67 Entrenchment Ratio (Wfpa / Wbkf) 1.85 1.55 2.15

Pool Dimensions

Pool Width (Wbkfp) 43.2 43.2 43.2 Pool Width to Riffle Width (Wbkfp / Wbkf) 0.946 0.946 0.946

Mean Pool Depth (dbfkp) 4.68 4.68 4.68 Mean Pool Depth to Mean Riffle Depth (dbkfp / dbkf) 1.212 1.212 1.212

Maximum Pool Depth (dmaxp) 6.97 6.97 6.97 Pool Area to Riffle Area (Abkfp / Abkf) 1.148 1.148 1.148 2 Pool Cross‐Sectional Area (Abkfp) (ft ) 202 202 202 Max Pool Depth to Mean Riffle Depth (dmaxp / dbkf) 1.806 1.806 1.806

Thief River Watershed Fluvial Geomorphology Report  Appendix A 8

Site 2

INTRODUCTION Site 2 was located in Section 32 of Thief Lake Township, just upstream of 400th St. NE (Figure 11). Site 1 was just upstream in Section 29. According to StreamStats, the drainage area is 251 mi2. The subwatershed that drains to site 2 is outlined in pink in Figure 11.

Figure 11. Watershed that drains to site 2.

The Thief River flows through its old riparian zone, but was channelized at some point (Figure 12). Data for this study was collected only one year at this site; therefore cross‐ section overlays and a Pfankuch stability rating were not completed.

Figure 12. Aerial view of site 2. Source: 2013 DNR aerial photo, 3m DEM, and ArcGIS Image Service best available resolution hillshade.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 9

LONGITUDINAL PROFILE The longitudinal profile at site 2 was 1,174 ft. long and covered about 1 meander length (Figure 13). The deepest pool was located at 4+16. The profile started and ended at riffles. There were only 2 bankfull data points, one at each cross‐section. In addition, there were only 3 water surface elevation points and the vertical change between those points ranged from 0.03‐0.14 ft. (0.9‐4.3cm). Using the two bankfull points the slope was near zero. The water surface slope from the survey points was 0.00032. For the velocity and discharge calculations a slope of 0.00022 was used from an estimate using LiDAR data. The slope through site 2 appeared to be more gradual than the water surface upstream, which had a slope around 0.0004. Thief River: Site 2 Longitudinal Profile, 2011

1160

1158 Thalweg

1156 Pool XS 4+16, 2011 4+16, XS Pool Riffle XS 6+99, 2011 6+99, XS Riffle Water Sfc 1154 GM2 Pool XS 4+16 SB 2011 4+16 SB XS GM2 Pool 1152 Bankfull

1150 P1

Elevation (ft) 1148

1146 P2

1144 P3

1142 P4 1140 0 150 300 450 600 750 900 1050 1200 Distance along stream (ft)

Figure 13. Longitudinal profile at site 2.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 10

CROSS‐SECTIONS One riffle cross‐section was completed at site 2. The bankfull area was 207.81 ft2 and the bankfull width was 50.49 ft. (Figure 14). The mean depth was 4.12 ft. and the maximum depth was 6.21 ft.

The entrenchment ratio, comparing flood‐prone width to bankfull width, came out as slightly entrenched. Riffle 6+99 had a flood‐prone width of approximately 383 ft., giving it an entrenchment ratio of 7.59.

Thief River: Site 2 Riffle 6+99, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 50.5 Dbkf = 4.12 Abkf = 207.8 1175 Bankfull Width = 50.49 ft. 1170 Mean Depth = 4.12 ft. Bankfull Area = 207.81 ft2 1165 Flood‐prone Elevation

1160 Bankfull Elevation Elevation (ft) 1155

1150

1145 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 Horizontal Distance (ft)

Figure 14: Riffle cross section 6+99 at site 2.

Grave‐sized particles made up 66.37% of the particles at riffle 6+99 (Figure 15). Clay and silt‐sized particles made up 10.62% of the sample. Sand and cobble particles were also tallied. The D84 at riffle cross‐section 6+99 was 52.78 mm. This value was used to estimate velocity and discharge.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 11

16.00% Active Riffle Particle Analysis 6+99 110.00% 14.00% 90.00% (%) 12.00%

(%) 70.00% 10.00% Volume 8.00% 50.00%

Volume 6.00% 30.00% 4.00% 10.00% 2.00% Cumulative 0.00% ‐10.00% 2 4 8 ‐ ‐ ‐ 16 32 45 64 90 .25 5.7 .50 1.0 ‐ ‐ 2 ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 1.0 4 5.7 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 15: Active riffle particle analysis at riffle cross section 6+99.

One pool cross‐section was completed at site 2 (4+16). The bankfull area was 242.8 ft2 and the bankfull width was 45.49 ft. (Figure 16). The pool area was larger than the riffle area, but the pool bankfull width was slightly narrower than the riffle.

Thief River: Site 2 Pool 4+16, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 45.5 Dbkf = 5.34 Abkf = 242.8 1180 Bankfull Width = 45.49 ft. Mean Depth = 5.34 ft. 1170 Bankfull Area = 242.8 ft2

1160 Bankfull Elevation

1150

Elevation (ft) 1140

1130

1120 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100105110115120 Horizontal Distance (ft)

Figure 16: Pool cross section at 4+16.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 12

DISCHARGE TABLE The estimated velocities and discharges at riffle cross‐section 6+99 were similar using Manning Limerinos n, Darcy‐Weisbach, and U/U* methods (Table 4). Generally the velocities were approximately 1.8 ft/sec and the estimated discharge was 375 cfs. Using LiDAR data to estimate water surface slope, it appeared that the slope through site 2 was slightly lower than the adjacent stream reaches. Table 4. Bankfull velocity and discharge estimates at site 2.

BANKFULL VELOCITY & DISCHARGE ESTIMATES Stream: Thief River Location: Site 2 Observers: Dave Friedl Date: 8/22/2011

Riffle Cross‐section: Riffle 6+99 Bankfull Slope: 0.00022 D84 of Active Bed: 52.78 Dia. (mm) Drainage Area: 251 Mi2

Estimation Methods Bankfull Velocity Bankfull Discharge Manning Limerinos n 1.813 ft / sec 376.76 cfs Darcy‐Weisbach 1.834 ft / sec 381.035 cfs U/U* 1.73 ft / sec 360.42 cfs Stream Type (medium to large) 2.862 ft / sec 594.752 cfs

A slope of 0.00022 was used for these calculations, but slope estimates ranged from 0.00022‐ 0.00046. StreamStats estimated the 1.5, 2, 5, 10, 25 and 50 year recurrence interval discharges at 387, 508, 828, 1040 1340 and 1500 cfs, respectively.

PARTICLE DISTRIBUTION The Web Soil Survey (2015) maps the riparian zone as a Fluvaquents, frequently flooded‐Hapludolls complex (map unit I16F) (Figure 17). The typical profile for this map unit is an A soil horizon from 0 to 16 inches is a fine sandy loam. The 16 to 80 inches of the profile is a Cg horizon with stratified loamy sand to silt loam. A C soil horizon contains glacial or post‐glacial material with little affected by soil forming processes. A lowercase g denotes the presents of gray colors, indicative of permanent or periodic intensive reduction due to saturation. The pebble count along the entire study reach was Figure 17. Soil map at site 2. Source: SSURGO Soils and 2013 DNR aerial photo.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 13

very similar to the site 1 reach count. The full reach particle analysis consisted of 38.03% sand and 56.34% gravel. Silt and clay, as well as some cobble sizes, were also present (Figure 18). At site 2 the D50 was 4.56 mm.

25.00% Reach Particle Analysis ‐ Site 2 110.00%

90.00% (%)

20.00% 70.00%

(%) 15.00%

Volume 50.00% 10.00% 30.00% Volume

5.00% 10.00% Cumulative 0.00% ‐10.00% 8 4 2 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 ‐ ‐ 2 ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 5.7 4 1.0 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 18: Reach particle count at site 2.

RIPARIAN ZONE The land adjacent to the Thief River at site 2 was a mix of tree and shrub canopies and an herbaceous layer (Figure 19). Tree species that were present included 45% Acer negundo (boxelder), 30% Fraxinus pennsylvanica (green ash), 10% Fraxinus nigra (black ash), 10% Ulmus americana (American elm) and 5% Quercus macrocarpa (bur oak). Understory shrubs covered approximately 20% of the riparian zone. Prunus virginiana (Chokecherry) made up 65% of the shrub layer. Other shrubs present included 10% Figure 19: Riparian zone at site 2. Photo taken on 5/3/2010. Cornus sericea (redosier dogwood), 10% Cornus alternifolia (pagoda dogwood), 5% Prunus americana (wild plum), 5% Rhamnus cathartica (buckthorn) and 5% Corylus spp. (hazelnut). The dominant grass and forbs present were 20% Phalaris arundinacea (reed canarygrass), 15% (burdock), 15% Urtica dioica (stinging nettle), 15% Carex rostrata (beaked sedge), 5% Cirsium arvense (Canada thistle) and 5% Galium boreale (northern bedstraw).

Thief River Watershed Fluvial Geomorphology Report  Appendix A 14

CLASSIFICATION Table 5. Stream classification input variables The Rosgen stream type through site 2 was classified as a for site 2. C4c‐ (Table 5). The number 4 denotes the 50th percentile size of particles. A D50 of 4.56 mm was fine gravel. The STREAM CLASSIFICATION ‐ Site 2 lowercase c‐ signifies that the energy slope at this site was less than 0.001. Site 2 was slightly entrenched with an STREAM TYPE C4c‐ entrenchment ratio of 7.59. C stream types usually have a VALLEY TYPE X width‐to‐depth (w/d) ratio of greater than 12.0. Riffle 6+99 was just above this at 12.25. Sinuosity was 1.12. Slope 0.00022

A modified (by stream type) Pfankuch stability rating was not collected at this site; however, DNR staff that collected Entrenchment Ratio 7.59 the survey data on 8/22/2011 stated that site 2 appeared to be in a very stable condition. C4 stream types are very Width/Depth Ratio 12.25 sensitive to disturbance but their recovery response when the disturbance is removed is good. C4 streams have a high sediment supply and are very susceptible to streambank Sinuosity 1.12 erosion. C4 stream types are extremely dependent on vegetation to maintain stability. 52.78 mm Channel Material (D ) 50 (GRAVEL)

DIMENSIONLESS RATIOS The mean, minimum and maximum were all the same because only one riffle and one pool cross‐ section was completed at site 2. The riffle was slightly wider than the pool, but the pool had a larger cross‐sectional area (Table 6). The pool was also deeper than the riffle. At higher flows, the river has access to a wide floodplain with an entrenchment ratio of 7.59.

Table 6. Channel morphology and dimensionless ratios at site 2.

CHANNEL MORPHOLOGY & DIMENSIONLESS RATIOS Stream: Thief River Site: 2 Observers: Dave Friedl Date: 8/22/2011

Riffle Dimensions Mean Min Max Dimensionless Ratios Mean Min Max Riffle Width (W ) 50.49 50.49 50.49 Width of Flood‐Prone Area (W ) 383 383 383 bkf pfa Mean Riffle Depth (dbfk) 4.12 4.12 4.12 Riffle Width/Depth Ratio (Wbkf / dbkf) 12.26 12.26 12.26

Maximum Riffle Depth (dmax) 6.21 6.21 6.21 Max Riffle Depth to Mean Riffle Depth (dmax / dbkf) 1.51 1.51 1.51 2 Riffle Cross‐Sectional Area (Abkf) (ft ) 207.81 207.81 207.81 Entrenchment Ratio (Wfpa / Wbkf) 7.59 7.59 7.59

Pool Dimensions

Pool Width (Wbkfp) 45.49 45.49 45.49 Pool Width to Riffle Width (Wbkfp / Wbkf)0.90.90.9

Mean Pool Depth (dbfkp) 5.34 5.34 5.34 Mean Pool Depth to Mean Riffle Depth (dbkfp / dbkf)1.31.31.3

Maximum Pool Depth (dmaxp) 7.44 7.44 7.44 Pool Area to Riffle Area (Abkfp / Abkf) 1.17 1.17 1.17 2 Pool Cross‐Sectional Area (Abkfp) (ft ) 242.8 242.8 242.8 Max Pool Depth to Mean Riffle Depth (dmaxp / dbkf) 1.81 1.81 1.81

Thief River Watershed Fluvial Geomorphology Report  Appendix A 15

Site 3

INTRODUCTION Site 3 was located in Section 8 of Whiteford Township, just south of Hwy. 6 (Figure 20). According to StreamStats, the drainage area is 257 mi2. The subwatershed that drains to site 3 is outlined in pink in Figure 20.

Figure 20. Watershed that drains to site 3.

At site 3, the Thief River is classified as a perennial drainage ditch. The river channel appears to be flowing in its original meanders, but periodic clean‐outs have created spoil piles adjacent to the channel (Figure 21).

Figure 21. Aerial view of site 3. Source: 2013 DNR aerial photo, 3m DEM, and ArcGIS Image Service best available resolution hillshade.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 16

LONGITUDINAL PROFILE

The first approximately 250 ft. of the longitudinal profile appeared steeper than the rest of the station (Figure 22). The water surface slope from the beginning of the station to the end was 0.00051. Measuring just from the second water surface data point to the last, the slope was 0.00015. Using 1m DEM derived from LiDAR data the water surface slope ranged from 0.00001 to 0.00032. A bankfull slope was not determined at site 3. The riffle and pool cross‐sections were the only locations that a bankfull call was made. Both bankfull calls on the downstream pool cross‐section created unrealistic slopes. The slope used for the velocity and discharge calculations was 0.00011.

Thief River: Site 3 Longitudinal Profile 2011

1150

Thalweg

1148 Water Sfc GM3 Pool XS 3+12, 2011 XS Pool GM3 GM3 Riffle XS 1+48, 2011 XS Riffle GM3

1146 2011 3+12 SB, XS Pool GM3 Bankfull

Bankfull 2

Elevation (ft) 1144 Water Surface Slope = 0.00015 GM3 Pool XS 3+12 SB, 2011_Adjusted Elev 2011_Adjusted 3+12 SB, XS Pool GM3 P2

1142 P3

P4 1140 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 Distance along stream (ft)

Figure 22. Longitudinal profile at site 3.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 17

CROSS‐SECTIONS One riffle cross‐section was completed at site 3 (Figure 23). The bankfull area was 204.12 ft2 and the bankfull width was 48.09 ft. The mean depth was 4.24 ft. and the maximum depth was 5.46 ft. The entrenchment ratio, comparing flood‐prone width to bankfull width, was 74.9.

Thief River: Site 3 Riffle 1+48, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 51.7 Dbkf = 3.95 Abkf = 204.3 1160 Bankfull Width = 48.09 ft. 1158 Mean Depth = 4.24 ft. 2 1156 Flood‐prone Elevation Bankfull Area = 204.12 ft 1154

1152

1150 Elevation (ft) Bankfull Elevation 1148

1146

1144

1142

1140 0 1020304050607080 Horizontal Distance (ft)

Figure 23. Active riffle particle analysis from riffle cross section 1+48.

The 84th percentile particle size (D84) at riffle 1+48 was 7.97 mm (fine gravel) (Figure 24). Fine particles were tallied at this location, with silt and clay particles making up 27.78% of the sample. Sand and gravel‐sized particles made up 22.22% and 50.0% of the particle analysis, respectively.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 18

30.00% Active Riffle Particle Analysis 1+48 110.00%

25.00% 90.00% (%)

20.00% 70.00% Volume 15.00% 50.00%

(%) 10.00% 30.00%

5.00% 10.00% Cumulative

Volume 0.00% ‐10.00% 4 2 8 ‐ ‐ ‐ 16 32 45 64 90 1.0 5.7 .50 .25 2 ‐ ‐ ‐ ‐ ‐ 180 256 362 512 128 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 1.0 4 5.7 ‐ <.062 32 45 64 ‐ 8 .50 .25 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm) Figure 24. Active riffle particles analysis at riffle cross section 1+48.

One pool cross‐section was completed at site 3 (3+12). The bankfull area was 226.74 ft2 and the bankfull width was 43.31 ft. (Figure 25). The mean depth was 5.24 ft. and the maximum depth was 7.3 ft. Thief River: Site 3 Pool 3+12, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 43.3 Dbkf = 5.24 Abkf = 226.7 1160

1158 Bankfull Width = 43.31 ft. 1156 Mean Depth = 5.24 ft. 2 1154 Bankfull Area = 226.74 ft 1152 1150 Bankfull Elevation 1148 Elevation (ft) 1146 1144 1142 1140 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 Horizontal Distance (ft)

Figure 25. Pool section 3+12.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 19

The study bank overlay at pool cross‐section 3+12 indicated an actual erosion rate of .0888 ft/yr from 2011 to 2012 (Figure 26). This rate closely matched the predicted erosion rate of .06 ft/yr derived from the BANCS model.

Thief River: Site 3 Pool Study Bank Overlay 2011-2012 Site 3 Pool 3+12 Study Bank, 2011 Site 3 Pool 3+12 Study Bank, 2012

1150

1149

1148

1147

1146 Elevation (ft) 1145

1144

Toe Pin 1143

1142

1141

1140 48 49 50 51 52 53 54 55 56 57 58 59 60

Horizontal Distance (ft)

Figure 26. Site 3 pool cross‐section 3+12 study bank overlay showing the study bank profile in 2011 and 2012.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 20

DISCHARGE TABLE The estimated velocities and discharges at riffle cross‐section 1+48 were similar using Manning Limerinos n, Darcy‐Weisbach, and U/U* methods (Table 7). Generally the velocities were approximately 1.8 ft/sec and the estimated discharge was 378 cfs. A slope of 0.00011 was measured using a 1m DEM derived from LiDAR data. The water surface slope from the longitudinal profile was 0.00015 and a bankfull slope was not determined. StreamStats estimated the 1.5, 2, 5, 10, 25 and 50 recurrence interval discharges at 257, 509, 821, 1030, 1280 and 1470 cfs, respectively.

