Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Prepared in cooperation with City of Tuscaloosa

Scientific Investigations Report 2013–5152

U.S. Department of the Interior U.S. Geological Survey Cover: Aerial photograph of Carroll Creek, a tributary to Lake Tuscaloosa. Photograph taken by Scott Sanderford, Lakes Manager, City of Tuscaloosa. Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

By K.G. Lee

Prepared in cooperation with the City of Tuscaloosa

Scientific Investigations Report 2013–5152

U.S. Department of the Interior U.S. Geological Survey U.S. Department of the Interior SALLY JEWELL, Secretary U.S. Geological Survey Suzette M. Kimball, Director

U.S. Geological Survey, Reston, Virginia: 2013

For more information on the USGS—the Federal source for science about the Earth, its natural and living resources, natural hazards, and the environment, visit http://www.usgs.gov or call 1–888–ASK–USGS. For an overview of USGS information products, including maps, imagery, and publications, visit http://www.usgs.gov/pubprod To order this and other USGS information products, visit http://store.usgs.gov

Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

Although this report is in the public domain, permission must be secured from the individual copyright owners to reproduce any copyrighted materials contained within this report.

Suggested citation: Lee, K.G., 2013, Estimation of sediment inflows to Lake Tuscaloosa, Alabama, 2009–11: U.S. Geological Survey Scientific Investigations Report 2013–5152, 65 p., http://pubs.usgs.gov/sir/2013/5152. iii

Contents

Abstract...... 1 Introduction...... 1 Purpose and Scope...... 3 Acknowledgments...... 3 Description of Study Area...... 3 Previous Investigations...... 4 Approach and Methods...... 6 Streamflow Data Collection...... 6 Suspended-Sediment Data Collection...... 6 Data Analysis...... 8 North River...... 10 Land Use...... 10 Historical Data...... 10 Data Collected...... 15 Streamflow...... 15 Suspended Sediment...... 15 Estimation of North River Suspended-Sediment Loads...... 15 Turkey Creek...... 18 Land Use...... 18 Historical Data...... 18 Data Collected...... 18 Streamflow...... 18 Suspended Sediment...... 26 Estimation of Suspended-Sediment Loads from Turkey Creek...... 26 Binion Creek...... 28 Land Use...... 28 Historical Data...... 28 Data Collected...... 28 Streamflow...... 28 Suspended Sediment...... 33 Estimation of Suspended-Sediment Loads from Binion Creek...... 33 Pole Bridge Creek...... 34 Land Use...... 36 Data Collected...... 36 Streamflow...... 36 Suspended Sediment...... 38 Estimation of Suspended-Sediment Loads from Pole Bridge Creek...... 38 Tierce Creek...... 40 Land Use...... 41 Historical Data...... 41 Data Collected...... 41 Streamflow...... 42 Suspended Sediment...... 44 iv

Estimation of Suspended-Sediment Loads from Tierce Creek...... 44 Carroll Creek...... 45 Land Use...... 45 Historical Data...... 48 Data Collected...... 50 Streamflow...... 50 Suspended Sediment...... 50 Estimation of Suspended-Sediment Loads from Carroll Creek...... 50 Brush Creek...... 51 Land Use...... 51 Historical Data...... 51 Data Collected...... 51 Streamflow...... 51 Suspended Sediment...... 57 Estimation of Suspended-Sediment Loads from Brush Creek...... 57 Estimation of Suspended-Sediment Loads to Lake Tuscaloosa...... 58 Summary...... 61 References Cited...... 63

Figures

1. Location of Lake Tuscaloosa contributing subwatersheds and tributaries, Fayette and Tuscaloosa Counties, Alabama...... 2 2. National Land Cover Database (NLCD) land use in the Lake Tuscaloosa watershed in 2006, Tuscaloosa and Fayette Counties, Alabama...... 5 3. Location of sampling sites and contributing drainage area at U.S. Geological Survey streamflow-gaging stations in the Black Warrior-North River watershed, Alabama...... 7 4. A diagram of the equal-width-increment sampling method used to collect sediment samples in tributaries to Lake Tuscaloosa, Alabama...... 8 5. Location of U.S. Geological Survey streamflow-gaging station 02464000 North River near Samantha, Alabama, and the contributing drainage area...... 11 6. Location of the North River in Alabama at Bull Slough Road as shown in aerial photography dated February 12, 1974 and September 29, 2011...... 12 7. Locations of sediment deposition near the upstream extent of the backwater of Lake Tuscaloosa on the North River, Alabama. Aerial photography dated 2009 shows sediment deposition to the northwest and southeast of Bull Slough Road...... 13 8. Percentage of land-use coverage at the U.S. Geological Survey streamflow- gaging station 02464000 North River near Samantha, Alabama. National Land Cover Database 2001and National Land Cover Database 2006...... 14 9. Annual mean streamflow for the U.S. Geological Survey streamflow-gaging station 02464000 North River near Samantha, Alabama...... 16 10. Flow duration curve and sampled streamflow for the U.S. Geological Survey streamflow-gaging station 02464000 North River near Samantha, Alabama, for 1940–54 and 1969–2011...... 16 v

11. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464000 North River near Samantha, Alabama...... 17 12. Location of U.S. Geological Survey streamflow-gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama, and the contributing drainage area...... 19 13. The upstream extent of the backwater of Lake Tuscaloosa on Turkey Creek, Alabama, is shown in aerial photography dated February 12, 1974, and September 29, 2011...... 20 14. Aerial photograph taken in 2009 shows sediment deposition near the upstream extent of the backwater of Lake Tuscaloosa on Turkey Creek, Alabama...... 21 15. Percentage of land-use coverage at the U.S. Geological Survey streamflow- gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama. National Land Cover Database 2001 and National Land Cover Database 2006...... 22 16. Streamflow hydrograph for the 2009 water year at U.S. Geological Survey streamflow-gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama...... 24 17. Annual mean streamflow for U.S. Geological Survey streamflow-gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama...... 24 18. Flow duration curve and sampled streamflow for U.S. Geological Survey streamflow-gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama, 1982–2011...... 25 19. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama...... 27 20. Location of the U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama, and the contributing drainage area...... 29 21. Percentage of land-use coverage at the U.S. Geological Survey streamflow- gaging station 02464360 Binion Creek near Samantha, Alabama. National Land Cover Database 2001 and National Land Cover Database 2006...... 30 22. Streamflow hydrograph for the 2010–11 water years at the U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama...... 31 23. Annual mean streamflow for U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama...... 32 24. Flow duration curve and sampled streamflow for U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama, 1987–2011...... 32 25. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama...... 34 26. Location of U.S. Geological Survey streamflow-gaging station 02464385 Pole Bridge Creek near Samantha, Alabama, and the contributing drainage area...... 35 27. Percentage of land-use coverage at U.S. Geological Survey streamflow- gaging station 02464385 Pole Bridge Creek near Samantha, Alabama. National Land Cover Database 2001 and National Land Cover Database 2006...... 36 28. Streamflow hydrograph for the 2010–11 water years at the U.S. Geological Survey streamflow-gaging station 02464385 Pole Bridge Creek near Samantha, Alabama...... 37 29. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464385 Pole Bridge Creek near Samantha, Alabama...... 39 30. Location of U.S. Geological Survey streamflow-gaging station 02464503 Tierce Creek near Northport, Alabama, and the contributing drainage area...... 40 vi

31. Percentage of land-use coverage at U.S. Geological Survey streamflow-gaging station 02464503 Tierce Creek near Northport, Alabama. National Land Cover Database 2001 and National Land Cover Database 2006...... 41 32. Streamflow hydrograph for the 2010–11 water years at U.S. Geological Survey streamflow-gaging station 02464503 Tierce Creek near Northport, Alabama...... 43 33. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464503 Tierce Creek near Northport, Alabama...... 45 34. Location of U.S. Geological Survey streamflow-gaging station 02464660 Carroll Creek near Northport, Alabama, and the contributing drainage area...... 46 35. Aerial photograph showing the upstream extent of the backwater portion of Carroll Creek near Northport, Alabama, November 29, 2007. Lake Tuscaloosa water-surface elevation is 221.04 feet (National Geodetic Vertical Datum of 1929)...... 47 36. Aerial photograph showing the upstream extent of the backwater portion of Carroll Creek near Northport, Alabama, March 24, 2009. Lake Tuscaloosa water-surface elevation is 223.74 feet (National Geodetic Vertical Datum of 1929)...... 47 37. Percentage of land-use coverage at U.S. Geological Survey streamflow-gaging station 02464660 Carroll Creek near Northport, Alabama. National Land Cover Database 2001 and National Land Cover Database 2006...... 48 38. Comparison of aerial photography (2009) and National Land Cover Database Impervious Cover (2006) for the Carroll Creek subwatershed draining into Lake Tuscaloosa, Alabama...... 49 39. Streamflow hydrograph for the 2009 water year at U.S. Geological Survey streamflow-gaging station 02464660 Carroll Creek near Northport, Alabama...... 52 40. Annual mean streamflow for U.S. Geological Survey streamflow-gaging station 02464660 Carroll Creek near Northport, Alabama...... 52 41. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464660 Carroll Creek near Northport, Alabama...... 53 42. Location of U.S. Geological Survey streamflow-gaging station 02464680 Brush Creek near Northport, Alabama, and the contributing drainage area...... 54 43. Percentage of land-use coverage at U.S. Geological Survey streamflow-gaging station 02464680 Brush Creek near Northport, Alabama. National Land Cover Database 2001 and National Land Cover Database 2006...... 55 44. Streamflow hydrograph for the 2011–12 water years at U.S. Geological Survey streamflow-gaging station 02464680 Brush Creek near Northport, Alabama...... 56 45. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464680 Brush Creek near Northport, Alabama...... 58 vii

Tables

1. Hydrologic Unit Code (HUC) designations for the drainage area contributing to Lake Tuscaloosa in the Upper Black Warrior subbasin, Alabama...... 3 2. Historical streamflow and suspended-sediment data for U.S. Geological Survey streamflow-gaging station 02464000 North River near Samantha, Alabama...... 14 3. Summary of suspended-sediment samples collected at U.S. Geological Survey streamflow-gaging station 02464000 North River near Samantha, Alabama...... 17 4. Historical streamflow and suspended-sediment data for U.S. Geological Survey streamflow-gaging stations 02464145 and 02464149 Turkey Creek near Tuscaloosa, Alabama...... 23 5. Summary of suspended-sediment samples collected at U.S. Geological Survey streamflow-gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama...... 26 6. Historical streamflow and suspended-sediment data for U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama...... 30 7. Summary of suspended-sediment samples collected at U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama...... 33 8. Summary of suspended-sediment samples collected at U.S. Geological Survey streamflow-gaging station 02464385 Pole Bridge Creek near Samantha, Alabama...... 38 9. Historical streamflow and suspended-sediment data for U.S. Geological Survey streamflow-gaging station 02464503 Tierce Creek near Northport, Alabama...... 42 10. Summary of suspended-sediment samples collected at U.S. Geological Survey streamflow-gaging station 02464503 Tierce Creek near Northport, Alabama...... 44 11. Historical streamflow and suspended-sediment data for U.S. Geological Survey streamflow-gaging station 02464660 Carroll Creek near Northport, Alabama...... 49 12. Summary of suspended-sediment samples collected at U.S. Geological Survey streamflow-gaging station 02464660 Carroll Creek near Northport, Alabama...... 53 13. Historical streamflow and suspended-sediment data for U.S. Geological Survey streamflow-gaging station 02464680 Brush Creek near Northport, Alabama...... 55 14. Summary of suspended-sediment samples collected at U.S. Geological Survey streamflow-gaging station 02464680 Brush Creek near Northport, Alabama...... 57 15. Summary of suspended-sediment data collection, transport curves, and loads for seven tributaries to the Lake Tuscaloosa reservoir, Alabama...... 59 16. Estimated suspended-sediment load for seven tributaries to the Lake Tuscaloosa reservoir, Alabama...... 60 viii

Conversion Factors Inch/Pound to SI

Multiply By To obtain Length inch (in.) 2.54 centimeter (cm) inch (in.) 25.4 millimeter (mm) foot (ft) 0.3048 meter (m) mile (mi) 1.609 kilometer (km) Area acre 4,047 square meter (m2) acre 0.4047 hectare (ha) acre 0.4047 square hectometer (hm2) acre 0.004047 square kilometer (km2) square foot (ft2) 929.0 square centimeter (cm2) square foot (ft2) 0.09290 square meter (m2) square mile (mi2) 259.0 hectare (ha) square mile (mi2) 2.590 square kilometer (km2) Volume cubic foot (ft3) 28.32 cubic decimeter (dm3) cubic foot (ft3) 0.02832 cubic meter (m3) cubic yard (yd3) 0.7646 cubic meter (m3) acre-foot (acre-ft) 1,233 cubic meter (m3) acre-foot (acre-ft) 0.001233 cubic hectometer (hm3) Flow rate cubic foot per second (ft3/s) 0.02832 cubic meter per second (m3/s) cubic foot per second per square 0.01093 cubic meter per second per mile [(ft3/s)/mi2] square kilometer [(m3/s)/km2] Mass ton, short (2,000 lb) 0.9072 megagram ton per day (ton/d) 0.9072 metric ton per day ton per year (ton/yr) 0.9072 metric ton per year SI to Inch/Pound Multiply By To obtain Length millimeter (mm) 0.03937 inch (in.)

Temperature in degrees Fahrenheit (°F) may be converted to degrees Celsius (°C) as follows: °C=(°F–32)/1.8 A water year is the 12-month period October 1 through September 30 designated by the calendar year in which it ends. Datums Elevation refers to distance above the vertical datum based on the National Geodetic Vertical Datum of 1929 (NGVD 29). Concentrations of chemical constituents in water are given in milligrams per liter (mg/L).

Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

By K.G. Lee

Abstract and historical suspended-sediment samples to estimate a long- term average suspended-sediment yield of 300 tons per year The U.S. Geological Survey, in cooperation with the per square mile in the North River watershed. Analysis of data City of Tuscaloosa, evaluated the concentrations, loads, collected in the North River watershed during the 2009 water year (October 2008 to September 2009) estimated a sediment yield of and yields of suspended sediment in the tributaries to Lake 2 Tuscaloosa in west-central Alabama, from October 1, 2008, 360 tons/mi . The North River watershed, a major portion of the to January 31, 2012. The collection and analysis of these data Lake Tuscaloosa drainage basin, has not experienced a substantial will facilitate the comparison with historical data, serve as increase in sedimentation rates. During the 2009 water year, the Turkey Creek watershed a baseline for future sediment-collection efforts, and help to 2 2 identify areas of concern. (6.16 mi ) and the Carroll Creek watershed (20.9 mi ) produced Lake Tuscaloosa, at the reservoir dam, receives runoff greater suspended-sediment yields than the North River watershed but contribute a much smaller drainage area to Lake Tuscaloosa. from a drainage area of 423 square miles (mi2). Basinwide in Aerial photography and bathymetric surveys indicate that Carroll 2006, forested land was the primary land cover (68 percent). Creek has experienced increased sediment deposition in the Comparison of historical imagery with the National Land upstream portions of the channel. Carroll Creek is also the only Cover Database (2001 and 2006) indicated that the greatest watershed in the current study that has a substantial percentage temporal land-use change was timber harvest. The land (11 percent) of urban land use. cover in 2006 was indicative of this change, with shrub/ scrub land (12 percent) being the secondary land use in the basin. Agricultural land use (10 percent) was represented predominantly by hay and pasture or grasslands. Urban land Introduction use was minimal, accounting for 4 percent of the entire basin. The remaining 6 percent of the basin has a land use of open Lake Tuscaloosa, constructed in 1969 on the North River water or wetlands. in Tuscaloosa County, Alabama, serves as the water supply for Storm and monthly suspended-sediment samples were Tuscaloosa, Northport, and other communities in Tuscaloosa collected from seven tributaries to Lake Tuscaloosa: North County (fig. 1). The lake is also used for recreational activities in River, Turkey Creek, Binion Creek, Pole Bridge Creek, Tierce western Alabama. Protection and monitoring of this water supply Creek, Carroll Creek, and Brush Creek. Suspended-sediment has been a concern for many years. Changes in land use, such concentrations and streamflow measurements were statistically as agriculture, timber clear-cutting, coal mining, and residential analyzed to estimate annual suspended-sediment loads and development in basins that drain into the lake have caused yields from each of these contributing watersheds. concern about possible changes in the rate of sedimentation in Estimated annual suspended-sediment yields in 2009 the reservoir. Sedimentation damages a reservoir if the decreased were 360, 540, and 840 tons per square mile (tons/mi2) storage capacity prevents supplying the full service for which it at the North River, Turkey Creek, and Carroll Creek was designed. Sufficient capacity must be maintained in domestic streamflow-gaging stations, respectively. Estimated annual water supply reservoirs to assure continuity of supply during suspended-sediment yields in 2010 were 120 and 86 tons/mi2 periods of prolonged drought and to meet expected increases in at the Binion Creek and Pole Bridge Creek streamflow-gaging water demand. stations, respectively. Estimated annual suspended-sediment The U.S. Geological Survey (USGS), in cooperation with yields in 2011 were 190 and 300 tons/mi2 at the Tierce Creek the City of Tuscaloosa, performed intensive suspended-sediment and Brush Creek streamflow-gaging stations, respectively. studies for seven tributaries in the Upper Black Warrior subbasin The North River watershed at the streamflow-gaging (hydrologic unit code (HUC) 03160112). The collection and analy- station contributes 53 percent of the drainage area for Lake sis of suspended-sediment data will allow a better understanding of Tuscaloosa. A previous study in the 1970s analyzed streamflow sedimentation rates and their effects on reservoir capacity. 2 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

88° 87°

Jasper !

FAYETTE COUNTY WALKER COUNTY

North LAMAR COUNTY

River

Binion

33°30' JEFFERSON COUNTY Dry Creek Creek Turkey Creek Pole Tierce Bridge Creek Creek Carroll

Brush Creek Creek PICKENS Lake COUNTY Tuscaloosa

River Northport ! ! Warrior Tuscaloosa Black TUSCALOOSA COUNTY

BIBB COUNTY

33°

Base from U.S. Geological Survey digital data, 1:100,000 02.5 5 10 MILES EXPLANATION Hydrography from National Atlas of the United States, 1:2,000,000 Lake Tuscaloosa subwatersheds 02.5 5 10 KILOMETERS Headwaters North River Upper North River Middle North River Binion Creek Lower North River Carroll Creek

Location of Fayette and Tuscaloosa Counties in Alabama

Figure 1. Location of Lake Tuscaloosa contributing subwatersheds and tributaries, Fayette and Tuscaloosa Counties, Alabama. Description of Study Area 3

Purpose and Scope Tuscaloosa Counties. Lake Tuscaloosa, at the reservoir dam, receives runoff from a drainage area of 423 square The purpose of this report is to provide estimates miles (mi2). The Black Warrior-North River watershed of suspended-sediment concentrations and loads deliv- (confluence of Black Warrior River and North River) ered to the Lake Tuscaloosa reservoir during 2009–11. has a total drainage area of 425 mi2. The Black Warrior- To accomplish this, (1) previous suspended-sediment North River watershed contains six subwatersheds; data and studies for the study area were reviewed; headwaters of the North River, upper North River, (2) land-use changes were investigated; (3) data were middle North River, Binion Creek, lower North River, collected at seven tributaries to the Lake Tuscaloosa and Carroll Creek, (fig. 1; table 1). reservoir; and (4) using current and historic data, Seven major streams flow into Lake Tuscaloosa: suspended-sediment loads to the Lake Tuscaloosa North River and Dry, Turkey, Binion, Tierce, Carroll, reservoir were estimated. and Brush Creeks (fig. 1). The normal pool elevation of Lake Tuscaloosa is 223.2 feet (ft)—National Geodetic Vertical Datum of 1929 (NGVD 29). The lake, Acknowledgments approximately 25 mi long as measured along the old river channel, has a surface area and capacity, at normal The assistance of Mr. Jimmy Junkin, Director, pool, of 5,250 acres and 168,917 acre feet (acre-ft), City of Tuscaloosa Water and Sewer Department, respectively (Charley Foster and Associates, Inc., and Mr. Scott Sanderford, Lakes Manager, City of 2004). Tuscaloosa, is greatly appreciated. Appreciation is also The area of study has a subtropical climate character- extended to Mr. Vic Stricklin and others who performed ized by warm, humid weather. According to long-term fieldwork or otherwise assisted with this study. The climatological records compiled by the National Oceanic field work and sample collection were performed by the and Atmospheric Administration (NOAA), the mean annual USGS Alabama Water Science Center Tuscaloosa Field air temperature is 62.2 °F (degrees Fahrenheit). Generally, Office. July is the hottest month with a mean temperature of 83.8 °F, and January is the coldest with a mean temperature of 52.5 °F (National Oceanic and Atmospheric Administra- Description of Study Area tion, 2002). Average annual precipitation, based on records for the 1971–2000 period at NOAA precipitation stations Lake Tuscaloosa is located in north-central in Tuscaloosa and Fayette Counties, is about 56 inches Tuscaloosa County, Alabama. The reservoir was (National Oceanic and Atmospheric Administration, 2002). created by the impoundment of the North River Higher streamflow is generally observed during December approximately 1.5 miles (mi) upstream from its conflu- through April and lower during May through November. ence with the Black Warrior River (fig. 1). The Lake Average annual runoff at long-term USGS gaging stations Tuscaloosa watershed (Black Warrior-North River in the watershed is about 22 inches or 1.6 cubic feet per HUC 0316011204) drains surface area from Fayette and second per square mile [(ft3/s)/mi2].

