You are here:
Improving Water Quality Assessments through a HierarchicalBayesian Analysis of Variability
Citation:
GRONEWOLD, A. AND M. E. Borsuk. Improving Water Quality Assessments through a HierarchicalBayesian Analysis of Variability. ENVIRONMENTAL SCIENCE AND TECHNOLOGY. John Wiley & Sons, Ltd., Indianapolis, IN, 44(20):7858-7864, (2010).
Impact/Purpose:
The United States Environmental Protection Agency (USEPA) TotalMaximumDaily Load (TMDL) program is the nation’s most comprehensive and far-reaching program governing protection and improvement of surface water quality (1–3). The TMDL program requires that states identify waters failing to meet water quality standards and then determine the maximum allowable pollutant load that can enter such waters and yet meet applicable water quality standards (4). Although the TMDL program was adopted into United States policy as part of the 1972 Amendments to the Federal Water Pollution Control Act (commonly referred to as the Clean Water Act), TMDL assessments were not completed on a large scale until the late 1990s (5). Since then, the number of TMDLs addressed by USEPA has increased almost every year (6).
Description:
Water quality measurement error and variability, while well-documented in laboratory-scale studies, is rarely acknowledged or explicitly resolved in most water body assessments, including those conducted in compliance with the United States Environmental Protection Agency (USEPA) Total Maximum Daily Load (TMDL) program. In particular, deterministic model forecasts are commonly compared directly to corresponding sample-based water quality standards without explicitly accounting for measurement error. Such error may be a result of variability in measurements made using a single procedure, or variability across measurements made using multiple procedures, each with its own sources of error. Consequently, model-based management decisions, including proposed pollutant loading reductions in TMDLs and similar studies, may be biased, resulting in either slower-than-expected rates of water quality restoration and designated use reinstatement or, in some cases, overly-conservative management decisions. To address this problem, we present an approach to quantifying variability within and between pollutant concentration estimates from different sampling and analysis procedures based on a Bayesian assessment of measurement uncertainty. We apply this method to recently approved TMDLs to investigate whether appropriate accounting for measurement variability will lead to different management decisions. We find that required pollutant loading reductions may in fact vary depending not only on how measurement variability is addressed, but also on which analysis-based standard is imposed and which water quality analysis procedure is used. As a general strategy, our approach may represent a solution to problems with current TMDL protocol, particularly the common practice of attempting to address all potential sources of uncertainty through a single margin of safety (MOS).