Augmenting Annual Runoff Records Using Tree-Ring Data

Abstract

INTRODUCTION: Any statistical work involving hydrologic records is handicapped when the records are of relatively short duration, as are most such records in the Southwestern United States. This is because the short records are not necessarily a random sample of the infinite population of events, and consequently any statistical descriptions are likely to be in error to some extent. Work recently completed at the Laboratory of Tree-Ring Research [Stockton, 1971] has shown that tree ring data can be used to extend available runoff records backward in time, thereby providing a longer record from which to more accurately estimate the three most common statistics used in hydrology, the mean, the variance, and the first order autocorrelation.