Accurate and representative sediment data are critical for assessing the potential effects of highway and urban runoff on receiving waters. The U.S. Environmental Protection Agency identified sediment as the most widespread pollutant in the Nation's rivers and streams, affecting aquatic habitat, drinking water treatment processes, and recreational uses of rivers, lakes, and estuaries. Representative sediment data are also necessary for quantifying and interpreting concentrations, loads, and effects of trace elements and organic constituents associated with highway and urban runoff. Many technical issues associated with the collecting, processing, and analyzing of samples must be addressed to produce valid (useful for intended purposes), current, complete, and technically defensible data for local, regional, and national information needs. All aspects of sediment data-collection programs need to be evaluated, and adequate quality-control data must be collected and documented so that the comparability and representativeness of data obtained for highway- and urban-runoff studies may be assessed.
Collection of representative samples for the measurement of sediment in highway and urban runoff involves a number of interrelated issues. Temporal and spatial variability in runoff result from a combination of factors, including volume and intensity of precipitation, rate of snowmelt, and features of the drainage basin such as area, slope, infiltration capacity, channel roughness, and storage characteristics. In small drainage basins such as those found in many highway and urban settings, automatic samplers are often the most suitable method for collecting samples of runoff for a variety of reasons. Indirect sediment-measurement methods are also useful as supplementary and(or) surrogate means for monitoring sediment in runoff. All of these methods have limitations in addition to benefits, which must be identified and quantified to produce representative data. Methods for processing raw sediment samples (including homogenization and subsampling) for subsequent analysis for total suspended solids or suspended-sediment concentration often increase variance and may introduce bias. Processing artifacts can be substantial if the methods used are not appropriate for the concentrations and particle-size distributions present in the samples collected.
Analytical methods for determining sediment concentrations include the suspended-sediment
concentration and the total suspended solids methods. Although the terms suspended-sediment concentration and total suspended solids are often used interchangeably to describe the total concentration of suspended solid-phase material, the analytical methods differ and can produce substantially different results. The total suspended solids method, which commonly is used to produce highway- and urban-runoff sediment data, may not be valid for studies of runoff water quality. Studies of fluvial and highway-runoff sediment data indicate that analyses of samples by the total suspended solids method tends to under represent the true sediment concentration, and that relations between total suspended solids and suspended-sediment concentration are not transferable from site to site even when grain-size distribution information is available. Total suspended solids data used to calculate suspended-sediment loads in highways and urban runoff may be fundamentally unreliable. Consequently, use of total suspended solids data may have adverse consequences for the assessment, design, and maintenance of sediment-removal best management practices. Therefore, it may be necessary to analyze water samples using the suspended-sediment concentration method. Data quality, comparability, and utility are important considerations in collection, processing, and analysis of sediment samples and interpretation of sediment data for highway- and urban-runoff studies. Results from sediment studies must be comparable and readily transf