Author ORCID Identifier

Date Available


Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation


Arts and Sciences


Earth and Environmental Sciences (Geology)

First Advisor

Dr. Alan E. Fryar

Second Advisor

Dr. Dwayne R. Edwards


There is an inherent difficulty in predicting source contributions of fine-grained fluvial sediment in mixed land-use watersheds. Over a 56-week period, the spatial and temporal variability in sediment-source contributions and water quality was monitored at three sites along Otter Creek in Hardin and Meade counties, Kentucky (USA). The 203-km2 study area drains rural and agricultural lands and includes Fort Knox Army Post’s tracked-vehicle training areas. The main objectives for sediment source apportionment were to 1) identify and differentiate characteristics of civilian and military source soils to Otter Creek and 2) to apportion sediment at locations along Otter Creek to different source-soil categories. We hypothesized that the primary sources of fine-grained sediment to Otter Creek were derived from Fort Knox’s tracked-vehicle training areas within the watershed. The water-quality objectives were to 1) draw inferences about spatial and temporal controls on water chemistry in Otter Creek and 2) compare measurements obtained during the study with historical data obtained from the U.S. Geological Survey (USGS) and the National Oceanic and Atmospheric Administration.

In civilian lands, which occupy 81.6% of the study area, primary land use/land cover is cropland (57.1%), with the remainder allocated to forest (29.6%), developed area (12.2%) and other (1.1%). Only 18.4% of the study area is composed of military training lands, of which forest is 63.6%, cropland is 0%, 23.8% is developed and 12.1% is other. The greatest proportion of military lands is located in the farthest downstream subcatchment, which had the largest amount of forested land (74.5%) and least amount of cropland (5.3%). Between summer (March–September) and winter (October–February) and between base- and stormflow, differences in sediment yield and composition were observed. The farthest downstream sampling site had the greatest sediment yield for summer, winter, baseflow and stormflow. The USGS program Sed_SAT identified Cu, Zn, Co, Ni, Al and Mg as conservative tracers that could distinguish between source-soil groups during the unmixing model process. Sed_SAT allocated target sediment to five source-soil groups: civilian near-stream, military near-stream, military forest, military average erosion and military extreme erosion. Because of geochemical similarities, those groups were combined into two simplified categories in two different scenarios. Sediment sources were dominated by streambank/forest soils (63-67%) in the first scenario and by military upland soils (57-66%) in the second scenario.

Water-quality monitoring included water pressure, electrical conductivity and specific conductance (EC and SC), water temperature, dissolved oxygen (DO) and pH, in addition to chloride (Cl-), nitrate (NO3-), phosphate (PO43-) and sulfate (SO42-) over a 6-month period. Seasonality was observed in fluid-pressure responses, with higher peaks in wetter seasons (December–mid-May) and lower peaks during drier months (mid-May–November). During storm events, a temporary EC decrease in response to dilution from runoff and corresponding increase in fluid pressure were observed at the two upstream sites. The downstream increase in pH between the first and last sampling sites corresponds to dissolution of carbonate bedrock that underlies the area. A significant positive relationship was observed at all three sites for Cl- and SO42-, which suggests a common source for both anions. During winter, elevated concentrations of Cl- were likely to result from deicer washing off roadways. The increases in NO3- observed from late fall through winter may be a result of higher-than-average precipitation and slow nitrification and leaching from soils. A comparison of historical water quality data (1994-98) to the study period (2015-16) shows there was a significant increase in SC, a significant decrease in DO, and a marginally significant increase in pH at the farthest downstream site.

Digital Object Identifier (DOI)

Funding Information

Funding for this project was provided by:

2016-2017: U.S. Geological Survey grant G16AP00055 through the Kentucky Water Resources Research Institute.

2016- 2017 Casner Fellowship from the University of Kentucky (UK) Tracy Farmer Institute for Sustainability and the Environment.

2014, 2015, 2016 University of Kentucky’s Department of Earth and Environmental Sciences (UKEES) Ferm Research Support Fund.

Peterman_Supplemental_Site_1_Datalogger.pdf (686 kB)

Peterman_Supplemental_Site_2_Datalogger.pdf (686 kB)