Table 7. Bankfull velocity and discharge estimates for site 3.

BANKFULL VELOCITY & DISCHARGE ESTIMATES Stream: Thief River Location: Site 3 Observers: Dave Friedl Date: 8/23/2011

Riffle Cross‐section: Riffle 1+48 Bankfull Slope: 0.00011 D84 of Active Bed: 7.97 Dia. (mm) Drainage Area: 257 Mi2

Estimation Methods Bankfull Velocity Bankfull Discharge Manning Limerinos n 1.86 ft / sec 379.663 cfs Darcy‐Weisbach 1.9 ft / sec 388.204 cfs U/U* 1.81 ft / sec 368.75 cfs Stream Type (medium to large ‐ C4) 2.065 ft / sec 421.508 cfs

PARTICLE DISTRIBUTION The Web Soil Survey (2015) maps the riparian zone as a Fluvaquents, frequently flooded‐Hapludolls complex ‐map unit I16F (Figure 27). The typical profile for this map unit is an A soil horizon from 0 to 16 inches is a fine sandy loam. The 16 to 80 inches of the profile is a Cg horizon with stratified loamy sand to silt loam. A C soil horizon contains glacial or post‐glacial material with little affected by soil forming processes. A lowercase g denotes the presents of gray colors, indicative of permanent or periodic intensive reduction due to saturation. The full reach particle analysis was Figure 27. Soil map at site 3. Source: SSURGO Soils and dominated by very fine gravel‐sized 2013 DNR aerial photo.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 21

particles (18.06%). The largest particle size tallied was 32 mm (Figure 28). The D50 (50th percentile) particle size was a fine gravel (2.75 mm).

20.00% Reach Particle Analysis ‐ Site 3 110.00% 18.00% (%) 16.00% 90.00% 14.00% 70.00%

(%) 12.00%

Volume 10.00% 50.00% 8.00% 30.00%

Volume 6.00% 4.00% 10.00% 2.00% Cumulative 0.00% ‐10.00% 4 2 8 ‐ ‐ ‐ 16 32 45 64 90 .25 5.7 .50 1.0 2 ‐ ‐ ‐ ‐ ‐ 180 256 362 512 128 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 1.0 4 5.7 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 28. Particle analys is for site 3 study reach.

RIPARIAN ZONE The land adjacent to Thief River at site 3 was a mix of tree and shrub canopies with an herbaceous layer (Figure 29). The tree canopy included Acer negundo (boxelder), Fraxinus pennsylvanica (green ash), Ulmus americana (American elm) and Quercus spp. (oak). Understory shrubs included Cornus sericea (redosier dogwood), Prunus virginiana (chokecherry), Rosa spp. (wild rose), Lonicera spp. (honeysuckle) and Viburnum trilobum (highbush cranberry).

Figure 29: Site 3 riparian cover. Photo taken on 8/23/2011. The dominant grass and forbs present were Phalaris arundinacea (reed canarygrass), Cirsium arvense (Canada thistle), Solidago spp. (goldenrod), Asclepias spp. (milkweed) and Echinacea spp. (coneflower).

Thief River Watershed Fluvial Geomorphology Report  Appendix A 22

CLASSIFICATION The Rosgen stream type through site 3 was classified as a Table 8. Stream classification input variables for site 3. C4c‐/E4 (Table 8). The number 4 denotes the 50th percentile size of particles. A D50 of 2.75mm is very fine STREAM CLASSIFICATION ‐ Site 3 gravel. The lowercase c‐ signifies that the energy slope at this site is less than 0.001. The entrenchment ratio was STREAM TYPE C4c‐/E4 74.9 and the width to depth (w/d) ratio was 11.34. An E VALLEY TYPE VIII(b) stream type has a w/d ratio less than 12.0 and a C is more than 12.0. The sinuosity was 1.38. Slope 0.00011

The modified (by stream type) Pfankuch stability rating Entrenchment Ratio 74.86 on 10/24/2012 was an 89. A good, or stable, rating range for a C4 stream type is 70‐90, and a fair rating is 91‐110. A good, or stable, rating range for an E4 is 50‐75, Width/Depth Ratio 11.34 and a fair rating is 76‐96. C4 and E4 stream types are very sensitive to disturbance but their recovery response when the disturbance is removed is good. C4 streams Sinuosity 1.38 have a high sediment supply and are very susceptible to streambank erosion. E4 streams have a moderate 2.75 mm Channel Material (D ) sediment supply. Both C4 and E4 stream types are 50 (GRAVEL) extremely dependent on vegetation to maintain stability.

DIMENSIONLESS RATIOS The mean, minimum and maximum dimensions are all the same because only 1 riffle and 1 pool cross‐section was completed at site 3 (Table 9). The riffle was slightly wider than the pool, but the pool had a larger cross‐sectional area. The pool was also deeper than the riffle. At higher flows, the river had access to a wide floodplain with an entrenchment ratio of 74.86. Table 9. Channel morphology and dimensionless ratios for site 3.

CHANNEL MORPHOLOGY & DIMENSIONLESS RATIOS Stream: Thief River Site: 3 Observers: Dave Friedl Date: 8/23/2011

Riffle Dimensions Mean Min Max Dimensionless Ratios Mean Min Max

Riffle Width (Wbkf) 48.09 48.09 48.09 Width of Flood‐Prone Area (Wpfa) 3600 3600 3600

Mean Riffle Depth (dbfk) 4.24 4.24 4.24 Riffle Width/Depth Ratio (Wbkf / dbkf) 11.34 11.34 11.34

Maximum Riffle Depth (dmax) 5.46 5.46 5.46 Max Riffle Depth to Mean Riffle Depth (dmax / dbkf) 1.29 1.29 1.29 2 Riffle Cross‐Sectional Area (Abkf) (ft ) 204.12 204.12 204.12 Entrenchment Ratio (Wfpa / Wbkf) 74.86 74.86 74.86

Pool Dimensions

Pool Width (Wbkfp) 43.31 43.31 43.31 Pool Width to Riffle Width (Wbkfp / Wbkf)0.90.90.9

Mean Pool Depth (dbfkp) 5.24 5.24 5.24 Mean Pool Depth to Mean Riffle Depth (dbkfp / dbkf) 1.24 1.24 1.24

Maximum Pool Depth (dmaxp) 7.3 7.3 7.3 Pool Area to Riffle Area (Abkfp / Abkf) 1.11 1.11 1.11 2 Pool Cross‐Sectional Area (Abkfp) (ft ) 226.74 226.74 226.74 Max Pool Depth to Mean Riffle Depth (dmaxp / dbkf) 1.72 1.72 1.72

Thief River Watershed Fluvial Geomorphology Report  Appendix A 23

Site 4

INTRODUCTION Site 4 was located on the Mud River, approximately 3 mi. NW of Grygla, MN,in section 20 of Valley Township (Figure 30). According to StreamStats, the drainage area is 153 mi2. The subwatershed that drains to site 4 is outlined pink in figure 30.

Figure 30. Watershed that drains to site 4.

This section of the Mud River was straightened prior to July 1939 (Figure 31). The old meander scroll can be seen in Figure 32 with aerial imagery and hillshade.

Figure 31. Aerial photo dated 7/10/1939. Figure 32. Aerial view of site 4. Source: 2013 DNR aerial photo, 3m DEM, and ArcGIS Image Service best available resolution hillshade.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 24

LONGITUDINAL PROFILE The longitudinal profile at site 4 was 1800 feet, or 41 bankfull channel widths, in length (Figure 33). The majority of the longitudinal profile was surveyed through a long, straight reach, with only two small bends included at the beginning and end of the survey. The water surface and bankfull slopes from the survey points were 0.00016 and 0.0002, respectively. As a comparison, water surface slopes were obtained using LiDAR and ranged from 0.00024 to 0.00029. A slope of 0.00024 was used for the velocity and discharge calculations.

Mud River: Site 4 Longitudinal Profile 2011

1165

1164

1163 Thalweg

1162 Bankfull Slope = 0.0002

1161 Water Surface GM4 Riffle XS 0+00, 2011 GM4 Pool XS 16+04, 2011

1160 GM4 Riffle XS 15+11, 2011

1159 Bankfull

1158 Water Surface Slope = 0.00016

1157 P1

Elevation (ft) 1156

1155 P2 1154

1153 P3 1152

1151 P4 1150 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 Distance along stream (ft)

Figure 33. Longitudinal profile at site 4 in 2011.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 25

CROSS‐SECTIONS Two riffle cross‐sections were completed at site 4. Riffle 0+00 was used for the velocity and discharge calculations (Figure 34). The bankfull area was 171.38 ft2 and the bankfull width was 44.01 ft. The second riffle was located 1,511 ft. from the start of the longitudinal profile. The bankfull area at riffle 15+11 was 146.91 ft2 and the bankfull width was 42.32 ft.

The width‐to‐depth ratios for 0+00 and 15+11 were 11.31 and 12.2, respectively. Both riffle cross‐ sections had access to their adjacent floodplains. The entrenchment ratio for 0+00 and 15+11 were 13.18 and 15.48, respectively.

Mud River: Site 4 Riffle 0+00, 2011 Ground Points Bankfull Indicators Water Surface Points Wbkf = 44 Dbkf = 3.89 Abkf = 171.4 1180 1177 Bankfull Width = 44.01 ft. 1174 Mean Depth = 3.89 ft. 2 1171 Bankfull Area = 171.38 ft 1168 Flood‐prone Elevation 1165 REP 1162 Bankfull Elevation

Elevation (ft) 1159 1156 1153 1150 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 Horizontal Distance (ft) Figure 34. Riffle cross section 0+00 at site 4.

The riffle cross‐section 0+00 was dominated by gravel‐sized particles. Gravel made up 68.63% of the total particle analysis (Figure 35). The 50th percentile particle size (D50) at the active riffle was 8 mm. Silt and clay made up 5.88% of the particles measured at this cross‐section. The D84 used for the velocity and discharge calculations was 29.6 mm.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 26

16.00% Active Riffle Particle Analysis 0+00 110.00%

14.00% 90.00% (%)

12.00% 70.00% 10.00% (%)

Volume 8.00% 50.00%

Volume 6.00% 30.00% 4.00%

10.00% Cumulative 2.00%

0.00% ‐10.00% 4 8 2 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 2 ‐ ‐ ‐ ‐ ‐ 180 256 362 512 128 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 5.7 4 1.0 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 35. Active riffle particle analysis at cross section 0+00.

One pool cross‐section (16+04) was completed at site 4. The bankfull area was 143.93 ft2 and the bankfull width was 37.38 ft. (Figure 36). The pool dimensions were smaller than both riffle cross‐ sections.

Mud River: Site 4 Pool 16+04, 2011 Ground Points Bankfull Indicators Water Surface Points Wbkf = 37.4 Dbkf = 3.85 Abkf = 143.9 1170 Bankfull Width = 37.38 ft. 1168 Mean Depth = 3.85 ft. 1166 Bankfull Area = 143.93 ft2 REP 1164

1162LEP Bankfull Elevation 1160 1158 TP Elevation (ft) 1156 1154 1152

-5 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 Horizontal Distance (ft)

Figure 36. 2011 Pool cross section 16+04 at site 4.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 27

The study bank overlay at pool cross‐section 16+04 indicated an actual erosion rate of .0837 ft/yr from 2011 to 2012 (Figure 37). This rate closely matched the predicted erosion rate of .036 ft/yr derived from the BANCS model.

Mud River: Site 4 Pool Study Bank Overlay 2011-2012 Site 4 Pool 16+04 Study Bank, 2011 Site 4 Pool 16+04 Study Bank, 2012

1163

1162

1161

1160 Elevation (ft) 1159

1158

1157 Toe Pin

1156

1155 62 63 64 65 66 67 68 69 70 71 72 73 74 75

Horizontal Distance (ft)

Figure 37. Site 4 pool cross‐section 16+04 study bank overlay showing the study bank profile in 2011 and 2012.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 28

DISCHARGE TABLE The estimated velocities and discharges at riffle cross‐section 0+00 were very similar between different methods (Table 10). Generally the velocities were approximately 2.0 ft/sec and the estimated discharge was 340 cfs. StreamStats estimated the 1.5, 2, 5, 10, 25 and 50 recurrence interval discharges at 361, 519, 983, 1320, 1760 and 2090 cfs, respectively.

Table 10. Velocity and discharge estimates for riffle cross‐section 0+00.

BANKFULL VELOCITY & DISCHARGE ESTIMATES Stream: Thief River Location: Site 4 Observers: Dave Friedl Date: 9/8/2011

Riffle Cross‐section: Riffle 0+00 Bankfull Slope: 0.00024 D84 of Active Bed: 29.6 Dia. (mm) Drainage Area: 153 Mi2

Estimation Methods Bankfull Velocity Bankfull Discharge Manning Limerinos n 2.029 ft / sec 347.73 cfs Darcy‐Weisbach 2.086 ft / sec 357.482 cfs U/U* 1.95 ft / sec 333.88 cfs Stream Type (E5 with vegetatve cover) 0.999 ft / sec 171.209 cfs

PARTICLE DISTRIBUTION The Web Soil Survey (2015) maps the riparian zone as a Fluvaquents, frequently flooded‐Hapludolls complex (Figure 38). The typical soil profile is:

A ‐ 0 to 16 in.: fine sandy loam

Cg ‐ 16 to 80 in.: stratified loamy sand to silt loam

A C soil horizon contains glacial or post‐glacial material with little affected by soil forming processes. A lowercase g denotes the presents of gray colors, indicative of permanent Figure 38. 2013 FSA aerial photo and SSURGO soils. or periodic intensive reduction due to saturation.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 29

Sand‐sized particles made up 81.31% of the full reach pebble count (Figure 39). Coarse sand (0.5‐ 1.0mm) was the most tallied size category. The largest particle size tallied was 64 mm. The D50 was 0.46 mm, which is medium‐sized sand.

35.00% Reach Particle Analysis ‐ Site 4 110.00%

30.00% 90.00% (%) 25.00% 70.00% (%) 20.00% Volume 50.00% 15.00%

Volume 30.00% 10.00%

5.00% 10.00% Cumulative

0.00% ‐10.00% 4 8 2 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 2 ‐ ‐ ‐ ‐ ‐ 180 256 362 512 128 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 5.7 4 1.0 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 39. Particle analysis for site 4.

RIPARIAN ZONE The riparian zone at site 4 had trees, shrubs, grass and forbs (Figure 40). The herbaceous layer (grass and forbs) covered 50% of the ground cover. The dominate species included Bromus inermis (smooth brome), Phalaris arundinacea (reed canarygrass), Urtica dioica (stinging nettle), Toxicodendron radicans (poison ivy) and aster.

Cornus sericea (redosier dogwood) was the only documented shrub species and covered 15% of the ground cover. The tree canopy covered approximately 25% of Figure 40. Site 4 vegetation cover. Photo taken on 5/4/2010. the aerial cover and 20% of the basal cover. Tree species included oak, ash and ironwood.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 30

CLASSIFICATION Table 11. Stream classification input factors The Rosgen stream type through site 4 was classified as a at site 4. borderline C5c‐/E5 (Table 11). The number five denotes STREAM CLASSIFICATION ‐ Site 4 the 50th percentile size of particles. With a D50 of 0.46 mm, it was considered a sand bed stream. The lowercase c STREAM TYPE C5c‐/E5 signifies that the energy slope at this site was less than 0.02. VALLEY TYPE X The entrenchment ratio of both riffle cross‐sections fell within the slightly entrenched category. The w/d ratio at Slope 0.00024 0+00 was 11.31 and at 15+11 it was 12.39. A w/d ratio of greater than 12 is a C stream type and a ratio less than 12 is an E stream type. Entrenchment Ratio 13.18

The modified (by stream type) Pfankuch stability rating on Width/Depth Ratio 11.31 11/01/2012 was a 91. This is a fair (moderately stable) stability rating. C4 and E4 stream types are very sensitive to disturbance but their recovery response when the Sinuosity 1.28 disturbance is removed is good. C4 streams have a high sediment supply and are very susceptible to streambank erosion. E4 streams have a moderate sediment supply. 0.46 mm Channel Material (D ) 50 (SAND) Both C4 and E4 stream types are extremely dependent on vegetation to maintain stability.

DIMENSIONLESS RATIOS The upstream and downstream riffle had similar morphology (Table 12). The upstream riffle (0+00) was slightly larger than the downstream riffle. Only one pool cross‐section was completed at this site. The mean pool depth to‐mean riffle depth ratio was 1.05, so the riffles were slightly shallower than the pools; however, the pool did have a smaller cross‐sectional area than both riffles. The pool was also narrower than the riffles.

Table 12. Channel morphology and dimensionless ratios for site 4.

CHANNEL MORPHOLOGY & DIMENSIONLESS RATIOS Stream: Thief River Site: 4 Observers: Dave Friedl Date: 9/8/2011

Riffle Dimensions Mean Min Max Dimensionless Ratios Mean Min Max

Riffle Width (Wbkf) 43.17 42.32 44.01 Width of Flood‐Prone Area (Wpfa) 617.5 580 655

Mean Riffle Depth (dbfk) 3.68 3.47 3.89 Riffle Width/Depth Ratio (Wbkf / dbkf) 11.76 11.31 12.2

Maximum Riffle Depth (dmax) 5.08 5.03 5.12 Max Riffle Depth to Mean Riffle Depth (dmax / dbkf) 1.39 1.32 1.45 2 Riffle Cross‐Sectional Area (Abkf) (ft ) 159.15 146.91 171.38 Entrenchment Ratio (Wfpa / Wbkf) 14.33 13.18 15.48

Pool Dimensions

Pool Width (Wbkfp) 37.38 37.38 37.38 Pool Width to Riffle Width (Wbkfp / Wbkf) 0.866 0.866 0.866

Mean Pool Depth (dbfkp) 3.85 3.85 3.85 Mean Pool Depth to Mean Riffle Depth (dbkfp / dbkf) 1.05 1.05 1.05

Maximum Pool Depth (dmaxp) 5.55 5.55 5.55 Pool Area to Riffle Area (Abkfp / Abkf) 0.9 0.9 0.9 2 Pool Cross‐Sectional Area (Abkfp) (ft ) 143.9 143.9 143.9 Max Pool Depth to Mean Riffle Depth (dmaxp / dbkf) 1.51 1.51 1.51

Thief River Watershed Fluvial Geomorphology Report  Appendix A 31

Site 5

INTRODUCTION Site 5 was located on the Mud River, just downstream of Grygla, MN, in section 27 of Valley Township (Figure 41). According to StreamStats, the drainage area is 139 mi2. The subwatershed that drains to site 5 is outlined pink in Figure 41.

