Estimating Sediment Discharge

Abstract

Sediment-discharge measurements usually are available on a discrete or periodic basis. However, estimates of sediment transport often are needed for unmeasured periods, such as when daily or annual sediment-discharge values are sought, or when estimates of transport rates for unmeasured or hypothetical fl ows are required. Selected methods for estimating suspended-sediment, bed-load, bedmaterial-load, and total-load discharges have been presented in some detail elsewhere in this volume. The purposes of this contribution are to present some limitations and potential pitfalls associated with obtaining and using the requisite data and equations to estimate sediment discharges and to provide guidance for selecting appropriate estimating equations. Records of sediment discharge are derived from data collected with suffi cient frequency to obtain reliable estimates for the computational interval and period. Most sedimentdischarge records are computed at daily or annual intervals based on periodically collected data, although some partial records represent discrete or seasonal intervals such as those for fl ood periods. The method used to calculate sedimentdischarge records is dependent on the types and frequency of available data. Records for suspended-sediment discharge computed by methods described by Porterfi eld (1972) are most prevalent, in part because measurement protocols and computational techniques are well established and because suspended sediment composes the bulk of sediment discharges for many rivers. Discharge records for bed load, total load, or in some cases bed-material load plus wash load are less common. Reliable estimation of sediment discharges presupposes that the data on which the estimates are based are comparable and reliable. Unfortunately, data describing a selected characteristic of sediment were not necessarily derived—collected, processed, analyzed, or interpreted—in a consistent manner. For example, bed-load data collected with different types of bed-load samplers may not be comparable (Gray et al. 1991; Childers 1999; Edwards and Glysson 1999). The total suspended solids (TSS) analytical method tends to produce concentration data from open-channel fl ows that are biased low with respect to their paired suspended-sediment concentration values, particularly when sand-size material composes more than about a quarter of the material in suspension. Instantaneous sediment-discharge values based on TSS data may differ from the more reliable product of suspendedsediment concentration values and the same water-discharge data by an order of magnitude (Gray et al. 2000; Bent et al. 2001; Glysson et al. 2000; 2001). An assessment of data comparability and reliability is an important fi rst step in the estimation of sediment discharges. There are two approaches to obtaining values describing sediment loads in streams. One is based on direct measurement of the quantities of interest, and the other on relations developed between hydraulic parameters and sedimenttransport potential. In the next sections, the most common techniques for both approaches are briefl y addressed.