DETERMINING AN OPTIMAL SAMPLING FREQUENCY FOR MEASURING BULK TEMPORALCHANGES IN GROUNDWATER QUALITY

In the Data Quality Objectives (DQO) process, statistical methods are used to determine an optimal sampling and analysis plan, When the DQO decision rule for instituting remedial actions is based on a critical change in water quality, the monitoring program design must ensure tha t this change can be detected and measured with a specified confidence , Usually the focus is on the change at a single monitoring location a nd the process is limited to addressing the uncertainty inherent in th e analytical methods and the variability at that location, However, ne w strategies that permit ranking the waste sites and prioritizing reme dial activities require the means for assessing overall changes for sm all regions over time, where both spatial and temporal variability exi st and where the uncertainty associated with these variations far exce eds measurement error, Two new methods for assessing these overall cha nges have been developed and are demonstrated by application to a wast e disposal site in Oak Ridge, Tennessee. These methods incorporate his torical data where available and allow the user to either test the sta tistical significance of a linear trend or of an annual change compare d to a baseline year for a group of water quality wells.