Many water quality monitoring programs, as well as ecosystem restoration projects, are continually challenged with providing relevant data and information under a diverse set of regulations, goals, and objectives with smaller and smaller budgets. These drivers demand that managers continually adapt to, and utilize, a range of techniques that ensure only relevant data are efficiently generated. Many approaches are gaining attention to allow scientists and managers to spatially and temporally optimize the monitoring designs for both smallscale and large-scale water quality monitoring programs. These optimization approaches ensure that redundant data collection is minimized while maintaining key data and statistical robustness. Effective optimization addresses the suite of parameters being monitored, the spatial distribution of sampling locations, and sample frequency of sample collection necessary to meet monitoring program goals and objectives.

Battelle has developed an optimization approach for evaluating trends in water quality monitoring parameters to meet South Florida Water Management District (SFWMD) program objectives. Seventeen distinct SFWMD water quality monitoring projects were evaluated to determine if the current sampling design was sufficient to meet the goals and objectives of each respective project. The approach incorporates clear articulation of the end-uses of the data, and thorough awareness of the management and policy decisions that the data will support.

The optimization of these projects incorporated the EPA Data Quality Objectives process to better define the monitoring hypotheses and identify the need for the information, the data to be collected, and the statistical analytical procedures that can be applied. For the monitoring projects evaluated during this study, the DQO process determined that the monitoring data were primarily used to detect trends and changes from those trends in selected water quality parameters. A non-parametric statistical procedure based on the Seasonal Kendall Tau trend analysis, developed by Reckhow et al. (1993), coupled with statistical modeling allowed an evaluation of the current monitoring design and a series of alternative designs to determine which design would provide the best estimate of an annual percent change in water quality parameter concentrations. The resulting statistical tool, written and performed using SAS, was delivered to the SFWMD and is being used for evaluation of other current and future monitoring designs.

This optimization approach and statistical tool requires a validated long-term-monitoring dataset, and allows program managers to determine how much change in a trend can be detected with specific monitoring designs. Predictions are based on statistical characterizations of historic trends and patterns present in the dataset. By employing this optimization, the SFWMD was able to assess alternative sampling designs to consider trade-offs between maximizing monitoring information obtained and minimizing monitoring costs.

For additional information about Battelle's water quality monitoring capabilities, contact Dr. Carlton Hunt at (781) 952-5374, huntc@battelle.org.