Table 1. Hydrologic Unit Code (HUC) designations for the drainage area contributing to Lake Tuscaloosa in the Upper Black Warrior subbasin, Alabama.

[mi2, square miles; —, no data]

HUC 8 HUC 10 HUC 12 Drainage area Name subbasin watershed subwatershed (mi2) 03160112 — — Upper Black Warrior — — 0316011204 — Black Warrior-North River (Lake Tuscaloosa) 425 — — 031601120401 Headwaters North River 96 — — 031601120402 Upper North River 44 — — 031601120403 Binion Creek 72 — — 031601120404 Middle North River 145 — — 031601120405 Carroll Creek 26 — — 031601120406 Lower North River 42 4 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Approximately 68 percent of the land in the study and water-quality data. Hubbard (1976a) estimated a long-term area is forested (Fry and others, 2011; fig. 2). Inspec- average sediment yield of 300 tons per year per square mile for tion of land-use trends indicate that the primary change North River. Hubbard also addressed the magnitude of potential in the Lake Tuscaloosa watershed has been forested sedimentation that could result from an increase in logging activi- areas. Three datasets were used for comparison: ties, coal mining, construction, or agriculture in the basin. Hubbard National Land Cover Database (NLCD) 1992–2001 (1976b) presented results of a water-quality reconnaissance study Retrofit Change Product (Fry and others, 2009), NLCD of the lake for the period March–June 1975 that included: standard 2001 (Homer and others, 2004), and NLCD 2006 (Fry chemical analyses of surface waters with analyses for nutrients and others, 2011). The results indicated 3 percent of the and trace elements, bacteria concentrations, chemical analyses of basin changed from forest to grassland/shrub from 1991 bottom deposits, and temperature and dissolved oxygen vertical to 2001. Similarly, the 2001 and 2006 NLCD showed a profiles. 3-percent decrease in forested land, changing from 71 Several studies of the effects of coal mining on hydrology to 68 percent, respectively. Direct comparison of the that are pertinent to this investigation have been made in the two datasets indicate a decrease in evergreen forest spe- Warrior Coal Field. Puente and others (1980) provided baseline cifically and an increase in shrub/scrub and grasslands. hydrologic information for selected basins. Harkins and others This decrease could be attributed to timber harvesting (1980) described the hydrology of part of the Warrior Coal Field without replanting, or to increased development. that included the North River basin. The U.S. Bureau of Land Although the relatively impermeable Pottsville Management (1980, 1983) assessed impacts of coal mining on Formation of Pennsylvanian age underlies all of the Federal coal-lease tracts in North River and adjacent basins. study area, it is exposed mainly in the northeastern part. Puente and Newton (1982) developed methods to estimate effects The Pottsville Formation consists primarily of sand- of surface mining on the hydrology of basins in the Warrior Coal stone, shale, and siltstone with shale being dominant Field. Puente and others (1982) described hydrologic conditions (Metzger, 1965). Beds of coal and underclay are present in four coal-lease tracts in the Warrior Coal Field. Cole (1985a, in some parts of the formation. Groundwater usually 1985b) sampled 14 sites in the North River basin to determine if occurs in openings along joints, fractures, and bedding surface coal mining had impacted the quality of water in the lake planes (Culbertson, 1964). and selected tributaries. Slack (1987) described the flow, water The more permeable Coker Formation of Creta- quality, and changes in the water quality by compiling historical ceous age lies above the Pottsville Formation and has data and collecting data at 17 sites in Lake Tuscaloosa and selected outcrops in the southern and western parts of the study tributaries. area. Although the upper 300 ft of the Coker Formation Since impoundment, several data collection efforts have been consists chiefly of clay (Metzger, 1965), the perme- undertaken to determine bathymetric changes in Lake Tuscaloosa. able sand and gravel beds in the lower 100 ft provide The bathymetric surveys were collected in 1973 (Hubbard, 1975), substantial quantities of base flow to streams and are 1982 (Cole, 1985b), 1986 (Slack and Prichett, 1988), 2000 (Strick- the principal source of water for wells in much of the lin, 2001), 2004 (Charley Foster and Associates, Inc., 2004), and Lake Tuscaloosa area (Cole, 1985b). 2010 (Lee and Kimbrow, 2011). For the 1982 bathymetric survey (Cole, 1985b), 17 cross sections were established in 7 principal tributaries to Lake Tuscaloosa, including North River, Dry Creek, Previous Investigations Turkey Creek, Binion Creek, Tierce Creek, Carroll Creek, and Brush Creek. Results revealed that sediment depositions ranged Since the impoundment of North River to form Lake from 2 to 20 ft in 14 of 17 cross sections (Cole, 1985b). These Tuscaloosa, there have been numerous efforts to monitor the cross sections were resurveyed in 1986 (Slack and Prichett, 1988) water quality, sedimentation rate, and bathymetric changes of the and in May 2000 (Stricklin, 2001) to determine whether any reservoir. Some investigations have been reconnaissance in nature additional sedimentation or scour occurred. Results from the 2000 and provide information to define the hydrologic conditions of the survey indicated that the maximum amount of sediment deposition reservoir since impoundment. Keener and others (1975) presented occurred in the upper end of Carroll Creek (Stricklin, 2001). geologic and hydrologic data for Lake Tuscaloosa, its tributaries, Charley Foster and Associates, Inc., conducted a survey of Lake and drainage basin. They described the geology in the general Tuscaloosa and its tributaries in 2004 and computed a capacity of area, soil associations and thickness, and provided basic data 190,495 acre-ft at a water-surface elevation of 223.2 ft NGVD 29. on groundwater and surface-water quality. They also presented As a byproduct of this investigation, a 3.5-mile reach of Carroll baseline sedimentation data collected from a fathometer survey Creek was surveyed in 2010 (Lee and Kimbrow, 2011) to prepare of 39 cross sections in the lake. Almon and Associates (1976) a current bathymetric map, determine storage capacities at addressed requirements of the Federal Water Pollution Control specified water-surface elevations, and compare current conditions Act Amendments of 1972, Public Law 92–500, as they relate to historical cross sections. The capacity for the 3.5-mile reach of to the Lake Tuscaloosa area. Included were descriptions of the Carroll Creek was estimated to be 7,100 acre-ft at a water-surface reservoir and drainage area, housing development information, elevation of 223 ft (NGVD 29). Previous Investigations 5

87°48' 87°36' 87°24'

33°48'

FAYETTE WALKER

33°36'

02464000

JEFFERSON 02464360 02464146

33°24' 02464385

PICKENS 02464503

02464680 TUSCALOOSA

0 4 8 MILES 02464660 0 0 4 8 KILOMETERS 02464800 Base from U.S. Geological Survey National Land Cover Data, EXPLANATION from LANDSAT-7 Thematic Mapper images Land Cover 30-meter resolution, 2006 Open water Mixed forest

Developed, open space Shrub/scrub

Developed, low intensity Grassland/herbaceous

Developed, medium intensity Pasture/hay

Developed, high intensity Cultivated crops

Barren land Woody wetlands

Deciduous forest Emergent herbaceous wetlands

Evergreen forest 02464800 USGS streamgage and number

Figure 2. National Land Cover Database (NLCD) land use in the Lake Tuscaloosa watershed in 2006, Tuscaloosa and Fayette Counties, Alabama. 6 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Approach and Methods sampler through the water at a constant rate. The samples from each equal-width segment were then combined into a single composite Streamflow and suspended-sediment data were collected sample for analysis. A composite water-sediment sample is horizon- for seven tributaries (fig. 3) to Lake Tuscaloosa (Black tally and vertically averaged throughout the stream cross section Warrior-North River watershed). These data were collected and is assumed to represent the average streamflow-weighted during water years (October to September) 2009 to 2012, and suspended-sediment concentration (Edwards and Glysson, used to develop sediment transport curves, estimate annual 1999; U.S. Geological Survey, 2006). The monthly low-flow loads, and evaluate the tributaries for areas of concern. The samples were used to evaluate conditions during normal flow approach and methodology used to measure streamflow and conditions. Samples were also collected during four flood collect suspended-sediment samples are described in the events. A total of five samples were taken during various following sections. flood events. Two were collected during the rising limb of the hydrograph, one around the peak of flood flow, and two on the Streamflow Data Collection falling limb of the hydrograph. These cross-section samples were collected and used to adjust the fixed-point automatic Streamflow data were collected for seven tributaries to samples collected during flooding. Lake Tuscaloosa. Of the seven collection sites, three are long- Isokinetic, depth-integrated (cross section) sediment samples term streamflow-gaging stations. At the remaining four sites, provide samples representative of stream conditions necessary to make streamflow-gaging stations were installed. The streamflow-gaging sediment load estimates for a stream. Manually collected isokinetic stations are equipped with data collection platforms (DCPs) that samples can be time consuming and expensive, and in some cases allow for near real-time access (data transmitted every hour) to the may not define the entire hydrograph. In order to collect more frequent hydrologic data from the USGS Alabama Water Science Center sediment samples, ISCO 6712 automatic water samplers (Teledyne (ALWSC) Web site (http://al.water.usgs.gov, accessed May 8, Isco, Inc., 2005) were deployed at selected sites. The ISCO sampler 2012). Continuous streamflow data were computed at all stream- features an electrically driven peristaltic pump, which is activated on flow-gaging stations by using standard USGS stage-discharge a predetermined schedule by an internal timer or in response to stage techniques (Carter and Davidian, 1968; Rantz and others, 1982; change. The intake tube is purged before and after each pumping Kennedy, 1984). Streamflow data are reviewed, approved, and period by automatic reversal of the pump. The automatic sampler stored in the USGS Automated Data-Processing System (ADAPS) can collect up to 24 discrete samples over the rise and fall of the of the National Water Information System (NWIS) database. hydrograph. The discrete sample frequency was set to collect samples Quality-assured surface-water data are available for retrieval on at a uniform time. The sample frequency was determined based on the Internet at http://waterdata.usgs.gov/al/nwis/sw, accessed the typical storm duration for each site. Samples for 10 storm events January 11, 2013. were collected for each site with an automatic sampler. Although the samples collected are efficient, they represent a point sample and are Suspended-Sediment Data Collection not necessarily indicative of average streamflow-weighted sediment concentrations. In order to adjust the point sample to represent Two methods of data collection were used for the sediment streamflow-weighted suspended-sediment concentrations, automatic samples, manual and automatic. The manually collected samples (fixed point) samples and manual (cross section) samples were taken were taken by USGS personnel and represent a streamflow-weighted simultaneously. These samples were collected for 4 of the 10 storm suspended-sediment concentration. The automatic samples were events. A regression was developed for each site, between concentra- collected using an ISCO 6712 automatic pump sampler (manufac- tions of the fixed-point automatic samples (dependent variable) and tured by Teledyne Isco, Inc.). These samples are taken at a fixed point the cross-section manual samples (independent variable). This relation and do not represent a streamflow-weighted suspended-sediment was used to adjust the automatic (fixed point) sample concentrations to concentration. Both methods are explained in detail in the following manual (cross section) sample concentrations. At all sampling locations paragraphs. the coefficient of determination (R2) for the regression equation was Monthly low-flow and various flood events were sampled greater than 0.7, with three sites having R2 values of 0.9. using the equal-width-increment method. Using standard USGS protocols and quality-control procedures (U.S. Geological Survey, Suspended-sediment samples were analyzed by the USGS 2006), an isokinetic sampler was used to collect suspended-sediment sediment laboratory in Louisville, Kentucky. Samples were samples. An isokinetic sampler collects a water-sediment sample analyzed for suspended-sediment concentration, and selected from the stream at a rate such that the velocity in the intake nozzle samples were analyzed for sand separation. Sand-separation is equal to the incident stream velocity at the nozzle entrance. The analysis gives the percentage of sediment, by weight, that is finer water-sediment sample collected is proportional to the instantaneous and coarser than 0.0625 millimeter (mm). Particle sizes smaller than stream velocity at the locus of the intake nozzle and, therefore, 0.0625 mm are defined as silt and clay; particle sizes 0.0625 milli- is representative of the sediment load at that point (Davis, 2005). meter or larger are defined as sand (Guy, 1969).Q uality-assured Samples were collected at equally spaced intervals (fig. 4) across the water-quality data are available for retrieval on the Internet at stream channel and were depth-integrated by lowering and raising the http://waterdata.usgs.gov/al/nwis/qw, accessed January 11, 2013. Approach and Methods 7

87°45' 87°30' 87°15'

¤£78 ¤£78 «¬4 ¤£78 Study «¬13 Jasper«¬269 69 area ! «¬ FAYETTE «¬102 «¬124 «¬69 «¬102 ¤£78 ¤£43 ¤£78

«¬102 ¬69 33°45' «

Location of study area in «¬13 Fayette and Tuscaloosa Counties «¬18 ¤£43

«¬18

«¬69 WALKER

¤£43

«¬171

«¬69

33°30' ¤£43 TUSCALOOSA

0 02464000 JEFFERSON

024643600 024641460 «¬171 02464385 0 02464503 ¤£43

«¬69 0 02464680 002464660

«¬171 ¤£82 «¬69 02464800 0 ¤£43 33°15' 0 2 4 6 8 MILES

¤£82 0 2 4 6 8 KILOMETERS

Base from U.S. Geological Survey digital data, ! 1:100,000, Universal Transverse Mercator Projection, EXPLANATION Zone 16. 02464800 USGS streamgage and number Black Warrior-North River watershed Contributing drainage area at USGS gaging stations North River Turkey Creek Binion Creek Tierce Creek Pole Bridge Creek Brush Creek Carroll Creek

Figure 3. Location of sampling sites and contributing drainage area at U.S. Geological Survey streamflow- gaging stations in the Black Warrior-North River watershed, Alabama. 8 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Samples collected at the center of each increment

W1 W2 W3 W4 Wn

Q1 Q2 Q3 Qd Qn

EXPLANATION W Width between verticals (equal, EWI) Q Discharge in each increment (not equal)

Figure 4. A diagram of the equal-width-increment sampling method (Edwards and Glysson, 1999) used to collect sediment samples in tributaries to Lake Tuscaloosa, Alabama. EWI, equal width increment.