Figure 41. Watershed that drains to site 5.

This area of the Mud River appears to be following the same meander pattern since July 1939 (Figure 42). The land to the north of the channel changed from a forested land use to pasture/hay land use prior to 1954 (Figure 43).

Figure 42. Historical photo date 7/10/1939. Figure 43. Aerial view of site 5. Source: 2013 DNR aerial photo, 3m DEM, and ArcGIS Image Service best available resolution hillshade.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 32

LONGITUDINAL PROFILE The longitudinal profile at site 5 was approximately 1375 feet, or 36 bankfull channel widths, in length (Figure 44). The thalweg, or channel bottom, through site 5 appeared to fluctuate between riffles and pools as expected, though the variability in depths at different facets (e.g., riffles, pools, glides) was minimal. The bankfull slope was 0.0003 and the water surface slope was 0.00002. As a comparison, water surface slopes were obtained from LiDAR and ranged from 0.00019 to 0.00048. A slope of 0.0003 was used for the velocity and discharge calculations.

Mud River: Site 5 Longitudinal Profile 2011

1175

1173 Thalweg

1171

GM5 Pool 5+38, XS 2011 Water Surface GM5 Riffle XS 3+46, 2011 1169 Bankfull Slope = 0.0003 GM5 Pool XS 5+38 SB, 2011

1167 Bankfull

1165 Water Surface Slope = 0.00002 P1 Elevation (ft) Elevation 1163

P2 1161

1159 P3

1157 P4 1155 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 1000 1050 1100 1150 1200 1250 1300 1350 1400

Distance along stream (ft)

Figure 44. Longitudinal profile at site 5 in 2011.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 33

CROSS‐SECTIONS One riffle cross‐section was completed at site 5 (Figure 45). Riffle 3+46 was used for the velocity and discharge calculations. The bankfull area was 152 ft2 and the bankfull width was 37.78 ft. The entrenchment ratio, comparing flood‐prone width to bankfull width, came out as slightly entrenched at 20.7.

Mud River: Site 5 Riffle 3+46, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 37.8 Dbkf = 4.02 Abkf = 152 1180 Bankfull Width = 37.78 ft. 1179 1178 Mean Depth = 4.02 ft. 1177 1176 Bankfull Area = 152 ft2 1175 Flood‐prone Elevation 1174 1173 1172LEP 1171 1170 1169 Bankfull Elevation

Elevation (ft) 1168 1167 1166 1165 1164 1163 1162 1161 1160 1159 1158 1157 1156

-15-10 -5 0 5 10152025303540455055606570 Horizontal Distance (ft)

Figure 45. Riffle cross section 3+46 at site 5.

The D50, or 50th percentile, particle size was 4.71 mm. Very fine‐sized particles made up 37.61% of the particles measured at this cross section (Figure 46). Gravel‐sized particles made up 55.05% of the particles measured. Silt, clay, sand, and cobble were also tallied. The D84 used for the velocity and discharge calculations was 26 mm.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 34

Active Riffle Particle Analysis 3+46 40.00% 110.00% 35.00%

90.00% (%) 30.00% (%)

70.00% 25.00% Volume 20.00% 50.00%

Volume 15.00% 30.00% 10.00% 10.00%

5.00% Cumulative 0.00% ‐10.00% 4 2 8 ‐ ‐ ‐ 16 32 45 64 90 .25 .50 1.0 5.7 2 ‐ ‐ ‐ ‐ ‐ 180 256 362 512 128 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 1.0 4 5.7 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm) Figure 46. Active riffle particle analysis at cross section 3+46.

One pool cross‐section, 5+38, was completed at site 5. The bankfull area was 174.17 ft2 and the bankfull width was 38.45 ft. (Figure 47). Compared to riffle cross‐section 3+46, the pool was slightly wider and had a greater cross‐sectional area.

Mud River: Site 5 Pool 5+38, 2011 Ground Points Bankfull Indicators Water Surface Points Wbkf = 38.4 Dbkf = 4.53 Abkf = 174.2 1175 Bankfull Width = 38.45 ft. 1173 Mean Depth = 4.53 ft. EP REP 1171 Bankfull Area = 174.17 ft2 1169 Bankfull Elevation 1167 1165 TP 1163 Elevation (ft) 1161 1159 1157 1155 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 Horizontal Distance (ft)

Figure 47. 2011 Pool cross section 5+38 at site 5.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 35

The study bank overlay at pool cross‐section 5+38 indicated an actual erosion rate of .3617 ft/yr from 2011 to 2012 (Figure 48). This rate closely matched the predicted erosion rate of .25 ft/yr derived from the BANCS model.

Mud River: Site 5 Pool Study Bank Overlay 2011-2012 Site 5 Pool 5+38 Study Bank, 2011 Site 5 Pool 5+38 Study Bank, 2012

1170

1169

1168

1167

1166

Elevation (ft) 1165

1164

1163 Toe Pin

1162

1161

1160 5 6 7 8 9 101112131415161718192021222324252627282930 Horizontal Distance (ft)

Figure 48. Site 5 pool cross‐section 5+38 study bank overlay showing the study bank profile in 2011 and 2012.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 36

DISCHARGE TABLE The estimated velocities and discharges at riffle cross‐section 3+46 were very similar between different methods, except for the Manning n by stream type (Table 13). Generally the velocities were approximately 2.4 ft/sec and the estimated discharge was 360 cfs. StreamStats estimated the 1.5, 2, 5, 10, 25 and 50 recurrence interval discharges at 336, 482, 912, 1220, 1630 and 1930 cfs, respectively.

Table 13. Velocity and discharge estimates for riffle cross‐section 3+46.

BANKFULL VELOCITY & DISCHARGE ESTIMATES Stream: Thief River Location: Site 5 Observers: Dave Friedl Date: 8/19/2011

Riffle Cross‐section: Riffle 3+46 Bankfull Slope: 0.0003 D84 of Active Bed: 23 Dia. (mm) Drainage Area: 139 Mi2

Estimation Methods Bankfull Velocity Bankfull Discharge Manning Limerinos n 2.412 ft / sec 366.648 cfs Darcy‐Weisbach 2.511 ft / sec 381.693 cfs U/U* 2.32 ft / sec 352.91 cfs Stream Type (small with vegetative influence) 0.872 ft / sec 132.553 cfs

PARTICLE DISTRIBUTION The Web Soil Survey (2015) maps the riparian zone as a Fluvaquents, frequently flooded‐Hapludolls complex (Figure 49). The typical profile for this map unit is an A soil horizon from 0 to 16 inches is a fine sandy loam. The 16 to 80 inches of the profile is a Cg horizon with stratified loamy sand to silt loam. A C soil horizon contains glacial or post‐ glacial material with little affected by soil forming processes. A lowercase g denotes the presents of gray colors, indicative of permanent or periodic intensive reduction due to saturation. Figure 49. Soil map of site 5. Source: 2013 DNR aerial photo and SSURGO soils. Very fine sand (0.062‐0.125mm) made up 23.39% of the full study reach pebble count (Figure 50). Combined, all of the various sand‐sized particles were tallied 52.42% of the time. Gravel‐sized particles made up 38.71% of the sample. The D50 at site 5 was 0.85mm, which is coarse sand.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 37

Reach Particle Analysis ‐ Site 5 25.00% 110.00% (%) 20.00% 90.00% 70.00% (%)

15.00% Volume

50.00% 10.00%

Volume 30.00%

5.00% 10.00% Cumulative

0.00% ‐10.00% 8 2 4 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 1.0 .50 ‐ ‐ 2 ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 4 5.7 1.0 ‐ <.062 32 45 64 ‐ 8 .50 .25 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 50. Particle analysis for site 5. RIPARIAN ZONE The riparian vegetation at site 5 was forested with an understory of a few shrubs, as well as grasses and forbs. The tree canopy covered approximately 50% of the aerial cover and 75% of the basal cover. Tree species included 40% Salix spp. (willow), 35% Fraxinus pennsylvanica (green ash), 15% Ulmus americana (American elm), 4% Tilia americana (basswood) and a few box elder, cottonwood, ironwood, aspen, white oak, and birch. Understory shrubs covered approximately 5% of the riparian zone and included Cornus sericea (redosier dogwood), Rhamnus cathartica (buckthorn), Prunus pensylvanica (pin cherry), and Rosa spp. (rose).

The dominant herbaceous species present were 40% Bromus inermis (smooth brome), 20% Solidago spp. (goldenrod), 15% Phalaris arundinacea (reed canarygrass), 5% Cirsium arvense (Canada thistle), 1% Toxicodendron radicans (poison ivy) and 19% other species.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 38

CLASSIFICATION Table 14. Stream classification input The Rosgen stream type through site 5 was classified as a factors at site 5. borderline B5c / E5 (Table 14). The number five denotes the 50% percentile size of particles. With a D50 of 0.85 mm, site 5 STREAM CLASSIFICATION ‐ Site 5 is considered a sand stream. The entrenchment ratio was 2.12, STREAM TYPE B5c/E5 which was moderately entrenched. The flood‐prone width was VALLEY TYPE X measured on LiDAR elevation to be 80 ft. If the floodprone width was a couple more feet, it would be considered an E Slope 0.0003 stream type. Sinuosity was 1.33.

Entrenchment Ratio 2.12 The modified (by stream type) Pfankuch stability rating on 10/11/2012 was 101. This is considered poor, or unstable for both B5c and E5 stream types. E stream types are low Width/Depth Ratio 9.4 gradient, meandering riffle/pool streams with a low width‐to‐ depth ratio. E5 stream types are very highly sensitive to Sinuosity 1.33 disturbance, but have a good recovery potential when the disturbance is removed. This stream type is very highly dependent on robust stream bank vegetation to remain stable 0.85 mm Channel Material (D ) 50 (SAND) and the potential for streambank erosion is high.

DIMENSIONLESS RATIOS The mean, minimum and maximum measurements of the riffle and pool did not vary because only one of each type was collected. The width of the riffle compared to the pool was approximately the same (Table 15). The pool had a slightly larger cross‐sectional area and was slightly deeper.

Table 15. Channel morphology and dimensionless ratios for site 5.

CHANNEL MORPHOLOGY & DIMENSIONLESS RATIOS Stream: Thief River Site: 5 Observers: Dave Friedl Date: 8/10/2011 Riffle Dimensions Mean Min Max Dimensionless Ratios Mean Min Max Riffle Width (Wbkf) 37.8 37.8 37.8 Width of Flood‐Prone Area (Wpfa)808080

Mean Riffle Depth (dbfk) 4.02 4.02 4.02 Riffle Width/Depth Ratio (Wbkf / dbkf) 9.4 9.4 9.4

Maximum Riffle Depth (dmax) 5.77 5.77 5.77 Max Riffle Depth to Mean Riffle Depth (dmax / dbkf) 1.44 1.44 1.44 2 Riffle Cross‐Sectional Area (Abkf) (ft ) 152 152 152 Entrenchment Ratio (Wfpa / Wbkf) 2.12 2.12 2.12

Pool Dimensions

Pool Width (Wbkfp) 38.5 38.5 38.5 Pool Width to Riffle Width (Wbkfp / Wbkf) 1.02 1.02 1.02

Mean Pool Depth (dbfkp) 4.53 4.53 4.53 Mean Pool Depth to Mean Riffle Depth (dbkfp / dbkf) 1.13 1.13 1.13

Maximum Pool Depth (dmaxp) 6.49 6.49 6.49 Pool Area to Riffle Area (Abkfp / Abkf) 1.15 1.15 1.15 2 Pool Cross‐Sectional Area (Abkfp) (ft ) 174.2 174.2 174.2 Max Pool Depth to Mean Riffle Depth (dmaxp / dbkf) 1.61 1.61 1.61

Thief River Watershed Fluvial Geomorphology Report  Appendix A 39

Site 6

INTRODUCTION Site 6 was located on the Mud River, approximately 3 mi. east of Grygla, MN, in section 29 of Benville Township (Figure 51). According to StreamStats, the drainage area is 94.4 mi2. The subwatershed that drains to site 6 is outlined pink in Figure 51.

Figure 51. Watershed that drains to site 6.

The Mud River flows in the same channel as it did in 1940 (Figure 52). It is a perennial drainage ditch at this site, as well as designated a public water. The longitudinal profile and cross‐sections were all located adjacent to the short road between Flintlock Rd. and Hwy 89 (Figure 53).

Figure 53. Aerial view of site 1. Source: 2013 Figure 52. Historical aerial photo dated 9/11/1940. DNR aerial photo, 3m DEM, and ArcGIS Image Service best available resolution hillshade.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 40

LONGITUDINAL PROFILE The longitudinal profile at site 6 was 540 feet, or 22 bankfull widths, in length (Figure 54). Because of its location in a channelized reach, this site lacked the typical and easily discernable riffle‐pool sequences found in meandering streams. Four survey points were used to determine both the water surface and bankfull slopes, which were 0.00062 and 0.00068, respectively. Water surface slopes of 0.00061 and 0.00067 obtained using LiDAR were in close agreement with these surveyed slopes. A slope of .00062 was used for the velocity and discharge calculations.

Mud River: Site 6 Longitudinal Profile 2011

1185

1184 Thalweg 1183

1182 GM6 Pool XS 3+43, 2011 3+43, XS GM6 Pool 2011 3+71, XS GM6 Pool GM6 Riffle XS 5+07, 2011 5+07, XS GM6 Riffle GM6 Riffle XS 0+93, 2011 0+93, XS GM6 Riffle Bankfull Slope = 0.00068 1181 Water Surface GM6 Pool XS 3+71 SB 2011 SB 3+71 XS GM6 Pool

1180

1179 Bankfull Water Surface Slope = 0.00062 1178

1177 P1 Elevation (ft) 1176

1175 P2

1174

1173 P3

1172

1171 P4 1170 0 25 50 75 100 125 150 175 200 225 250 275 300 325 350 375 400 425 450 475 500 525 550

Distance along stream (ft)

Figure 54. Longitudinal profile at site 6 in 2011.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 41

CROSS‐SECTIONS Two riffle cross‐sections were completed at site 6. Riffle 0+93 was used for the velocity and discharge calculations. The bankfull area at this cross‐section was 71.66 ft2 and the bankfull width was 25.24 ft (Figure 55). The second riffle was located 507 ft. downstream from the start of the longitudinal profile. The bankfull area at this riffle was 72.17 ft2 and the bankfull width was 26.13 ft.

The entrenchment ratio, comparing flood‐prone width to bankfull width, came out as slightly entrenched at the upstream riffle and moderately entrenched at the downstream riffle. Riffle 0+93 had a flood‐prone width of 58 ft., giving it an entrenchment ratio of 2.3. Riffle 5+07 had a flood‐ prone width of 48.75 ft., giving it a ratio of 1.87.

Mud River: Site 6 Riffle 0+93, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 25.2 Dbkf = 2.84 Abkf = 71.7 1190

1188

1186 EP Flood‐prone Elevation 1184 REP

1182 Bankfull Elevation 1180 Elevation (ft) 1178 Bankfull Width = 25.24 ft. 1176 Mean Depth = 2.84 ft. 2 1174 Bankfull Area = 71.66 ft

1172

1170 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 Horizontal Distance (ft)

Figure 55. Riffle cross‐section 0+93 at site 6.

The D50, or 50th percentile, particle size was 0.12 mm. Very fine‐sized particles made up 34.45% of the total particle analysis (Figure 56). All the gravel sizes combined made up 31.93% of the sample. Silt, clay, and cobble were also tallied. The D84 used for the velocity and discharge calculations was 6.44 mm.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 42

40.00% Active Riffle Particle Analysis 0+93 110.00%

35.00% 90.00% (%)

30.00% 70.00%

(%) 25.00%

Volume 20.00% 50.00%

Volume 15.00% 30.00% 10.00%

10.00% Cumulative 5.00%

0.00% ‐10.00% 4 2 8 ‐ ‐ ‐ 16 32 45 64 90 .25 5.7 .50 1.0 2 ‐ ‐ ‐ ‐ ‐ 180 256 362 512 128 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ 1.0 5.7 ‐ 4 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 56. Active riffle particle analysis at cross‐section 0+93.

Two pool cross‐sections were completed at site 6. The bankfull areas at 3+43 and 3+71 were 91.51 ft2 and 90.5 ft2, respectively (Figure 57), and the bankfull widths differed by 0.28 ft. Compared to the riffle cross‐sections, both pools had a larger cross‐sectional area, but a slightly smaller width.

Mud River: Site 6 Pool 3+71, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 25 Dbkf = 3.62 Abkf = 90.5 1190 Bankfull Width = 25 ft. 1188 Mean Depth = 3.62 ft. LEP 1186 Bankfull Area = 90.5 ft2 1184 REP

1182 Bankfull Elevation 1180 Elevation (ft) 1178 TP 1176

1174

1172

-30-20 -10 0 10 20 30 40 50 60 70 Horizontal Distance (ft)

Figure 57. 2011 Pool cross‐section 3+71 at site 6.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 43

The study bank overlay at pool cross‐section 3+71 indicated an actual erosion rate of .6273 ft/yr from 2011 to 2012 (Figure 58). This rate was slightly higher than the predicted erosion rate of .153 ft/yr derived from the BANCS model.

