Date of this Version
QUANTITATIVE ANALYSIS AND MODELING OF URBAN STORM WATER MASS LOADINGS IN THE CITY OF LINCOLN, NEBRASKA
Patrick Ryan Hartman, M. S. University of Nebraska, 2010
Advisors: Bruce I. Dvorak, David M Admiraal
The main goal of this study is to monitor, and then compare results related to the water quality and mass loadings of several constituents at two sites in Lincoln, NE. Differences in water quality were assessed using matched-pair t-tests. Mass loadings were examined using cumulative mass plots, and a predictive model for total suspended solids (TSS) was developed at both sites using real-time data obtained from a USGS data probe. Statistical comparison tests were conducted on the collected water quality data to detect differences based on sample type (auto vs. grab) and sample location (Taylor Park vs. Colonial Hills) for both wet and dry weather events. Preliminary results indicate that there is statistically no difference between concentrations in the auto and grab samples at either site.
For flows during dry weather periods, the Nitrate, Phosphorous, and E. Coli concentrations at Taylor Park are higher than the concentrations at Colonial Hills with 95% confidence. The Turbidity and Chlorine levels at Colonial Hills are higher than the concentrations at Taylor Park with 90% confidence. These differences, particularly the higher concentrations at Taylor Park may be related to best management practices (BMPs) in the Colonial Hills watershed. The same match-pair analysis was conducted on data collected during wet weather flows to detect general differences in the water quality between the two sites. Conductivity concentrations were found to be statistically greater at the Colonial Hills site with 95% confidence.
The mass loadings of several contaminants were examined through the use of cumulative mass plots (CMP’s). CMP’s were developed for Turbidity, Dissolved Oxygen, and Conductivity for the 2009 sampling season. Results indicate that about 90% of the mass for Turbidity occurs at flow stages higher than the average annual flow. This suggests that the sampling focus in subsequent seasons should be placed on wet weather monitoring in order to accurately describe the mass loading relationship at the highest flows.
A statistical model was developed using log transformations to predict the TSS concentration as a function of the turbidity and flow rate. Preliminary results for the TSS models have an R2 value of 0.711 at Taylor Park, and 0.906 at Colonial Hills. The average error generated using the models on a log/log scale are about 12% in both cases, and maximum errors were about 40% for both sites.