Alternative principal components regression procedures for dendrohydrologic reconstructions

Streamflow reconstruction using tree ring information (dendrohydrology) has traditionally used principal components analysis (PCA) and stepwise regres sion to form a transfer function. However, PCA has several procedural choic es that may result in very different reconstructions. This study assesses t he different procedures in PCA-based regression and suggests alternative pr ocedures for selection of variables and principal components. Cross-validat ion statistics are presented as an alternative for independently testing an d identifying the optimal model. The objective is to use these statistics a s a measure of the model's performance to find a conceptually acceptable mo del with a low prediction error and the fewest number of variables. The res ults show that a parsimonious model with a low mean square error can be obt ained by using strict rules for principal component selection and cross-val idation statistics, Additionally, the procedure suggested in this study res ults in a model that is physically consistent with the relationship between the predictand and the predictor. The alternative PCA-based regression mod els presented here are applied to the reconstruction of the Upper Colorado River Basin streamflow and compared with results of a previous reconstructi on using traditional procedures. The streamflow reconstruction proposed in this study shows more intense drought periods, which may influence the futu re allocation of water supply in the Colorado River Basin.