September 2008, Vol. 20, No.9

Research Notes

Preventing Ecological Deterioration of Streams

A recent study funded and managed by the Water Environment Research Foundation (WERF; Alexandria, Va.) details a protocol that could prevent severe ecological deterioration of urban streams.

According to the executive summary of the study, changes in land use, especially urbanization, can affect runoff characteristics and the aquatic environments of streams receiving the runoff. A pilot study of eight watersheds with a gradient of urban development in the North Carolina Piedmont demonstrated that the hydrologic metric (a measure of time over a multiyear record during which the flow is greater than the average flow) responds to changes in land use and to alternative runoff control scenarios. The research team also found that the ecological health of streams is responsive to the hydrologic metric.

These findings imply that land use and stormwater management criteria can be set using what is known from the links among biological parameters, hydrologic and geomorphic metrics, and land use patterns and runoff control practices, according to the study. From these findings, an 8-step process for urban stormwater management agencies was created:

  1. identify resources for protection and preservation, and establish goals;
  2. identify parameters relevant to targeted resources;
  3. identify development gradient of sites used to generate relationships between hydrologic and geomorphic metrics and biologic parameters;
  4. obtain biologic data and compute metrics, then compute values for hydrologic and geomorphic metrics;
  5. establish relationships between hydrologic and geomorphic metrics and biologic data;
  6. evaluate impact of runoff and control strategies on hydrologic and geomorphic metrics;
  7. establish management criteria for stream types of interest;
  8. proceed with ongoing monitoring.

“This procedure provides a scientific platform in establishing suitable rules and design standards for real estate development plans,” said Jane Casteline, WERF program manager. “Using the protocol early in the process may lessen damage to river ecology before a problem develops.”

Researchers recommended that experiments should be conducted on an urbanizing watershed of 2.6 km² to 7.7 km² (1 mi² to 3 mi²) and should use runoff controls and best management practices to relate macroinvertebrate health to the hydrologic metric as the watershed develops.

This protocol was developed to help city planners find a way to measure the impact of land changes on the health of the watershed before these changes are made to prevent or reduce harm to rivers from urban development, according to a WERF press release. Principal research investigators for this study were Larry Roesner and Brian Bledsoe of Colorado State University (Fort Collins), and Christine Pomeroy from the University of Utah (Salt Lake City).

Statistical Models for Forecasting Microbial Concentration

A new study published in the American Chemical Society’s journal Environmental Science & Technology indicates that using the Virtual Beach (VB) tool can optimize the development of statistical models used for forecasting microbial concentrations at recreational water sites.

Researchers used VB, public-domain software that prescribes site-specific predictive models, as a tool to evaluate statistical modeling for predicting Escherichia coli levels at Huntington Beach, Ohio, on Lake Erie.

According to the study, persistence models typically are used to predict healthy water levels. However, because this model takes about 24 hours to analyze indicator bacteria samples, its success depends on steady concentrations of pollutants. In highly variable conditions, persistence models can produce false predictions, which can result in a waterbody with microbial concentrations above the threshold for safety, predicted to be safe for recreational use.

Huntington Beach was found to have highly variable conditions daily. Using the persistence model on a 2006 data set resulted in only seven correct water quality predictions out of a total of 26.

Statistical regression models, such as the multiple linear regression (MLR) model, have been used increasingly to predict contamination at beaches. MLR models use the known bacteria concentrations and independent data to produce mathematical models to predict bacteria concentrations. In the Huntington Beach case, MLR models outperform persistence models in the accuracy of health advisories, which was reconfirmed by this study.

The study was initiated to demonstrate the efficacy of the VB MLR model development tool and to evaluate and assess the feasibility of dynamic models, and to evaluate 24-hour forecasts of microbial contamination at beaches.

The experimental forecast period covered 42 days. Predictions of current but unknown bacteria levels, known as nowcasts, compared to actual observations in the study verified the effectiveness of VB and showed that dynamic models are about as accurate as long-term static models.

The results of the study show that VB performed as intended, and the automated functionalities of VB allow its users to perform daily model updates in about one hour.

The researchers found that the Virtual Beach tool can facilitate and optimize the development of statistical models used for nowcasting and forecasting microbial concentrations at recreational water sites, particularly for dynamic models based on short-term data sets.

The researchers recommended increasing the frequency of bacterial measurements to both establish statistical models and increase the demand for data at additional sites because, as this study has shown, bacteria concentrations can be forecast with reasonable accuracy, which will assist in predicting the time and location of waters suitable for recreational uses.