Data Analysis The shape of the sediment transport curve is strongly affected by time offsets in the streamflow hydrograph and Suspended-sediment and streamflow data were collected suspended-sediment time series for specific runoff events. at seven tributaries to Lake Tuscaloosa and analyzed using Although suspended-sediment concentration generally graphical and statistical techniques. The data-analysis effort increases with streamflow, the relation is more complex. focused on developing daily sediment-transport curves and Streamflow and instantaneous sediment concentration may estimating sediment loads and yields at each site. A sediment- not have a stationary relation during a single storm flow. The transport curve is the curve that defines the average relation tendency for sediment concentration to have different values at between suspended-sediment discharge and streamflow. identical stream discharges (a hysteresis effect) is the primary Several factors can affect the shape, slope, and intercept of the drawback to application of a single transport curve during sediment transport curve. Some of the major factors are (1) storm flow (Walling, 1977; Williams, 1989; Dinehart, 1998). seasons, (2) timing between sediment concentration peak and When substantial timing variations exist between the sus- streamflow peak, and (3) extreme high-water events (Glysson, pended-sediment concentration peak and the streamflow peak, 1987). Seasons can have a major effect on sediment yield, separate regression curves may need to be developed for rising especially in the more humid areas. In the region of study, and falling data. The sediment hydrographs for each storm during the summer when high-intensity storms are prevalent, event were graphically analyzed to determine if they preceded, raindrop impact is high and thus sediment concentrations are tracked, or lagged the streamflow hydrographs. The sites that higher, particularly in nonforested land-use areas. An addi- have preceding peaks have hysteresis that are clockwise. This tional complication may also occur when a large area of the can be attributed to depletion of available sediment before drainage basin is used for agricultural purposes. Typically, the the streamflow hydrograph peaks or formation of an armored fields are bare during the winter and spring, but as the crops layer prior to the occurrence of the streamflow peak (Williams, begin to grow the soil becomes protected from erosion by the 1989). The sites that have lagging peaks have hysteresis plants (Glysson, 1987). that are counterclockwise. This can result from at least three Suspended-sediment samples were collected at each possible causes: relative travel times of the flood wave and streamflow-gaging station, as described in the Suspended- the sediment flux, high soil erodibility in conjunction with Sediment Data Collection section. Samples were collected prolonged erosion during the flood, and seasonal variability for 10 storm events; during 4 of the events, samples were of rainfall distribution and of sediment production within the collected by both manual and automatic methods. The samples drainage basin (Williams, 1989). In the event of the hysteresis were collected in efforts to reflect seasonal variations and their effect, separate instantaneous sediment transport curves were variable antecedent conditions. However, streamflow is gener- developed for the rising and falling data. ally higher during December through April than during May Another variable affecting the shape and applicability through November, and therefore the storm event samples of the sediment-transport curve is streamflow magnitude. typically were collected during the December to April time- Once the available sediment is exhausted, additional rainfall frame. In addition to the storm events, samples were collected does not increase suspended sediment and the transport curve on a monthly basis to represent normal daily streamflows. will flatten out. A comparison was made with streamflow to Approach and Methods 9 determine how climatic conditions during this investigation regression equation was used to calculate the mean suspended- may have influenced the data collected and may limit the sediment concentration from the concentrations of the point applicability of the regression equations developed by this samples. A linear regression equation (ordinary least squares) investigation to periods of differing climatic conditions. provides a line about which the sum of the square deviations For existing streamflow-gaging stations (streamgages), of observed values from the regression line was minimized. At annual mean streamflows for the period of record and the all sampling locations the coefficient of determination (R2) for individual water years were examined. Flow duration curves the regression equation was greater than 0.7, with three sites were developed to evaluate the average streamflow conditions having R2 values of 0.9. for the site compared to the period of investigation. For sites The resulting suspended-sediment concentrations (SSC) that were newly installed and lack a long-term record (10 were converted to suspended-sediment discharge (SSQ). or more years), the peak streamflow during data collection Suspended-sediment discharge is defined as the quantity was evaluated based on regional flood-frequency relations. of sediment per unit time carried past any cross section of A flood-frequency relation was developed based on regres- a stream (Vanoni, 1977) and will be reported in tons per sion equations developed in “Magnitude and Frequency of day. Each suspended-sediment sample was associated with Floods in Alabama, 2003” (Hedgecock and Feaster, 2007). A a streamflow value. Streamflow data were obtained from flood-frequency relation is the relation of peak streamflow to the stage-discharge rating curve at the streamflow-gaging probability of exceedance. Probability of exceedance refers to station where suspended-sediment samples were collected. the chance that a given peak streamflow will be exceeded in The availability of streamflow data and suspended-sediment any one year. For example, a 50-percent chance exceedance concentration data allowed computation of suspended- flood corresponds to the flow magnitude that has a probability sediment discharge, according to equation 1: of 0.50 of being equaled or exceeded in any given year. Typi- cally, the channel forming (bankfull) flow has a magnitude in SSQ = Q(SSC)(ks) (1) the range of a 67- to 50-percent chance exceedance flood. where Sediment-transport curves were developed for each site SSQ is the suspended-sediment discharge, in tons after evaluating seasonality, flow magnitude, and hysteresis. per day; The first step taken was to adjust the suspended-sediment Q is the streamflow, in cubic feet per second; samples to reflect average streamflow-weighted suspended- SSC is the suspended-sediment concentration, in sediment concentrations. As described in the Suspended milligrams per liter; and Sediment Data Collection section of this report, two types ks is a conversion factor of 0.0027, which re- of samples were collected, point samples and horizontally sults in a suspended-sediment discharge in tons per day, given and vertically averaged samples. To adjust the point sample streamflow in cubic feet per second and suspended-sediment to represent streamflow-weighted suspended-sediment concentration in milligrams per liter. concentrations, automatic (point) samples and manual samples (horizontally and vertically averaged) were taken The suspended-sediment concentration (SSC) was simultaneously. These samples were collected for 4 of the 10 graphically and statistically compared with streamflow Q( ) storm events. A cross-section coefficient, automatic sample to evaluate the relation and determine if any seasonal bias concentration divided by manual sample concentration, exists. The suspended-sediment discharge (SSQ) was then was computed for each sample. Many factors can affect the graphically and statistically compared with streamflow Q( ) cross-section coefficient; for example, size distribution of to develop sediment-transport curves. A linear regression sediment, channel alignment, bank stability, source of sedi- equation (ordinary least squares) was developed to provide a ment, streamflow, and location of sampling intake. According line about which the sum of the square deviations of observed to Stokes Law, suspended sediment considered to be fine values from the regression line was minimized. A composite material will have a fairly even distribution throughout the regression equation was developed to include all of the data water column, whereas coarse material will have a larger and separate regression equations were developed for the concentration closer to the bed. The smaller channel slopes in rising and falling limb of the hydrograph for sites that have the study area provide a good indication that the suspended documented hysteresis effects. The relation of log-transformed sediment will be considered fine. Chanel slopes were investi- suspended-sediment discharge and streamflow was linear and gated on a site-by-site basis by analyzing a sand/fines split for was expressed in the form of equation 2: some of the suspended-sediment samples. The cross-section coefficients were graphically compared with respect to time SSQ = aQb or Log(SSQ) = Log (a) +b Log(Q) (2) and streamflow to determine if any trends exist. The other where factors affecting the cross-section coefficient were accounted SSQ is the suspended-sediment discharge, in tons for by developing a linear regression equation (ordinary least per day; squares) for each site. The regression equation describes Q is the streamflow, in cubic feet per second; the relation between the automatic samples (dependent a is the intercept; and variable) and the manual samples (independent variable). The b is the slope. 10 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Because the regression equations were developed in log space Land Use and the predictions are made in original engineering units, retransformation of the data is required. This retransformation results in a In general, land use within the North River subwatershed in systematic distortion of the statistic, a bias. The regression equation 2006 (Fry and others, 2011) can be classified as rural. Forested will provide a value that is closer to the median response as opposed land (cumulative total of mixed, deciduous, and evergreen) to the mean, and the resulting suspended-sediment load will be was the predominant basinwide land use in 2006 at 72 percent. too low due to the skewed distribution. Several methods exist to Shrub/scrub, represented predominantly by shrubs less than remove this bias and have been tested (Cohn and others, 1989) to 16 ft tall with shrub canopy typically greater than 20 percent of determine their applicability. For this study the smearing estimator, total vegetation, accounted for approximately 11 percent of the a nonparametric estimator, was used to remove the potential bias. subwatershed. Low-, medium-, and high-intensity residential and A full description of the smearing estimator is described by Duan developed open-space land coverages were combined to compute (1983). total urban land use. Urban land use was approximately 3 percent The resulting sediment-transport regression equation with bias in 2006 (fig. 8). Inspection of aerial photography indicates little to correction may be classified according to either the period of the no urban development since completion of the 2006 NLCD. basic data that defined a curve or the kind of sediment discharge that General changes in land use were evaluated by inspecting a curve represents. Sediment-transport curves based on the period the NLCD 1992–2001 Land Cover Change Retrofit product of the basic data may be classified as instantaneous, daily, monthly, (Fry and others, 2009) and direct comparison of the NLCD seasonal, annual, or flood- or storm-period curves (Colby, 1956). 2001(Homer and others, 2004) and the NLCD 2006 (Fry and The suspended-sediment samples collected for the seven tributaries others, 2011). One of the largest reductions in land use from 1992 to Lake Tuscaloosa represent sediment concentrations at a given to 2001 was forested land that decreased by 4 percent. During time and streamflow conditions, and produce a sediment-transport this period, 2 percent of the basin was converted to forested land. curve that is classified as instantaneous. These curves are useful in Direct comparison of NLCD 2001 and 2006 (fig. 8), shows a estimating an average annual sediment load expressed in tons and small shift from forest categories to the shrub/scrub category, based on 15-minute streamflow values. This process can be labori- indicating that land use in the basin has been stable and the inclu- ous without the use of an automated program. For ease of use, these sion of historical data for analysis with recent data is acceptable. curves were used to develop daily sediment-transport curves. The instantaneous sediment-transport curves were used in conjunction with the 15-minute streamflow data to compute an average daily Historical Data suspended-sediment discharge for every day during the period of Historical data can provide various benefits in assessing data collection. These average daily suspended-sediment discharges current datasets. If the basin is relatively stable, historical data were graphically and statistically compared to the average daily can be used in conjunction with current data to develop sediment streamflow to produce a regression equation in the same manner transport curves. If the basin is dynamic, the historical and current that the instantaneous sediment-transport curves were developed. data can be used to monitor temporal changes. Periodic monthly The resulting daily sediment-transport curve was used to compute sediment samples (table 2) were collected at the same location an average annual sediment load and yield. The specific process and as the North River streamgage from 1979 to 1983. The smallest results are described for each site in later sections. low-flow and largest storm event sampled had a streamflow of 4 and 2,390 cubic feet per second (ft3/s), respectively. The historical data were included in the dataset to develop the suspended- North River sediment transport curve because the land-use changes described above indicate a relatively stable basin. Suspended-sediment data North River was impounded in 1969 to form the Lake were also collected at this streamgage from March to September Tuscaloosa reservoir. The USGS streamflow-gaging station at North 1975 to estimate the long-term average sediment yield of 300 tons River near Samantha, Ala. (station number 02464000, also referred per year per square mile (Hubbard, 1976a). This estimation was to as the North River streamgage) (fig. 5) is one of seven locations based on daily and flood event samples. The suspended-sediment for sampling suspended-sediment inflow into Lake Tuscaloosa. concentrations were available for inclusion in the dataset, but the The North River streamgage was established in December 1938, at corresponding streamflows were not published. For comparison, County Road 38, about 11 mi upstream from the backwater of Lake the corresponding average daily streamflows were taken from the Tuscaloosa. The drainage area for this streamgage is 223 mi2, about published annual data report and correlated with the suspended- 53 percent of the drainage area to Lake Tuscaloosa. The upstream sediment concentrations. The resulting dataset was plotted and area of the lake (near Bull Slough Road) has experienced sediment compared with other historical data and current data, but not deposition. A comparison of historical (1974) and recent (2011) included in the suspended-sediment transport curve. Even though aerial photography illustrates the changes in the shoreline (fig. 6). the 1975 dataset is an approximation of what was collected, it A closer aerial photograph (fig. 7) shows newly formed sediment followed the same trend as the data used to develop the current islands. sediment transport curve. North River 11

87°40' 87°30'

FAYETTE 102« 33°50' WALKER 13

124« 102« 102 FAYETTE 102«

102 TUSCALOOSA 102« Location of subwatersheds

WALKER 13 18

18 33°40'

r t h R i v e r 69 43

33°30' TUSCALOOSA

02464000 171

Base from U.S. Geological Survey digital data, 0 1 2 3 4 MILES 1:100,000, Universal Transverse Mercator Projection, EXPLANATION Zone 16. 0 1 2 3 4 KILOMETERS North River subwatershed

02464000 USGS streamgage and number

Figure 5. Location of U.S. Geological Survey streamflow-gaging station 02464000 North River near Samantha, Alabama, and the contributing drainage area. 12 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

87°36' 87°32'

North River Location shown in aerial photographs

33°24'

Treasure Island County Park

69 B 33°20' LAKE TUSCALOOSA

171

43 69 171

33°16' 43

0 1.5 3 MILES BULL SLOUGH

0 1.5 3 KILOMETERS Treasure Island County Park

ROAD

C EXPLANATION 1974 shoreline

Figure 6. Location of the North River in Alabama at (A) Bull Slough Road as shown in aerial photography dated (B) February 12, 1974 and (C) September 29, 2011. North River 13

D C

EXPLANATION

Sediment deposition northwest of Bull Slough Road

Sediment deposition southeast of Bull Slough Road

Figure 7. Locations of sediment deposition near the upstream extent of the backwater of Lake Tuscaloosa on the North River, Alabama. (A) Aerial photography dated 2009 shows sediment deposition to the (B) northwest and (C) southeast of Bull Slough Road. 14 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

A Table 2. Historical streamflow and suspended-sediment data for Land use coverage, in percent U.S. Geological Survey streamflow-gaging station 02464000 North River near Samantha, Alabama. 3 3 2 [Data are available at http://waterdata.usgs.gov/al/nwis/qw. Dates are by 5 month/day/year. Abbreviations: ft3/s, cubic foot per second; mg/L, milligram 4 per liter]

8 Suspended Mean Suspended sediment 35 Date streamflow sediment discharge (ft3/s) (mg/L) (tons/day)

14 6/5/1979 98 16 4.2 8/28/1979 59 41 6.5 11/28/1979 896 35 85 1/30/1980 410 7 7.7 2/28/1980 105 7 2.0 26 3/27/1980 874 14 33 4/29/1980 830 16 36 B 5/27/1980 360 22 21

2 3 3 6/23/1980 93 15 3.8 5 7/17/1980 15 20 0.79 4 10/9/1980 24 14 0.91 11/20/1980 234 18 11 11 35 1/20/1981 63 7 1.2 3/18/1981 170 12 5.5 4/14/1981 221 10 6 5/14/1981 30 4 0.33 13 5/28/1981 69 66 12 10/5/1982 4 13 0.13 11/2/1982 10 3 0.08 24 2/8/1983 1,290 41 143 3/11/1983 800 35 76 EXPLANATION 4/8/1983 2,390 265 1,710 Land cover Developed, open space 5/4/1983 243 7 4.6 Barren land 5/31/1983 247 9 6.0 Deciduous forest Evergreen forest 7/8/1983 56 6 0.91 Mixed forest 8/4/1983 23 58 3.6 Shrub/scrub Grassland/herbaceous 9/9/1983 41 43 4.8 Pasture/hay Cultivated crops Woody wetlands

Figure 8. Percentage of land-use coverage at the U.S. Geological Survey streamflow-gaging station 02464000 North River near Samantha, Alabama. (A) National Land Cover Database 2001 and (B) National Land Cover Database 2006. North River 15

Data Collected sediment-transport curves. A linear regression equation (ordinary least squares) was developed, and a smearing estimator applied, using all Streamflow and suspended-sediment data were collected data (composite). The composite instantaneous sediment transport at the North River streamgage. Streamflow is computed from curve was used to compute an average daily sediment discharge for stage values that are continuously measured and recorded every every day in the 2009 water year. These values were regressed with 15 minutes. Periodic and event suspended-sediment samples were the average daily flow to provide a daily transport curve (fig. 11; eq. 3) collected during the 2009 water year using isokinetic and equal- using the following equation: width increment methods.

SSQ =0.0135 (Q)1.4483 for Q ≤ 15,400 ft3/s (3) Streamflow where SSQ is the average daily suspended-sediment Climatic conditions have a direct effect on streamflow. Aver- age annual mean streamflow for the period of record (1940–54; discharge, in tons per day; and 1969–2011) and the individual water years were examined and are Q is the average daily streamflow, in cubic feet shown in figure 9. The suspended-sediment samples collected in 2009 per second. were collected during a period when annual streamflow was above average. The North River streamgage has a long period of record (10 The application of the daily suspended-sediment transport curve or more years), which was used to evaluate the average streamflow is limited to the maximum measured flow, 15,400 ft3/s, for which conditions for the site. A flow duration curve (fig. 10) was constructed suspended-sediment samples were collected. Suspended-sediment and used to evaluate how well the sampled streamflows represented transport curves have the tendency to flatten out once the available the range of possible flow conditions. Flow duration curves provide sediment is exhausted and additional rainfall no longer increases the the percentage of the time a certain streamflow can be expected to be suspended sediment. Because this exact point is not known, extrapola- equaled or exceeded for a site based on the daily mean streamflows tion of the suspended-sediment transport curve should be based on for the period of record at that site. The streamflows at the time of engineering judgment and analysis. sampling were overlain on the curve to determine if a fairly representative range of streamflows were sampled. Approximately 60 percent Annual load and yield values were estimated for suspended sedi- of the samples were collected at median or higher flow conditions ment at North River for the 2009 water year. The suspended-sediment (exceedance percentiles equal to or less than 50). The peak streamflow discharge was estimated for each day of the water year using the daily of record was also evaluated based on regional regression relations. transport curve (eq. 3). Load computations were performed by applica- A flood-frequency relation was developed (Hedgecock and Feaster, tion of the mid-interval method (Porterfield, 1972). The estimated 2007), and the 67- and 50-percent chance exceedance flood flows for average annual load for the 2009 water year, based on average daily the North River streamgage are 6,600 and 8,230 ft3/s, respectively. The streamflow and the North River daily transport curve is 81,400 tons peak streamflow during the suspended-sediment data collection period and the average yield is 360 tons per square mile (tons/mi2). 3 (2009) was 15,600 ft /s. The largest peak streamflow that was sampled It is worth noting that the daily mean sediment concentration is during the collection period was 15,400 ft3/s, which is greater than the a time-weighted mean value. Thus, calendar days can be divided and streamflow corresponding to 0.1 annual exceedance probability. analyzed in shorter periods of time when water or sediment discharge exceed certain limits. The term “subdivide” refers to the division of Suspended Sediment data for a calendar day into shorter periods of time to obtain correct daily mean values of streamflow or suspended-sediment discharge Suspended-sediment samples were collected at the North when one or both change beyond certain limits during the day. River streamgage isokinetically and horizontally and vertically averaged throughout the stream cross section. Because of the The values for these periods are then summed to obtain values of extensive historical data, samples were collected for two storm suspended-sediment discharge. Computation of a daily mean sediment events (table 3). The streamflows sampled were close in magni- discharge requires subdivision of the day when both water discharge tude to the streamflows sampled in the1975 dataset. In addition and concentration are changing because the average of the products of to the storm events, samples were collected on a monthly basis to two variable quantities is not the same as the product of the averages represent normal daily streamflows. of the quantities (Porterfield, 1972). For comparison, the instantaneous suspended-sediment transport curve, streamflow, and the mid-interval Estimation of North River Suspended-Sediment method were used to compute an estimated average annual load for Loads the 2009 water year, based on a 15-minute interval. This computation resulted in an average annual load that is 1 percent more than the load The suspended-sediment concentrations were converted calculated using the average daily values. It was determined that the to suspended-sediment discharges (SSQ) and graphically and use of the average daily streamflow and the daily transport curve is statistically compared with streamflow Q( ) to develop instantaneous more practical for estimating average annual loads. 16 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

800 EXPLANATION Annual mean streamflow 700 Average annual mean streamﬂow (1940–54, 1969–2011)

600

500

400

300

200

Annual meanstreamflow, in cubic feet per second 100

0 1940 1947 1954 1961 1968 1975 1982 1989 1996 2003 2010 Water year, October–September Figure 9. Annual mean streamflow for the U.S. Geological Survey streamflow-gaging station 02464000 North River near Samantha, Alabama.

100,000

10,000

1,000

100 Streamflow, in cubic feet per second Streamflow,

EXPLANATION 10 Flow duration curve

Sampled streamflows

1 0 10 20 30 40 50 60 70 80 90 100

Percentage of the time streamflow is equaled or exceeded

Figure 10. Flow duration curve and sampled streamflow for the U.S. Geological Survey streamflow-gaging station 02464000 North River near Samantha, Alabama, for 1940–54 and 1969–2011. North River 17

Table 3. Summary of suspended-sediment samples collected at U.S. Geological Survey streamflow-gaging station 02464000 North River near Samantha, Alabama.

[Data are available at http://waterdata.usgs.gov/al/nwis/qw. Dates are by month/day/year. Abbreviations: ft3/s, cubic foot per second; <, less than]

Peak Annual Collection Date streamflow exceedance type* (ft3/s) probability 1/6/2009 Storm 15,400 10 3/26/2006 Storm 7,620 <50 10/24/2008 Monthly 7 <67 11/25/2008 Monthly 7 <67 12/29/2008 Monthly 1,130 <67 1/27/2009 Monthly 141 <67 2/26/2009 Monthly 103 <67 4/28/2009 Monthly 135 <67 5/27/2009 Monthly 244 <67 6/30/2009 Monthly 30 <67 7/30/2009 Monthly 45 <67 * Storm indicates that the suspended-sediment samples were collected during a significant rainfall event. Monthly indicates that the suspended- sediment sample was collected during average daily flow.

100,000

SSQ = 0.0135 (Q)1.4483 for Q ≤ 15,400 cubic feet per second 10,000

1,000

100

Supended-sediment discharge, in tons per day EXPLANATION Daily transport equation 1983 1979 2009 monthly 0.1 1980 1/6/2009 storm 1981 3/26/2009 storm 1982

0.01 1 10 100 1,000 10,000 100,000

Streamflow, in cubic feet per second Figure 11. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464000 North River near Samantha, Alabama. SSQ, suspended sediment concentration; Q, streamflow, in cubic feet per second. 18 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Turkey Creek low-flow conditions are a variable function of the available sediment fines (wash load), the samples above a streamflow of 3 The USGS streamflow-gaging station on Turkey Creek 30 ft /s were graphically and statistically compared with the 2009 near Tuscaloosa, Ala. (station number 02464146, also data. Comparison of the historical data and the 2009 data showed a referred to as the Turkey Creek streamgage) was established slight upward trend of the measured suspended-sediment dis- in February 1981 on Turkey Creek at Turkey Creek Road, charge. The historical data were not included in the development near Tuscaloosa. The drainage area for this site (fig. 12) is of the suspended-sediment transport equation. 6.16 mi2 and is located 3 mi upstream of the backwater of Lake Tuscaloosa. The upstream extent of the pooled portion Data Collected of Turkey Creek (near Treasure Island County Park) has experienced increased sediment deposition. A comparison Streamflow and suspended-sediment data were collected of historical and current aerial photography illustrates the at the Turkey Creek streamgage at Turkey Creek Road, near changes in the shoreline (fig. 13). A close aerial photograph Tuscaloosa (station number 02464145). A streamflow-gaging (fig. 14) shows a newly formed sediment island. station was installed in February 1981 and has a well-established stage-discharge rating curve. Stage values are recorded every Land Use 15 minutes and suspended-sediment samples were collected isokenetically using equal-width-increment methods. Suspended- In general, land use within the Turkey Creek subwatershed sediment samples were collected during the 2009 water year. in 2006 (Fry and others, 2011) can be classified as rural. Forested Sediment samples were automatically collected using an automatic land (cumulative total of mixed, deciduous, and evergreen) was water sampler. Streamflow data have been collected at this site the predominant basinwide land use in 2006 at 64 percent. Barren from installation to present. land and shrub/scrub land use accounted for approximately 21 percent of the subwatershed. Low-, medium-, and high- Streamflow intensity residential, and developed open space land coverages were combined to compute total urban land use. Urban land use Climatic conditions have a direct effect on streamflow. A was approximately 2 percent in 2006 (fig. 15). Inspection of comparison was made with streamflow fig.( 16) and the dates of aerial photography indicates little to no urban development since data collection to determine how climatic conditions during this completion of the 2006 NLCD. investigation may have influenced the data collected and may limit General changes in land use were evaluated by inspecting the the applicability of the regression equations developed by this NLCD 1992–2001 Land Cover Change Retrofit product (Fry and investigation to periods of differing climatic conditions. The Turkey others, 2009) and direct comparison of the NLCD 2001 (Homer Creek streamgage has a long period of record (10 or more years) and others, 2004) and the NLCD 2006 (Fry and others, 2011). that was used to evaluate the average streamflow conditions for One of the largest reductions in land use from 1992 to 2001 was the site. Average annual mean streamflow for the period of record a 12-percent change from forested land to barren or grasslands. (1982–2011) and the individual water years were examined and During this period, 3 percent of the basin was converted to for- are shown in figure 17. The suspended-sediment samples collected ested land. Direct comparison of NLCD 2001 and 2006 (fig. 15) in 2009 were collected during a period when annual streamflow shows a decrease in forested land of 7 percent and an increase in magnitude was above average. A flow duration curve (fig. 18) was shrub/scrub and grassland of 3- and 5- percent, respectively. The constructed and used to evaluate how well the sampled streamflows temporal trend indicates that land use in the basin is not stable and represented the range of possible flow conditions at each site. the inclusion of historical data for analysis with recent data is not Flow duration curves provide the percentage of the time a certain acceptable. streamflow can be expected to be equaled or exceeded for a site based on the daily mean streamflows for the period of record at that Historical Data site. The streamflows at the time of sampling were overlain on the curve to determine if a fairly representative range of streamflows Monthly sediment samples (table 4) were collected at two were sampled. Approximately 96 percent of the samples were col- discontinued streamflow-gaging stations. Data were collected lected at median or higher flow conditions (exceedance percentiles for the 1977 to 1979 water years (October 1 to September 30) at equal to or less than 50). The peak streamflow of record was also Turkey Creek near Tuscaloosa (station number 02464145), located evaluated based on regional regression relations. A flood-frequency 1,000 ft downstream at State Highway 69. Additionally, various relation was developed (Hedgecock and Feaster, 2007) and the 67- samples were collected from Turkey Creek near the Patterson and 50-percent chance exceedance flood flows for the Turkey Creek Chapel streamflow-gaging station (station number 02464149). The streamgage are 390 and 640 ft3/s, respectively. The peak streamflow drainage areas at these sites are comparable to the 2009 sampling during the suspended-sediment data collection period (2009) was location and therefore the data were used for comparison. The 1,060 ft3/s. The largest streamflow that was sampled during the maximum and minimum streamflows sampled were 410 and collection period was 1,020 ft3/s, which is less than the streamflow 1.1 ft3/s, respectively. Because samples taken during normal corresponding to a 20-percent chance exceedance flood. Turkey Creek 19

87°33' 87°30'

FAYETTE WALKER 33°26'

TUSCALOOSA Location of subwatersheds 02464146

Tur key C reek 69

33°24'

k e e r C e c r ie T 02464503

Base from U.S. Geological Survey digital data, 1:100,000, Universal Transverse Mercator Projection, EXPLANATION 0 0.5 1 1.5 2 MILES Zone 16. Turkey Creek subwatershed 0 0.5 1 1.5 2 KILOMETERS 02464503 USGS streamgage and number

Figure 12. Location of U.S. Geological Survey streamflow-gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama, and the contributing drainage area. 20 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

A 87°36' 87°32'

Location shown in aerial photographsphotographs Turkey Creek

33°24' 69

Treasure Island County Park

69 33°20' LAKE B TUSCALOOSA

171

43 69 171

33°16' 43

0 1.5 3 MILES

0 1.5 3 KILOMETERS Treasure Island County Park

C EXPLANATION 1974 shoreline

Figure 13. The upstream extent of the backwater of (A) Lake Tuscaloosa on Turkey Creek, Alabama, is shown in aerial photography dated (B) February 12, 1974, and (C) September 29, 2011. Turkey Creek 21

87°36' 87°32'

Location shown in photographphotographs Turkey Creek

33°24' 69

LAKE TUSCALOOSA

69 33°20'

171

43 69 171

33°16'N33°16' 43

0 1.5 3 MILES

0 1.5 3 KILOMETERS

Figure 14. Aerial photograph taken in 2009 shows sediment deposition near the upstream extent of the backwater of Lake Tuscaloosa on Turkey Creek, Alabama. 22 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

A B <1 2 Land use coverage, in percent 2

8 2 3 13

16 16 16

12 3636 4400

EXPLANATION Land cover Open water Developed, open space Barren land Deciduous forest Evergreen forest Mixed forest Shrub/scrub Grassland/herbaceous

Figure 15. Percentage of land-use coverage at the U.S. Geological Survey streamflow-gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama. (A) National Land Cover Database 2001 and (B) National Land Cover Database 2006. Turkey Creek 23

Table 4. Historical streamflow and suspended-sediment data for U.S. Geological Survey streamflow-gaging stations 02464145 and 02464149 Turkey Creek near Tuscaloosa, Alabama.