Mud River: Site 6 Pool Study Bank Overlay 2011-2012 Site 6 Pool 3+71 Study Bank, 2011 Site 6 Pool 3+71 Study Bank, 2012

1185

1184

1183

1182

1181

1180

Elevation (ft) 1179

1178

1177 Toe Pin

1176

1175

1174

1173 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55

Horizontal Distance (ft)

Figure 58. Site 6 pool cross‐section 3+71 study bank overlay showing the study bank profile in 2011 and 2012.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 44

DISCHARGE TABLE The estimated velocities and discharges at riffle cross‐section 0+93 were very similar between different methods (Table 16). Generally the velocities were approximately 3.3 ft/sec and the estimated discharge was 240 cfs. StreamStats estimated the 1.5, 2, 5, 10, 25 and 50 recurrence interval discharges at 240, 335, 602, 787, 1020 and 1190 cfs, respectively.

Table 16. Velocity and discharge estimates for riffle cross section 0+93.

BANKFULL VELOCITY & DISCHARGE ESTIMATES Stream: Thief River Location: Site 6 Observers: Dave Friedl Date: 8/19/2011

Riffle Cross‐section: Riffle 0+93 Bankfull Slope: 0.00062 D84 of Active Bed: 6.44 Dia. (mm) Drainage Area: 94.4 Mi2

Estimation Methods Bankfull Velocity Bankfull Discharge Manning Limerinos n 3.331 ft / sec 238.7 cfs Darcy‐Weisbach 3.584 ft / sec 256.799 cfs U/U* 3.23 ft / sec 231.23 cfs Stream Type (small without vegetative influence ‐ E4) 1.756 ft / sec 125.835 cfs

PARTICLE DISTRIBUTION The Web Soil Survey (2015) maps the riparian zone as a Kratka fine sandy loam (Figure 59). This is a poorly drained soil type.

The typical soil profile is:

Ap ‐ 0 to 11 in.: fine sandy loam Bg ‐ 11 to 18 in.: loamy fine sand

Cg ‐ 18 to 25 in.: fine sand 2Cg2 ‐ 25 to 79 in.: loam

A lowercase g denotes the presents of gray colors, indicative of permanent or periodic intensive reduction due to Figure 59. Soil map at site 6. Source: SSURGO soils and 2013 DNR saturation. aerial photo.

The full reach particle analysis within the stream channel was dominated by clay and silt particles (Figure 60). The D50 was four mm, which is gravel‐sized. Clay and silt particles made up 31.4% and gravel‐sized particles made up 50.5% of the overall sample.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 45

35.00% Reach Particle Analysis ‐ Site 6 110.00%

30.00% 90.00% (%)

25.00% 70.00% (%)

20.00% Volume 50.00% 15.00% Volume 30.00% 10.00% Cumulative 5.00% 10.00%

0.00% ‐10.00% 8 4 2 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 ‐ ‐ 2 ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 5.7 4 1.0 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 60. Particle analysis for site 6.

RIPARIAN ZONE The riparian zone adjacent to the Mud River at site 6 was dominated by reed canarygrass (Phalaris arundinacea). Goldenrod and a thistle species were also present (Figure 61).

Figure 61. Site 6 vegetation cover. Photo taken on 5/4/2010.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 46

CLASSIFICATION Table 17. Stream classification input The Rosgen stream type through site 6 was classified as an E4 factors at site 6. Table 17). The number four denotes the 50% percentile size of particles. With a D50 of 4 mm, site 6 is considered a gravel STREAM CLASSIFICATION ‐ Site 6 stream. The width‐to‐depth ratio of both riffles was less than STREAM TYPE E4 12.0. The entrenchment ratio, comparing flood‐prone width to VALLEY TYPE X bankfull width decreased moving downstream. Near the top of Slope 0.00062 the study reach it was over 2.2 and at riffle 5+07 it was 1.87.

The modified (by stream type) Pfankuch stability rating on Entrenchment Ratio 2.3 10/11/2012 was a 63. A good, or stable, rating range for an E4 is 50‐75. E stream types are low gradient, meandering riffle/pool Width/Depth Ratio 8.89 streams with a low width‐to‐depth ratio. E5 stream types are very highly sensitive to disturbance, but have a good recovery potential Sinuosity 1.01 when the disturbance is removed. This stream type is very highly dependent on robust stream bank vegetation to remain stable and 4 mm the potential for streambank erosion is high. Channel Material (D50) (GRAVEL)

DIMENSIONLESS RATIOS The upstream and downstream riffles had similar morphology, as did the two pools (Table 18). The mean pool depth‐to‐mean riffle depth ratio was 1.307, so the riffles were shallower than the pools. The pools also had greater cross‐sectional area.

Table 18. Channel morphology and dimensionless ratios for site 6.

CHANNEL MORPHOLOGY & DIMENSIONLESS RATIOS Stream: Thief River Site: 6 Observers: Dave Friedl Date: 8/10/2011

Riffle Dimensions Mean Min Max Dimensionless Ratios Mean Min Max

Riffle Width (Wbkf) 25.69 25.24 26.13 Width of Flood‐Prone Area (Wpfa) 53.38 48.75 58

Mean Riffle Depth (dbfk) 2.8 2.76 2.84 Riffle Width/Depth Ratio (Wbkf / dbkf) 9.18 8.89 9.47

Maximum Riffle Depth (dmax) 3.82 3.74 3.9 Max Riffle Depth to Mean Riffle Depth (dmax / dbkf) 1.364 1.355 1.373 2 Riffle Cross‐Sectional Area (Abkf) (ft ) 71.92 71.66 72.17 Entrenchment Ratio (Wfpa / Wbkf) 2.082 1.866 2.298

Pool Dimensions

Pool Width (Wbkfp) 24.86 24.72 25 Pool Width to Riffle Width (Wbkfp / Wbkf) 0.968 0.962 0.973

Mean Pool Depth (dbfkp) 3.66 3.62 3.7 Mean Pool Depth to Mean Riffle Depth (dbkfp / dbkf) 1.307 1.293 1.321

Maximum Pool Depth (dmaxp) 4.78 4.63 4.93 Pool Area to Riffle Area (Abkfp / Abkf) 1.265 1.258 1.272 2 Pool Cross‐Sectional Area (Abkfp) (ft ) 91.01 90.5 91.51 Max Pool Depth to Mean Riffle Depth (dmaxp / dbkf) 1.707 1.654 1.761

Thief River Watershed Fluvial Geomorphology Report  Appendix A 47

Site 7

INTRODUCTION Site 7 was located on the Thief River, downstream of the outlet from Agassiz National Wildlife Refuge, in section 5 of Agder Township (Figure 62). According to StreamStats, the drainage area is 641 mi2. The subwatershed that drains to site 7 is outlined pink in Figure 62.

Figure 62. Watershed that drains to site 7.

This area of the Thief River was ditched and is fairly straight (Figure 63). The valley walls are confined by spoil piles from clean‐outs. Areas of State Ditch 83 are still periodically spot cleaned where sedimentation or snags are major flow restrictions or deflect flow into streambanks.

Figure 63. Aerial view of site 7. Source: 2013 DNR aerial photo, 3m DEM, and ArcGIS Image Service best available resolution hillshade.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 48

LONGITUDINAL PROFILE The longitudinal profile at site 7 was approximately 1800 feet, or 28 bankfull channel widths, in length (Figure 64). The bankfull and water surface slopes were in close agreement at 0.0003 and 0.00027, respectively; however, the bankfull slope included only two survey points. As a comparison, a water surface slope was obtained using LiDAR and was 0.0002, which was in close agreement with the surveyed slopes. A slope of 0.00022 was used for the velocity and discharge calculations.

Thief River: Site 7 Longitudinal Profile 2011

1140

1138

1136 Thalweg

1134 Bankfull Slope = 0.0003 1132 Water Surface

1130 GM7 Pool XS 4+02, 2011 GM7 Riffle XS 16+05, 2011 1128 GM7 Pool XS 4+02 SB 2011 Bankfull Water Surface Slope = 0.00027 1126 1124 P1 Elevation (ft) 1122

1120 P2 1118

1116 P3 1114

1112 P4 1110 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900

Distance along stream (ft)

Figure 64. Longitudinal profile at site 1 in 2011.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 49

CROSS‐SECTIONS One riffle cross‐section was completed at site 7. The bankfull area was 314.98 ft2 and the bankfull width was 64.95 ft (Figure 65). The entrenchment ratio, comparing flood‐prone width to bankfull width, came out as moderately entrenched. Riffle 16+05 had a flood‐prone width of 101.5 ft., giving it an entrenchment ratio of 1.56.

Thief River: Site 7 Riffle 16+05, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 65 Dbkf = 4.85 Abkf = 315 1155 Bankfull Width = 64.95 ft. 1150 Mean Depth = 4.85 ft. Bankfull Area = 314.98 ft2 1145 EP

REP 1140 Flood‐prone Elevation

Elevation (ft) 1135 Bankfull Elevation

1130

1125

1120

1115 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100105110115120125130135140145150155160 Horizontal Distance (ft) Figure 65. Riffle cross‐section 16+05 at site 7.

The 50th percentile (D50) particle size was 1.08 mm. Silt and clay made up 35.66% of the particles measured at this cross‐section (Figure 66). Sand‐ and gravel‐sized particles made up another 18.6% and 44.96% of the sample, respectively. The D84 used for the velocity and discharge calculations was 13.27 mm.

Thief River Watershed Fluvial Geomorphology Report  Appendix A

40.00% Active Riffle Particle Analysis 16+05 110.00%

35.00% 90.00% (%) 30.00% 70.00%

(%) 25.00%

Volume

20.00% 50.00%

Volume 15.00% 30.00% 10.00% Cumulative 10.00% 5.00%

0.00% ‐10.00% 2 4 8 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 ‐ ‐ 2 ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ 5.7 ‐ 1.0 4 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 66. Active riffle particle analysis at cross‐section 16+05.

One pool cross‐section (4+02) was completed at site 7. The bankfull area was 349.43 ft2 and the bankfull width was 72.29 ft. (Figure 67).

Thief River: Site 7 Pool 4+02, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 72.3 Dbkf = 4.83 Abkf = 349.4 1155 1153 1151 Bankfull Width = 72.29 ft. 1149 Mean Depth = 4.83 ft. 1147 1145 Bankfull Area = 349.43 ft2 1143 1141 1139LEP REP 1137 1135 1133 Bankfull Elevation Elevation (ft) 1131 1129 1127 TP 1125 1123 1121 1119 1117

-110 -100 -10-20-30-40-50-60-70-80-90 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 Horizontal Distance (ft)

Figure 67. 2011 Pool cross‐section 4+02 at site 7.

Thief River Watershed Fluvial Geomorphology Report  Appendix A

The study bank overlay at pool cross‐section 4+02 indicated an actual erosion rate of .0768 ft/yr from 2011 to 2012 (Figure 68). This rate closely matched the predicted erosion rate of .06 ft/yr derived from the BANCS model.

Thief River: Site 7 Pool Study Bank Overlay 2011-2012 Site 7 Pool 4+02 Study Bank, 2011 Site 7 Pool 4+02 Study Bank, 2012

1135

1134

1133

1132

1131

1130 Elevation (ft) 1129

1128 Toe Pin 1127

1126

1125

1124

1123 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93

Horizontal Distance (ft)

Figure 68. Site 7 pool cross‐section 4+02 study bank overlay showing the study bank profile in 2011 and 2012.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 52

DISCHARGE TABLE The estimated velocities and discharges at riffle cross‐section 16+05 were very similar between different methods. Generally the velocities were approximately 2.6 ft/sec and the estimated discharge was 840 cfs (Table 19). StreamStats estimated the 1.5, 2, 5, 10, 25 and 50 recurrence interval discharges at 796, 1050, 1730, 2160, 2700 and 3070 cfs, respectively.

Table 19. Velocity and discharge estimates for riffle cross section 16+05.

BANKFULL VELOCITY & DISCHARGE ESTIMATES Stream: Thief River Location: Site 7 Observers: Dave Friedl Date: 8/19/2011

Riffle Cross‐section: Riffle 16+05 Bankfull Slope: 0.00022 D84 of Active Bed: 13.27 Dia. (mm) Drainage Area: 641 Mi2

Estimation Methods Bankfull Velocity Bankfull Discharge Manning Limerinos n 2.677 ft / sec 843.202 cfs Darcy‐Weisbach 2.703 ft / sec 851.379 cfs U/U* 2.59 ft / sec 816.61 cfs Stream Type (medium to large) 1.426 ft / sec 449.162 cfs

PARTICLE DISTRIBUTION The Web Soil Survey (2015) maps the riparian zone as a Fluvaquents, frequently flooded‐Hapludolls complex, Boash clay loam and Foxlake loam (Figure 69). All 3 types are poorly drained soils. The typical profile of the Foxlake loam is:

Ap ‐ 0 to 16 in.: loam A – 10 to 19 in: loam

Bg – 19 to 38 in.: silty clay Bkg – 38 to 49 in: silty clay Cg – 49 to 80 in.: silty clay

The full reach particle analysis was dominated by clay and silt‐sized particles Figure 69. Soil map of site 7. Source: SSURGO soils and 2013 DNR aerial photo. (31.55%). Sand, gravel and very small cobble were also present (Figure 70). Though there was a dominant presence of clay and silt, the D50 for the full reach count was 0.14mm (fine sand). The active riffle, pool, and full reach particle analyses all were very similar for size distribution.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 53

35.00% Reach Particle Analysis ‐ Site 7 110.00%

30.00% 90.00% (%)

25.00% 70.00% (%)

20.00% Volume

50.00% 15.00% Volume 30.00% 10.00% Cumulative 5.00% 10.00%

0.00% ‐10.00% 4 8 2 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 2 ‐ ‐ ‐ ‐ ‐ 180 256 362 512 128 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 5.7 4 1.0 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 70. Particle analysis for riffle cross section 16+05.

CLASSIFICATION Table 20. Stream classification input factors The Rosgen stream type through site 7 was classified as a B5c at site 7. (Table 20). With a D50 of 0.14 mm, site 7 was considered a STREAM CLASSIFICATION ‐ Site 7 sand bed stream. The lowercase c signifies that the energy slope at this site is less than 0.02. The width‐to‐depth ratio STREAM TYPE B5c was 13.39 and the entrenchment ratio was 1.56. Sinuosity VALLEY TYPE X was 1.2. The number five denotes the 50% percentile size of particles. Slope 0.00022

The modified (by stream type) Pfankuch stability rating on Entrenchment Ratio 1.56 10/12/2012 was 111. A good, or stable, rating range for a B5 is 48‐68. A poor, or unstable, rating is 89 or greater. B stream types are moderately entrenched with a moderate gradient. Width/Depth Ratio 13.39 They usually are riffle dominated with infrequently spaced pools. B5 stream types are moderately sensitive to Sinuosity 2 disturbance, but have an excellent recovery potential when the disturbance is removed. This stream type is moderately 0.14 mm dependent on robust stream bank vegetation to remain stable Channel Material (D ) 50 (SAND) and the potential for streambank erosion is low.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 54

DIMENSIONLESS RATIOS Only one riffle and one pool cross‐sections were completed at site 7; therefore, there was no variation between the mean, minimum and maximum values (Table 21). Comparing the riffle to the pool, the mean pool depth‐to‐mean riffle depth ratio was 0.996, so the pool was shallower than the riffle, but the pool area was slightly larger than the riffle.

Table 21. Channel morphology and dimensionless ratios for site 7.

CHANNEL MORPHOLOGY & DIMENSIONLESS RATIOS Stream: Thief River Site: 7 Observers: Dave Friedl Date: 10/05/2011

Riffle Dimensions Mean Min Max Dimensionless Ratios Mean Min Max

Riffle Width (Wbkf) 656565Width of Flood‐Prone Area (Wpfa) 102 102 102

Mean Riffle Depth (dbfk) 4.85 4.85 4.85 Riffle Width/Depth Ratio (Wbkf / dbkf) 13.39 13.39 13.39

Maximum Riffle Depth (dmax) 6.55 6.55 6.55 Max Riffle Depth to Mean Riffle Depth (dmax / dbkf) 1.351 1.351 1.351 2 Riffle Cross‐Sectional Area (Abkf) (ft ) 314.98 314.98 314.98 Entrenchment Ratio (Wfpa / Wbkf) 1.563 1.563 1.563

Pool Dimensions

Pool Width (Wbkfp) 72.3 72.3 72.3 Pool Width to Riffle Width (Wbkfp / Wbkf) 1.113 1.113 1.113

Mean Pool Depth (dbfkp) 4.83 4.83 4.83 Mean Pool Depth to Mean Riffle Depth (dbkfp / dbkf) 0.996 0.996 0.996

Maximum Pool Depth (dmaxp) 7.38 7.38 7.38 Pool Area to Riffle Area (Abkfp / Abkf) 1.109 1.109 1.109 2 Pool Cross‐Sectional Area (Abkfp) (ft ) 349 349 349 Max Pool Depth to Mean Riffle Depth (dmaxp / dbkf) 1.522 1.522 1.522

Thief River Watershed Fluvial Geomorphology Report  Appendix A 55

Site 8

INTRODUCTION Site 8 was located on County Ditch 20, immediately upstream from its confluence with the Thief River, in section 24 of Excel Township (Figure 71). According to StreamStats, the drainage area is 210 mi2. The subwatershed that drains to site 8 is outlined pink in Figure 71.

Figure 71. Watershed that drains to site 8.

At this site, CD 20 is classified as a perennial ditch. After the first year of field assessment, grade control structures were installed within the longitudinal profile, consequently several of the cross‐sections surveyed in 2011 were not resurveyed in 2012 (Figure 72).

Figure 72. Aerial view of site 8. Source: 2013 DNR aerial photo, 3m DEM, and ArcGIS Image Service best available resolution hillshade.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 56

LONGITUDINAL PROFILE The longitudinal profile at site 8 was 1200 feet, or 25 bankfull channel widths, in length (Figure 73). Given its location within a highly altered ditch system, this site had poorly defined riffle and pool features. Its proximity to the Thief River may have also had an impact on the longitudinal profile and accompanying bed features. The bankfull and water surface slopes were 0.00016 and 0.00018, respectively. The water surface slope (0.00018) was used for the velocity and discharge calculations.

