Water Quality Sampling Programs in Rivers

A central decision in the collection of water quality data from a river is the choice of the locations and frequencies of sampling. This paper proposes a method for the selection of these design variables, for given water quality constituents, based on the use of a stochastic model of the river. The design method is demonstrated for the coupled constituents of BOD and DO in the Shenango-Beaver River, Pennsylvania. Cost-effective sampling designs are identified for two principal objectives: (1) Detection of violations of in-stream water quality standards; and (2) assessment of the present state of water quality in the river. The results indicate that data collection efforts for the detection of in-stream violations should emphasize intensive sampling at a few critical locations.