Analysis of soil chemistry data for an EPA (Environmental Protection Agency) acid rain assessment

The Direct/Delayed Response Project (DDRP), implemented by the US Environmental Protection Agency (EPA), as part of the National Acid Precipitation Assessment Program (NAPAP), was designed to project the long-term response of selected eastern US watersheds and surface waters to acidic deposition. Specific DDRP tasks include (i) to characterize variability of soil and other watershed attributes, (ii) to determine which of the soil and watershed characteristics are most strongly related to surface water chemistry, (iii) to estimate the relative importance of key watershed processes in controlling surface water chemistry, and (iv) to classify a sample of watersheds with regard to their response time and extrapolate the classification to the population of lakes and streams in the region. An overview of the DDRP is presented relating it to other EPA studies and describing its implementation. The main problems addressed are collinearity and variable selection in multiple regression when the number of variables exceeds the number of observations.