A data-driven index for evaluating BMP water quality performance

A data-driven stormwater best management practice (BMP) performance index (PI) developed herein provides a simple, unbiased method to interpret the water quality treatment performance of a structural BMP, and provides actionable information for managers in planning, implementing and maintaining BMPs. The PI is derived from field-monitored influent-effluent pollutant event mean concentrations normalized by a user-specified water quality benchmark that may be adapted to reflect watershed-specific objectives. Benchmarking allows performance of any BMP to be investigated regardless of the treatment mechanisms, climate conditions, hydrologic performance, or site-specific pollutant concentrations. Quantitative monitoring data normalized by the benchmark are subsequently binned into performance categories of Success, Excess, Marginal, Insufficient, and Failure, which are indicators of the relative potential to achieve downstream receiving water goals. A single PI score per analyte (PI_analyte) is derived by compositing the categorical distribution. A simplified analytical hierarchy procedure is adapted to combine multiple PI_analyte scores, if/when BMP selection for future applications must address a range of pollutants, e.g., where multiple water quality objectives are present. Data from the International Stormwater BMP Database are used to demonstrate PI applications such as selecting the "right" BMP to address specific water quality concerns, comparing amongst similar types of BMP to identify beneficial design features, and the clear interpretation offered by the score providing actionable information compared to a percent-removal assessment method. An index interpretation guide provides managers, maintenance teams, and designers with a feedback mechanism to refine future project designs, operations, or maintenance needs, and informs progress towards achieving successful watershed management plans.