Filmon Araya

Date of Graduation


Document Type


Degree Type



Statler College of Engineering and Mineral Resources


Civil and Environmental Engineering

Committee Chair

Omar I Abdul-Aziz

Committee Co-Chair

Seung Ho Hong

Committee Member

P V Vijay


Dissolved oxygen (DO) is a key indicator for water quality and the general health of stream ecosystems. However, the dynamics of stream DO is controlled by myriad interacting environmental drivers. This thesis investigates the major environmental drivers, emergent similitude (similarity leading to parametric reductions), and scaling of stream DO across the Pacific Coast of U.S.A. The research was conducted by assembling a dataset for 1998--2015 at 35 water quality monitoring stations of U.S. Geological Survey and Environmental Protection Agency. Pearson correlation matrix, principal component analysis, and factor analysis were employed to identify the interrelations and groupings of stream DO and the environmental drivers. Power-law based partial least squares regression (PLSR) models, with a bootstrap Monte Carlo procedure (1000 iterations), were developed to reliably estimate the dominant controls of DO by resolving multi-collinearity among the environmental drivers. In the Pacific Northwest (Washington and Oregon) and southern California, water temperature was the most dominant driver of DO in majority of the streams, suggesting an environmental regime dictated by climatic control. However, in central through northern California, the in-stream DO was controlled by multiple drivers (water temperature, pH, stream flow, and total phosphorus), exhibiting a mixed environmental regime controlled by stream metabolism. Dimensional analysis was then conducted with the dominant variables to investigate any emergent similitude, and formulate mechanistically meaningful dimensionless numbers to represent the collective as well as contrasting environmental controls of stream DO. Plot of the response and predictor dimensionless number was found to collapse observations from diverse streams into a single dimensionless curve, indicating a remarkable emergent pattern along the Pacific Coast. A threshold value of the driver dimensionless number emphasized the existence of two environmental regimes, representing the climatic versus metabolic controls of stream DO. The dimensionless curve was estimated with observed data by using a power-law scaling function, which was then used as a non-linear empirical model to robustly predict stream DO in different environmental regimes along the U.S. Pacific Coast (Nash-Sutcliffe Efficiency, NSE=0.72). The research findings and tools would facilitate a robust prediction and assessment of stream water quality and ecosystem health across the Pacific Coast of U.S.A. and beyond.