Thief GM8 Longitudinal Profile 2011

1135

1134

1133 Thalweg

1132

1131 Water Surface GM8 Pool XS 8+57, 2011 GM8 Riffle XS 0+00, 2011 GM8 Riffle XS 3+12, 2011 GM8 Riffle XS 11+13, 2011

Bankfull Slope = 0.00016 GM8 Pool XS 8+57 SB 2011 1130 GM8 Pool XS 8+57, 2011 Trimble GM8 Riffle XS 3+12, 2011 Trimble 1129 Bankfull

1128

1127 P1 Elevation (ft)

1126 Water Surface Slope = 0.00018 1125 P2

1124

1123 P3

1122

1121 P4 1120 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 1000 1050 1100 1150 1200 1250

Distance along stream (ft)

Figure 73. Longitudinal profile at site 8 in 2011.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 57

CROSS‐SECTIONS Three riffle cross‐sections were completed at site 8. Riffle 3+12 was used for the velocity and discharge calculations. The bankfull area was 187.19 ft2 and the bankfull width was 48.64 ft. (Figure 74). The most upstream riffle, located at the start of the longitudinal profile, had a bankfull area of 203.78 ft2 and width of 46.69 ft. The last riffle cross‐section was the smallest with a bankfull width of 40.29 ft. and cross‐sectional area of 179.16 ft2.

The entrenchment ratio, comparing flood‐prone width to bankfull width, came out as moderately entrenched for both riffle cross‐sections 0+00 and 3+12. Riffle 11+13 had a wider flood‐prone width. This riffle had a maximum depth that was one ft. greater than the upstream riffles. Since the flood‐prone width is calculated by multiplying the maximum depth by two, this flood‐prone elevation for riffle 11+13 was able to get over the spoil piles.

County Ditch 20: Site 8 Riffle 3+12, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 48.6 Dbkf = 3.85 Abkf = 187.2 1145 1143

1141 REP 1139 1137 LEP Flood‐prone Elevation 1135 1133 1131 Bankfull Elevation 1129 Elevation (ft) 1127 1125 Bankfull Width = 48.64 ft. 1123 Mean Depth = 3.85 ft. 1121 Bankfull Area = 187.19 ft2 1119 1117

-5 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 125 130 135 Horizontal Distance (ft)

Figure 74. Riffle cross‐section 3+12 at site 8.

The 50th percentile particle size (D50) was 0.61 mm. Silt and clay made up 25.42% of the particles measured at this cross‐section (Figure 75). Gravel, cobble and boulder‐sized particles were also tallied. The D84 used for the velocity and discharge calculations was 7.28 mm.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 58

Active Riffle Particle Analysis 3+12 30.00% 110.00%

25.00% 90.00% (%)

(%) 20.00% 70.00%

Volume 15.00% 50.00% Volume 10.00% 30.00%

5.00% 10.00% Cumulative

0.00% ‐10.00% 8 2 4 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 ‐ ‐ 2 ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 4 5.7 1.0 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER

Particle Size (mm) Figure 75. Active riffle particle analysis at cross‐section 3+12.

One pool cross‐section (8+57) was completed at site 8. The bankfull area was 212.85 ft2 and the bankfull width was 47.43 ft. (Figure 76).

County Ditch 20: Site 8 Pool 8+57, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 47.4 Dbkf = 4.49 Abkf = 212.8 1145 1143 Bankfull Width = 47.43 ft. 1141 1139 EP Mean Depth = 4.49 ft. Bankfull Area = 212.85 ft2 1137 REP 1135 1133 1131 Bankfull Elevation 1129

Elevation (ft) 1127 1125 TP 1123 1121 1119 1117 1115 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 Horizontal Distance (ft)

Figure 76. 2011 Pool cross‐section 8+57 at site 8.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 59

DISCHARGE TABLE The estimated velocities and discharges at riffle cross‐section 3+12 were approximately 2.25 ft/sec and 420 cfs, respectively (Table 22). The Manning’s n by stream type method underestimated velocity and discharge values. StreamStats estimated the 1.5, 2, 5, 10, 25 and 50 recurrence interval discharges at 448, 639, 1200, 1590, 2110 and 2480 cfs, respectively.

Table 22. Velocity and discharge estimates for riffle cross‐section 3+12.

BANKFULL VELOCITY & DISCHARGE ESTIMATES Stream: Thief River Location: Site 8 Observers: Dave Friedl Date: 8/18/2011

Riffle Cross‐section: Riffle 3+12 Bankfull Slope: 0.00018 D84 of Active Bed: 7.28 Dia. (mm) Drainage Area: 210 Mi2

Estimation Methods Bankfull Velocity Bankfull Discharge Manning Limerinos n 2.269 ft / sec 424.734 cfs Darcy‐Weisbach 2.316 ft / sec 433.538 cfs U/U* 2.21 ft / sec 412.92 cfs Stream Type (medium to large) ‐‐ ft / sec ‐‐ cfs

PARTICLE DISTRIBUTION The Web Soil Survey (2015) maps the riparian zone as a Roliss loam and Clearwater clay (Figure 77). Both are poorly drained soil types. The typical Roliss loam contains loam soils from 0 to 80 in. The typical Clearwater clay profile is:

Ap ‐ 0 to 8 in.: ‐ clay Bss – 8 to 35 in.: ‐ clay Cg – 35 to 80 in.: ‐ clay

The lower case p denotes plowing present and the ss denotes slickensides on the clay soils. Figure 77. Soils map at site 8. Source: SSURGO soils and 2013 DNR aerial photo. Clay and silt‐sized particles made up 29.2% of the full reach particle count, and sand and gravel made up 48.5% and 21%, respectively (Figure 78). The 50th percentile particle size was 0.12 mm, which is very fine sand.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 60

35.00% Reach Particle Analysis ‐ Site 8 110.00%

30.00% 90.00% (%) 25.00% 70.00% (%) 20.00% Volume 50.00% 15.00% Volume 30.00% 10.00%

5.00% 10.00% Cumulative

0.00% ‐10.00% 4 2 8 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 2 ‐ ‐ ‐ ‐ ‐ 180 256 362 512 128 ‐ ‐ ‐ ‐ 11.3 .125 22.6 ‐ 1024 2048 ‐ ‐ 5.7 1.0 4 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 78. Particle analysis for site 8.

RIPARIAN ZONE The land adjacent to CD 20 at site 8 was a mixture of trees, shrubs and groundcover (Figure 79). Trees covered 45% of the aerial cover and 15% of the ground cover. Species included 40% Fraxinus pennsylvanica (green ash), 15% Quercus macrocarpa (bur oak), 10% Ulmus americana (American elm), 25% Acer neguno (box elder), 5% Salix spp. (willow) and 5% Populus deltoides (cottonwood).

Understory shrubs covered approximately 13% of the riparian zone. Shrubs present included 20% dogwood, 20% rose, 20% wild Figure 79. Site 8 vegetation cover. Photo taken on plum, 20% beaked hazelnut, 10% willow and 8/18/2011. 10% honeysuckle. The herbaceous layer was dominanted by 90% Phalaris arundinacea (reed canarygrass) and 5% Bromus inermis (smooth brome). Other species included 1% Solidago spp. (goldenrod), Toxicodendron radicans (poison ivy), 1% Arctium minus (burdock), 1% Asclepias spp. (milkweed) and 1% Cirsium arvense (Canada thistle).

Thief River Watershed Fluvial Geomorphology Report  Appendix A 61

CLASSIFICATION Table 23. Stream classification input factors at The Rosgen stream type through site 8 was classified as a site 8. B5c (Table 23). The number five denotes the 50th percentile size of particles. With a D50 of 0.12 mm, site 8 STREAM CLASSIFICATION ‐ Site 8 is considered a very fine sand stream. The lowercase c STREAM TYPE B5c signifies that the energy slope at this site is less than 0.02. VALLEY TYPE X The entrenchment ratio of both riffle cross‐sections 0+00 and 3+12 fell within the moderately entrenched category. Slope 0.00018 Riffle cross‐section 11+13 had an entrenchment ratio of 5.21, which is slightly entrenched.

Entrenchment Ratio 1.71 The modified (by stream type) Pfankuch stability rating on 10/10/2012 was a 120. A good, or stable, rating range for a B5 is 48‐68. A poor, or unstable, rating is 89 or Width/Depth Ratio 12.63 greater. B stream types are moderately entrenched with a moderate gradient. They usually are riffle dominated with infrequently spaced pools. B5 stream types are Sinuosity 1.04 moderately sensitive to disturbance, but have an excellent recovery potential when the disturbance is removed. This stream type is moderately dependent on 0.12 mm Channel Material (D ) robust stream bank vegetation to remain stable and the 50 (V. FINE SAND) potential for streambank erosion is moderate.

DIMENSIONLESS RATIOS There was some variation in dimensions between the 3 riffle cross‐sections (Table 24). Riffle 11+13 had the widest flood‐prone width, but it also had the deepest riffle. Flood‐prone width is the stream width at a discharge that is 2 x the maximum bankfull depth. The one pool cross‐section was only slightly larger and wider than mean riffle values. Table 24. Channel morphology and dimensionless ratios for site 8.

CHANNEL MORPHOLOGY & DIMENSIONLESS RATIOS Stream: Thief River Site: 8 Observers: Dave Friedl Date: 8/18/2011

Riffle Dimensions Mean Min Max Dimensionless Ratios Mean Min Max

Riffle Width (Wbkf) 45.2 40.3 48.6 Width of Flood‐Prone Area (Wpfa) 120 65.4 210

Mean Riffle Depth (dbfk) 4.22 3.85 4.45 Riffle Width/Depth Ratio (Wbkf / dbkf) 10.8 9.05 12.63

Maximum Riffle Depth (dmax) 5.89 5.54 6.57 Max Riffle Depth to Mean Riffle Depth (dmax / dbkf) 1.4 1.27 1.5 2 Riffle Cross‐Sectional Area (Abkf) (ft ) 190.04 179.16 203.78 Entrenchment Ratio (Wfpa / Wbkf) 2.78 1.4 5.2

Pool Dimensions

Pool Width (Wbkfp) 47.4 47.4 47.4 Pool Width to Riffle Width (Wbkfp / Wbkf) 1.05 1.05 1.05

Mean Pool Depth (dbfkp) 4.49 4.49 4.49 Mean Pool Depth to Mean Riffle Depth (dbkfp / dbkf) 1.06 1.05 1.06

Maximum Pool Depth (dmaxp) 6.57 6.57 6.57 Pool Area to Riffle Area (Abkfp / Abkf) 1.12 1.12 1.12 2 Pool Cross‐Sectional Area (Abkfp) (ft ) 213 213 213 Max Pool Depth to Mean Riffle Depth (dmaxp / dbkf) 1.56 1.56 1.56

Thief River Watershed Fluvial Geomorphology Report  Appendix A 62

Site 9

INTRODUCTION Site 9 was located on County Ditch 20, about 1.4 mi. upstream of the angle ditch, in section 27 of Agder Township (Figure 80). According to StreamStats, the drainage area is 174 mi2. The subwatershed that drains to site 9 is outlined pink in Figure 80.

Figure 80. Watershed that drains to site 9.

Site 9 was located south of Elm Lake WMA where CD 20 flows adjacent to 240th St. NE (Figure 81). CD 20 is a perennial drainage ditch at this location.

Figure 81. Aerial view of site 1. Source: 2013 DNR aerial photo, 3m DEM, and ArcGIS Image Service best available resolution hillshade.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 63

LONGITUDINAL PROFILE The longitudinal profile at site 9 was approximately 690 feet, or 18 bankfull channel widths, in length (Figure 82). Because of its location in a ditched system, this site lacked the typical and easily discernable riffle‐pool sequences found in meandering streams. The bankfull and water surface slopes were 0.0003 and 0.0001, respectively. The bankfull slope (0.0003) was used for velocity and discharge calculations, as it gave more realistic estimates.

County Ditch 20: Site 9 Longitudinal Profile 2011

1138

1137 Bankfull Slope = 0.0003 Thalweg 1136

1135 Water Surface

1134 GM9 Pool XS 4+36, 2011 XS 4+36, GM9 Pool GM9 Riffle XS 6+84, 2011 XS 6+84, GM9 Riffle GM9 Riffle XS 0+00, 2011 XS 0+00, GM9 Riffle 2011 XS 1+88, GM9 Riffle

GM9 Pool XS 4+36 SB 2011 XS 4+36 GM9 Pool Bankfull 1133 Water Surface Slope = 0.0001

1132 P1 Elevation (ft) 1131

1130 P2

1129

1128 P3

1127 P4 1126 0 35 70 105 140 175 210 245 280 315 350 385 420 455 490 525 560 595 630 665 700

Distance along stream (ft)

Figure 82. Longitudinal profile at site 9 in 2011.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 64

CROSS‐SECTIONS Three riffle cross‐sections were completed at site 9. Riffle 6+84 was used for the velocity and discharge calculations (Figure 83). The bankfull area was 167.2 ft2 and the bankfull width was 39.1 ft. Riffle cross‐sections 0+00 and 1+88 were both smaller than 6+84. Their bankfull areas were 155.5 and 134.37 ft2 and widths are 37.95 and 36.64 ft., respectively. The entrenchment ratio, comparing flood‐prone width to bankfull width, came out as moderately entrenched for all riffle cross‐sections. County Ditch 20: Site 9 Riffle 6+84, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 39.1 Dbkf = 4.28 Abkf = 167.2 1148

Bankfull Width = 39.1 ft. REP 1146 Mean Depth = 4.28 ft. 1144 EP Bankfull Area = 167.2 ft2 Flood‐prone Elevation 1142

1140

1138 Bankfull Elevation Elevation (ft) 1136

1134

1132

1130

1128 0 102030405060708090 Horizontal Distance (ft)

Figure 83. Riffle cross‐section 6+84 at site 9.

The 50th percentile (D50) particle size was 6.39 mm. Silt and clay made up 28.04% of the particles measured at this cross section (Figure 84). Gravel, cobble and boulder‐sized particles were also tallied. The D84 used for the velocity and discharge calculations was 20.73 mm.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 65

30.00% Active Riffle Particle Analysis 110.00%

25.00% 90.00% (%)

20.00% 70.00% (%)

Volume 15.00% 50.00%

Volume 10.00% 30.00%

5.00% 10.00% Cumulative

0.00% ‐10.00% 2 4 8 ‐ ‐ ‐ 16 32 45 64 90 .25 5.7 .50 1.0 ‐ ‐ 2 ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ 1.0 5.7 ‐ 4 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 84. Active riffle particle analysis at cross‐section 6+84.

One pool cross‐section (4+36) was completed at site 9. The bankfull area was 148.98 ft2 and the bankfull width was 35.74 ft. (Figure 85). Compared to riffle cross‐sections 0+00 and 6+84, the pool was slightly smaller and narrower.

County Ditch 20: Site 9 Pool 4+36, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 35.7 Dbkf = 4.17 Abkf = 149 1150

Bankfull Width = 35.74 ft. REP Mean Depth = 4.17 ft. 1145 2 EP Bankfull Area = 148.98 ft

1140

Elevation (ft) Bankfull Elevation 1135

TP 1130

1125 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 125 130 135 Horizontal Distance (ft) Figure 85. 2011 Pool cross‐section 4+36 at site 9.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 66

The study bank overlay at pool cross‐section 4+36 indicated an actual erosion rate of .2982 ft/yr from 2011 to 2012 (Figure 86). This rate was slightly higher than the predicted erosion rate of .153 ft/yr derived from the BANCS model.

County Ditch 20: Site 9 Pool Study Bank Overlay 2011-2012 Site 9 Pool 4+36 Study Bank, 2011 Site 9 Pool 4+36 Study Bank, 2012

1138

1137

1136

1135

1134

Elevation (ft) 1133

1132 Toe Pin 1131

1130

1129

1128 34 35 36 37 38 39 40 41 42 43 44 45 46 47 Horizontal Distance (ft)

Figure 86. Site 9 pool cross‐section 4+36 study bank overlay showing the study bank profile in 2011 and 2012.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 67

DISCHARGE TABLE The estimated velocities and discharges at riffle cross‐section 6+84 were very similar between different methods (Table 25). Generally the velocities were approximately 2.5 ft/sec and the estimated discharge was 420 cfs. StreamStats estimated the 1.5, 2, 5, 10, 25 and 50 recurrence interval discharges at 389, 554, 1030, 1370, 1810, and 2140 cfs, respectively.

Table 25. Velocity and discharge estimates for riffle cross section 6+84.

BANKFULL VELOCITY & DISCHARGE ESTIMATES Stream: Thief River Location: Site 9 Observers: Dave Friedl Date: 8/19/2011

Riffle Cross‐section: Riffle 6+84 Bankfull Slope: 0.0003 D84 of Active Bed: 20.73 Dia. (mm) Drainage Area: 174 Mi2

Estimation Methods Bankfull Velocity Bankfull Discharge Manning Limerinos n 2.55 ft / sec 426.53 cfs Darcy‐Weisbach 2.68 ft / sec 447.33 cfs U/U* 2.46 ft / sec 411.21 cfs Stream Type (medium to large) ‐‐ ft / sec ‐‐ cfs

PARTICLE DISTRIBUTION The Web Soil Survey (2015) maps the riparian zone as Berner and Cathros soils (Figure 87). This soil type is very poorly drained. The typical profile for this map unit is:

Oa ‐ 0 to 28 in.: muck A ‐ 28 to 31 in.: sandy loam Bg ‐ 31 to 44 in.: sand 2CBkg ‐ 44 to 60 in.: loam

This site is the only one that had an organic layer within the soil profile. The lowercase g denotes the presents of gray colors, indicative of permanent or periodic Figure 87. Soil map of site 9. Source: SSURGO soils and 2013 DNR photo. intensive reduction due to saturation.

