Impacts of Californian dams on flow regime and maximum/minimum flow probability distribution

Abstract

Dams have major impacts on river hydrology with a general tendency to decrease annual maximum flows and increase annual minimum flows. The analysis of 41 streamflow series in California, USA are examined, and the results show that, as expected, the mean values and variations of annual peak flows and maximum flows of different durations are reduced for almost all sites after dam use, and the larger the ratio of total reservoir capacity to pre-dam annual runoff, the larger the rate of peak flow reduction. However, the impacts on minimum flow are mixed. For five out of seven cases with long data records for periods before and after dam use, the average annual minimum flow as well as its variation increased, but for the other two cases, they decreased. No significant changes are detected for various extreme precipitation indices; therefore, dam construction is believed to be the major reason for flow regime changes. The probability distribution of extreme flows also changed, due to the impacts of dams. The Log-Pearson Type III (LP3) distribution is best for peak flow series and one-day maximum series at sites with or without the impact of dams; the three-parameter Weibull (W3) distribution is the best model for the seven-day minimum flow at sites with no or minor dam impacts, whereas at sites with major dam impacts, the best model is the generalized extreme value (GEV) model for the seven-day minimum flow.