Laserfiche WebLink
<br />The model allows for the input of multiple rain-gage data for <br />soil-moisture accounting. With multiple input, pairs of distance-area <br />curves for pervious and impervious areas are required for each subbasin. <br />The combination of distributed moisture accounting (multiple-input) and <br />distributed routing make it possible to simulate the effects of both <br />rainfall variability and urban development. <br /> <br />The measured impervious area was reduced and the pervious area was <br />correspondingly increased to account for the flow from impervious sur- <br />faces into pervious areas as described subsequently. This reduced im- <br />perviousness is called effective imperviousness. <br /> <br />Model-parameter values are determined by a "hill climbing" optimi- <br />zation technique (Dawdy, Lichty, Bergmann, 1972), which uses bounded pa- <br />rameters (to constrain the parameters within reasonable limits) and an <br />objective function. The objective function is the sum of the squared <br />deviations of the logarithms of observed and simulated peak flows and <br />(or) volumes. The combined objective functions give weight to both the <br />storm peaks and storm volumes, while the single function weights either <br />storm peaks or storm volumes as specified. <br /> <br />The parameters are automatically adjusted to minimize the differ- <br />ences between observed and simulated data, as specified in the objective <br />function. A set of parameters can be greater than, less than, or nearly <br />an average of the "true" values. Because of interactions and assump- <br />tions, some error always exists. If the differences between observed <br />and simulated data are assumed to be random, then the standard error of <br />estimate, according to statistical analysis, is a meaningful measure of <br />the error (Riggs, 1968). <br /> <br />The input data required for model calibration are rainfall, pan <br />evaporation, storm rainfall, discharge, initial-parameter estimates, <br />base-flow estimates, and an appropriate number of distance-area curves. <br /> <br />DATA-COLLECTION METHODS AND USE OF DATA <br /> <br />Data for this study were collected during 1962-70 (water years) <br />in seven principal basins and six subbasins (fig. 3), which ranged from <br />rural to fully urbanized. <br /> <br />Urbanization within a basin may occur in the lower or upper parts, <br />and construction and drainage improvement may occur concurrently or at <br />differing times and rates. The states of urbanization that existed in <br />1968 are assumed as mean conditions for the period of record of rainfall <br />and runoff. <br /> <br />-8- <br />