Coarse sand and very fine gravels were tallied the most during the in‐channel particle count, at/with 14.8% and 17.8% respectively (Figure 88). Clay and silt made up 13.6% of the sample. The 50th percentile particle size (D50) was 1.86 mm, which is very coarse sand.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 68

20.00% Reach Particle Analysis ‐ Site 9 110.00% 18.00% 90.00% 16.00% (%)

14.00% 70.00% (%) 12.00% Volume 10.00% 50.00% 8.00% Volume 30.00% 6.00% 4.00% 10.00% Cumulative 2.00% 0.00% ‐10.00% 2 4 8 ‐ ‐ ‐ 16 32 45 64 90 .25 5.7 .50 1.0 ‐ ‐ 2 ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 1.0 4 5.7 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm) Figure 88. Particle analysis for site 9.

CLASSIFICATION Table 26. Stream classification input factors The Rosgen stream type through site 9 was classified as a at site 9. B5c (Table 26). The number five denotes the 50% percentile size of particles. With a D50 of 1.86 mm, site 9 is STREAM CLASSIFICATION ‐ Site 9 considered a sand stream. The lowercase c signifies that STREAM TYPE B5c the energy slope at this site is less than 0.02. The VALLEY TYPE X entrenchment ratios of all the riffle cross‐sections fell within the moderately entrenched category. B stream types Slope 0.0003 usually have a width‐to‐depth (w/d) ratio of greater than 12.0. All of the riffle cross‐sections had w/d ratios ranging from 9.14‐9.98. Sinuosity was 1.00. Entrenchment Ratio 1.81 The modified (by stream type) Pfankuch stability rating on 10/10/2012 was a 74. A good, or stable, rating range for a Width/Depth Ratio 9.14 B5 is 48‐68. A poor, or unstable, rating is 89 or greater. B stream types are moderately entrenched with a moderate gradient. They usually are riffle dominated with Sinuosity 1.00 infrequently spaced pools. B5 stream types are moderately sensitive to disturbance, but have an excellent recovery 1.86 mm potential when the disturbance is removed. This stream Channel Material (D50) (V. COARSE type is moderately dependent on robust stream bank SAND) vegetation to remain stable and the potential for streambank erosion is moderate.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 69

DIMENSIONLESS RATIOS The three riffle cross‐sectional areas varied by 33 ft2, while the widths only varied by 2.5 ft (Table 27). One pool cross‐section was completed at this site. The mean pool depth‐to‐mean riffle depth ratio was 1.04, so there was very little variation between the pool and riffles. The pool was slightly smaller than the riffles.

Table 27. Channel morphology and dimensionless ratios at site 9.

CHANNEL MORPHOLOGY & DIMENSIONLESS RATIOS Stream: Thief River Site: 9 Observers: Dave Friedl Date: 8/8/2011

Riffle Dimensions Mean Min Max Dimensionless Ratios Mean Min Max

Riffle Width (Wbkf) 37.9 36.6 39.1 Width of Flood‐Prone Area (Wpfa) 68.4 66.9 70.9

Mean Riffle Depth (dbfk) 4.02 3.67 4.28 Riffle Width/Depth Ratio (Wbkf / dbkf) 9.46 9.14 9.98

Maximum Riffle Depth (dmax) 5.37 4.83 5.74 Max Riffle Depth to Mean Riffle Depth (dmax / dbkf) 1.34 1.32 1.35 2 Riffle Cross‐Sectional Area (Abkf) (ft ) 152.36 134.37 167.2 Entrenchment Ratio (Wfpa / Wbkf) 1.8 1.76 1.84

Pool Dimensions

Pool Width (Wbkfp) 35.7 35.7 35.7 Pool Width to Riffle Width (Wbkfp / Wbkf) 0.94 0.94 0.94

Mean Pool Depth (dbfkp) 4.17 4.17 4.17 Mean Pool Depth to Mean Riffle Depth (dbkfp / dbkf) 1.04 1.04 1.04

Maximum Pool Depth (dmaxp) 6.14 6.14 6.14 Pool Area to Riffle Area (Abkfp / Abkf) 0.98 0.98 0.98 2 Pool Cross‐Sectional Area (Abkfp) (ft ) 149 149 149 Max Pool Depth to Mean Riffle Depth (dmaxp / dbkf) 1.53 1.53 1.53

Thief River Watershed Fluvial Geomorphology Report  Appendix A 70

Site 10

INTRODUCTION Site 10 was located on CD 20, about 5 miles south of Grygla, MN,in section 26 of Moylan Township (Figure 89). According to StreamStats, the drainage area is 86.4 mi2. The subwatershed that drains to site 10 is outlined pink in Figure 89.

Figure 89. Watershed that drains to site 10.

Site 10 was in a straight, channelized reach, with a sinuosity of 1.003 (Figure 90). The upstream drainage area is approximately 29.5% wetlands, based on the DNR 100K wetland layer. This area is a complex of various Wildlife Management Areas. Beltrami State Forest is on the eastern side of this subwatershed, near Fourtown. The southern portion is located within the boundary of the Red Lake Indian Reservation.

Figure 90. Aerial view of site 1. Source: 2013 DNR aerial photo, 3m DEM, and ArcGIS Image Service best available resolution hillshade.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 71

LONGITUDINAL PROFILE The longitudinal profile at site 10 was approximately 600 feet, or 21 bankfull channel widths, in length (Figure 91). The thalweg, or channel bottom, through site 10 appeared to follow a normal pattern—the river became shallower through the glides and riffles and deeper in the pools. However, because of its location in a ditched system, the bed features were not easily discernable. The bankfull slope was 0.00065 using 3 bankfull calls, one at each cross‐section. The water surface slope was 0.00156. The bankfull slope (0.00065) was used for velocity and discharge calculations, as it provided more reasonable estimates.

County Ditch 20: Site 10 Longitudinal Profile 2011

1185

1183 Thalweg

1181

GM10 Pool XS 4+25, 2011 4+25, XS GM10 Pool Water Sfc GM10 Riffle XS 6+02, 2011 6+02, XS GM10 Riffle GM10 Riffle XS 0+00, 2011 0+00, XS GM10 Riffle GM10 Pool XS 4+25, 2011_2 4+25, XS GM10 Pool

1179 2011 SB, 4+25 XS GM10 Pool GM10 Riffle XS 6+02, 2011_Original 6+02, XS GM10 Riffle 1177 Bankfull Slope = 0.00065 Bankfull

1175

Elevation (ft) P1 Water Surface Slope = 0.00156 1173

P2 1171

1169 P3

1167 P4

1165 0 25 50 75 100 125 150 175 200 225 250 275 300 325 350 375 400 425 450 475 500 525 550 575 600 625

Distance along stream (ft)

Figure 91. Longitudinal profile at site 10 in 2011.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 72

CROSS‐SECTIONS Two riffle cross‐sections were completed at site 10. Riffle 0+00 was used for the velocity and discharge calculations, because a pebble count was located at this location. The bankfull area was 88.28 ft2 and the bankfull width was 28.34 ft. (Figure 92). The second riffle was located at 602 ft. along the longitudinal profile. The bankfull area of this cross‐section was 97.79 ft2 and the bankfull width was 32.18 ft.

The entrenchment ratio, comparing flood‐prone width to bankfull width, came out as moderately entrenched for both riffle cross‐sections. Riffle 0+00 had a flood‐prone width of 55 ft., giving it an entrenchment ratio of 1.94. Riffle 6+02 had a floodplain width of 49.15 ft., giving it a ratio of 1.53.

County Ditch 20: Site 10 Riffle 0+00, 2011 Ground Points Bankfull Indicators Water Surface Points Wbkf = 28.3 Dbkf = 3.12 Abkf = 88.3 1190 Bankfull Width = 28.34 ft. 1188 Mean Depth = 3.12 ft. 2 1186 Bankfull Area = 88.28 ft REP 1184 EP 1182 Flood‐prone Elevation

1180

1178 Bankfull Elevation

Elevation (ft) 1176

1174

1172

1170 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Horizontal Distance (ft)

Figure 92. Riffle cross‐section 0+00 at site 10.

The D50, or 50th percentile, particle size was 4.99 mm (Figure 93). Silt and clay made up 33.94% of the particles measured at this cross‐section. Sand, gravel, and cobble‐sized particles were also tallied. The D84 used for the velocity and discharge calculations was 19.33 mm.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 73

40.00% Active Riffle Particle Analysis 0+00 110.00% 35.00% 90.00% (%) 30.00% 70.00% 25.00% Volume (%)

20.00% 50.00%

15.00% 30.00%

Volume 10.00%

10.00% Cumulative 5.00%

0.00% ‐10.00% 2 4 8 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 ‐ ‐ 2 ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ 5.7 ‐ 1.0 4 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 93. Active riffle particle analysis at cross‐section 0+00.

One pool cross section (4+25) was completed at site 10. The bankfull area was 111.97 ft2 and the bankfull width was 28.79 ft. (Figure 94). Compared to riffle cross‐section 0+00, the pool was the same width, but had a greater cross‐sectional area.

County Ditch 20: Site 10 Pool 4+25, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 28.8 Dbkf = 3.89 Abkf = 112 1190

1188 Bankfull Width = 28.79 ft. Mean Depth = 3.89 ft. 1186 LEP Bankfull Area = 111.97 ft2 1184 REP 1182

1180 Elevation (ft)

1178 Bankfull Elevation 1176

1174 TP 1172

-10 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 Horizontal Distance (ft)

Figure 94. 2011 Pool cross‐section 4+25 at site 10.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 74

The study bank overlay at pool cross‐section 4+25 indicated an actual erosion rate of .37 ft/yr from 2011 to 2012 (Figure 95). This rate was slightly higher than the predicted erosion rate of .153 ft/yr derived from the BANCS model.

County Ditch 20: Site 10 Pool Study Bank Overlay 2011-2012 Site 10 Pool 4+25 Study Bank, 2011 Site 10 Pool 4+25 Study Bank, 2012

1178

1177

1176

1175 Elevation (ft) 1174

1173

Toe Pin 1172

1171

1170 25 26 27 28 29 30 31 32 33 34 35

Horizontal Distance (ft)

Figure 95. Site 10 pool cross‐section 4+25 study bank overlay showing the study bank profile in 2011 and 2012.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 75

DISCHARGE TABLE The estimated velocities and discharges at riffle cross‐section 0+00 were very similar between different methods (Table 28). Generally the velocities were approximately 3.0 ft/sec and the estimated discharge was 270 cfs. StreamStats estimated the 1.5, 2, 5, 10, 25 and 50 recurrence interval discharges at 244, 354, 683, 927, 1250 and 1500 cfs, respectively.

Table 28. Velocity and discharge estimates for riffle cross‐section 0+00.

BANKFULL VELOCITY & DISCHARGE ESTIMATES Stream: County Ditch 20 Location: Site 10 Observers: Dave Friedl Date: 8/9/2011

Riffle Cross‐section: Riffle 0+00 Bankfull Slope: 0.00065 D84 of Active Bed: 19.33 Dia. (mm) Drainage Area: 86.4 Mi2

Estimation Methods Bankfull Velocity Bankfull Discharge Manning Limerinos n 3.03 ft / sec 267.842 cfs Darcy‐Weisbach 3.206 ft / sec 282.992 cfs U/U* 2.91 ft / sec 257.04 cfs Stream Type (small with vegetative influence) 1.309 ft / sec 115.559 cfs

PARTICLE DISTRIBUTION The Web Soil Survey (2015) maps the riparian zone as a Reiner fine sandy loam, Smiley loam and Eckvoll loamy fine sand (Figure 96). The Smiley loam is poorly drained, while the other 2 loams are moderately well drained. The typical soil profile for the Reiner fine sandy loam is:

Ap ‐ 0 to 7 in.: fine sandy loam Bt ‐ 7 to 17 in.: clay loam Bw ‐ 17 to 21 in.: loam Bk – 21 to 35 in.: loam C ‐ 35 to 80 in.: loam

Very fine sand made up 17.48% of the full Figure 96. Soil map at site 10. Source: SSURGO soils and 2013 reach particle analysis, while clay and silt DNR aerial photo. made up 7.8% and all the gravel sizes combined made up 52.38% (Figure 97). The 50th percentile (D50) particle size was 3 mm, which is very fine gravel.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 76

20.00% Reach Particle Analysis ‐ Site 10 110.00% 18.00% 90.00% 16.00% (%) 14.00% 70.00%

(%) 12.00%

Volume 10.00% 50.00% 8.00% Volume 30.00% 6.00%

4.00% Cumulative 10.00% 2.00% 0.00% ‐10.00% 8 4 2 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 ‐ ‐ 2 ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 5.7 4 1.0 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 97. Particle analysis for site 10.

RIPARIAN ZONE A riparian vegetation worksheet was not completed at site 10. The site had a mix of grasses and some shrubs (Figure 98). Measuring off of aerial imagery, the width from bankfull to County Hwy 2 was about 47 ft. From the left bank to the adjacent field was about 35 ft.

Figure 98. Riparian zone along site 10. Photo taken on 6/29/11.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 77

CLASSIFICATION

Table 29. Stream classification input The Rosgen stream type through site 10 was classified as a B4c factors at site 1. (Table 29). The number four denotes the 50th percentile size of particles. With a D50 of 3mm, site 10 was considered a STREAM CLASSIFICATION ‐ Site 10 gravel stream. The lowercase c signifies that the energy slope STREAM TYPE B4c at this site is less than 0.02. The entrenchment ratios at both VALLEY TYPE XIIIb riffle cross‐sections were within the moderately entrenched category. B stream types usually have a width‐to‐depth (w/d) Slope 0.00065 ratio of greater than 12.0. Riffle 0+00 width‐to‐depth ratio was 9.08 and riffle 6+02 was 10.59. Sinuosity was 1.0. Entrenchment Ratio 1.81 The modified (by stream type) Pfankuch stability rating on 10/11/2012 was a 73. A good, or stable, rating range for a B4 Width/Depth Ratio 9.08 is 40‐64. A poor, or unstable, rating is 85 or greater. B stream types are moderately entrenched with a moderate gradient. Sinuosity 1.0 They usually are riffle dominated with infrequently spaced pools. B4 stream types are moderately sensitive to disturbance, but have an excellent recovery potential when the 3.0 mm Channel Material (D ) 50 (GRAVEL) disturbance is removed. This stream type is moderately dependent on robust stream bank vegetation to remain stable and the potential for streambank erosion is low.

DIMENSIONLESS RATIOS The upstream and downstream riffle had similar morphology, but the downstream riffle (6+02) was slightly larger than the upstream riffle (Table 30). Only 1 pool cross‐section was completed at this site. The mean pool depth‐to‐mean riffle depth ratio was 1.26, so the riffles were shallower than the pools. Table 30. Channel morphology and dimensionless ratios at site 10.

CHANNEL MORPHOLOGY & DIMENSIONLESS RATIOS Stream: Thief River Site: 10 Observers: Dave Friedl Date: 8/9/2011

Riffle Dimensions Mean Min Max Dimensionless Ratios Mean Min Max

Riffle Width (Wbkf) 30.26 28.34 32.18 Width of Flood‐Prone Area (Wpfa) 50.3 49.2 51.3

Mean Riffle Depth (dbfk) 3.083.043.12Riffle Width/Depth Ratio (Wbkf / dbkf) 9.84 9.08 10.59

Maximum Riffle Depth (dmax) 4.39 4.28 4.5 Max Riffle Depth to Mean Riffle Depth (dmax / dbkf) 1.431.411.44 2 Riffle Cross‐Sectional Area (Abkf) (ft ) 93.04 88.28 97.79 Entrenchment Ratio (Wfpa / Wbkf) 1.67 1.58 1.81

Pool Dimensions

Pool Width (Wbkfp) 28.79 28.79 28.79 Pool Width to Riffle Width (Wbkfp / Wbkf) 0.950.950.95

Mean Pool Depth (dbfkp) 3.893.893.89Mean Pool Depth to Mean Riffle Depth (dbkfp / dbkf)1.261.261.26

Maximum Pool Depth (dmaxp) 5.345.345.34Pool Area to Riffle Area (Abkfp / Abkf)1.21.21.2 2 Pool Cross‐Sectional Area (Abkfp) (ft ) 112 112 112 Max Pool Depth to Mean Riffle Depth (dmaxp / dbkf) 1.73 1.73 1.73

Thief River Watershed Fluvial Geomorphology Report  Appendix A 78

Site 12

INTRODUCTION Site 12 was located on the Moose River, adjacent to the Moose River Rd., in Section 5 of Northwood Township (Figure 99). According to StreamStats, the drainage area is 84.4 mi2. The subwatershed that drains to site 12 is outlined pink in Figure 99.

Figure 99. Watershed that drains to site 12.

The Moose River starts at the outlet of the north pool of the Moose River Impoundment, and once the Moose River Rd. begins, the Moose River flows adjacent to the road through Sections 1‐5 of Northwood Township. It is a channelized watercourse that is designated a perennial drainage ditch and a public watercourse (Figure 100).

Figure 100. Aerial view of site 12. Source: 2013 DNR aerial photo, 3m DEM, and ArcGIS Image Service best available resolution hillshade.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 79

LONGITUDINAL PROFILE The longitudinal profile at site 12 was approximately 480 feet, or 18 bankfull channel widths, in length (Figure 101). This site was located in a channelized reach, so the riffle and pool features were generally poorly defined. Both the bankfull and water surface slopes were derived from only three survey points, and were 0.00041 and 0.00145, respectively. Because of the large discrepancy in slopes, LiDAR was used to verify which slope was more accurate. The LiDAR derived slope values were 0.00014, 0.00031, and 0.00052, which lead to the bankfull slope (0.00041) being used for the velocity and discharge calculations.

Moose River: Site 12 Longitudinal Profile 2011

1180

1178 Thalweg

1176 Bankfull Slope = 0.00041 Water Surface 1174 GM12 Pool XS 1+62, 2011 GM12 Riffle XS 0+00, 2011 GM12 Riffle XS 0+91, 2011 1172 Bankfull GM12 Pool XS 1+62 SB, 2011

1170 P1

Elevation (ft) 1168

P2 1166

1164 P3

1162

P4 1160 0 100 200 300 400 500 Distance along stream (ft)

Figure 101. Longitudinal profile at site 12 in 2011.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 80

CROSS‐SECTIONS Two riffle cross‐sections were completed at site 12. Riffle 0+00 was used for the velocity and discharge calculations and had a bankfull area of 88.04 ft2 and a bankfull width of 26.22 ft. (Figure 102). The second riffle was located 91 ft. from the start of the longitudinal profile and had a bankfull area of 83.97 ft2 and a bankfull width of 23.74 ft.

