A DYNAMIC, PHYSICALLY-BASED METHOD FOR ESTIMATING EROSION OF PERVIOUS URBAN LAND

The erosion dynamics of pervious urban surfaces are described through field observation and by a deterministic model. Sampling was done at 8 sites (7 pervious; 1 impervious) representing different surface characteristics. In general, erosion rates were greater for pervious aggregate surfaces (e.g. parking lots, railway land) than grassed surfaces and also increased with slope length and steepness. Total solids yield rates from an aggregate lot were 1 to 2 orders of magnitude greater than from adjacent suburban land (lawns, roofs, and roads) for sampled events. The aggregate lot therefore contributed 2 to 25 times more total solids to the sewer system than did the suburban land with 10 times the area. A modified version of the deterministic CREAMS model was used to estimate erosion rates and eroded particle size distribution for 6 events at an aggregate parking lot and 5 events at a grassed playing field. Prediction errors in event yields averaged 63% and 12% for the aggregate lot and grassed field, respectively. Pervious land can be a source of particulates and pollutants to the sewer system and should be considered in detail when modelling urban stormwater runoff quality.