Effects of Toxic Pollutants on Aquatic Resources Using Statistical Models and Techniques to Extrapolate Acute and Chronic Effects Benchmarks

摘要

Data extrapolations are required to assess the effects of toxic pollutants on aquatic resources. We present a procedure that uses maximum likelihood methods of estimating a functional linear relationship for extrapolation. Optimal estimates for the variance of the extrapolation parameters as well as the fitted values do not exist. We propose several alternative variance estimates which are used for computations of the extrapolated chronic benchmark and its distribution. Improved estimates can be obtained by applying computer-intensive resampling methods.