The entrenchment ratio, comparing flood‐prone width to bankfull width, came out as slightly entrenched for both riffle cross‐sections. Riffle 0+00 had a flood‐prone width of 319 ft., giving it an entrenchment ratio of 12.2. Riffle 0+91 has a flood‐prone width of 317 ft., giving it a ratio of 13.4.

Moose River: Site 12 Riffle 0+00, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 26.2 Dbkf = 3.36 Abkf = 88 1190 Bankfull Width = 26.22 ft. 1188 Mean Depth = 3.36 ft. 1186 2 1184 Bankfull Area = 88.04 ft REP 1182 Flood‐prone Elevation 1180 EP 1178 1176 Bankfull Elevation 1174

Elevation (ft) 1172 1170 1168 1166 1164 1162 1160 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 Horizontal Distance (ft)

Figure 102. Riffle cross‐section 0+00 at site 12.

The D50, or 50th percentile, particle size was 0.16 mm. Very fine‐sized particles made up 46.53% of the particles measured at this cross‐section (Figure 103). Gravel, cobble and boulder‐sized particles were also tallied. No silt or clay was measured. The D84 used for the velocity and discharge calculations was 4.71 mm.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 81

50.00% Active Riffle Particle Analysis 0+00 110.00% 45.00% 90.00% 40.00% (%)

35.00% 70.00% (%)

30.00% Volume 25.00% 50.00% 20.00% Volume 30.00% 15.00% 10.00% 10.00% Cumulative 5.00% 0.00% ‐10.00% 8 2 4 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 ‐ ‐ 2 ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 5.7 1.0 4 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 103. Active riffle particle analysis at cross‐section 0+00.

One pool cross‐section (1+62) was completed at site 12. The bankfull area was 93.95 ft2 and the bankfull width was 25.33 ft. (Figure 104). Compared to riffle cross‐section 0+00, the pool was slightly narrower, but had a greater cross‐sectional area.

Moose River: Site 12 Pool 1+62, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 25.3 Dbkf = 3.71 Abkf = 94 1190 1188 Bankfull Width = 25.33 ft. 1186 Mean Depth = 3.71 ft. 1184 REP Bankfull Area = 93.95 ft2 1182 1180 EP 1178 1176 Bankfull Elevation Elevation (ft) 1174 1172 1170 TP 1168 1166 1164 1162 1160 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 Horizontal Distance (ft)

Figure 104. 2011 Pool cross‐section 1+62 at site 12.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 82

The study bank overlay at pool cross‐section 1+62 indicated an actual erosion rate of .2373 ft/yr from 2011 to 2012 (Figure 105). This rate closely matched the predicted erosion rate of .253 ft/yr derived from the BANCS model.

Moose River: Site 12 Pool Study Bank Overlay 2011-2012 Site 12 Pool 1+62 Study Bank, 2011 Site 12 Pool 1+62 Study Bank, 2012

1175

1174

1173

1172 Elevation (ft) 1171

1170 Toe Pin

1169

1168

1167 16 17 18 19 20 21 22 23

Horizontal Distance (ft)

Figure 105. Site 12 pool cross‐section 1+62 study bank overlay showing the study bank profile in 2011 and 2012.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 83

DISCHARGE TABLE The estimated velocities and discharges at riffle cross‐section 0+00 were very similar between different methods (Table 30). Generally the velocities were approximately 3.3 ft/sec and the estimated discharge was 270 cfs. StreamStats estimated the 1.5, 2, 5, 10, 25 and 50 recurrence interval discharges at 389, 554, 1030, 1370, 1810 and 2140 cfs, respectively.

Table 30. Velocity and discharge estimates for riffle cross section 0+00.

BANKFULL VELOCITY & DISCHARGE ESTIMATES Stream: Moose River Location: Site 12 Observers: Dave Friedl Date: 8/9/2011

Riffle Cross‐section: Riffle 0+00 Bankfull Slope: 0.00041 D84 of Active Bed: 4.71 Dia. (mm) Drainage Area: 174 Mi2

Estimation Methods Bankfull Velocity Bankfull Discharge Manning Limerinos n 3.276 ft / sec 275.086 cfs Darcy‐Weisbach 3.541 ft / sec 297.33 cfs U/U* 3.19 ft / sec 267.61 cfs Stream Type (small with vegetative influence) 1.177 ft / sec 98.833 cfs

PARTICLE DISTRIBUTION The Web Soil Survey (2015) maps the riparian zone as a Pengilly very fine sandy loam (Figure 106). This type of soil has a poorly drained natural drainage class. The typical profile for this soil type is:

A – 0 to 4 in.: very fine sandy loam Cg1..Cg4 – 4 to 60 in.: stratified loamy very fine sand to silt loam

The full reach particle analysis was dominated by sand‐sized particles (Figure 107). Sand made up 69.9% of the in‐ channel particles. Clay, silt and gravel were also tallied. The 50th percentile particle size Figure 106. Soil map at site 12. Source: SSURGO soils and 2013 DNR aerial photo. (D50) was 0.58 mm, which is coarse sand.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 84

25.00% Reach Particle Analysis ‐ Site 10 110.00%

90.00% 20.00% (%)

70.00% 15.00% (%)

Volume 50.00% 10.00% 30.00% Volume

5.00% 10.00% Cumulative

0.00% ‐10.00% 4 8 2 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 2 ‐ ‐ ‐ ‐ ‐ 180 256 362 512 128 ‐ ‐ ‐ ‐ 11.3 .125 22.6 ‐ 1024 2048 ‐ ‐ 5.7 1.0 4 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 107. Particle analysis for site 12.

RIPARIAN ZONE The vegetation along the river at site 12 was a mixture of small trees, shrubs and groundcover (Figure 108). The dominant herbaceous species present were 40% Phalaris arundinacea (reed canarygrass), 35% Bromus inermis (smooth brome), 15% Cirsium arvense (Canada thistle), 5% Asclepias spp. (milkweed), 3% Solidago spp. (goldenrod) and 5% Taraxacum officinale (common dandelion). Understory shrubs included Salix interior (sandbar wilow), Rosa spp. (Rose), Cornus sericea (redosier dogwood), and Rubus idaeus (American red raspberry). Figure 108. Site 12 vegetation cover. Photo taken on 5/3/2010. Tree species included Fraxinus pennsylvanica (green ash), Ulmus americana (American elm), Salix pentandra (laurel leaf willow), Acer neguno (box elder), Cornus spp. (dogwood) and Prunus virginiana (chokecherry).

Thief River Watershed Fluvial Geomorphology Report  Appendix A 85

CLASSIFICATION Table 31. Stream classification input The Rosgen stream type through site 12 was classified as an factors at site 12. E5 (Table 31). The number four denotes the 50th percentile size of particles. With a D50 of 0.58 mm, site 12 is considered STREAM CLASSIFICATION ‐ Site 12 a sand stream. The entrenchment ratio of both riffle cross‐ STREAM TYPE E5 sections fell within the slightly entrenched category. The VALLEY TYPE X width‐to‐depth ratios of cross sections 0+00 and 0+91 were 7.8 and 6.71, respectively. Sinuosity was 1.1. Slope 0.00041 The modified (by stream type) Pfankuch stability rating on 10/11/2012 was 111. A good, or stable, rating range for an E5 Entrenchment Ratio 12.17 is 50‐75. A poor, or unstable, rating is 97 or greater. E stream types are low gradient, meandering riffle/pool streams with a Width/Depth Ratio 7.8 low width‐to‐depth ratio. E5 stream types are highly sensitive to disturbance, but have a good recovery potential when the disturbance is removed. This stream type is also highly Sinuosity 1.1 dependent on robust stream bank vegetation to remain stable, and the potential for streambank erosion is high. 0.58 mm Channel Material (D ) 50 (SAND)

DIMENSIONLESS RATIOS The upstream and downstream riffle had similar morphology (Table 32). The upstream riffle (0+00) was slightly larger than the downstream riffle (0+91). Only 1 pool cross‐section was completed at this site. The mean pool depth‐to‐mean riffle depth ratio was 1.08, so the riffles were slightly shallower than the pools.

Table 32. Channel morphology and dimensionless ratios at site 12.

CHANNEL MORPHOLOGY & DIMENSIONLESS RATIOS Stream: Thief River Site: 12 Observers: Dave Friedl Date: 8/9/2011

Riffle Dimensions Mean Min Max Dimensionless Ratios Mean Min Max

Riffle Width (Wbkf) 24.98 23.74 26.22 Width of Flood‐Prone Area (Wpfa) 318 317 319

Mean Riffle Depth (dbfk) 3.45 3.36 3.54 Riffle Width/Depth Ratio (Wbkf / dbkf) 7.26 6.71 7.8

Maximum Riffle Depth (dmax) 5.54 5.31 5.77 Max Riffle Depth to Mean Riffle Depth (dmax / dbkf) 1.61 1.5 1.72 2 Riffle Cross‐Sectional Area (Abkf) (ft ) 86.01 83.97 88.04 Entrenchment Ratio (Wfpa / Wbkf) 12.76 12.17 13.35

Pool Dimensions

Pool Width (Wbkfp) 25.33 25.33 25.33 Pool Width to Riffle Width (Wbkfp / Wbkf) 1.01 1.01 1.01

Mean Pool Depth (dbfkp) 3.71 3.71 3.71 Mean Pool Depth to Mean Riffle Depth (dbkfp / dbkf) 1.08 1.08 1.08

Maximum Pool Depth (dmaxp) 6.19 6.19 6.19 Pool Area to Riffle Area (Abkfp / Abkf) 1.09 1.09 1.09 2 Pool Cross‐Sectional Area (Abkfp) (ft ) 93.95 93.95 93.95 Max Pool Depth to Mean Riffle Depth (dmaxp / dbkf) 1.79 1.79 1.79

Thief River Watershed Fluvial Geomorphology Report  Appendix A 86

Site 14

INTRODUCTION Site 14 was located on the Thief River, about four miles north of Thief River Falls,in Section three of North Township (Figure 109). According to StreamStats, the drainage area is 967 mi2. The subwatershed that drains to site 14 is outlined pink in Figure 109.

Figure 109. Watershed that drains to site 14.

The longitundinal profile was completed just upstream of 140th Ave NE (Figure 110). There is an USGS stream gage located about 1200 ft. upstream of the road. The river takes a more natural, meandering path through site 14 and is designated a perennial watercourse at this location.

Figure 110. Aerial view of site 1. Source: 2013 DNR aerial photo, 3m DEM, and ArcGIS Image Service best available resolution hillshade.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 87

LONGITUDINAL PROFILE The longitudinal profile at site 14 was approximately 3,240 feet, or 34 bankfull channel widths, in length (Figure 111). The thalweg, or channel bottom, through site 14 appeared to follow a normal pattern—the river became shallower through the glides and riffles and deeper in the pools. The pool near the beginning of the study reach contained the deepest hole, but it was not consistently deep, rather the bottom had some variability of depths. The downstream river bend was a “compound bend” with a separating riffle feature from about 2400 to 2450 ft. The bankfull slope was 0.00035 using just two bankfull calls, one at pool cross‐section 3+64 and the other at riffle cross‐section 21+07. LiDAR derived water surface slopes of 0.0004 and 0.00041 were in close agreement with the surveyed bankfull slope, so it was used for velocity and discharge estimates. Thief River: Site 14 Longitudinal Profile 2011

1130

1128 Thalweg

1126 Bankfull Slope = 0.00035 Water Surface

1124 GM14 Pool XS 3+64, 2011 GM14 Riffle XS 21+07, 2011 GM14 Pool XS 3+64 SB 2011

1122 Bankfull

1120 P1 Elevation (ft) 1118

P2 1116

1114 P3

1112 P4 1110 0 150 300 450 600 750 900 1050 1200 1350 1500 1650 1800 1950 2100 2250 2400 2550 2700 2850 3000 3150 3300 Distance along stream (ft) Figure 111. Longitudinal profile at site 14 in 2011.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 88

CROSS‐SECTIONS One riffle cross‐section was completed at site 14. The bankfull area was 423.85 ft2 and the bankfull width was 95.99 ft. (Figure 112). The entrenchment ratio, comparing flood‐prone width to bankfull width, came out as slightly entrenched at 2.22.

Thief River: Site 14 Riffle 21+07, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 96 Dbkf = 4.42 Abkf = 423.9 1135 Bankfull Width = 95.99 ft. 1133 Mean Depth = 4.42 ft. Bankfull Area = 423.85 ft2 1131

1129 Flood‐prone Elevation

1127LEP

1125

REP

Elevation (ft) 1123 Bankfull Elevation

1121

1119

1117

-45 -5-10-15-20-25-30-35-40 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100105110115120125130135140145 Horizontal Distance (ft) Figure 112. Riffle cross‐section 21+07 at site 14.

The D50, or 50th percentile, particle size was 9.03 mm. Silt and clay made up 9.84% of the particles measured at this cross‐section, but gravel‐sized particles were the most dominant category size tallied, making up 65.57% of the particles measured (Figure 113). The D84 used for the velocity and discharge calculations was 38.39 mm.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 89

14.00% Active Riffle Particle Analysis 21+07 110.00%

12.00% 90.00% (%) 10.00% 70.00% (%)

8.00% Volume 50.00% 6.00%

Volume 30.00% 4.00%

2.00% 10.00% Cumulative

0.00% ‐10.00% 2 4 8 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 ‐ ‐ 2 ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ 5.7 ‐ 1.0 4 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 113. Active riffle particle analysis at cross‐section 21+07.

One pool cross‐section (3+64) was completed at site 14.. The bankfull area was 473.5 ft2 and the bankfull width was 80.69 ft. (Figure 114). Compared to riffle cross‐section 21+07, the pool was narrower, but had a greater cross‐sectional area.

Thief River: Site 14 Pool 3+64, 2011 Ground Points Bankfull Indicators Water Surface Points Wbkf = 80.7 Dbkf = 5.87 Abkf = 473.5 1140

1138 Bankfull Width = 80.69 ft. 1136 Mean Depth = 5.87 ft. 1134 REP 2 1132 Bankfull Area = 473.5 ft

1130

1128 LEP 1126

1124 Bankfull Elevation

Elevation (ft) 1122

1120 TP 1118

1116

1114

1112

-85 -5-10-15-20-25-30-35-40-45-50-55-60-65-70-75-80 0 5 101520253035404550556065707580859095100105110115 Horizontal Distance (ft)

Figure 114. 2011 Pool cross‐section 3+64 at site 14.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 90

The study bank overlay at pool cross‐section 3+64 indicated an actual erosion rate of .4542 ft/yr from 2011 to 2012 (Figure 115). This rate was higher than the predicted erosion rate of .03 ft/yr derived from the BANCS model.

Thief River: Site 14 Pool Study Bank Overlay 2011-2012 Site 14 Pool 3+64 Study Bank, 2011 Site 14 Pool 3+64 Study Bank, 2012

1125

1124

1123

1122

1121 Elevation (ft) 1120

1119

Toe Pin 1118

1117

1116

1115 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92

Horizontal Distance (ft)

Figure 115. Site 14 pool cross‐section 3+64 study bank overlay showing the study bank profile in 2011 and 2012.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 91

DISCHARGE TABLE The estimated velocities and discharges at riffle cross‐section 3+64 were very similar using Manning Limerinos n, Darcy‐Weisbach and U/U* methods (Table 33). Generally the velocities were approximately 2.55 ft/sec and the estimated discharge was 1075 cfs. StreamStats estimated the 1.5, 2, 5, 10, 25 and 50 reoccurrence interval discharges at 1110, 1480, 2450, 3080, 3850, and 4380 cfs, respectively.

Table 33. Velocity and discharge estimates for riffle cross section 21+07.

BANKFULL VELOCITY & DISCHARGE ESTIMATES Stream: Thief River Location: Site 14 Observers: Dave Friedl Date: 10/04/2011

Riffle Cross‐section: Riffle 3+64 Bankfull Slope: 0.00035 D84 of Active Bed: 67.19 Dia. (mm) Drainage Area: 967 Mi2

Estimation Methods Bankfull Velocity Bankfull Discharge Manning Limerinos n 2.597 ft / sec 1100.739 cfs Darcy‐Weisbach 2.612 ft / sec 1106.89 cfs U/U* 2.49 ft / sec 1056.68 cfs Stream Type (medium to large ‐ C4) 3.813 ft / sec 1616.14 cfs

PARTICLE DISTRIBUTION The Web Soil Survey (2015) maps the riparian zone at site 14 as a Fluvaquents, frequently flooded‐Hapludolls complex (map unit I16F) (Figure 116). The typical profile for this map unit is a fine sandy loam (A horizon) from 0 to 16 inches. The 16 to 80 inches of the profile is a Cg horizon with stratified loamy sand to silt loam. A C soil horizon contains glacial or post‐glacial material with little affected by soil forming processes. A lowercase g denotes the presents of gray colors, indicative of permanent or periodic intensive reduction due to saturation. The adjacent soil types were clay or clay loam. Figure 116. Soil map at site 14. Source: SSURGO soils and 2013 FSA aerial photo.

Clay and silt‐sized particles were the most abundant sediment size tallied during the full study reach pebble count (Figure 117). Sand and gravel made up 31.67% and 45.42%, respectively. The 50th percentile particle size (D50) was 3.45 mm, which is gravel. Site 14 had the largest particles

Thief River Watershed Fluvial Geomorphology Report  Appendix A 92

tallied, compared to the other study sites. The largest particle size tallied over 1024 mm, which is considered a boulder‐sized particle.