[Data are available at http://waterdata.usgs.gov/al/nwis/qw. Dates are by month/day/year. Abbreviations: ft3/s, cubic foot per second; mg/L, milligram per liter]

Mean Suspended Suspended sediment Station Date streamflow sediment discharge number (ft3/s) (mg/L) (tons/day) 02464145 11/3/1976 3.5 1 0.01 02464145 12/7/1976 4.8 1 0.01 02464145 1/7/1977 12 5 0.16 02464145 1/7/1977 13 5 0.18 02464145 1/9/1977 106 141 40 02464145 1/9/1977 106 141 40 02464145 1/14/1977 60 28 4.5 02464145 3/12/1977 135 215 78 02464145 3/12/1977 135 215 78 02464145 3/12/1977 225 360 219 02464145 3/12/1977 390 476 501 02464145 4/4/1977 275 234 174 02464145 4/4/1977 275 234 174 02464145 4/4/1977 340 245 225 02464145 4/4/1977 340 245 225 02464145 4/4/1977 410 481 532 02464145 4/5/1977 220 236 140 02464145 4/11/1977 3.7 16 0.16 02464145 4/11/1977 11 5 0.15 02464145 5/16/1977 3.7 16 0.16 02464145 6/22/1977 2.4 7 0.05 02464145 8/18/1977 2.3 11 0.07 02464145 9/19/1977 4.0 4 0.04 02464145 10/17/1977 3.8 12 0.12 02464145 11/14/1977 5.5 0 0 02464145 1/5/1978 6.2 0 0 02464145 4/4/1978 5.0 3 0.04 02464145 5/1/1978 14 26 0.98 02464145 5/3/1978 30 69 5.6 02464145 5/8/1978 193 168 88 02464145 5/31/1978 4.5 11 0.13 02464145 7/10/1978 1.9 12 0.06 02464145 10/2/1978 1.1 4 0.01 02464145 11/6/1978 1.7 1 0 02464145 12/6/1978 2.5 17 0.11 02464145 1/3/1979 15 1 0.04 02464145 1/30/1979 15 2 0.08 02464145 2/27/1979 20 9 0.49 02464145 3/28/1979 8.7 6 0.14 02464145 4/27/1979 12 5 0.16 02464145 5/23/1979 5.8 7 0.11 02464145 6/4/1979 4.7 19 0.24 02464145 9/4/1979 2.9 10 0.08 02464145 1/20/1981 3.9 1 0.01 02464149 10/14/1982 5.6 13 0.20 02464149 4/20/1983 21 5 0.28 24 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

1,000

500

200

100

10 in cubic feet per second 5

1 Mean discharge , 0.5 EXPLANATION Period of suspended-sediment 0.2 data collection

0.1 O N D J F M A M J J A S 2008 2009

Figure 16. Streamflow hydrograph for the 2009 water year at U.S. Geological Survey streamflow-gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama.

18 EXPLANATION 16 Annual mean streamflow Average annual mean streamﬂow (1982–2011)

10 in cubic feet per second

, w o l f 8 a m e r t s 6 a n e m

a l

u 4 n n A

0 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 Water year, October–September

Figure 17. Annual mean streamflow for U.S. Geological Survey streamflow-gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama. Turkey Creek 25

10,000

1,000

100 Streamflow, in cubic feet per second Streamflow, 10 EXPLANATION

Flow duration curve

Sampled streamflows

1 0 10 20 30 40 50 60 70 80 90 100

Percentage of the time streamflow is equaled or exceeded

Figure 18. Flow duration curve and sampled streamflow for U.S. Geological Survey streamflow-gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama, 1982–2011. 26 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Suspended Sediment Estimation of Suspended-Sediment Loads from Suspended sediment samples were collected at the Turkey Creek Turkey Creek streamgage as described in the Suspended- The suspended-sediment concentrations of the automatic Sediment Data Collection section (see Approach and pump samples were adjusted to reflect average streamflow- Methods). Samples were collected for 11 storm events weighted suspended-sediment concentrations as described (table 5); during 4 of the events, samples were collected in the Data Analysis section (see Approach and Methods). by both manual and automatic methods. In addition to The suspended-sediment concentrations were converted the storm events, samples were collected on a monthly to suspended-sediment discharges (SSQ) and graphically basis to represent normal daily streamflows. The samples and statistically compared with streamflow Q( ) to develop were collected to reflect seasonal variations and their instantaneous sediment-transport curves. A linear regression variable antecedent conditions. Figure 16 illustrates the equation (ordinary least squares) was developed, and a smear- periods of data collection and how they relate to the annual ing estimator applied, for the rise and fall of the hydrograph hydrograph. The ISCO 6712 automatic pump sampler was and using all data (composite). The resulting instantaneous set to collect samples at a time interval ranging from 15 to sediment-transport curves converged for lower values of 30 minutes to ensure that suspended-sediment concentra- streamflow and slightly varied on the upper end of the curve. tions were collected over the rise, peak, and fall of the The composite instantaneous sediment transport curve was hydrograph. The sediment hydrographs for Turkey Creek used to compute an average daily sediment discharge for were graphically compared to the streamflow hydrographs. every day in the 2009 water year. These values were regressed The majority of the runoff events sampled had a sedi- with the average daily flow to provide a daily transport curve ment hydrograph whose peak preceded the streamflow (eq. 4; fig. 19) using the following equation: hydrograph (clockwise hysteresis), which is often ascribed to resuspension of sediment from the stream channel at the SSQ =0.1858 (Q)1.3919 for Q ≤ 1,020 ft3/s (4) initiation of storm runoff, and to relatively limited sedi- where ment supply on the stormflow recession. SSQ is the average daily suspended-sediment discharge, in tons per day; and Q is the average daily streamflow, in cubic feet Table 5. Summary of suspended-sediment samples collected at per second. U.S. Geological Survey streamflow-gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama. The application of the daily suspended-sediment transport curve is limited to the maximum measured flow, 1,020 ft3/s, [Data are available at http://waterdata.usgs.gov/al/nwis/qw. Dates are by month/day/year. Abbreviations: ft3/s, cubic foot per second; <, less than] that suspended-sediment samples collected. Suspended- sediment transport curves have the tendency to flatten out once Peak Annual the available sediment is exhausted and additional rainfall no Collection Date streamflow exceedance longer increases the suspended sediment. Because this exact type* (ft3/s) probability point is not known, extrapolation of the suspended-sediment transport curve should be based on engineering judgment and 12/10/2008 Manual/Automatic 240 <67 analysis. 2/27/2009 Manual/Automatic 1,020 <20 Annual load and yield values were estimated for 3/14/2009 Automatic 99 <67 suspended sediment at Turkey Creek for the 2009 water 3/25/2009 Automatic 266 <67 year. The suspended-sediment discharge was estimated for each day of the water year using the daily transport curve 4/2/2009 Automatic 22 <67 (eq. 4). Load computations were performed by application of 5/16/2009 Automatic 50 <67 the mid-interval method (Porterfield, 1972). The estimated 5/26/2009 Automatic 34 <67 average annual load for the 2009 water year, based on average 6/4/2009 Automatic 588 <67 daily streamflow and the Turkey Creek daily transport curve, is 3,350 tons and the average yield is 540 tons/mi2. 7/16/2009 Automatic 295 <67 It is worth noting that the daily mean sediment concentra- 8/20/2009 Manual/Automatic 82 <67 tion is a time-weighted mean value. Thus, calendar days 9/20/2009 Manual/Automatic 359 <67 can be divided and analyzed in shorter periods of time when * Automatic—suspended-sediment samples collected by an automatic water or sediment discharge exceeds certain limits. The term water sampler. Manual—manually collected isokenetic suspended-sediment “subdivide” refers to the division of data for a calendar day samples. into shorter periods of time to obtain correct daily mean values of streamflow or suspended-sediment discharge when one or both change beyond certain limits during the day. The values for these periods are then summed to obtain values of Turkey Creek 27 suspended-sediment discharge. Computation of a daily mean load for the 2009 water year, based on a 15-minute interval. sediment discharge requires subdivision of the day when both This computation resulted in an average annual load that is water discharge and concentration are changing because the 4 percent more than the load calculated using the average daily average of the products of two variable quantities is not the values. It was determined that the use of the average daily same as the product of the averages of the quantities (Porter- streamflow and the daily transport curve is more practical for field, 1972). For comparison, the instantaneous suspended- estimating average annual loads. sediment transport curve, streamflow, and the mid-interval method were used to compute an estimated average annual

10,000 1.3919 SSQSSQ = =0.1858 0.1858 (Q) 1.3919(Q) forfor Q Q ≤ ≤1,020 1,020 cubic cubic feet perfeet second per second

1,000

100

Supended-sediment discharge, in tons per day 10 EXPLANATION Daily transport curve Instantaneous falling data Instantaneous rising data Historical data

1 10 100 1,000 10,000 Streamflow, in cubic feet per second

Figure 19. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464146 Turkey Creek near Tuscaloosa, Alabama. SSQ, suspended sediment concentration; Q, streamflow in cubic feet per second. 28 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Binion Creek collected at high streamflow fall within the range of data collected in 2010 during the rising limb of the hydrograph. The USGS streamflow-gaging station on Binion Because the 1983 samples were collected prior to the instal- Creek below Gin Creek, near Samantha, Ala. (station lation of the streamgage, the hydrographs are not available number 02464360, also referred to as the Binion Creek to determine when the samples were collected relative to the streamgage) was established in October 1986 on Binion Creek peak streamflow. at Old Fayette Road, near Samantha. The drainage area for this site (fig. 20) is 57.2 mi2, and it contributes 14 percent of Data Collected the surface drainage area to Lake Tuscaloosa. The streamgage is located 2 mi upstream of the backwater region of Lake Streamflow and suspended-sediment data were collected Tuscaloosa. at Binion Creek at Old Fayette Road near Samantha, Ala. (station number 02464360). This streamflow-gaging Land Use station was established in 1986 and has a well-established stage-discharge rating curve. Stage values are recorded In general, land use within the Binion Creek subwater- every 15 minutes, and suspended-sediment samples were shed in 2006 (Fry and others, 2011) can be classified as rural. collected isokenetically using equal-width-increment methods. Forested land (cumulative total of mixed, deciduous, and Suspended-sediment samples were collected during the 2010 evergreen) was the predominant basinwide land use in 2006 water year. at 65 percent. Barren land and shrub/scrub land use accounted for approximately 14 percent of the subwatershed. Low-, Streamflow medium-, and high-intensity residential and developed open space land coverages were combined to compute total urban Climatic conditions have a direct effect on streamflow. A land use. Urban land use was approximately 3 percent in 2006 comparison was made with streamflow (fig. 22), and the dates (fig. 21). Inspection of aerial photography indicates little to no of data collection, to determine how climatic conditions during urban development since completion of the 2006 NLCD. this investigation may have influenced the data collected General changes in land use were evaluated by inspecting and may limit the applicability of the regression equations the NLCD 1992–2001 Land Cover Change Retrofit product developed by this investigation to periods of differing climatic (Fry and others, 2009) and direct comparison of the NLCD conditions. The Binion Creek streamgage has a long period 2001 (Homer and others, 2004) and the NLCD 2006 (Fry of record (10 or more years) that was used to evaluate the and others, 2011). Comparison of the 1992 and 2001 datasets average streamflow conditions for the site. Average annual showed little change. The only notable change was a 1-percent mean streamflow for the period of record (1987–2011) and increase in grasslands. Direct comparison of NLCD 2001 the individual water years were examined and are shown and 2006 (fig. 21) indicates a decrease in forested land of in figure 23. The suspended-sediment samples collected in 7 percent and an increase in shrub/scrub and grassland of 2010 were collected during a period when annual streamflow 8 percent. Current conditions were examined using 2009 and magnitude was above average. A flow duration curve (fig. 24) 2010 aerial photography in conjunction with the NLCD 2006. was constructed and used to evaluate how well the sampled Comparison of the two timeframes indicated between 2006 streamflows represented the range of possible flow conditions 2 and 2009 and 2009 and 2010, approximately 2 and 1.4 mi of at each site. Flow duration curves provide the percentage of timber were harvested, respectively. The clear-cut areas closest the time a certain streamflow can be expected to be equaled or to the streamflow-gaging station were further inspected. exceeded for a site based on the daily mean streamflows for Approximately 1 mi upstream of the Binion Creek streamgage, the period of record at that site. The streamflows at the time 33 acres of timber were harvested. The timber company of sampling were overlain on the curve to determine if a fairly confirmed that the timber was removed from September 7 to representative range of streamflows were sampled. Approxi- October 15, 2010, which was during the collection period. mately 99 percent of the samples were collected at median or higher flow conditions (exceedance percentiles equal to or less Historical Data than 50). The peak streamflow of record was also evaluated based on regional regression relations. A flood-frequency Monthly sediment samples (table 6) were collected at relation was developed (Hedgecock and Feaster, 2007), and the same location as the Binion Creek streamflow-gaging the 67- and 50-percent chance exceedance flood flows for the station (station number 02464360) for the 1983 water year Binion Creek streamgage are 1,130 and 2,000 ft3/s, respec- (October 1, 1982, to September 30, 1983). The historical tively. The peak streamflow during the suspended-sediment data were used to show temporal trends, because the land-use data-collection period (2010–11) was 3,210 ft3/s. The largest changes described above indicate a decrease in forested streamflow that was sampled during the collection period was land. Comparison of the 1983 and 2010 datasets indicate a 2,900 ft3/s, which is larger than the streamflow corresponding similar trend for the lower values of streamflow. The samples to a 50-percent chance exceedance flood. Binion Creek 29

87°45' 87°40' 87°35'

FAYETTE FAYETTE WALKER

TUSCALOOSA

Location of subwatersheds

33°30' 43

02464000

TUSCALOOSA B i ni on C re ek 02464360

33°25'

171 43

osa 02464385 alo usc e T Lak

Base from U.S. Geological Survey digital data, 1:100,000, Universal Transverse Mercator Projection, EXPLANATION 0 0.5 1 1.5 2 MILES Zone 16. Binion Creek subwatershed 0 0.5 1 1.5 2 KILOMETERS 02464385 USGS streamgage and number

Figure 20. Location of the U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama, and the contributing drainage area. 30 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

A Land use coverage, in percent <1

6 3 4 5

Table 6. Historical streamflow and suspended-sediment data for 19 U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama.

[Data are available at http://waterdata.usgs.gov/al/nwis/qw. Dates are by 1133 month/day/year. Abbreviations: ft3/s, cubic foot per second; mg/L, milligram per liter]

Suspended B Mean Suspended <1 sediment Date streamflow sediment 3 discharge 6 3 (ft /s) (mg/L) 3 (tons/day) 5 10/5/1982 16 10 0.43 4 11/6/1982 30 7 0.57 12/3/1982 132 27 9.6 1/4/1983 106 18 5.2 14 39 2/7/1983 370 94 94 3/4/1983 152 491 202 4/5/1983 98 23 6.1 5/3/1983 84 111 25 6/6/1983 79 17 3.6 16 7/6/1983 52 20 2.8 1100 8/3/1983 34 23 2.1 9/6/1983 34 12 1.1 EXPLANATION Land cover Open water Developed, open space Deciduous forest Evergreen forest Mixed forest Shrub/scrub Grassland/herbaceous Pasture/hay Cultivated crops Woody wetlands

Figure 21. Percentage of land-use coverage at the U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama. (A) National Land Cover Database 2001 and (B) National Land Cover Database 2006. Binion Creek 31

A 10,000

5,000 EXPLANATION Period of suspended-sediment data collection 2,000

1,000

500

200

100

1 OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. 2009 2010 B 10,000

5,000 EXPLANATION Period of suspended-sediment data collection 2,000 Mean discharge, in cubic feet per second 1,000

500

200

100

1 OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. 2010 2011

Figure 22. Streamflow hydrograph at U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama. (A) 2010 water year. (B) 2011 water year. 32 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

140 EXPLANATION Annual mean streamflow 120 Average annual mean streamﬂow (1987–2011)

100

Annual meanstreamflow, in cubic feet per second 20

0 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Water year, October–September

Figure 23. Annual mean streamflow for U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama.

