Date of Award

Fall 2014

Degree Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Linda S. Lee

Committee Chair

Linda S. Lee

Committee Member 1

P. Suresh C. Rao

Committee Member 2

Chad T. Jafvert

Committee Member 3

Nandita Basu


A passive surface water flux meter (PSFM) for measurement of contaminant concentration/flux in rivers and streams is described and tested. The novel PSFM design was developed for portability and ease of adaptability for a variety of contaminant classes. Although previous designs have been evaluated under constant flow conditions, the PSFM has never been used for measurement of pesticides or hormones and this is the first time that it has been tested under transient flow. Discharge through the PSFM is assessed by measuring miscible displacement of alcohol tracers from granular activated carbon (GAC). The tracer retardation factors (R) measured by miscible displacement were typically smaller (within 25%) than those estimated from batch studies with sorption of the larger tracers being more nonlinear. For calibration of the ratio of PSFM water flux to external flow velocity, water flux through the PSFM was measured in a flume under constant flow conditions at a range of velocities representative of those in streams and rivers. The relationship between PSFM water flux and external flow velocity in a flume was non-linear as predicted by Bernoulli's equation for velocity potential flow. However, in samples deployed in a natural stream, the relationship between PSFM water flux and external flow was weak with less flow passing through the PSFM under field conditions than predicted by measurements in a flume. The sorption and degradation of the contaminants of interest on surfactant modified zeolite (SMZ), the sorbent used for contaminant capture, were evaluated in laboratory and field experiments. PSFM performance was evaluated in a stream network at the Purdue Animal Sciences Research and Education Center (ASREC). Sampling was focused on the steroid hormones and pesticides, which are present at trace concentrations in the stream network as a result of agricultural activity. Estimates of contaminant flow weighted average concentration obtained using the PSFM were compared to concentrations measured in water samples taken at regular intervals using automated sampling equipment. Concentrations of both estrogens and pesticides measured using the PSFM were generally higher than those measured in water samples. This difference was attributed primarily to the high temporal variability of contaminant concentration and flow in the stream resulting in large temporal inequality of transport which is not adequately sampled using discreet methods. Furthermore, it was hypothesized that non-equilibrium tracer desorption during periods of high stream velocity may cause underestimation of PSFM specific discharge and consequent overestimation of contaminant flux in some cases. The PSFM was used to measure contaminant concentration at five strategically located sampling stations over the course of two months. The flow weighted average concentrations of steroid estrogens measured using the PSFM were generally in the low ng/Lrange while that of the pesticides was in the ug/L range. Estrogen concentrations were not correlated with manure application and were more highly variable relative to the pesticides. The highest estrogen concentrations were measured nearest the source zone following a prolonged period of high discharge. Conversely, the concentration of the pesticides atrazine, desethyl-atrazine, and metolachlor were not correlated with distance from the source zone, but increased dramatically with the first precipitation event following pesticide application suggesting disparate transport mechanisms for the compound classes.