Rainfall simulation studies were conducted in conjunction with the Denver Regional Urban Runoff Program to: (1) Compare runoff quantity and quality from two different intensities of rainfall on impervious plots having identical antecedent conditions, (2) document a first flush of constituent loads in runoff from l,000-square-foot street-surface plots, (3) compare runoff characteristics from a street surface subjected to simulated rainfall with those from a 69-acre urban basin of mixed land use subjected to natural rainfall, (4) perform statistical analysis of constituent loads in the runoff with several independent variables, and (5) compare the quantity and quality of runoff from 400-square-foot plots of native grasses used for pasture and subjected to simulated rainfall with that from a 405-acre basin covered with native grasses used for pasture and subjected to natural rainfall.
The rainfall simulations conducted on the street surface showed that higher intensity simulated rainfall produced a higher percentage of runoff than lower intensity rainfall. A first flush of constituent loads occurred for most constituents in the runoff from most rainfall simulations on the street surface; however, a first flush did not occur in the runoff from simulated rainfall on the pasture. The event mean concentrations of constituents in the runoff from simulated storms on the street surface were generally much smaller than the event mean concentrations of constituents in the runoff from an adjacent urban basin.
Analysis of the data from the rainfall simulations on a street surface indicates that intensity of rainfall and total rainfall are important variables determining constituent loads. The design of the experiment was such that intensity of rainfall and total rainfall were highly correlated, thus precluding the development of useful regression equations to predict washoff loads.
The quality of runoff from the simulated rainfall on the pasture was influenced by the disturbed perimeters of the plots; however, the runoff-to-rainfall ratios for the simulated storms fell within the range of ratios measured for natural storms over the adjacent 405-acre basin.