10,000

1,000

100 Streamflow, in cubic feet per second Streamflow, 10 EXPLANATION

Flow duration curve

Sampled streamflows

1 0 10 20 30 40 50 60 70 80 90 100

Percentage of the time streamflow is equaled or exceeded

Figure 24. Flow duration curve and sampled streamflow for U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama, 1987–2011. Binion Creek 33

Suspended Sediment Estimation of Suspended-Sediment Loads from Suspended-sediment samples were collected at the Binion Binion Creek Creek streamgage as described in the Suspended-Sediment The suspended-sediment concentrations of the automatic Data Collection section (see Approach and Methods). pump samples were adjusted to reflect average streamflow-weighted Samples were collected for 11 storm events (table 7); during suspended-sediment concentrations as described in the Data Analysis 3 of the events, samples were collected by both manual and section (see Approach and Methods). The suspended-sediment automatic methods. In addition to the storm events, samples concentrations were converted to suspended-sediment discharges were collected on a monthly basis to represent normal daily (SSQ) and graphically and statistically compared with streamflow streamflows. The samples were collected to reflect seasonal (Q) to develop instantaneous sediment-transport curves. A linear variations and their variable antecedent conditions. Figure 22 regression equation (ordinary least squares) was developed, and a illustrates the periods of data collection and how they relate smearing estimator applied, for the rise and fall of the hydrograph to the annual hydrograph. The ISCO 6712 automatic pump and using all data (composite). Visual inspection of the data points sampler was set to collect samples at a time interval of showed significant scatter in the data collected on the rising limb 2 hours. This was done to ensure that suspended-sediment of the hydrographs. This could be attributed to the anthropogenic concentrations were collected over the rise, peak, and fall of changes in the basin or be a natural characteristic of the stream. the hydrograph. The sediment hydrographs for Binion Creek Comparison of current and historical data indicates that poor were graphically compared to the streamflow hydrographs. correlation on the rising limb of the hydrograph is a natural The majority of the runoff events sampled had a sediment characteristic of the stream. In order to improve the statistical hydrograph whose peak preceded the streamflow hydrograph correlation between suspended-sediment discharge and streamflow, (clockwise hysteresis), which is often ascribed to resuspen- the group averaging process as described by Glysson (1987) was sion of sediment from the stream channel at the initiation of used. The dataset was divided into 15 classes based on a logarithmic storm runoff, and to relatively limited sediment supply on the distribution of streamflow. The average, maximum, and minimum stormflow recession. streamflow and suspended-sediment discharge were determined for each class. A suspended-sediment transport curve was developed using linear regression (ordinary least squares), with a smearing Table 7. Summary of suspended-sediment samples collected at estimator applied, for the group averaged data. Comparison of the U.S. Geological Survey streamflow-gaging station 02464360 Binion group average and non-group averaged data indicated that the group Creek near Samantha, Alabama. averaging did not provide any additional benefits. [Data are available at http://waterdata.usgs.gov/al/nwis/qw. Dates are by Because the intercepts of the rising and falling instantaneous month/day/year. Abbreviations: ft3/s, cubic foot per second; <, less than; sediment-transport curves are substantially different, the unit value >, greater than; ≈, approximately equal to] (15 minute) streamflows for the entire water year were visually and statistically inspected and the appropriate rising/falling instantaneous Peak Annual Collection equation was used to compute a suspended-sediment discharge. Date streamflow exceedance These values were used to compute an average daily sediment type* 3 (ft /s) probability discharge for every day in the 2010 water year. These values were 10/30/09 Automatic 259 <67 regressed with the average daily flow to develop a daily transport 11/10/09 Automatic 1,010 ≈67 curve (eq. 5; fig. 25). The resulting equation has a small intercept value similar to the falling equation: 12/24/09 Automatic 558 <67 01/21/10 Manual/Automatic 835 <67 SSQ =0.043 (Q)1.2689 for Q ≤ 2,900 ft3/s (5) 03/10/10 Manual/Automatic 921 <67 where 03/25/10 Automatic 468 <67 SSQ is the average daily suspended-sediment discharge, in tons per day; and 05/02/10 Automatic 1,510 >67 Q is the average daily streamflow, in cubic feet 05/21/10 Automatic 818 <67 per second. 03/05/11 Manual/Automatic 935 <67 03/09/11 Manual 2,900 >50 The application of the daily suspended-sediment transport curve is limited to the maximum measured flow, 2,900 ft3/s, 03/30/11 Automatic 124 <67 at which suspended-sediment samples were collected. * Automatic—suspended-sediment samples collected by an automatic Suspended-sediment transport curves have the tendency water sampler. Manual—manually collected isokenetic suspended-sediment samples. to flatten out once the available sediment is exhausted and additional rainfall no longer increases the suspended sediment. Because this exact point is not known, extrapolation of the suspended-sediment transport curve should be based on engineering judgment and analysis. 34 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

10,000

SSQ = 0.043 (Q)1.2689 for Q ≤ 2,900 cubic feet per second

1,000

100

Supended-sediment discharge, in tons per day 10

EXPLANATION Daily transport curve Instantaneous falling data Instantaneous rising data

1 10 100 1,000 10,000 Streamflow, in cubic feet per second

Figure 25. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464360 Binion Creek near Samantha, Alabama. SSQ, suspended sediment concentration; Q, streamflow in cubic feet per second.

Annual load and yield values were estimated for are changing because the average of the products of suspended sediment at Binion Creek for the 2010 water two variable quantities is not the same as the product year. The suspended-sediment discharge was estimated of the averages of the quantities (Porterfield, 1972). for each day of the water year using the daily transport For comparison, the instantaneous suspended-sediment curve (eq. 5). Load computations were performed by transport curve, streamflow, and the mid-interval method application of the mid-interval method (Porterfield, were used to compute an estimated average annual load 1972). The estimated average annual load for the 2010 for the 2010 water year, based on a 15-minute interval. water year, based on average daily streamflow and the This computation resulted in an average annual load that Binion Creek daily transport curve, is 6,860 tons, and the is 4 percent more than the load calculated using the aver- average yield is 120 tons/mi2. age daily values. It was determined that the use of the It is worth noting that the daily mean sediment average daily streamflow and the daily transport curve is concentration is a time-weighted mean value. Thus, more practical for estimating average annual loads. calendar days can be divided and analyzed in shorter periods of time when water or sediment discharge exceed certain limits. The term “subdivide” refers to Pole Bridge Creek the division of data for a calendar day into shorter periods of time to obtain correct daily mean values of The USGS streamflow-gaging station on Pole Bridge streamflow or suspended-sediment discharge when one Creek near Samantha, Ala. (station number 02464385, also or both change beyond certain limits during the day. referred to as the Pole Bridge Creek streamgage) was estab- The values for these periods are then summed to obtain lished in January 2010 on Pole Bridge Creek at Old Fayette values of suspended-sediment discharge. Computation Road, near Samantha. The drainage area for this site (fig. 26) of a daily mean sediment discharge requires subdivision is 4.67 mi2, and it is located 1.3 mi upstream of the backwater of the day when both water discharge and concentration portion of Lake Tuscaloosa. Pole Bridge Creek 35

87°44' 87°40' 87°36'

Bin ion Cr 02464360 eek

33°24'

171

Pole Bri dge Creek 02464385 43

FAYETTE WALKER 33°22'

sa o lo a sc u T e ak L TUSCALOOSA

Location of subwatersheds 33°20'

Base from U.S. Geological Survey digital data, EXPLANATION 1:100,000, Universal Transverse Mercator Projection, 0 0.5 1 1.5 2 MILES Zone 16. Pole Bridge Creek subwatershed 0 0.5 1 1.5 2 KILOMETERS 02464385 USGS streamgage and number

Figure 26. Location of U.S. Geological Survey streamflow-gaging station 02464385 Pole Bridge Creek near Samantha, Alabama, and the contributing drainage area. 36 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Land Use A Land use coverage, in percent In general, land use within the Pole Bridge Creek subwatershed 1 in 2006 (Fry and others, 2011) can be classified as rural. Forested 2 land (cumulative total of mixed, deciduous, and evergreen) was the 3 2 predominant basinwide land use in 2006 at 75 percent. Shrub/scrub land use accounted for approximately 18 percent of the subwatershed. 11 Low-, medium-, and high-intensity residential, and developed open space land coverages were combined to compute total urban land use. Urban land use was approximately 1 percent in 2006 (fig. 27). General changes in land use were evaluated by inspecting the NLCD 1992–2001 Land Cover Change Retrofit product (Fry and 17 others, 2009) and direct comparison of the NLCD 2001 (Homer 51 and others, 2004) and the NLCD 2006 (Fry and others, 2011). Comparison of the 1992 and 2001datasets indicates little change. The only notable change was a 4-percent change from forest to grassland and agriculture. Direct comparison of NLCD 2001 and 2006 (fig. 27) 13 shows a decrease in forested land of 6 percent and an increase in shrub/scrub and grassland of 7 percent. Current conditions were examined using 2009 and 2010 aerial photography in conjunction with the NLCD 2006. Comparison of the two timeframes indicated that between 2009 and 2010, approximately 43 acres of timber were B 2 1 1 1 harvested. 2 Data Collected

Streamflow and suspended-sediment data were collected at 18 Pole Bridge Creek at Old Fayette Road near Samantha, Ala. (station number 02464385). A streamflow-gaging station was installed in January 2010. Stage values are recorded every 15 minutes and a stage-discharge rating curve was developed. Sediment samples 53 were automatically collected using an automatic water sampler. 13 Suspended-sediment samples were collected during the 2010–11 water year.

Streamflow 9 Climatic conditions have a direct effect on streamflow. A comparison was made with streamflow (fig. 28) and the dates of data EXPLANATION collection (2010–11) to determine how climatic conditions during Land cover this investigation may have influenced the data collected and may Developed, open space Deciduous forest limit the applicability of the regression equations developed by this Evergreen forest investigation to periods of differing climatic conditions. The Pole Mixed forest Bridge Creek streamgage does not have a long period of record (10 Shrub/scrub or more years) and therefore assumptions cannot be made about Grassland/herbaceous the average streamflow conditions for the site. However, the peak Pasture/hay Cultivated crops streamflow of record was evaluated based on regional regression Woody wetlands relations. A flood-frequency relation was developed (Hedgecock and Feaster, 2007) and the 67- and 50-percent chance exceedance flood flows for the Pole Bridge Creek streamgage are 510 and Figure 27. Percentage of land-use coverage at U.S. Geological 690 ft3/s, respectively. The peak streamflow experienced during the Survey streamflow-gaging station 02464385 Pole Bridge Creek suspended-sediment data-collection period (2010–11) was 431 ft3/s. near Samantha, Alabama. (A) National Land Cover Database 2001. The largest streamflow that was sampled during the collection period (B) National Land Cover Database 2006. was 96 ft3/s, which is significantly less than the streamflow that corresponds to a 67-percent chance exceedance flood. Pole Bridge Creek 37

A 1,000

500 EXPLANATION Period of suspended-sediment data collection 200

100

0.5

0.2

0.1 OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. 2009 2010 B 1,000

500 EXPLANATION Period of suspended-sediment data collection 200 Mean discharge, in cubic feet per second 100

0.5

0.2 0.1 OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. 2010 2011

Figure 28. Streamflow hydrograph at U.S. Geological Survey streamflow-gaging station 02464385 Pole Bridge Creek near Samantha, Alabama. (A) 2010 water year. (B) 2011 water year. 38 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Suspended Sediment Estimation of Suspended-Sediment Loads from Suspended-sediment samples were collected at Pole Bridge Creek the Pole Bridge Creek streamgage as described in the The suspended-sediment concentrations of the automatic Suspended-Sediment Data Collection section (see Approach pump samples were adjusted to reflect average streamflow- and Methods). Samples were collected for 11 storm events weighted suspended-sediment concentrations as described in (table 8); during 4 of the events, samples were collected by the Data Analysis section (see Approach and Methods). The both manual and automatic methods. In addition to the storm suspended-sediment concentrations were converted to suspended- events, samples were collected on a monthly basis to represent sediment discharges (SSQ) and graphically and statistically normal daily streamflows. The samples were collected to compared with streamflow Q( ) to develop instantaneous sediment- reflect seasonal variations and their variable antecedent transport curves. The graphical comparison of all the data conditions. Figure 28 illustrates the periods of data collection indicated significant scatter among the dataset. The points were and how they relate to the annual hydrograph. The automatic compared to the unit value (15 minute) streamflow hydrographs pump sampler was set to collect samples at a time interval of to determine if they were collected on the rising or falling limb of 30 to 45 minutes to ensure that suspended-sediment con- the hydrograph and they were plotted accordingly. The dataset was centrations were collected over the rise, peak, and fall of the also plotted by date of storm event. The plot indicated grouping hydrograph. The sediment hydrographs for Pole Bridge Creek among individual rainfall events. In an effort to determine if any were graphically compared to the streamflow hydrographs. other biases exist, the rising and falling plots were color coded The majority of the runoff events sampled had a sediment to indicate season and identify rainy months. The data collected hydrograph whose peak preceded the streamflow hydrograph on the rising limb of the hydrograph indicated no general trend. (clockwise hysteresis), which is often ascribed to resuspension The data collected on the falling limb of the hydrograph showed a of sediment from the stream channel at the initiation of storm trend of higher concentrations for the summer months. During the runoff, and to a relatively limited sediment supply on the summer when high-intensity storms are prevalent, raindrop impact stormflow recession. is high and thus sediment concentrations are higher (Glysson, 1987). A portion of the trend may be attributed to the grouping among storm events. It was determined that not enough data exist Table 8. Summary of suspended-sediment samples collected at to develop seasonal curves for the falling limb of the hydrograph, U.S. Geological Survey streamflow-gaging station 02464385 Pole but bias may exist. Bridge Creek near Samantha, Alabama. In order to improve the statistical correlation between [Data are available at http://waterdata.usgs.gov/al/nwis/qw. Dates are by suspended-sediment discharge and streamflow, the group averag- month/day/year. Abbreviations: ft3/s, cubic foot per second; <, less than] ing process as described by Glysson (1987) was used. The dataset was divided into 15 classes based on a logarithmic distribution of Peak Annual streamflow. The average, maximum, and minimum streamflow Collection Date streamflow exceedance for suspended-sediment discharge were determined for each type* (ft3/s) probability class. A suspended-sediment transport curve was developed 03/10/10 Manual 62 <67 using linear regression (ordinary least squares), and a smearing estimator applied, for the group-averaged data. The results of both 03/25/10 Manual/Automatic 43 <67 curves (rising and falling) were higher for the group average as 04/08/10 Automatic 13 <67 opposed to using all data points. Inspection of the rising curve 04/23/10 Automatic 46 <67 showed no obvious benefit using the group averaging process. 05/02/10 Automatic 46 <67 Inspection of the falling data indicated that the group averaging process improved the correlation between suspended-sediment 05/21/10 Automatic 51 <67 discharge and streamflow. Because of the grouping among 05/30/10 Automatic 36 <67 individual events for the falling data, the group averaging process 07/26/10 Automatic 96 <67 provides a better representation of the data and is less influenced 10/26/10 Automatic 23 <67 by how many falling data points were collected for each storm. In order to provide consistency, the group averaging process was 11/30/10 Manual/Automatic 44 <67 used to determine an instantaneous suspended-sediment transport 02/01/11 Manual/Automatic 46 <67 equation for rising, falling, and all data (composite). The unit value * Automatic—suspended-sediment samples collected by an automatic (15 minute) streamflows for the entire water year were visually water sampler. Manual—manually collected isokenetic suspended-sediment and statistically inspected and the appropriate rising/falling (group samples. averaged) equation was used to compute a suspended-sediment discharge. These values were used to compute an average daily sediment discharge for every day in the 2010 water year. These values were regressed with the average daily flow to provide a daily transport curve (eq. 6; fig. 29): Pole Bridge Creek 39

SSQ =0.0169 (Q)1.9611 for Q ≤ 96 ft3/s (6) It is worth noting that the daily mean sediment concentration where is a time-weighted mean value. Thus, calendar days can be divided SSQ is the average daily suspended-sediment and analyzed in shorter periods of time when water or sediment discharge, in tons per day; and discharge exceed certain limits. The term “subdivide” refers to Q is the average daily streamflow, in cubic feet the division of data for a calendar day into shorter periods of time per second. to obtain correct daily mean values of streamflow or suspended- sediment discharge when one or both change beyond certain limits The application of the daily suspended-sediment transport during the day. The values for these periods are then summed to curve is limited to the maximum measured flow, 96 ft3/s, at which suspended-sediment samples were collected. Suspended-sediment obtain values of suspended-sediment discharge. Computation transport curves have the tendency to flatten out once the available of a daily mean sediment discharge requires subdivision of the sediment is exhausted and additional rainfall no longer increases day when both water discharge and concentration are changing the suspended sediment. Because this exact point is not known, because the average of the products of two variable quantities extrapolation of the suspended-sediment transport curve should be is not the same as the product of the averages of the quantities based on engineering judgment and analysis. (Porterfield, 1972). For comparison, the instantaneous suspended- Annual load and yield values were estimated for suspended sediment transport curve, streamflow, and the mid-interval sediment at Pole Bridge Creek for the data collection period method were used to compute an estimated average annual load (2010–11). The suspended-sediment discharge was estimated for the data collection period, based on a 15-minute interval. This for each day of the water year using the daily transport curve computation resulted in an average annual load that is 6 percent (eq. 6). Load computations were performed by application of the less than the load calculated using the average daily values. It mid-interval method (Porterfield, 1972). The estimated average was determined that the use of the average daily streamflow and annual load for the data collection period, based on average daily streamflow and the Pole Bridge Creek daily transport curve, is the daily transport curve is more practical for estimating average 400 tons, and the average yield is 86 tons/mi2. annual loads.

1,000

SSQ = 0.0169 (Q)1.9611 for Q ≤ 96 cubic feet per second

100

1 Supended-sediment discharge, in tons per day

0.1 EXPLANATION Daily transport curve Instantaneous falling data Instantaneous rising data 0.01 1 10 100 1,000 Streamflow, in cubic feet per second

Figure 19 146). Figure 29. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464385 Pole Bridge Creek near Samantha, Alabama. SSQ, suspended sediment concentration; Q, streamflow in cubic feet per second. 40 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Tierce Creek Northport. The drainage area for this site (fig. 30) is 1.03 mi2. The Tierce Creek streamgage is located 2 mi upstream of the The USGS streamflow-gaging station on Tierce Creek backwater of Lake Tuscaloosa. Although the Tierce Creek above Northport, Ala. (station number 02464503, also streamgage was not established until 2010, various historical referred to as the Tierce Creek streamgage) was established in water-quality and discharge measurements have been collected November 2010 on Tierce Creek at Tierce Creek Road, near 1.2 mi downstream from the gage.

87°34' 87°33' 87°32' 87°31' 87°30' 87°29'

Creek 02464146 Tur key FAYETTE WALKER

33°24'

TUSCALOOSA 69 Location of subwatersheds k e e r C e c r ie T

33°23' 02464503

69 0 0.25 0.5 0.75 1 MILE

33°22' 0 0.25 0.5 0.75 1 KILOMETER

Base from U.S. Geological Survey digital data, 1:100,000, Universal Transverse Mercator Projection, EXPLANATION Zone 16. Tierce Creek subwatershed

02464503 USGS streamgage and number

Figure 30. Location of U.S. Geological Survey streamflow-gaging station 02464503 Tierce Creek near Northport, Alabama, and the contributing drainage area. Tierce Creek 41

Land Use A Land use coverage, in percent In general, land use within the Tierce Creek <1 subwatershed in 2006 (Fry and others, 2011) can be classified as rural. Forested land (cumulative total of 2 4 2 mixed, deciduous, and evergreen) was the predominant 9 basinwide land use in 2006 at 66 percent. Agricultural 4 land use, represented predominantly by hay and pasture or grasslands, accounted for about 18 percent of 7 the subwatershed. Low-, medium-, and high-intensity residential, and developed open space land coverages were 41 combined to compute total urban land use. Urban land use 8 was approximately 4 percent in 2006 (fig. 31). General changes in land use were evaluated by inspecting the NLCD 1992–2001 Land Cover Change Retrofit product (Fry and others, 2009) and direct comparison of the NLCD 2001 (Homer and others, 2004) and 23 the NLCD 2006 (Fry and others, 2011). One of the largest reductions in land use from 1992 to 2001 was forested land that decreased by 6 percent. Direct comparison of NLCD 2001 and 2006 (fig. 31) shows a decrease in forested land B of 6 percent and an increase in agricultural lands of 5 percent. Inspection of aerial photography indicates mini- 2 4 mal land-use changes since completion of the 2006 NLCD. 9 2 Comparison of the NLCD 2006 and aerial photography (2010–11) shows an increase in forested land of 22 acres. 9

Historical Data 8 40 Monthly sediment samples (table 9) were collected 1.2 mi downstream of the current sampling location on Tierce Creek near the Northport streamgage (station num- 8 ber 02464505) for the 1983 water year (October 1, 1982– September 30, 1983). Comparison of the 1983 and 2011 datasets indicates similar suspended-sediment concentra- 18 tion values for lower streamflow. Because samples taken during normal low-flow conditions are a variable function of the available sediment fines, no quantitative conclusions EXPLANATION were drawn from the comparison. Land cover Developed, open space Barren land Deciduous forest Data Collected Evergreen forest Mixed forest Streamflow and suspended-sediment data were Shrub/scrub collected at Tierce Creek at Tierce Creek Road, near Grassland/herbaceous Northport, Ala. (station number 02464503). A streamflow- Pasture/hay Cultivated crops gaging station was installed in November 2010. Stage Woody wetlands values are recorded every 15 minutes and a stage-discharge rating curve was developed. Sediment samples were automatically collected using an automatic water sampler. Figure 31. Percentage of land-use coverage at U.S. Geological Survey streamflow-gaging station 02464503 Tierce Creek near Streamflow data was collected at this site from installation Northport, Alabama. (A) National Land Cover Database 2001. (B) to May 2012. Suspended-sediment samples were collected National Land Cover Database 2006. during the 2011–12 water year. 42 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Table 9. Historical streamflow and suspended-sediment data for U.S. Geological Survey streamflow-gaging station 02464505 Tierce Creek near Northport, Alabama.

