One of the central problems in regional monitoring is choosing a station array that accurately reflects the distribution of values for the entire region of interest. For time-consuming or expensive measurements, an additional goal is to make the number of sampling locations and times as small as possible. These problems are probably most difficult in estuaries because of the relatively large variability on many different spatial and temporal scales. This high variability often means that comprehensive historical data are not available for accurately assessing sparser sampling efforts. It also implies that a higher frequency of sampling in space and time is required compared to other water bodies.
In this report, we focus on selection of an “optimum” station configuration for the channel of San Francisco Bay for vertical profiling of water quality. Our analysis is based on the monthly cruises conducted by the USGS under the auspices of the Regional Monitoring Program for Trace Substances (Caffrey et al. 1994; SFEI 1994). The underlying rationale for undertaking the analysis is that the distribution of trace substances is structured, at least in part, by the same forces acting on water quality parameters. This must be true to some extent, as trace substance concentrations are partially dependent on water quality characteristics such as salinity. On the other hand, the quantitative importance of these parameters in accounting for overall variability in individual trace substances is unknown. Furthermore, trace substances have their own unique sources, and these sources may dominate their distribution.
