Estimation and analysis of the uncertainty in streamflow and change in reservoir-content data at selected streamflow-gaging stations in the lower Colorado River network, 1995-99

Abstract

New methods were used to estimate the standard error of the annual discharge and the annual change in reservoir content at stations in the lower Colorado River streamflowgaging network. The Moss and Gilroy (1980) method estimates the standard error of the annual discharge by using information contained in the periodic discharge measurements made at a streamflow-gaging station and was modified to be more site-specific and reflective of actual conditions during the year of interest. The two main improvements are using only discharge measurements pertinent to the year of interest rather than all discharge measurements, and using a semivariogram to estimate three discharge-uncertainty parameters: measurement variance, process variance, and the one-day serial correlation coefficient of the discharge-rating residuals. The new method for estimating the standard error of the annual change in reservoir content uses the standard error of the two reservoir stage readings and the corresponding reservoir surface areas for the readings.