[Data are available at http://waterdata.usgs.gov/al/nwis/qw. Dates are by month/day/year. Abbreviations: ft3/s, cubic foot per second; mg/L, milligram per liter]

Mean Suspended Suspended sediment Date streamflow sediment discharge (ft3/s) (mg/L) (tons/day) 10/4/1982 0.98 3 0.01 11/4/1982 1.6 5 0.02 12/2/1982 9.2 16 0.40 1/3/1983 4.8 7 0.09 2/3/1983 11 5 0.15 3/2/1983 7.9 4 0.08 4/5/1983 4.6 3 0.04 5/4/1983 4.9 9 0.12 6/3/1983 4.2 3 0.03 7/6/1983 3.2 8 0.07 8/2/1983 2.6 10 0.07 9/2/1983 2.1 5 0.03

Streamflow streamflow of record was evaluated based on regional regression relations. A flood-frequency relation was Climatic conditions have a direct effect on stream- developed (Hedgecock and Feaster, 2007) and the 67- flow. A comparison was made with streamflow (fig. 32) and 50-percent chance exceedance flood flows for the and the dates of data collection (2011–12) to determine Tierce Creek streamgage are 190 and 260 ft3/s, respec- how climatic conditions during this investigation may tively. The peak streamflow experienced during the have influenced the data collected and may limit the suspended-sediment data-collection period (2011–12) applicability of the regression equations developed was 110 ft3/s. The largest peak streamflow that was 3 by this investigation to periods of differing climatic sampled during the collection period was 110 ft /s, conditions. The Tierce Creek streamgage does not which is less than the streamflow that corresponds to a have a long period of record (10 or more years) and 67-percent chance exceedance flood. therefore assumptions cannot be made on the average streamflow conditions for the site. However, the peak Tierce Creek 43

A 10,000

5,000 EXPLANATION Period of suspended-sediment data collection 2,000

1,000

500

200

100

1 OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. 2010 2011 B 10,000

5,000 EXPLANATION Period of suspended-sediment data collection 2,000 Mean discharge, in cubic feet per second

1,000

500

200

100

1 OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. 2011 2012

Figure 32. Streamflow hydrograph at U.S. Geological Survey streamflow-gaging station 0246503 Tierce Creek near Northport, Alabama. (A) 2011 water year and (B) 2012 water year. 44 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Suspended Sediment Estimation of Suspended-Sediment Loads from Suspended-sediment samples were collected at the Tierce Tierce Creek Creek streamgage as described in the Suspended-Sediment The suspended-sediment concentrations of the automatic Data Collection section (see Approach and Methods). pump samples were adjusted to reflect average streamflow- Samples were collected for 10 storm events (table 10) during weighted suspended-sediment concentrations as described 4 of the events, samples were collected by both manual and in the Data Analysis section (see Approach and Methods). automatic methods. In addition to the storm events, samples The suspended-sediment concentrations were converted were collected on a monthly basis to represent normal daily to suspended-sediment discharges (SSQ) and graphically streamflows. The samples were collected to reflect seasonal and statistically compared with streamflow Q( ) to develop variations and their variable antecedent conditions. Figure 32 instantaneous sediment-transport curves. A linear regression illustrates the periods of data collection and how they relate equation (ordinary least squares) was developed, with a smear- to the annual hydrograph. The automatic pump sampler was ing estimator applied, for the rise and fall of the hydrograph set to collect samples at a time interval ranging from 15 to and using all data (composite). The resulting instantaneous 25 minutes to ensure that suspended-sediment concentrations sediment-transport curves had almost the same slope. The were collected over the rise, peak, and fall of the hydrograph. composite instantaneous sediment transport curve was used The sediment hydrographs for Tierce Creek were graphically to compute an average daily sediment discharge for every day compared to the streamflow hydrographs. The majority of (November 2010–October 2011). These values were regressed the runoff events sampled had a sediment hydrograph whose with the average daily flow to provide a daily transport curve peak tracked the streamflow hydrograph, which could be an (eq. 7; fig. 33): indication of a uniform source of sediment and (or) a lack of sediment stored in the channel. SSQ = 0.2249 (Q)1.3927 for Q ≤ 110 ft3/s (7) where SSQ is the average daily suspended-sediment discharge, in tons per day; and Q is the average daily streamflow, in cubic feet Table 10. Summary of suspended-sediment samples collected per second. at U.S. Geological Survey streamflow-gaging station 02464503 Tierce Creek near Northport, Alabama. The application of the daily suspended-sediment transport [Data are available at http://waterdata.usgs.gov/al/nwis/qw. Dates are by curve is limited to the maximum measured flow, 110 ft3/s, month/day/year. Abbreviations: ft3/s, cubic foot per second; <, less than] at which suspended-sediment samples were collected. Suspended-sediment transport curves have the tendency Peak Annual to flatten out once the available sediment is exhausted and Collection Date streamflow exceedance type* additional rainfall no longer increases the suspended sediment. (ft3/s) probability Because this exact point is not known, extrapolation of the 3/5/11 Manual/Automatic 21 <67 suspended- sediment transport curve should be based on 3/8/11 Manual/Automatic 57 <67 engineering judgment and analysis. Annual load and yield values were estimated for 4/11/11 Automatic 7 <67 suspended sediment at Tierce Creek for the collection period 4/15/11 Automatic 19 <67 (November 2010–October 2011). The suspended-sediment 9/5/11 Manual/Automatic 110 <67 discharge was estimated for each day of the water year using the 9/19/11 Automatic 26 <67 daily transport curve (eq. 7). Load computations were performed by application of the mid-interval method (Porterfield, 1972). The 11/27/11 Automatic 17 <67 estimated average annual load for the collection period, based 1/11/12 Automatic 32 <67 on average daily streamflow and the Tierce Creek daily transport 1/23/12 Automatic 18 <67 curve is 200 tons, and the average yield is 190 tons/mi2. 1/26/12 Manual/Automatic 25 <67 It is worth noting that the daily mean sediment concentration is a time-weighted mean value. Thus, calendar days can be divided * Automatic—suspended-sediment samples collected by an automatic water sampler. Manual—manually collected isokenetic suspended-sediment and analyzed in shorter periods of time when water or sediment samples. discharge exceed certain limits. The term “subdivide” refers to the division of data for a calendar day into shorter periods of time to obtain correct daily mean values of streamflow or suspended- sediment discharge when one or both change beyond certain limits during the day. The values for these periods are then summed to obtain values of suspended-sediment discharge. Computation of a daily mean sediment discharge requires subdivision of the day Carroll Creek 45

1,000

SSQ = 0.2249 (Q)1.3927 for Q ≤ 110 cubic feet per second

100

Supended-sediment discharge, in tons per day 1

EXPLANATION Daily transport curve Instantaneous falling data Instantaneous rising data

0.1 1 10 100 1,000 Stream ow (ft Streamflow, in cubic feet per second Figure 19 . Daily suspended-sediment transport curve fo Tr urkey Creek stream ow gaging station (USGS 02464146). Figure 33. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464503 Tierce Creek near Northport, Alabama. SSQ, suspended sediment concentration; Q, streamflow in cubic feet per second. when both water discharge and concentration are changing because investigation, a 3.5-mi reach of Carroll Creek was surveyed to the average of the products of two variable quantities is not the same prepare a current bathymetric map, determine storage capacities at as the product of the averages of the quantities (Porterfield, 1972). For specified water-surface elevations, and compare current conditions comparison, the instantaneous suspended-sediment transport curve, to historical cross sections. streamflow, and the mid-interval method were used to compute an The USGS streamflow-gaging station on Carroll Creek near estimated average annual load for the collection period, based on a Northport, Ala. (station number 02464660, also referred to as the 15-minute interval. This computation resulted in an average annual Carroll Creek streamgage) was established in November 2008 on load that is 15 percent more than the load calculated using the average Carroll Creek at State Highway 69, near Northport. Although the daily values. It was determined that the use of the average daily streamgage was established in 2008, various historical water qual- streamflow and the daily transport curve is more practical for estimat- ity and discharge measurements have been made at this location. ing average annual loads. The drainage area for this site (fig. 36) is 20.9 mi2 and is 3.75 mi upstream from the mouth of Carroll Creek and the reservoir.

Carroll Creek Land Use

Carroll Creek is one of six subwatersheds (Carroll Creek, In general, land use within the Carroll Creek subwatershed Binion Creek, headwaters of the North River, upper North River, in 2006 (Fry and others, 2011) can be classified as rural. Forested middle North River, and lower North River) that drain into Lake land (cumulative total of mixed, deciduous, and evergreen) was the Tuscaloosa (fig. 1, table 1). Carroll Creek at the mouth of the main predominant basinwide land use in 2006 at 59 percent. Agricultural body of the reservoir has a drainage area of 25.6 mi2 and contrib- land use, represented predominantly by hay and pasture or grasslands, utes approximately 5 percent of the surface drainage area to Lake accounted for about 11 percent of the subwatershed. Low-, medium-, Tuscaloosa (fig 34). Carroll Creek is in backwater of the reservoir and high-intensity residential, and developed open space land for approximately 3.5 mi. The upstream extent of the backwater coverages were combined to compute total urban land use. Urban land portion of Carroll Creek has experienced increased sediment use was approximately 11 percent in 2006 (fig. 37). Inspection of aerial deposition (figs. 35 and 36). The City of Tuscaloosa dredged photography indicates urban development since completion of the portions of Carroll Creek in 2009. As a byproduct of this scientific 2006 NLCD. 46 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

87°40' #0 87°36'

33°22'

sa o lo a sc u Carroll Creek T e ak L 43 171

33°20' 69

FAYETTE WALKER

33°18' 171 02464660 43 TUSCALOOSA 69 Location of subwatersheds

Base from U.S. Geological Survey digital data, 1:100,000, Universal Transverse Mercator Projection, EXPLANATION 0 0.5 1 1.5 2 MILES Zone 16. Carroll Creek subwatershed 0 0.5 1 1.5 2 KILOMETERS 02464660 USGS streamgage and number

Figure 34. Location of U.S. Geological Survey streamflow-gaging station 02464660 Carroll Creek near Northport, Alabama, and the contributing drainage area. Carroll Creek 47

Figure 35. Aerial photograph showing the upstream extent of the backwater portion of Carroll Creek near Northport, Alabama, November 29, 2007. Lake Tuscaloosa water-surface elevation is 221.04 feet (National Geodetic Vertical Datum of 1929).

FigureFigure 36. 36. AerialAerial photographphotograph showingshowing thethe upstreamupstream extentextent ofof thethe backwaterbackwater portionportion ofof CarrollCarroll CreekCreek near near Northport, Northport, Alabama, Alabama, March March 24, 24, 2009. 2009. Lake Lake Tuscaloosa Tuscaloosa water-surface water-surface elevation elevation is is 223.74223.74 feet feet (National (National Geodetic Geodetic Vertical Vertical Datum Datum of of 1929). 1929).

kgl_SedInflows_fig36.ai 48 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

A B Land use coverage, in percent 1 <1 <1 1 1 5 5 2 5 6 1 2 3 4 1 8 9

13 12 35 35

17 18

9 7 EXPLANATION Land cover Open water Mixed forest Developed, open space Shrub/scrub Developed, low intensity Grassland/herbaceous Developed, medium intensity Pasture/hay Barren land Cultivated crops Deciduous forest Woody wetlands Evergreen forest

Figure 37. Percentage of land-use coverage at U.S. Geological Survey streamflow-gaging station 02464660 Carroll Creek near Northport, Alabama. (A) National Land Cover Database 2001 and (B) National Land Cover Database 2006.

General changes in land use were evaluated by Historical Data inspecting the NLCD 1992–2001 Land Cover Change Retrofit product (Fry and others, 2009) and direct com- Monthly sediment samples (table 11) were collected at parison of the NLCD 2001(Homer and others, 2004) and the same location as the Carroll Creek streamgage (station the NLCD 2006 (Fry and others, 2011). One of the largest number 02464660) for the 1983 water year (October 1982– reductions in land use from 1992 to 2001 was forested land September 1983). The historical data were used to show that decreased by 3 percent. During this period, urbaniza- temporal trends because the land-use changes described tion increased by 0.5 percent. Direct comparison of NLCD above indicate an increase in urbanization. Comparison 2001 and 2006 (fig. 38), show a decrease in forested land of the 1983 and 2009 datasets show a slight increase in of 3 percent and an increase in urbanization of 3 percent. Comparison of the NLCD Impervious Cover (2006) and suspended-sediment concentrations. Because samples aerial photography (2009) (fig. 38) shows new develop- taken during normal low-flow conditions are a variable ment in the southeastern portion of the basin, as indicated function of the available sediment fines, no quantitative by impervious cover. conclusions were drawn from the comparison. Carroll Creek 49

87°40' FAYETTE WALKER

TUSCALOOSA

Location of subwatersheds 33°20' 33°20'

87°36' 87°40'

Base from U.S. Geological Survey digital data, 1:100,000, Universal Transverse Mercator Projection, Zone 16.

2464660 0 0.5 1 1.5 2 MILES #0 02464660

0 0.5 1 1.5 2 KILOMETERS 87°36' EXPLANATION

National Land Cover Database Impervious Cover

Carroll Creek subwatershed and drainage divide

02464660 USGS streamgage and number Figure 38. Comparison of aerial photography (2009) and National Land Cover Database Impervious Cover (2006) for the Carroll Creek subwatershed draining into Lake Tuscaloosa, Alabama.

Table 11. Historical streamflow and suspended-sediment data for U.S. Geological Survey streamflow-gaging station 02464660 Carroll Creek near Northport, Alabama.

[Data are available at http://waterdata.usgs.gov/al/nwis/qw. Dates are by month/day/year. Abbreviations: ft3/s, cubic foot per second; mg/L, milligram per liter]

Mean Suspended Suspended sediment Date streamflow sediment discharge (ft3/s) (mg/L) (tons/day) 10/4/1982 1.80 34 0.16 11/4/1982 14 12 0.45 12/3/1982 74 39 7.8 1/3/1983 57 35 5.4 2/4/1983 94 19 4.8 3/2/1983 111 31 9.3 4/4/1983 38 9 0.92 6/2/1983 22 13 0.77 7/5/1983 13 14 0.49 8/1/1983 9.80 17 0.45 9/1/1983 4.90 9 0.12 50 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Data Collected and fall of the hydrograph. The sediment hydrographs for Carroll Creek were graphically compared to the streamflow Streamflow and suspended-sediment data were collected hydrographs. The majority of the runoff events sampled had at Carroll Creek at State Highway 69 near Northport, Ala. a sediment hydrograph whose peak preceded the streamflow (station number 02464660). A streamflow-gaging station was hydrograph (clockwise hysteresis), which is often ascribed to installed in November 2008. Stage values are recorded every resuspension of sediment from the stream channel at the initia- 15 minutes and a stage-discharge rating curve was developed. tion of storm runoff and to a relatively limited sediment supply Sediment samples were automatically collected using an on the stormflow recession. automatic water sampler. Streamflow data have been collected at this site from installation to present. Suspended-sediment samples were collected during the 2009 water year. Estimation of Suspended-Sediment Loads from Carroll Creek

Streamflow The suspended-sediment concentrations of the automatic pump samples were adjusted to reflect average streamflow- Climatic conditions have a direct effect on streamflow. A weighted suspended-sediment concentrations as described comparison was made with streamflow (fig. 39) and the dates in the Data Analysis section (see Approach and Methods). of data collection to determine how climatic conditions during The suspended-sediment concentrations were converted this investigation may have influenced the data collected to suspended-sediment discharges (SSQ) and graphically and may limit the applicability of the regression equations and statistically compared with streamflow Q( ) to develop developed by this investigation to periods of differing climatic instantaneous sediment-transport curves. A linear regression conditions. Average annual mean streamflow (2009–11) was equation (ordinary least squares) was developed, and a smear- examined for the period of record and for the individual water ing estimator applied, for the rise and fall of the hydrograph years (fig. 40). The suspended-sediment samples collected in and using all data (composite). The resulting instantaneous 2009 were collected during a period when annual streamflow sediment-transport curves had almost the same slope, with magnitude was above average. The Carroll Creek streamgage the rising equation having a higher intercept. The composite does not have a long period of record (10 or more years) and instantaneous sediment-transport curve was used to compute therefore assumptions cannot be made about the average an average daily sediment discharge for every day in the 2009 streamflow conditions for the site. However, the peak stream- water year. These values were regressed with the average daily flow of record was evaluated based on regional regression flow to provide a daily transport curve (eq. 8; fig. 41): relations. A flood-frequency relation was developed (Hedge- cock and Feaster, 2007) and the 67- and 50-percent chance SSQ =0.0627 (Q)1.6113 for Q ≤ 670 ft3/s (8) exceedance flood flows for the Carroll Creek streamgage where 3 are 1,390 and 1,840 ft /s, respectively. The peak streamflow SSQ is the average daily suspended-sediment experienced during the suspended-sediment data-collection discharge, in tons per day; and 3 period (2009) was 1,630 ft /s. The largest peak streamflow that Q is the average daily streamflow, in cubic feet 3 was sampled during the collection period was 670 ft /s, which per second. is less than the streamflow that corresponds to a 67-percent chance exceedance flood. The application of the daily suspended-sediment transport curve is limited to the maximum measured flow, 670 ft3/s, at Suspended Sediment which suspended-sediment samples were collected. Suspended- sediment transport curves have the tendency to flatten out once Suspended-sediment samples were collected at the the available sediment is exhausted and additional rainfall no Carroll Creek streamgage as described in the Suspended- longer increases the suspended sediment. Because this exact point Sediment Data Collection section (see Approach and is not known, extrapolation of the suspended-sediment transport Methods). Samples were collected for nine storm events curve should be based on engineering judgment and analysis. (table 12); during four of the events, samples were collected Annual load and yield values were estimated for suspended by both manual and automatic methods. In addition to the sediment at Carroll Creek for the 2009 water year. The storm events, samples were collected on a monthly basis to suspended-sediment discharge was estimated for each day of the represent normal daily streamflows. The samples were col- water year using the daily transport curve (eq. 8). The maximum lected to reflect seasonal variations and their variable ante- average daily streamflow for the period of estimation is 1,070 cedent conditions. Figure 39 illustrates the periods of data ft3/s (February 28, 2009). This exceeds the maximum streamflow collection and how they relate to the annual hydrograph. The at which sediment was sampled. The daily transport curve was automatic pump sampler was set to collect samples at a time extended linearly to estimate suspended-sediment discharge for interval ranging from 15 to 45 minutes to ensure that suspend- this period of high flow. Load computations were performed by ed-sediment concentrations were collected over the rise, peak, application of the mid-interval method (Porterfield, 1972). Brush Creek 51

The estimated average annual load for the 2009 water year, others, 2011). The NLCD 1992–2001 Land Cover Change based on average daily streamflow and the Carroll Creek daily Retrofit product indicated no substantial change in land use transport curve, is 17,600 tons, and the average yield is 840 from 1992 to 2001. Direct comparison of NLCD 2001 and tons/mi2. 2006 (fig. 43) indicates a decrease in forested land of 10 It is worth noting that the daily mean sediment concentra- percent and an increase in shrub/scrub land use of 10 percent. tion is a time-weighted mean value. Thus, calendar days can Inspection of aerial photography indicates a decrease in be divided and analyzed in shorter periods of time when water forested land since completion of the 2006 NLCD. Compari- or sediment discharge exceed certain limits. The term “sub- son of the NLCD and aerial photography indicates that in the divide” refers to the division of data for a calendar day into time period of 2006 to 2010, approximately 74 acres of the shorter periods of time to obtain correct daily mean values of drainage basin were clear cut. streamflow or suspended-sediment discharge when one or both change beyond certain limits during the day. The values for these periods are then summed to obtain values of suspended- Historical Data sediment discharge. Computation of a daily mean sediment discharge requires subdivision of the day when both water Monthly sediment samples (table 13) were col- discharge and concentration are changing because the average lected at the Brush Creek near Northport streamgage of the products of two variable quantities is not the same (station number 02464680) for the 1983 water year as the product of the averages of the quantities (Porterfield, (October 1, 1982–September 30, 1983). Comparison of the 1972). For comparison, the instantaneous suspended-sediment 1983 and 2011 datasets indicates a similar trend for the lower transport curve, streamflow, and the mid-interval method were values of streamflow. Because samples taken during normal used to compute an estimated average annual load for the 2009 low-flow conditions are a variable function of the available water year, based on a 15-minute interval. This computation sediment fines, no quantitative conclusions were drawn from resulted in an average annual load that is 12 percent less the comparison. than the load calculated using the average daily values. It was determined that the use of the average daily streamflow Data Collected and the daily transport curve is more practical for estimating average annual loads. Streamflow and suspended-sediment data were collected at Brush Creek at Turner Road, near Northport, Ala. (station number 02464680). A streamflow-gaging station was installed Brush Creek in November 2010. Stage values are recorded every 15 minutes and a stage-discharge rating curve was developed. The USGS streamflow-gaging station on Brush Creek Sediment samples were automatically collected using an near Northport, Ala. (station number 02464680, also automatic water sampler. Streamflow data were collected at referred to as the Brush Creek streamgage) was established this site from installation to present. Suspended-sediment in November 2010 on Brush Creek at Turner Road, near samples were collected during the 2011–12 water years. Northport. The drainage area (fig. 42) for this site is 0.92 mi2, and it is located 0.3 mi upstream of the backwater of Lake Streamflow Tuscaloosa. Although the streamgage was established in 2010, various historical water-quality and discharge measurements Climatic conditions have a direct effect on streamflow. A have been collected at this site. comparison was made with streamflow (fig. 44) and the dates of data collection (2011–12) to determine how climatic conditions during this Land Use investigation may have influenced the data collected and may limit the applicability of the regression equations developed by this investiga- In general, land use within the Brush Creek subwatershed tion to periods of differing climatic conditions. The Brush Creek in 2006 (Fry and others, 2011) can be classified as rural. streamgage does not have a long period of record (10 or more years) Forested land (cumulative total of mixed, deciduous, and and therefore assumptions cannot be made on the average streamflow evergreen) was the predominant basinwide land use in 2006 at conditions for the site. However, the peak streamflow of record was 84 percent. Shrub/scrub, represented predominantly by shrubs evaluated based on regional regression relations. A flood-frequency less than 16 ft tall with shrub canopy typically greater than relation was developed (Hedgecock and Feaster, 2007) and the 67- and 20 percent of total vegetation, accounted for about 10 percent 50-percent chance exceedance flood flows for the Brush Creek of the subwatershed (fig. 43). streamgage are 170 and 240 ft3/s, respectively. The peak streamflow General changes in land use were evaluated by inspecting experienced during the suspended-sediment data-collection period the NLCD 1992–2001 Land Cover Change Retrofit product (2011–12) was 140 ft3/s. The largest peak streamflow that was sampled (Fry and others, 2009) and direct comparison of the NLCD during the collection period was 130 ft3/s, which is less than the 2001 (Homer and others, 2004) and the NLCD 2006 (Fry and streamflow that corresponds to a 67-percent chance exceedance flood. 52 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

10,000

5,000

2,000

1,000

500

200

100

10 Mean discharge, in cubic feet per second

1 OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. 2008 2009 YEAR EXPLANATION Period of suspended-sediment data collection Figure 39. Streamflow hydrograph for the 2009 water year at U.S. Geological Survey streamflow- gaging station 02464660 Carroll Creek near Northport, Alabama.

