Community Partner

Oregon Department of Environmental Quality (DEQ)

First Advisor

Eugene Foster

Date of Award


Document Type


Degree Name

Master of Environmental Management (MEM)


Environmental Science and Management


Willamette River Watershed (Or.) -- Environmental conditions, Water quality -- Oregon -- Willamette River Watershed, Water -- Pollution -- Total maximum daily load, Water -- Pollution -- Statistical analysis




Point and nonpoint sources contribute toxic pollutants to surface waters, degrading water quality and impairing aquatic and human life. As of 2012, 51 stream segments totaling approximately 6,000 kilometers (3,750 miles) in Oregon’s Willamette River Basin (the Basin) were listed as impaired on the Clean Water Act’s (CWA) Section 303(d) list with 48 toxic pollutants. These toxic pollutants exceed the Oregon Department of Environmental Quality’s (DEQ) state water quality standards, requiring Total Maximum Daily Loads (TMDLs) to restore water quality and protect beneficial uses. Identifying toxic pollutant point and nonpoint sources, and the total loads a waterbody receives, are critical components in TMDL development and a CWA requirement. However, developing a single pollutant TMDL is data- and time-intensive.

In this Project, I investigated: 1) point and nonpoint sources potentially contributing to the Basin’s impairments of metals, polycyclic aromatic hydrocarbons (PAHs), volatile organic compounds (VOCs), polychlorinated biphenyls (PCBs), organochlorine insecticides, and organophosphorus insecticides; and 2) developed a strategy for DEQ to approach TMDL modeling based on similarities in geographical source areas (by subbasin) and chemical characteristic similarities among the toxic pollutants (n=48). Characteristics considered included water solubility, vapor pressure, Henry’s Law Constant, octanol-water partition coefficient (Kow), and organic carbonwater distribution coefficient (Koc). These characteristics are primary mechanisms influencing environmental behavior and key inputs to surface water quality models used to determine toxic pollutant loads for TMDL allocation. These characteristics were evaluated using multivariate statistical techniques of cluster analysis and Nonmetric multidimensional scaling (NMDS).

Point source analysis of Basin stormwater/wastewater discharges revealed 2,050 national pollutant discharge elimination system (NPDES)-permitted facilities discharging 13 toxic pollutants, 81 NPDES-permitted Department of Geology and Mineral Industries mining sites, and 522 Oregon Department of Transportation outfalls. Nonpoint source analysis of Basin activities revealed 937 toxic pollutant-impacted contaminated waste sites (927 DEQ Environmental Cleanup Site Information sites and 10 Environmental Protection Agency Superfund Sites) and the application of 1.5 million kilograms (3.3 million pounds) of toxic pollutant-containing pesticides (metals, organochlorine and organophosphorus insecticides) to agricultural land (2000 to 2016). The Project results revealed that metals, PAHs, and VOCs can be attributed to both point and nonpoint sources while PCBs, organochlorine insecticides, and organophosphorus insecticides can be attributed to nonpoint sources.

Exploratory analysis of cluster analysis, in conjunction with Analysis of Similarity statistical evaluations, revealed four statistically different toxic pollutant groups discriminated by chemical characteristics (Clusters 1 through 4). NMDS analysis indicated clusters were associated with chemical characteristics and generally not to pollutant class, suggesting similarities in environmental behavior: Cluster 1 toxic pollutants (VOCs and metals) exhibiting soluble behavior, Cluster 2 toxic pollutants (VOCs and metals) exhibiting more volatile behavior, Cluster 3 toxic pollutants (PAHs, organochlorine and organophosphorus insecticides) exhibiting less volatile behavior, and Cluster 4 toxic pollutants (PAHs, organochlorine insecticides, and PCBs) exhibiting hydrophobic behavior. The Project results suggested that statistically evaluating chemical characteristic data for TMDL modeling of multiple pollutants could address multiple surface water impairments concurrently, efficiently meeting CWA requirements and DEQ needs, resulting in improvement of the Basin’s water quality.


A thesis submitted in partial fulfillment of the requirements for the degree of Master of Environmental Management.

Appendices A-D (tables, figures, subbasin profiles, and toxic pollutant profiles) are included below as supplemental files.

Persistent Identifier

Appendix A_Tables.pdf (1278 kB)

Appendix B_Figures.pdf (15543 kB)

Appendix C_Subbasin Profiles.pdf (12125 kB)
Subbasin profiles

Appendix D_Toxic Pollutant Profiles.pdf (1694 kB)