The use of surface water monitoring data for the estimation of potential extremes of pesticide annual concentration profiles is often challenged in pesticide regulatory assessments due to non-daily sampling (i.e., missing the peaks). Typical sampling frequencies are every 28-days, every 14-days, or every 7-days.  Because of this, direct estimators of extremes (e.g. high percentiles, maximum m-day rolling averages) are often both highly variable and downwardly biased.  To address this, a method known as bias factors has been proposed, which upwardly adjusts sample estimates by a protective multiple.  In this presentation we discuss the bias factor approach and its application for both estimation and decision making.  Alternative methods including multiplicative factors and kriging approaches are proposed and compared in performance to bias factors using extensive monitoring datasets having every-day (or near-daily) sampling from streams and lakes highly vulnerable to storm-induced runoff in agricultural watersheds.

Consideration of using bias factors and other methods to estimate potential maximum concentrations in monitoring data

254th National Meeting (2017)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.