35 3,500

32.9 30 3,000 30.6

25 2,500 EXPLANATION Annual mean streamﬂow, in cubic feet per second 20 2,000 in cubic feet per second

Peak ﬂow, in cubic feet per second , w

o Average annual mean streamﬂow (2009–2011) l

f 17.7

15 1,500 in cubic feet per second 1,630 a m , 50 percent chance exceedance probability e w r t o l s f

a n a k e 10 1,000 e P m

a l 970 u n n

A 5 500 564

0 0 2009 2010 2011 Water year, October–September

Figure 40. Annual mean streamflow for U.S. Geological Survey streamflow-gaging station 02464660 Carroll Creek near Northport, Alabama. Brush Creek 53

Table 12. Summary of suspended-sediment samples collected at U.S. Geological Survey streamflow-gaging station 02464660 Carroll Creek near Northport, Alabama.

[Data are available at http://waterdata.usgs.gov/al/nwis/qw. Dates are by month/day/year. Abbreviations: ft3/s, cubic foot per second; <, less than]

Peak Annual Collection Date streamflow exceedance type* (ft3/s) probability 12/10/2008 Manual/Automatic 250 <67 2/27/2009 Manual/Automatic 670 <67 3/14/2009 Automatic 83 <67 3/25/2009 Automatic 436 <67 4/2/2009 Automatic 65 <67 5/14/2009 Automatic 44 <67 5/24/2009 Automatic 221 <67 8/11/2009 Manual/Automatic 134 <67 8/20/2009 Manual/Automatic 285 <67 * Automatic—suspended-sediment samples collected by an automatic water sampler. Manual—manually collected isokenetic suspended-sediment samples.

10,000

SSQ = 0.0627 (Q)1.6113 for Q ≤ 670 cubic feet per second

1,000

100

Supended-sediment discharge, in tons per day 10

EXPLANATION Daily transport curve Instantaneous falling data Instantaneous rising data

1 10 100 1,000

Streamflow, in cubic feet per second Figure 19 . Daily suspended-sediment transport curve fo Tr urkey Creek stream ow gaging station (USGS 02464146). Figure 41. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464660 Carroll Creek near Northport, Alabama. SSQ, suspended sediment concentration; Q, streamflow in cubic feet per second. 54 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

87°32' 87°30' 87°28'

FAYETTE WALKER

33°20' TUSCALOOSA

Location of subwatersheds

Brush Creek 02464680

33°19'

SA OO CAL TUS KE EXPLANATION LA Brush Creek subwatershed

02464680 USGS streamgage and number

Base from U.S. Geological Survey digital data, 0 0.25 0.5 0.75 1 MILE 1:100,000, Universal Transverse Mercator Projection, Zone 16. 0 0.25 0.5 0.75 1 KILOMETER

Figure 42. Location of U.S. Geological Survey streamflow-gaging station 02464680 Brush Creek near Northport, Alabama, and the contributing drainage area. Brush Creek 55

Land use coverage, in percent 1

2 3

14 17

Table 13. Historical streamflow and suspended-sediment data for U.S. Geological Survey streamflow-gaging station 02464680 Brush Creek near Northport, Alabama.

[Data are available at http://waterdata.usgs.gov/al/nwis/qw. Dates are by 6633 month/day/year. Abbreviations: ft3/s, cubic foot per second; mg/L, milligram per liter]

B Mean Suspended Suspended sediment <1 Date streamflow sediment discharge (ft3/s) (mg/L) (tons/day) 3 3 10/4/1982 0.88 11 0.03 10 14 11/4/1982 1.2 0 0 12/6/1982 2.8 5 0.04 1/4/1983 2 10 0.05 2/3/1983 3.7 4 0.04 16 3/2/1983 3.4 3 0.03 4/4/1983 2.6 1 0.01 5/2/1983 2.6 7 0.05 6/2/1983 2.7 11 0.08 7/5/1983 2.5 20 0.14 5544 8/2/1983 1.9 10 0.05 9/2/1983 1.6 8 0.03

EXPLANATION Land cover Open water Developed, open space Deciduous forest Evergreen forest Mixed forest Shrub/scrub Grassland/herbaceous Pasture/hay

Figure 43. Percentage of land-use coverage at U.S. Geological Survey streamflow-gaging station 02464680 Brush Creek near Northport, Alabama. (A) National Land Cover Database 2001. (B) National Land Cover Database 2006. 56 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

A 10,000

5,000 EXPLANATION Period of suspended-sediment data collection 2,000

1,000

500

200

100

1 OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. 2010 2011 B 10,000 EXPLANATION 5,000 Period of suspended-sediment data collection 2,000 Mean discharge, in cubic feet per second 1,000

500

200

100

1 OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. 2011 2012

Figure 44. Streamflow hydrograph at U.S. Geological Survey streamflow-gaging station 02464680 Brush Creek near Northport, Alabama. (A) 2011 water year. (B) 2012 water year. Brush Creek 57

Suspended Sediment and fall of the hydrograph and using all data (composite). The composite instantaneous sediment transport curve was Suspended-sediment samples were collected at the Brush used to compute an average daily sediment discharge for Creek streamgage as described in the Suspended-Sediment Data every day (November 2010–October 2011). These values Collection section (see Approach and Methods). Samples were were regressed with the average daily flow to provide a daily collected for six storm events (table 14) during four of the events, transport curve (eq. 9; fig. 45): samples were collected by both manual and automatic methods. In addition to the storm events, samples were collected on a monthly SSQ =0.2798 (Q)1.5774 for Q ≤ 130 ft3/s (9) basis to represent normal daily streamflows. The samples were where collected to reflect seasonal variations and their variable anteced- SSQ is the average daily suspended-sediment ent conditions. Figure 44 illustrates the periods of data collection and how they relate to the annual hydrograph. The automatic discharge, in tons per day; and pump sampler was set to collect samples at a time interval of Q is the average daily streamflow, in cubic 15 minutes to ensure that suspended-sediment concentrations feet per second. were collected over the rise, peak, and fall of the hydrograph. The sediment hydrographs for Brush Creek were graphically The application of the daily suspended-sediment compared to the streamflow hydrographs. The majority of the rain transport curve is limited to the maximum measured flow, events sampled had a sediment hydrograph whose peak tracked 130 ft3/s, at which suspended-sediment samples were the streamflow hydrograph, which could be an indication of a collected. Suspended-sediment transport curves have the uniform source of sediment and (or) a lack of sediment stored in tendency to flatten out once the available sediment is the channel. exhausted and additional rainfall no longer increases the suspended sediment. Because this exact point is not known, Table 14. Summary of suspended-sediment samples collected at extrapolation of the suspended-sediment transport curve U.S. Geological Survey streamflow-gaging station 02464680 Brush should be based on engineering judgment and analysis. Creek near Northport, Alabama. Annual load and yield values were estimated for [Data are available at http://waterdata.usgs.gov/al/nwis/qw. Dates are by suspended sediment at Brush Creek for the collection period month/day/year. Abbreviations: ft3/s, cubic foot per second; <, less than] (November 2010–October 2011). The suspended-sediment discharge was estimated for each day of the water year Peak Annual Collection using the daily transport curve (eq. 9). Load computations Date streamflow exceedance type* were performed by application of the mid-interval method (ft3/s) probability (Porterfield, 1972). The estimated average annual load for 2/25/11 Automatic 18 <67 the collection period, based on average daily streamflow and 2/28/11 Manual/Automatic 7.3 <67 the Brush Creek daily transport curve, is 280 tons, and the 2 3/5/11 Manual/Automatic 35 <67 average yield is 300 tons/mi . It is worth noting that the daily mean sediment concentration 3/9/11 Manual/Automatic 70 <67 is a time-weighted mean value. Thus, calendar days can be divided 9/5/11 Manual/Automatic 130 <67 and analyzed in shorter periods of time when water or sediment 1/26/12 Automatic 18 <67 discharge exceed certain limits. The term “subdivide” refers to * Automatic—suspended-sediment samples collected by an automatic the division of data for a calendar day into shorter periods of time water sampler. Manual—manually collected isokenetic suspended-sediment to obtain correct daily mean values of streamflow or suspended- samples. sediment discharge when one or both change beyond certain limits during the day. The values for these periods are then summed to Estimation of Suspended-Sediment Loads from obtain values of suspended-sediment discharge. Computation Brush Creek of a daily mean sediment discharge requires subdivision of the day when both water discharge and concentration are changing The suspended-sediment concentrations of the because the average of the products of two variable quantities automatic pump samples were adjusted to reflect average is not the same as the product of the averages of the quantities streamflow-weighted suspended-sediment concentrations (Porterfield, 1972). For comparison, the instantaneous suspended- as described in the Data Analysis section (see Approach sediment transport curve, streamflow, and the mid-interval method and Methods). The suspended-sediment concentrations were used to compute an estimated average annual load for the were converted to suspended-sediment discharges and collection period, based on a 15-minute interval. This computation graphically and statistically compared with streamflow resulted in an average annual load that is 6 percent more than the (Q) to develop instantaneous sediment-transport curves. load calculated using the average daily values. It was determined A linear regression equation (ordinary least squares) was that the use of the average daily streamflow and the daily transport developed, and a smearing estimator applied, for the rise curve is more practical for estimating average annual loads. 58 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

1,000 SSQ = 0.2798 (Q)1.5774 for Q ≤ 130 cubic feet per second

100

1 Supended-sediment discharge, in tons per day

EXPLANATION Daily transport curve Instantaneous falling data Instantaneous rising data

0.1 1 10 100 1,000 Streamflow, in cubic feet per second

146). Figure 45. Daily suspended-sediment transport curve for U.S. Geological Survey streamflow-gaging station 02464680 Brush Creek near Northport, Alabama. SSQ, suspended sediment concentration; Q, streamflow in cubic feet per second,.

Estimation of Suspended-Sediment period of streamflow data record for the long-term stations (North River, Turkey Creek, Binion Creek, and Carroll Loads to Lake Tuscaloosa Creek) shows that annual runoff for the period during data collection was well above the average values for 2009 and Quantification of sediment inflows from the contribut- 2010 and below the average values for 2011. This differing watersheds is useful to help focus management efforts ence is illustrated by the North River streamgage. The in targeting areas around the reservoir. The comparison average annual runoff for the period of record (1939–2011) of the seven tributaries monitored in this study can be is 22.9 in. The average runoff for the 2009–11 water years made by examining the loads and resulting yields for each is 28.6, 32.4, and 19.7 in., respectively. Because of the site. Loads were computed for each site for the period of close proximity of the sites, it can be assumed that the sediment data collection (table 15). The estimated load is North River streamgage is a good representative for all computed from the daily average flow and daily sediment seven sites. transport curve for each site. Because the installation of Because the sampling sites vary in drainage area some streamgages was not finished by the beginning of 2 the water year, the estimated load corresponds to a 1-year from 0.92 to 223 mi , the best comparison is the load per period beginning at the time of installation (table 15). The unit area, or yield. The 2009 estimated yields for North estimated load is dependent on the hydrologic conditions River, Turkey Creek, and Carroll Creek are 360, 540, and 2 during the period for which it is estimated. The study area 840 tons/mi , respectively. Based on this comparison, was under drought conditions during portions of the data Carroll Creek contributes more sediment per square mile to collection period. Therefore, the loads/yields estimated are the Lake Tuscaloosa reservoir than North River or Turkey compared based on the year for which they were estimated. Creek. Inspection of table 15 shows that the Carroll Creek This assumes that the hydrologic conditions were roughly streamgage had a smaller peak flow (higher probability the same at the selected sites during the year of data of exceedance) during data collection than North River or collection. Comparison of the annual runoff for the period Turkey Creek, and yet it still had a higher yield than the of data collection and the average runoff for the entire other sites. Estimation of Suspended-Sediment Loads to Lake Tuscaloosa 59

) 2

86 360 540 120 190 840 300 yield sediment (tons/mi Estimated suspended-

of load Time Time period Oct 2011 Oct 2011 Oct 2009 Dec 2010 estimated Sept 2009 Sept 2009 Sept 2010 Jan 2010– Oct 2008– Oct 2008– Oct 2009– Nov 2010– Nov 2008– Nov 2010–

400 200 280 3,350 6,860 load 81,400 17,600 (tons) sediment Estimated suspended-

>10 % <20% >50% <67 % <67 % <67 % <67 % ance of sampled maximum of exceed Probability streamflow

/s) 3 ≤ 96 ≤ 110 ≤ 670 ≤ 130 (ft Daily curve limits ≤ 1,020 ≤ 2,900 Q ≤ 15,400 Q Q Q transport Q Q Q streamfllow

1.4483 1.3919 1.9611 1.3927 1.6113 1.5774 1.2689 ) ) ) ) ) ) ) Q Q Q Q Q Q Q , streamflow in cubic feet per second; %, percent; <, less than; >, greater ≤, than , streamflow in cubic feet per second; =0.0135 ( =0.1858 ( =0.043 ( =0.0169 ( =0.2249 ( =0.0627 ( =0.2798 ( Q Daily transport curve equation SSQ SSQ SSQ SSQ SSQ SSQ SSQ

period Sediment July 2009 collection March 2011 August 2009 March 2011– March 2010– January 2012 January 2012 February 2011 Jaunary 2009– October 2009– February 2011– September 2009 December 2008– December 2008–

— was year 2011 2011 2009 2010 2010 2009 ISCO Water Water sampler installed , suspended-sediment concentration;

SSQ — — — — date gage Streamflow March 2011 March 2012 March 2012 discontinued

date gage 1938 1981 1986 2010 2010 2008 2010 January , tons per square mile; October February 2 December November November November installation Streamflow

) 2 6.16 4.67 1.03 0.92 57.2 20.9 (mi area 223 Drainage

/s, cubic foot per second; tons/mi 3 Station name Samantha AL Samantha AL Tuscaloosa, Creek near Samantha, AL AL Samantha, Northport, AL Hwy 69 nr Northport, AL Northport, AL Summary of suspended-sediment data collection, transport curves, and loads for seven tributaries to the Lake Tuscaloosa reservoir, Alabama. reservoir, Summary of suspended-sediment data collection, transport curves, and loads for seven tributaries to the Lake Tuscaloosa North River near Creek near Turkey Binion Creek below Gin Pole Bridge Creek near Creek above Tierce Carroll Creek at St Brush Creek near

, square mile; ft 2 Station number 02464000 02464146 02464360 02464385 02464503 02464660 02464680 Table 15. Table [mi or equal to; ≥, greater than to] 60 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

The 2010 estimated yields for Binion Creek and Pole Turkey Creek, Binion Creek, and Carroll Creek for the entire Bridge Creek are 120 and 86 tons/mi2, respectively. The study period (2009–11). The daily transport curves were used Binion Creek streamgage was established 1986; flow to estimate sediment loads and yields for each site where data statistics indicate that the annual mean streamflow was were available; however, it should be noted that the Turkey above average in 2010 and was higher than 2009 and 2011. Creek and Binion Creek drainage basins had ongoing timber The largest peak flow recorded at the streamgage during data harvest operations, and so these numbers are for comparison collection was greater than the 50-percent chance exceedance. purposes only. Comparison of the sampling sites over the It should be noted that both basins had timber harvested during 3 years shows that Carroll Creek consistently had the highest the data collection period. estimated sediment yield during the study period. The sam- The 2011 estimated yields for Tierce Creek and pling site with the second highest sediment yield was Turkey Brush Creek are 190 and 300 tons/mi2, respectively. These Creek. streamgages were newly installed and do not have flow Comparison of the sites monitored indicates that Carroll statistics to qualify the hydrologic conditions based on histori- Creek is the largest producer of suspended-sediment under cal data. However, based on flood-frequency relations, both similar hydrologic conditions and the largest producer per sites had peak streamflows less than the 67-percent probability square mile for the timeframes estimated. Aerial photography of exceedance, indicating drier hydrologic conditions for the and bathymetric surveys also support that Carroll Creek has 2011 estimation period. experienced increased sediment deposition in the upstream Another quantitative approach for comparing the seven extent of the backwater. Carroll Creek is also the only basin suspended-sediment sampling sites is to estimate the sediment that has a substantial percentage (11 percent) of urban land load/yield for each year where streamflow data are available use. It is also the only basin that experienced an increase in (table 16). Streamflow data are available for North River, urban land use from 2001 to present.

Table 16. Estimated suspended-sediment load for seven tributaries to the Lake Tuscaloosa reservoir, Alabama.

