Uncertainty in Water Quality Data

Abstract Water quality data are collected to provide information to assist in the understanding and managing of water resources. The usefulness of water quality data collected is inversely related to the amount of uncertainty in the data. Data uncertainty may be defined as a state of doubt in how representative observed values are of the true population characteristics. Data uncertainty may be estimated as a function of both sampling and nonsampling errors. Sampling errors result from the sampling network design (location and frequency of sample collection) which samples only a subset of the total population. Nonsampling errors result from the process of measuring the amount of water quality material present. The measurement process may be divided into sample collection and laboratory analysis. Sample collection includes the physical procedure for obtaining, storing, and transporting a water sample for later analysis. Laboratory analysis consists of some method of estimating the amount (concentration) of a given material in the water sample. Presented here is a general discussion of the sources of water quality data uncertainty, a method to estimate data uncertainty in water quality variables, and the implications of uncertainty in water quality data.