The basic hydrologic data required to determine the water yield are usually unavailable for small basins and streams while increasing emphasis is being placed on their development. Therefore, some methods and techniques for estimating the amount of water available for development of these small units is needed. The purpose of this study is to use the concepts and techniques of statistical analysis to develop equations which are useful in estimating the water yield of watersheds for which no stream flow records are available. The approach is an extension of earlier studies at Utah State University (1, 10) in which physiographic and topographic parameters were related to mean annual runoff of Utah watersheds. Previous studies used multiple regression techniques primarily. The work reported herein utilizes the same data as in the earlier work but analysis is based on the multivariate technique of principal component analysis. Results and evaluations derived from the principal component analysis are compared with those obtained from multiple regression analysis.
Wang, Leei-Luoh and Huber, A. Leon, "Estimating Water Yields in Utah by Principal Component Analysis" (1967). Reports. Paper 21.