[mi2, square mile; tons/mi2, tons per square mile; —, no data]

Estimated suspended-sediment load in tons Drainage Station (Estimated suspended-sediment yield in Station name area 2 number tons/mi ) (mi2) 2009 2010 2011 02464000 North River near Samantha, AL 223 81,400 78,100 51,100 (360) (350) (230) 02464146 Turkey Creek near Tuscaloosa, AL 6.16 3,370 2,670 2,160 (550) (430) (350) 02464360 Binion Creek below Gin Creek near Samantha, AL 57.2 6,820 6,860 4,060 (120) (120) (71) 02464385 Pole Bridge Creek near Samantha, AL 4.67 — 400 — — (86) — 02464503 Tierce Creek above Northport, AL 1.03 — — 200 — — (190) 02464660 Carroll Creek at St Hwy 69 nr Northport, AL 20.9 17,600 10,800 7,270 (840) (520) (350) 02464680 Brush Creek near Northport, AL 0.92 — — 280 — — (300) Summary 61

Summary basin had an approximately 4- and 3-percent reductions in forested land, respectively. The minor changes indicate that The U.S. Geological Survey (USGS), in cooperation overall the basin is stable; therefore, the inclusion of historical with the City of Tuscaloosa, conducted an investigation of data would be acceptable. Historical data include streamflow suspended-sediment loads to the Lake Tuscaloosa reservoir, and periodic monthly sediment samples collected from 1979 which is formed by the impoundment of the North River to 1983. Suspended-sediment data were also collected at this in Fayette and Tuscaloosa Counties, Alabama. One of the streamgage from March to September 1975 to estimate the objectives of the investigation was to develop sediment long-term average sediment yield of 300 tons per year per transport curves and estimate annual loads for seven tributaries square mile. Streamflow and isokinetic equal-width-increment to the reservoir. Quantification of sediment inflows from the suspended-sediment data were collected during the 2009 contributing watersheds is needed by the City of Tuscaloosa water year. The combination of the historical and current data because this information will augment ongoing collaboration was used to develop a daily sediment transport curve. The transport curve is rated to a streamflow value of 15,400 cubic with local and State groups and help focus their management 3 efforts to target areas around the reservoir. Sedimentation feet per second (ft /s), the maximum streamflow sampled. This damages a reservoir if the decreased storage capacity prevents value corresponds to a greater than 10-percent probability of the reservoir from supplying the full service for which it was exceedance. The daily transport curve was used in conjunction designed. Sufficient capacity must be maintained in domestic with the daily average streamflow to estimate a load for the water-supply reservoirs to assure continuity of supply during 2009 water year. The load estimated for the 2009 water year is 81,400 tons, which corresponds to a yield of 360 tons per periods of prolonged drought and to meet expected increases 2 in water demand. square mile (tons/mi ). The average yield estimated compares Storm and monthly suspended-sediment samples were closely with the long-term average sediment yield of 300 tons collected at seven tributaries to Lake Tuscaloosa from October per square mile per year estimated in 1975. 2008 to January 2012. Suspended-sediment concentrations and The Turkey Creek streamgage (station number 02464146) streamflow measurements were statistically analyzed and cor- was established in February 1981 on Turkey Creek at Turkey Creek Road, near Tuscaloosa. The drainage area upstream related to develop daily suspended-sediment transport curves. 2 The individual basins were investigated for temporal changes of the Turkey Creek streamgage is 6.16 mi . Inspection of in land use. Based on the stability of the basin and anthropo- land use (1992, 2001, and 2006) indicates that the only major genic changes, historical measurements were compared to, or changes in the basin occurred in forested land. Comparison included in, the resulting sediment transport curves. of the two timeframes indicated that between 1992 and 2001 Basinwide in 2006 (Fry and others, 2011), forested land and between 2001 and 2006, the basin had an approximately (68 percent) was the primary land cover. Comparison of his- 12- and 7-percent reductions in forested land, respectively. torical imagery and the National Land Cover Database (2001 Monthly sediment samples were collected 1,000 feet (ft) and 2006) indicated that the greatest temporal land-use change downstream at the State Highway 69 streamgage (station was attributed to timber harvest. The land cover in 2006 was number 02464145) for the 1977 to 1979 water years. Addi- indicative of this change with shrub/scrub land (12 percent) tionally, various samples were collected at Turkey Creek near being the secondary land use in the basin. Agricultural land Patterson Chapel (station number 02464149). Because of the use (10 percent) was represented predominantly by hay and changes in forested land, the historical data were not included pasture or grasslands. Urban land use was minimal, accounting in the development of current transport curves. Streamflow, for 4 percent of the entire basin. automatically collected suspended-sediment samples, and Data collection at the seven tributaries (North River, isokinetic equal-width-increment suspended-sediment samples Turkey Creek, Binion Creek, Pole Bridge Creek, Tierce Creek, were collected during the 2009 water year. Comparison of Carroll Creek, and Brush Creek) was spread out over the 2009, the historical data and the 2009 data showed an upward 2010, and 2011 water years. Of the seven sites, three (North trend of the measured suspended-sediment discharge. The River, Turkey Creek, and Binion Creek) had existing USGS 2008–09 data were used to develop a daily sediment transport curve. The transport curve is rated to a streamflow value of streamflow-gaging stations (streamgages). At the remaining 3 four sites (Pole Bridge Creek, Tierce Creek, Carroll Creek, and 1,020 ft /s, the maximum streamflow sampled. This value cor- Brush Creek), streamgages were installed. responds to a less than 20-percent probability of exceedance. The North River streamgage (station number 02464000) The daily transport curve was used in conjunction with the was established in December 1938 on North River at County daily average streamflow to estimate a load for the 2009 water year. The load estimated for the 2009 water year is 3,350 tons, Road 38, near Samantha. The drainage area upstream of 2 the North River streamgage is 223 square miles (mi2), and which corresponds to a yield of 540 tons/mi . it contributes 53 percent of the total drainage area to Lake The Binion Creek streamgage (station number 02464360) Tuscaloosa. Review of land-use data (1992, 2001, and 2006) was established in October 1986 on Binion Creek at Old Fayette Road, near Tuscaloosa. The drainage area for this site indicates that the only major changes in the basin occurred in 2 forested land. Comparison of the two timeframes indicated is 57.2 mi , and it contributes 14 percent of the surface drain- that between 1992 and 2001 and between 2001 and 2006, the age area to Lake Tuscaloosa. Inspection of land use (1992, 62 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

2001, and 2006) indicates no major change in land use from 2006, the basin had approximately a 6-percent reduction in 1992 to 2001 and a decrease in forested land (7 percent) from forested land during both timeframes. Streamflow, automatic 2001 to 2006. Current conditions were examined using 2009 suspended-sediment samples, and isokinetic equal-width- and 2010 aerial photography in conjunction with the National increment suspended-sediment samples were collected during Land Cover Database 2006 (NLCD 2006). Comparison of the the 2011–12 water years. The data were used to develop a two timeframes indicated that between 2006 and 2009 and daily sediment transport curve. The transport curve is rated between 2009 and 2010, approximately 2 and 1.4 mi2, respec- to a streamflow value of 110 ft3/s, the maximum streamflow tively, of timber were harvested. The deforested areas closest sampled. This value corresponds to a less than 67-percent to the streamgage were further inspected. Approximately 1 mi probability of exceedance. The daily transport curve was used upstream of the Binion Creek streamgage, 33 acres of timber in conjunction with the daily average streamflow to estimate a were harvested. load for November 2010 to October 2011. The load estimated Monthly sediment samples were collected at the Binion for the timeframe is 200 tons, which corresponds to a yield of Creek streamgage for the 1983 water year. The historical 190 tons/mi2. data were used to show temporal trends because the land-use The Carroll Creek streamgage (station number 02464660) changes described above indicate a decrease in forested land. was established in November 2008 on Carroll Creek at State Streamflow, automatic suspended-sediment samples, and Highway 69, near Northport. The drainage area for this site isokinetic equal-width-increment suspended-sediment samples is 20.9 mi2. Inspection of land use (1992, 2001, and 2006) were collected during the 2010 water year. Comparison of the indicates that the only major changes in the basin were 1983 and 2010 datasets indicates a similar trend for the lower decreases in forested land and urbanization. Comparison of values of streamflow. The 2009–10 data were used to develop the two timeframes indicated that between 1992 and 2001 a daily sediment transport curve. The transport curve is rated and between 2001 and 2006, the basin had approximately a to a streamflow value of 2,900 ft3/s, the maximum streamflow 3-percent reduction in forested land during both timeframes. sampled. This value corresponds to a greater than 50-percent During these same timeframes, urbanization increased 0.5 probability of exceedance. The daily transport curve was used and 3 percent, respectively. Streamflow, automatic suspended- in conjunction with the daily average streamflow to estimate sediment samples, and isokinetic equal-width-increment a load for the 2010 water year. The load estimated for the suspended-sediment samples were collected during the 2009 2010 water year is 6,860 tons, which corresponds to a yield of water year. The 2008–09 data were used to develop a daily 120 tons/mi2. sediment transport curve. The transport curve is rated to The Pole Bridge Creek streamgage (station number a streamflow value of 670 ft3/s, the maximum streamflow 02464385) was established in January 2010 on Pole Bridge sampled. This value corresponds to a less than 67-percent Creek at Old Fayette Road, near Samantha. The drainage area probability of exceedance. The daily transport curve was used for this site is 4.67 mi2. Inspection of land use (1992, 2001, in conjunction with the daily average streamflow to estimate a and 2006) indicates no major change in land use from 1992 to load for November 2008 to October 2009. The load estimated 2001 and a decrease in forested land (6 percent) from 2001 to for the timeframe is 17,600 tons, which corresponds to a yield 2006. Current conditions were examined using 2009 and 2010 of 840 tons/mi2. aerial photography in conjunction with the NLCD 2006. Com- The Brush Creek streamgage (station number 02464680) parison of the two timeframes indicated that between 2009 was established in November 2010 on Brush Creek at Turner and 2010, approximately 43 acres of timber were harvested. Road, near Northport. Inspection of land use (1992, 2001, Streamflow, automatic suspended-sediment samples, and and 2006) indicates no major change in land use from 1992 to isokinetic equal-width-increment suspended-sediment samples 2001 and a decrease in forested land (10 percent) from 2001 were collected during the 2010 water year. The 2009–10 data to 2006. Streamflow, suspended-sediment samples collected were used to develop a daily sediment transport curve. The using an automatic pump sampler, and isokinetic equal-width- transport curve is rated to a streamflow value of 96 ft3/s, the increment suspended-sediment samples were collected during maximum streamflow sampled. This value corresponds to the 2011and 2012 water years. The data were used to develop a less than 67-percent probability of exceedance. The daily a daily sediment transport curve. The transport curve is rated transport curve was used in conjunction with the daily average to a streamflow value of 130 ft3/s, the maximum streamflow streamflow to estimate a load for the 2010 calendar year. The sampled. This value corresponds to a less than 67-percent load estimated for 2010 is 400 tons, which corresponds to a probability of exceedance. The daily transport curve was used yield of 86 tons/mi2. in conjunction with the daily average streamflow to estimate a The Tierce Creek streamgage (station number 02464503) load for November 2010 to October 2011. The load estimated was established in November 2010 on Tierce Creek at Tierce for the timeframe is 280 tons, which corresponds to a yield of Creek Road, near Northport. The drainage area for this site 300 tons/mi2. is 1.03 mi2. Inspection of land use (1992, 2001, and 2006) Quantification of sediment inflows from the contributing indicates that the only major changes in the basin were a watersheds is useful to help focus management efforts in decrease in forested land. Comparison of the two timeframes targeting areas around the reservoir. Loads were computed indicated that between 1992 and 2001 and between 2001 and for each site for the period of sediment data collection. The References Cited 63 estimated load is dependent on the hydrologic conditions Colby, B.R., 1956, Relation of sediment discharge to stream- during the period for which it is estimated. The study area was flow: U.S. Geological Survey Open-File Report 56–27, under drought conditions during portions of the data collection 170 p, http://pubs.er.usgs.gov/publication/ofr5627. period. Therefore, the loads/yields estimated are compared based on the year for which they were estimated. This assumes Cole, E.F., 1985a, Effects of coal mining on the water quality that the hydrologic conditions were roughly the same at the and sedimentation of Lake Tuscaloosa and selected tributar- selected sites during the year of data collection. Because the ies, North River basin, Alabama: U.S. Geological Survey sampling sites vary in drainage area from 0.92 to 223 mi2, the Water-Resources Investigations Report 84–4310, 53 p., best comparison is the load per unit area, or yield. http://pubs.er.usgs.gov/publication/wri844310. The 2009 estimated yields for North River, Turkey Creek, Cole, E.F., 1985b, Water-quality and sedimentation 2 and Carroll Creek are 360, 540, and 840 tons/mi , respectively. data from Lake Tuscaloosa and selected tributar- Based on this comparison, Carroll Creek contributes more ies, North River Basin, Alabama—Water year 1983: sediment per square mile to the Lake Tuscaloosa reservoir U.S. Geological Survey Open-File Report 85–91, 102 p., than either North River or Turkey Creek. The Carroll Creek http://pubs.er.usgs.gov/publication/ofr8591. streamgage had a smaller peak flow (higher probability of exceedance) during data collection than North River or Turkey Culbertson, W.C., 1964, Geology and coal resources of the Creek, yet it still had a higher yield than the other sites. coal-bearing rocks of Alabama: U.S. Geological Survey The 2010 estimated yields for Binion Creek and Pole Bulletin 1182–B, 79 p. Bridge Creek were 120 and 86 tons/mi2, respectively. The Binion Creek streamgage was established in 1986, and flow Davis, B.E., 2005, A guide to the proper selection and use of statistics indicate that the annual mean streamflow was above federally approved sediment and water-quality samplers: average in 2010 and was higher than 2009 and 2011. The U.S. Geological Survey Open File Report 2005–1087, 20 p., largest peak flow recorded on the streamgage during data http://pubs.usgs.gov/of/2005/1087/. collection was greater than a 50-percent chance exceedance. It Dinehart, R.L., 1998, Sediment transport at gaging stations should be noted that both basins had timber harvested during near Mount St. Helens, Washington, 1980–90—Data col- the data collection period. lection and analysis: U.S. Geological Survey Professional The 2011 estimated yields for Tierce Creek and Paper 1573, 105 p., http://pubs.er.usgs.gov/publication/ 2 Brush Creek were 190 and 300 tons/mi , respectively. These pp1573. streamgages were both newly installed and do not have flow statistics to qualify the hydrologic conditions based on Duan, Naihua, 1983, Smearing estimate—A nonparametric historical data. However, based on flood-frequency relations, retransformation method: Journal of the American Statisti- both sites had peak streamflows with less than a 67-percent cal Association, v. 78, no. 383, p. 605–610 probability of exceedance, indicating drier hydrologic conditions for the 2011 estimation period. Edwards, T.K., and Glysson, G.D., 1999, Field methods for measurement of fluvial sediment: U.S. Geological Survey Techniques of Water- Resources Investigations, book 3, chap. C2, 89 p., http://pubs.er.usgs.gov/publication/twri03C2. References Cited Fry, J.A., Coan, M.J., Homer, C.G., Meyer, D.K., and Wickham, J.D., 2009, Completion of the National Almon and Associates, 1976, Water-quality management in Land Cover Database (NLCD) 1992–2001 Land the Lake Tuscaloosa watershed: Report to the City of Tusca- Cover Change Retrofit product: U.S. Geological loosa, Alabama, in response to Public Law 92–500, 110 p. Survey Open-File Report 2008–1379, 18 p., Carter, R.W., and Davidian, Jacob, 1968, General procedure http://pubs.er.usgs.gov/publication/ofr20081379. for gaging streams: U.S. Geological Survey Techniques of Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Water-Resources Investigations, book 3, chap. A6, 13 p., Barnes, C., Herold, N., and Wickham, J., 2011, Completion http://pubs.usgs.gov/twri/twri3-A6/. of the 2006 National Land Cover Database for the conter- Charley Foster and Associates, Inc., 2004, Surveyor’s report minous United States, Photogrammetric Engineering and for Lake Tuscaloosa bathymetric survey: Tuscaloosa, Ala., Remote Sensing, v. 77, no. 9, p. 858–864. 5 p. Glysson, G.D., 1987, Sediment-transport curves: Cohn, T.A., DeLong, L.L., Gilroy, E.J., Hirsch, R.M., and U.S. Geological Survey Open File Report 87–218, 47 p., Wells, D.K., 1989, Estimating constituent loads: Water http://pubs.er.usgs.gov/publication/ofr87218. Resources Research, v. 25, no. 5, p. 937–942. 64 Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11

Guy, H.P., 1969, Laboratory theory and methods for sedi- Porterfield, George, 1972, Computation of fluvial-sediment analysis: U.S. Geological Survey Techniques of ment discharge: U.S. Geological Survey Techniques of Water-Resources Investigations, book 5, chap. C1, 58 p., Water-Resources Investigations, book 3, chap. C3, 66 p., http://pubs.er.usgs.gov/publication/twri05C1. http://pubs.usgs.gov/twri/twri3-c3/. Harkins, J.R., and others, 1980, Hydrologic assessment, Puente, Celso, and Newton, J.G., 1982, Hydrology of poten- Eastern Coal Province Area 23, Alabama: U.S. Geologi- tial mining areas in the Warrior coal field, Alabama: U.S. cal Survey Water-Resources Investigations Report 80–683, Geological Survey Open-File Report 82–105, 109 p., 76 p., http://pubs.er.usgs.gov/publication/ofr80683. http://pubs.er.usgs.gov/publication/ofr82105. Hedgecock, T.S., and Feaster, T.D., 2007, Magnitude and fre- Puente, Celso, Newton, J.G., and Hill, T.J., 1980, Hydrol- quency of floods in Alabama, 2003: U.S. Geological Survey ogy of selected basins in the Warrior coal field, Ala- Scientific Investigations Report 2007–5204, 28 p., + app., bama—A progress report: U.S. Geological Survey http://pubs.water.usgs.gov/sir2007-5204. Water-Resources Investigations Report 80–22, 62 p., http://pubs.er.usgs.gov/publication/wri8022. Homer, C., Huang, C., Yang, L., Wylie, B., and Coan, M., 2004. Development of a 2001 National Landcover Database Puente, Celso, Newton, J.G., and Bingham, R.H., 1982, for the United States: Photogrammetric Engineering and Assessment of hydrologic conditions in potential coal- Remote Sensing, v. 70, no. 7, p. 829–840. lease tracts in the Warrior Coal field, Alabama: U.S. Geological Survey Open-File Report 81–540, 43 p., Hubbard, E.F, 1975, Fathometer survey of Lake Tuscaloosa, http://pubs.er.usgs.gov/publication/ofr81540. Alabama: U.S. Geological Survey Open-File Report, 94 p. Rantz, S.E., and others, 1982, Measurement and computation of streamflow—Volume 1. Measurement of stage and dis- Hubbard, E.F., 1976a, Sedimentation in Lake Tuscaloosa, charge: U.S. Geological Survey Water-Supply Paper 2175, Alabama: U.S. Geological Survey Open-File Report 284 p., http://pubs.er.usgs.gov/publication/wsp2175_vol1. 76–158, 35 p. Slack, L.J., 1987, Water quality of Lake Tuscaloosa and Hubbard, E.F., 1976b, Water-quality reconnaissance of Lake streamflow and water quality of selected tributaries to Lake Tuscaloosa, Alabama, March–June 1975: U.S. Geological Tuscaloosa, Alabama, 1982–86: U.S. Geological Survey Survey Open-File Report 76–160, 34 p. Water-Resources Investigations Report 87–4002, 64 p., Keener, M.J., Moore, J.D., and Hubbard, E.F., 1975, Geologic http://pubs.er.usgs.gov/publication/wri874002. and hydrologic investigation of the Lake Tuscaloosa drain- Slack, L.J., and Pritchett, J.L., Jr., 1988, Sedimentation in age basin: Geological Survey of Alabama, Draft report to Lake Tuscaloosa, Alabama, 1982–86: U.S. Geological Sur- the City of Tuscaloosa (contract no. GSA–74–15), 196 p. vey Water-Resources Investigations Report 87–4256, 24 p., Kennedy, E.J., 1984, Discharge ratings at gaging stations: U.S. http://pubs.er.usgs.gov/publication/wri874256. Geological Survey Techniques of Water-Resources Inves- Stricklin, V.E., 2001, Sedimentation profiles in Lake Tusca- tigations, book 3, chap. A10, 59 p., http://pubs.er.usgs.gov/ loosa, Alabama, 2000: U.S. Geological Survey Open-File publication/twri03A10. Report 01–317, 19 p., http://pubs.er.usgs.gov/publication/ Lee, K.G., and Kimbrow, D.R., 2011, Bathymetric survey of ofr01317. Carroll Creek Tributary to Lake Tuscaloosa, Tuscaloosa Teledyne Isco, Inc., 2005, 6712 full-size portable sampler, County, Alabama, 2010: U.S. Geological Survey Scientific accessed August 22, 2011, http://www.isco.com/products/ Investigations Map 3176, 1 sheet, http://pubs.er.usgs.gov/ products3.asp?PL=201101010. publication/sim3176. U.S. Bureau of Land Management, 1980, Southern Appala- Metzger, W.J., 1965, Pennsylvanian stratigraphy of the chian regional coal environmental impact statement: U.S. Warrior basin, Alabama: Geological Survey of Alabama Bureau of Land Management, 447 p. Circular 30, 80 p. U.S. Bureau of Land Management, 1983, Southern Appala- National Oceanic and Atmospheric Administration, 2002, chian regional coal environmental impact statement II: U.S. Monthly station normals of temperature, precipitation, and Bureau of Land Management, 188 p. heating and cooling degree days 1971–2000: Ashville, N.C., National Climatic Data Center, Climatography of the U.S. Geological Survey, 2006, National field manual for United States No. 81, 26 p., http://nsstc.uah.edu/aosc/files/ the collection of water-quality data—Collection of water ALnorm.pdf. samples (ver. 2.0): U.S. Geological Survey Techniques of Water-Resources Investigations, book 9, chap. A4, http://pubs.water.usgs.gov/twri9A4/. References Cited 65

Vanoni, A.V., ed., 1977, Sedimentation engineering: New York, American Society of Civil Engineers, Manuals and Reports on Engineering Practice no. 54, 745 p. Walling, D.E., 1977, Assessing the accuracy of suspended sediment rating curves for a small basin, Water Resources Research, v. 13, p. 531–538. Williams, G.P., 1989, Sediment concentration versus water discharge during single hydrologic events in rivers: Journal of Hydrology, v. 111, p. 89–106.

For further information about this publication contact:

Director U.S. Geological Survey Alabama Water Science Center 75 TechnaCenter Drive Montgomery, AL 36117

Or visit the Alabama Water Science Center Web site at http://al.water.usgs.gov/

Prepared by the Raleigh Publishing Service Center Lee—Estimation of Sediment Inflows to Lake Tuscaloosa, Alabama, 2009–11–Scientific Investigations Report 2013–5152