Table 1: Pollutant Concentrations by Land Use: Total Suspended Solids (mg/l)

Land Use
Source Residential Commercial Roadway Industrial Notes
Schueler, 1987 mean 1001 - - - This value reflects an estimate based on 25 data points from a wide range of watershed sizes. Data reflect instream concentrations. A small watershed size (i.e., 10 acres) was assumed to minimize the influence of the channel erosion component.
Gibb et al., 1991 mean 150 - 220 - These values represent recommended estimates for planning purposes and are based on an analysis of mean concentrations from over 13 studies from the US and British Columbia.
Smullen and Cave, 1998 median 55 55 55 55 This study probably represents the most comprehensive data set, with 3,047 event samples being included from across the nation. Data includes pooled NURP, USGS, and NPDES sources. The value is a median of EMCs and applies to general urban runoff (i.e., mixed land uses). The low concentration relative to other data can be attributed to the fact that, while NURP data represent small watersheds where channel erosion may play a role, NPDES data are collected as "end of the pipe" concentrations for very small drainage areas of a uniform land use. The NPDES concentrations were approximately 70% lower than concentrations from NURP or USGS..
US EPA, 1983 median 101 69 - - These values represent NURP data for residential and commercial land use. NURP data were collected in the early 1980s in over 28 different metropolitan areas across the US.
Claytor and Schueler, 1996 - - 142 124 The roadway value is the un-weighted mean of 8 studies conducted by the FHWA. The industrial value is the mean value from 6 storms monitored at a heavy industrial site in Auckland, NZ.
Barrett and Malina, 1998 - - 173 - This data reflects a study of vegetative swales treating highway runoff in Austin, TX. Value represents average of the mean inflow concentrations measured at 2 sites. Data were collected over 34 storm events.
Caraco and Schueler (1999). Arid Climates 242 242 242 242 This value represents an average of EMC data collected from 3 arid climate locales (Phoenix, Boise, and Denver). A total of 90 data points are used, with each site having at least 16 data points. Value applies to general urban runoff (i.e., mixed land uses).
Driscoll, 1986 - - 242 - This value is the average of 4 median EMCs collected from highway sites in Nashville, Denver, Milwaukee, and Harrisburg. A total of 93 data points were used to develop value, with each site having at least 16 data points.
Shelley and Gaboury, 1986 - - 220 - This value is the median value of 8 highway studies from across the US. Some of the data from the Driscoll study (1986) is included.
Whalen and Cullum, 1988 228 168 - 108 These data are from an assessment of urban runoff quality that looked at NURP and State of Florida data. The NURP data are presented. Residential and commercial values are mean values for specified land uses and reflect between 200 and 1,100 sampling events depending on the parameter and land use. Industrial values are from 4 NURP sites and generally represent light industrial land use.
Model Default Value2 100 75 150 120
1: Concentration based on a 10-acre drainage area

  • The model default values represent best professional judgement, and give additional weight to studies conducted at a national level. Data do not incorporate studies on arid climates.