When characterizing water quality constituents, either from a measured dataset or modeling output, it is common to report only the peak concentration or total load. In addition to these basic statistics, there are many ways to quantify observed or computed water quality timeseries data. A broader classification (and the corresponding type of analysis involved) would include the: amount of pollutant over a specified period (loading analysis); magnitude of pollutant concentrations and mass loading rates (statistical analysis); cumulative time of threshold exceedance (duration analysis); and the frequency of occurrence (event analysis). Water quality analyses have direct parallels to water quantity analysis or design applications, including a suggested Quality Response Unit approach that can be applied to water quality constituents, by generating landscape-specific pollutant concentrations and loads in response to rainfall for individual surface cover types. This presentation begins with an overview of stormwater quality modeling using EPA SWMM5, methods for determining total pollutant loads and event mean concentrations, and highlights key concepts including hysteresis and first-flush effects. Analytical methods and performance metrics are also presented that characterize water quality timeseries. Case studies make use of sediment concentrations and loading rates for the following applications: a 32-year observed record of suspended sediment in the Atchafalaya River in Louisiana; and SWMM5 model results for a 21-year simulation at a development site in Florida, including a pre-development scenario and proposed alternatives that represent uncontrolled (piped discharge), dry, and wet retention pond.
