Mean Estimate Deficiencies in Water Quality Studies

Mathematical models are employed in the design, performance prediction, and evaluation of alternative water quality management programs. This paper examines the relative influence of deterministic and stochastic models on water quality management decisions to demonstrate some deficiencies of decisions based on mean estimates produced by deterministic models. Water quality management problems are examined which consider the implications of variability in waste generation, treatment plant performance, and receiving water behavior on the resulting variability of water quality. An example is provided in an evaluation of regional wastewater management alternatives given by the size, number, and location of regional treatment plants. It is concluded that an evaluation of water quality management alternatives without consideration for their performance variability may be so deficient as to negate the evaluation.