14.00% Reach Particle Analysis ‐ Site 14 110.00%

12.00% 90.00% (%) 10.00% 70.00% 8.00% (%)

Volume 50.00% 6.00% 30.00% Volume 4.00%

2.00% 10.00% Cumulative

0.00% ‐10.00% 8 4 2 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 ‐ ‐ 2 ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 5.7 4 1.0 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 117. Particle analysis for site 14.

RIPARIAN ZONE The riparian zone at site 14 contained trees, shrubs, and groundcover (Figure 118). The tree species present included Quercus macrocarpa (bur oak), Fraxinus pennsylvanica (green ash), Acer neguno (box elder), Salix spp. (willow species), Populus deltoides (cottonwood) and Picea spp. (spruce species). The shrub species present included Lonicera spp. (honeysuckle) and Crataegus spp. (Hawthorne).

The herbaceous layer included Phalaris arundinacea (reed canarygrass), Asclepias Figure 118. Site 14 vegetation cover. Photo taken on 10/4/2011. spp. (milkweed), Cirsium arvense (Canada thistle), Solidago spp. (goldenrod), Elymus repens (quackgrass), Xanthium strumarium (Common cocklebur).

Thief River Watershed Fluvial Geomorphology Report  Appendix A 93

CLASSIFICATION Table 34. Stream classification input factors The Rosgen stream type through site 14 was classified as a at site 14. C4c‐ (Figure 34). The number four denotes the 50% STREAM CLASSIFICATION ‐ Site 14 percentile size of particles. With a D50 of 3.45 mm, site 14 is considered a gravel stream. The lowercase c‐ signifies STREAM TYPE C4c‐ that the energy slope at this site is less than 0.001. The VALLEY TYPE X entrenchment ratio was just over the threshold into a C stream type at 2.22. C stream types usually have an Slope 0.00035 entrenchment ratio greater than 2.2 and a width‐to‐depth (w/d) ratio of greater than 12.0. The w/d ratio at riffle 21+07 was 21.72. Sinuosity was 2.85. Entrenchment Ratio 2.22

The modified (by stream type) Pfankuch stability rating on 10/12/2012 was an 86. A good, or stable, rating range for Width/Depth Ratio 21.72 a C4 is 70‐90. C stream types are slightly entrenched with a low gradient. Their channels are cut by movement of Sinuosity 2.85 sediment by water and have broad, well‐defined floodplains. C4 stream types have very high sensitivity to disturbance, but have a good recovery potential when the 3.45 mm Channel Material (D ) disturbance is removed. This stream type is highly 50 (GRAVEL) dependent on robust stream bank vegetation to remain stable and the potential for streambank erosion is very high.

DIMENSIONLESS RATIOS Only one riffle and one pool cross‐sections were completed at site 14; therefore, there was no variation between the mean, minimum and maximum values (Table 35). Comparing the riffle to the pool, the mean pool depth‐to‐mean riffle depth ratio was 1.328, so the riffle was shallower than the pool. The pool had a greater cross‐sectional area, but a smaller width, compared to the riffle. Table 35. Channel morphology and dimensionless ratios for site 14.

CHANNEL MORPHOLOGY & DIMENSIONLESS RATIOS Stream: Thief River Site: 14 Observers: Dave Friedl Date: 10/04/2011

Riffle Dimensions Mean Min Max Dimensionless Ratios Mean Min Max

Riffle Width (Wbkf) 969696Width of Flood‐Prone Area (Wpfa) 213 213 51.3

Mean Riffle Depth (dbfk) 4.42 4.42 4.42 Riffle Width/Depth Ratio (Wbkf / dbkf) 21.72 21.72 21.72

Maximum Riffle Depth (dmax) 5.38 5.38 5.38 Max Riffle Depth to Mean Riffle Depth (dmax / dbkf) 1.217 1.217 1.217 2 Riffle Cross‐Sectional Area (Abkf) (ft ) 423.85 423.85 423.85 Entrenchment Ratio (Wfpa / Wbkf) 2.219 2.219 2.219

Pool Dimensions

Pool Width (Wbkfp) 80.7 80.7 80.7 Pool Width to Riffle Width (Wbkfp / Wbkf) 0.841 0.841 0.841

Mean Pool Depth (dbfkp) 5.87 5.87 5.87 Mean Pool Depth to Mean Riffle Depth (dbkfp / dbkf) 1.328 1.328 1.328

Maximum Pool Depth (dmaxp) 7.78 7.78 7.78 Pool Area to Riffle Area (Abkfp / Abkf) 1.117 1.117 1.117 2 Pool Cross‐Sectional Area (Abkfp) (ft ) 474 474 474 Max Pool Depth to Mean Riffle Depth (dmaxp / dbkf) 1.76 1.76 1.76

Thief River Watershed Fluvial Geomorphology Report  Appendix A 94

Site 15

INTRODUCTION Site 15 was located in Section 8 of Whiteford Township, just north of Aggasiz NWR (Figure 119). Sites 1, 2 and 3 were located upstream of 15, between Thief Lake and Aggasiz NWR.

Figure 119. . Watershed that drains to site 15.

The study reach at site 15 started in a natural meander and transitioned into a straightened ditch. The reach of the Thief River is designated as a perennial stream. Based on aerial imagery (Figure 120), the river appears to flow within its natural floodplain, but was previously straightened and the adjacent spoil piles prevent the river from flowing within its historic meanders.

Figure 120. Aerial view of site 15. Source: 2013 DNR aerial photo, 3m DEM, and ArcGIS Image Service best available resolution hillshade.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 95

LONGITUDINAL PROFILE The longitudinal profile does show a few deeper locations, but overall, there was little variation is the depths along site 15 (Figure 121). The water surface slope, using water surface survey points, was 0.00039. The bankfull slope was created by interpolating a bankfull call at the bottom riffle cross‐section. Using LiDAR and ArcGIS, water surface slopes were also estimated at 0.00012 and 0.00086. The velocity and discharge calculations did not seem realistic using the higher slopes, so a slope of 0.0001 was used, based on the LiDAR data.

Thief River: Site 15 Longitudinal Profile, 2011 1150 Water Surface Slope = 0.00044 Water Surface Slope = 0.00044 Thalweg

1145 Water Sfc Water Surface Slope = 0.00039 Bankfull Slope = 0.00039 Bankfull 1140 P1 Elevation (ft) P2 1135 P3

1130 P4 0 250 500 750 1000 Distance along stream (ft)

Figure 121. Longitudinal profile at site 15.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 96

CROSS‐SECTIONS Two riffle cross‐sections were completed at site 15, at 0+16 and 10+03. The upstream riffle had a well‐defined bankfull bench at 5.42 ft. above the water surface (Figure 122). The downstream riffle had a more trapezoidal shape and it was difficult to determine bankfull. A bankfull elevation was interpolated 5.40 ft. above the water surface at riffle 10+03. The two riffle cross‐sections had similar morphology. The widths of cross‐sections 0+16 and 10+03 were 51.95 ft. and 47.58 ft., respectively. The bankfull areas were 231.18 ft2 and 239.77 ft2, and the mean depths were 4.45 ft. and 5.04 ft.

Thief River: Site 15 Riffle 0+16, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 51.9 Dbkf = 4.45 Abkf = 231.2 1160 1158 Bankfull Width = 51.95 ft. 1156 Mean Depth = 4.45 ft. Flood‐prone Elevation 1154 Bankfull Area = 231.18 ft2 1152 1150 1148 Bankfull Elevation 1146 1144 1142 Elevation (ft) 1140 1138 1136 1134 1132 1130 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 Horizontal Distance (ft)

Figure 122. Upstream riffle 0+16 at site 15.

Two sizes of particles were dominant in the riffle pebble count. Clay and silt‐sized particles were tallied 21.9% and very coarse sand (1‐2 mm) was tallied 20% (Figure 123). Various sized sand and gravel particles were also tallied. No cobble or boulder was present at this riffle. The D84 at riffle 0+16 was 12.24 mm. The largest particle measured was 64 mm. The pool cross‐section pebble count at 1+60 was very similar to this riffle count. The D84 at pool 1+60 was 12.13 mm and the D100 was 64 mm.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 97

25.00% Active Riffle Particle Analysis 0+16 110.00%

90.00% 20.00% (%)

70.00% 15.00% 50.00% (%)

Volume

10.00% 30.00%

5.00% Volume 10.00%

0.00% ‐10.00% Cumulative 4 2 8 ‐ ‐ ‐ 16 32 45 64 90 .25 .50 1.0 5.7 2 ‐ ‐ ‐ ‐ ‐ 128 180 256 362 512 ‐ ‐ ‐ ‐ .125 11.3 22.6 ‐ 1024 2048 ‐ ‐ 1.0 4 5.7 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 123. Pebble count results for riffle cross‐section 0+16.

One pool cross‐section was measured at site 15. Compared to the riffle at 0+16, the pool width was slightly narrower at 43.41 ft., but the cross‐sectional area was slightly larger at 248.49 ft. (Figure 124). The mean depth was 5.72 ft. and the maximum depth was 8.19 ft.

Thief River: Site 15 Pool 1+60, 2011 Ground Points Bankfull Indicators Water Surface Points

Wbkf = 43.4 Dbkf = 5.72 Abkf = 248.5 1160 1158 Bankfull Width = 43.41 ft. 1156 Mean Depth = 5.72 ft. 1154 Bankfull Area = 248.49 ft2 1152 1150 1148 Bankfull Elevation 1146 1144 1142 Elevation (ft) 1140 1138 1136 1134 1132

-15-10 -5 0 5 1015202530354045505560657075808590 Horizontal Distance (ft)

Figure 124. Pool cross‐section 1+60.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 98

The study bank overlay at pool cross‐section 1+60 indicated an actual erosion rate of .0046 ft/yr from 2011 to 2012 (Figure 125). This rate closely matched the predicted erosion rate of .03 ft/yr derived from the BANCS model.

Thief River: Site 15 Pool Study Bank Overlay 2011-2012 Site 15 Pool 1+60 Study Bank, 2011 Site 15 Pool 1+60 Study Bank, 2012

1148

1147

1146

1145

1144 Elevation (ft) Elevation

1143

1142

1141 Toe Pin

1140 59 60 61 62 63 64 65 66 67 68 69 70

Horizontal Distance (ft)

Figure 125. Site 15 pool cross‐section 1+60 study bank overlay showing the study bank profile in 2011 and 2012.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 99

DISCHARGE TABLE Velocity and discharge estimates were calculated using riffle cross‐section 0+16 (Table 36). The D84 was 12.24mm. The bankfull slope was 0.00039 and the water surface slope was 0.00044. These slopes clearly over‐estimated velocity and discharge, with velocities near 3.4 ft/sec and discharge near 780 cfs. Using LiDAR data a water surface slope was estimated at 0.0001. This slope produced more realistic flow estimates.

Table 36. Bankfull velocity and discharge estimates for site 15.

BANKFULL VELOCITY & DISCHARGE ESTIMATES Stream: Thief River Location: Site 15 Observers: Dave Friedl Date: 9/9/2011

Riffle Cross‐section: Riffle 0+16 Bankfull Slope: 0.0001 D84 of Active Bed: 12.24 Dia. (mm) Drainage Area: 258 Mi2

Estimation Methods Bankfull Velocity Bankfull Discharge Manning Limerinos n 1.713 ft / sec 396.011 cfs Darcy‐Weisbach 1.744 ft / sec 403.142 cfs U/U* 1.66 ft / sec 383.53 cfs Stream Type (medium to large) 1.21 ft / sec 279.728 cfs

PARTICLE DISTRIBUTION The Web Soil Survey (2015) maps the riparian zone as a Fluvaquents, frequently flooded‐Hapludolls complex (map unit I16F) (Figure 126). The typical profile for this map unit is a fine sandy loam (A soil horizon) from 0 to 16 inches. The 16 to 80 inches of the profile is a Cg horizon with stratified loamy sand to silt loam. A C soil horizon contains glacial or post‐glacial material with little affected by soil forming processes. A lowercase g denotes the presents of gray Figure 126. Soil map at site 15. Source: SSURGO Soils and 2013 DNR aerial photo.

Thief River Watershed Fluvial Geomorphology Report  Appendix A 100

colors, indicative of permanent or periodic intensive reduction due to saturation.

The full reach particle analysis was dominated by clay/silt and very fine sand‐sized particles, making up 25.93% of the count (Figure 127). An assortment of gravel sizes was also present. At site 15, the D50 was 1.5 mm, which is very coarse sand. Most of the particles were a medium‐sized gravel and smaller.

16.00% Reach Particle Analysis ‐ Site 15 110.00% 14.00% 90.00% (%) 12.00% 70.00% 10.00% (%)

Volume 8.00% 50.00% 6.00% 30.00% Volume 4.00% 10.00% 2.00% Cumulative 0.00% ‐10.00% 4 8 2 ‐ ‐ ‐ 16 32 45 64 90 5.7 .25 .50 1.0 2 ‐ ‐ ‐ ‐ ‐ 180 256 362 512 128 ‐ ‐ ‐ ‐ 11.3 .125 22.6 ‐ 1024 2048 ‐ ‐ 5.7 4 1.0 ‐ <.062 32 45 64 ‐ 8 .25 .50 90 11.3 16 22.6 .125 ‐ 128 ‐ 180 ‐ 256 ‐ 362 .062 ‐ 512 1024 SAND GRAVEL COBBLE BOULDER Particle Size (mm)

Figure 127. Reach particle analysis at site 15.

RIPARIAN ZONE The land adjacent to Thief River at site 15 was a mix of tree and shrub canopies and an herbaceous layer (Figure 128). The tree canopy covered 50% of the aerial canopy and 30% of the basal cover. Species included Salix nigra (black willow), Fraxinus pennsylvanica (green ash), Ulmus americana (American elm) and Acer spp. (maple). Understory shrubs covered 10% of the ground cover and included Cornus sericea (redosier dogwood) and Rosa spp. (wild rose). The dominant grass and forbs present were Phalaris arundinacea (reed canarygrass), Parthenocissus quinquefolia (Virginia Figure 128. Riparian zone at site 15. Photo taken on 9/9/2011. creeper), Equisetum spp. (horsetail) and Sagittaria latifolia (arrowhead).

Thief River Watershed Fluvial Geomorphology Report  Appendix A 101

CLASSIFICATION Table 37. Stream classification input variables for The Rosgen stream type through site 15 was classified site 15. as an E5 (Table 37). The number 5 denotes the 50th percentile size of particles. A D50 of 1.5 mm is very STREAM CLASSIFICATION ‐ Site 15 coarse sand. It was slightly entrenched with an STREAM TYPE E5 entrenchment ratio of 96.25. The width‐to‐depth ratio was 11.67. Sinuosity was 1.1. VALLEY TYPE X

The modified (by stream type) Pfankuch stability rating Slope 0.0001 on 10/24/2012 was a 92. A good, or stable, rating range for an E5 stream type is 50‐75. A fair rating is 76‐96. Entrenchment Ratio 96.25

E5 stream types are highly sensitive to disturbance, but they have a good potential for recovery when the Width/Depth Ratio 11.67 disturbance is removed. They have a high potential for streambank erosion and are very dependent on vegetation to limit streambank erosion. Sinuosity 1.1

1.5 mm Channel Material (D50) (SAND)

DIMENSIONLESS RATIOS Riffle cross‐sections 0+16 and 1+60 had very similar ratios (Table 38). The downstream riffle had a slightly narrower width, but a slightly larger cross‐sectional area. The pool cross‐section was narrower than both riffle cross‐sections, but had a larger cross‐sectional area. Table 38. Channel morphology and dimensionless ratios for site 15.

CHANNEL MORPHOLOGY & DIMENSIONLESS RATIOS Stream: Thief River Site: 15 Observers: Dave Friedl Date: 9/9/2011

Riffle Dimensions Mean Min Max Dimensionless Ratios Mean Min Max

Riffle Width (Wbkf) 49.77 47.58 51.95 Width of Flood‐Prone Area (Wpfa) 500050005000

Mean Riffle Depth (dbfk) 4.754.455.04Riffle Width/Depth Ratio (Wbkf / dbkf) 10.56 9.44 11.67

Maximum Riffle Depth (dmax) 6.556.426.67Max Riffle Depth to Mean Riffle Depth (dmax / dbkf) 1.38 1.32 1.44 2 Riffle Cross‐Sectional Area (Abkf) (ft ) 235.48 231.18 239.77 Entrenchment Ratio (Wfpa / Wbkf) 100.67 96.25 105.09

Pool Dimensions

Pool Width (Wbkfp) 43.41 43.41 43.41 Pool Width to Riffle Width (Wbkfp / Wbkf) 0.87 0.87 0.87

Mean Pool Depth (dbfkp) 5.725.725.72Mean Pool Depth to Mean Riffle Depth (dbkfp / dbkf) 1.2 1.2 1.2

Maximum Pool Depth (dmaxp) 8.198.198.19Pool Area to Riffle Area (Abkfp / Abkf) 1.01 1.01 1.01 2 Pool Cross‐Sectional Area (Abkfp) (ft ) 248.49 248.49 248.49 Max Pool Depth to Mean Riffle Depth (dmaxp / dbkf) 1.724 1.724 1.724

Thief River Watershed Fluvial Geomorphology Report  Appendix A 102

APPENDIX B PFANKUCH STABILITY FORM

APPENDIX C NEAR‐CHANNEL STREAM POWER INDEX

NEAR‐CHANNEL STREAM POWER INDEX In several locations within the Thief River Watershed, identifying upland sources of sediment is a recommendation to reduce sediment loading to the watercourses. This figure is an example of identifying near‐channel gully erosion, using LiDAR data.

A flow accumulation raster layer was created from a 5‐meter DEM, derived from the LiDAR data. A SPI raster layer was also created. A 200 foot buffer zone was applied to areas where the flow accumulation raster values were between 1 square mile and 200 square miles. The SPI values were then clipped to this 400 foot wide zone near watercourses. The raster cells making up individual SPI lines were grouped together and given one SPI value, which was equal to the mean of the group of cells. A higher mean SPI value means there is the more drainage area draining to that point and/or the steeper the slope of the cells. Information like this layer should be able to assist with targeting and prioritizing areas for near‐channel sediment erosion.