The goal of this project was to develop, test, and assess the effectiveness of a comprehensive set of laboratory, field, and modeling approaches in characterizing the role of urban stormwater in contamination of sediments and remediated sites. The research focused on the development and application of techniques to assess the magnitude and characteristics of episodic distributed sources (i.e., stormwater) and the effects of these sources on sediments and benthos. Baseline conditions, discharges of contaminants during storm events, relationship of these loads to other potential sources of sediment contamination, and the potential for recontamination following remediation were considered. The bulk chemical contamination was integrated with site-specific bioavailability, the potential for bioaccumulation, and identification of key stressors and ecological risk of stormwater and stormwater–related sediment contaminants. This provides a foundation for a decision-making framework for identifying stormwater sources and their consequences, designing effective source controls, and identifying the remedial goals realistically achievable without such controls.

The study efforts were conducted in four coupled and integrated phases:

1. Baseline sampling to characterize dry weather conditions;
2. Stormwater loading assessment by direct sampling;
3. Receiving waters assessment to link stormwater loads to sediment recontamination; and
4. Stormwater and receiving water modeling to predict stormwater loading over time.

Stormwater loads and the resulting impacts on surficial sediments were characterized physically, chemically, and biologically considering the spatial and temporal dynamics of the dominant stressors. The characterization of stormwater sources of contaminants was informed via a review and summary of existing data on stormwater source characterizations and loadings (the International Stormwater BMP Database and the National Stormwater Quality Database) as well as targeted sampling of stormwater sources to complement existing databases and source characterization. Measurements were used to calibrate a stormwater model that built on previous DoD modeling efforts and coupled with measurements of receiving sediment impacts and ecological effects. Intensive field stormwater monitoring at one site was conducted.

The calibrated stormwater modeling enabled predictions of stormwater discharge as determined by specific drainage area characteristics and activities. These stormwater loading predictions, along with information affecting the fate of the discharged suspended and bedload sediments (e.g., particle size distributions and related settling rates) were used to help quantify the recontamination potential of the sediments by stormwater discharges and to compare to the receiving sediment recontamination measurements. Contaminant dynamics in the receiving waters were interpreted using models developed by SERDP project ER-1746 and ESTCP project ER-201031.

The bulk of the method development, testing, and data acquisition was conducted at Paleta Creek, an urban watershed partially encompassing Naval Base San Diego and draining to San Diego Bay. Stormwater discharges at a secondary site at Puget Sound Naval Shipyard also were studied to identify whether the methods were applicable and whether general characteristics noted at Paleta Creek were reproduced at an additional site.

The primary goal of the project was to identify the best methods to assess sediment recontamination as a result of stormwater discharges. Key methods needed to assess sediment recontamination that were identified include:

• Assessing stormwater loads requires collecting stormwater during storms of different intensity and at different stages during the storm. The specific requirements will differ depending upon the frequency, duration and intensity of events in a particular location but storms representative of an area are needed. Sampling should be automated in such a manner that the effect of non-stormwater flows, e.g. tides, can be excluded from the samples. In general, it is believed that intensive monitoring of a few representative events (as outlined below including size segregated loads of a broad range of contaminants and monitoring at different locations in the watershed and in receiving waters) is more useful that more modest monitoring efforts of a large number of events.
• Equivalent sampling of stormwater entering and exiting the naval base portion of a watershed can provide differentiation between Navy and other sources of stormwater contaminants.
• Assessing the potential for sediment recontamination requires that the stormwater contaminant loads in collected samples be characterized as a function of settling characteristics or particle size. The memo provides a description of the methods used in this project to determine contaminant distribution in key suspended solid size ranges. The measurements collected during this project indicate that contaminants may often be associated with relative coarse particles that settle quickly and the solids distribution and contaminant distribution do not necessarily correlate.
• Receiving water sampling is best conducted using settling traps during storm events or over a period of storm events. Segments of sediment cores are influenced by a variety of longterm processes and do not necessarily indicate the effects of recent storm events.
• Sediment recontamination as indicated by stormwater loads and collection of contaminants in sediment traps does not necessarily indicate the ecological effects of the contaminants. In particular, contaminants associated with large, rapidly settling particles may contribute to rapid recontamination of sediment bulk solid concentration but may contribute little to negative effects on benthic or other organisms
• Stormwater and receiving water modeling provide opportunities to better link stormwater discharge measurements with source areas and the discharges to sediment recontamination. These models are most useful to calibrate to a finite set of measurements and then used to extrapolate to other conditions or to predict longer term average discharges or sediment recontamination rates. 

Linking and ranking stormwater inputs of specific stressor types (i.e., habitat and siltation, nutrients, anoxia, metals, and organic chemicals) with ecological impacts (e.g., site toxicity, food web contamination, degraded benthic macroinvertebrate communities) allows for more effective assessment of recontamination and its consequences. This study of significantly different sediment environments allows the development of broadly applicable tools for assessing the effect of stormwater loads and inputs on sediment recontamination and determining which stormwater-related stressors are most important. This study and the methodology support more effective source control and corrective action planning and advance the state of the practice for forecasting and quantifying sediment recontamination risk where continued long-term stormwater pollutant loading is present.