Water quality in the lower Colorado River and the effect of reservoirs,

Abstract

Comparison of the power spectra of TDS time series form different locations on the Lower Colorado River is useful in showing changes in salinity and for indicating physical factors influencing salinity. Similarities between the power spectra of the Lee Ferry and Grand Canyon time series indicated that lateral inputs and evaporation are not greatly influencing the salinity cycle. The salinity change within this reach was approximated by a constant concentration change of 66.6 ppm. A similar model form was used for the Hoover Dam to Parker Dam reach. Dissimilarities between power spectra indicated that additional inputs are significant and must be accounted for in any model of such reaches. The model for Lake Mead required compensation for evaporation and for the inputs of the Virgin River and Las Vegas Wash. The modeled salinity increase between Parker Dam and Yuma contained a trend factor to allow for the effect of irrigation return flows and seepage. The crosscovariance function was used to approximate the time lag between data stations. Time series statistics, including coherence, response function spectra, and overall unit response, were used and are of utility in estimating salinity in a river system.