Graphical and Statistical Methods to Assess the Effect of Landfills on Groundwater Quality

Abstract: Wisconsin’s groundwater protection law requires establishment of numerical standards for potential groundwater contaminants at landfill sites. For contaminants identified as being of public health or welfare concern, the standard is based on the assessed risk of the contaminant to human health. The standards for groundwater quality indicator parameters are based on changes in these constituents relative to background water quality conditions. Present methods used to set standards for indicator parameters are based on parametric statistical analyses of the data. Statistical characteristics of groundwater quality data may violate assumptions of common parametric methods and lead to incorrect inference. To determine if alternative statistical analyses might be preferable, I tested existing landfill monitoring data for skewness, serial correlation and seasonality and worked to develop meaningful and applicable methods for evaluating inorganic and COD data typically available from landfill monitoring. I found that skewness was most commonly observed, followed by positive serial correlation, and that nonparametric tests for assessing these characteristics performed better than corresponding parametric tests.