Sediment yield and spurious correlation—toward a better portrayal of the annual suspendend-sediment load of rivers

Abstract

Bivariate relations between annual sediment yield (tons per year per unit drainage area) and drainage-basin area are spurious because drainage-basin area is common to both axes. Two alternative methods for portraying the annual suspended-sediment load of a river are suggested. One method consists of plotting suspended-sediment load (tons per year) against distance downstream. Such plots indicate that annual suspended-sediment load does not necessarily have a linear relationship with distance. The second method consists of plotting annual suspended-sediment load against drainage-basin area. Both methods more accurately portray fundamental relations between annual sediment load and drainage-basin characteristics than does the yield-area relation because spurious correlation is avoided. Plots were made of annual suspended-sediment load versus time for several stations along each of eight rivers for the 10–15 years of available data. The plots are in-phase with respect to relative magnitude of annual sediment loads.