Event-based dimensionless models for runoff of Chabagou watersheds

In order to establish the prediction models of the event-based runoff in small watersheds, the observed event-based runoff data of 10 years from 4 Chabagou watersheds were used to analyze the effects of rainfalls and landforms on runoff. It was found that such factors as the maximum rainfalls and intensities in 30 minutes, the gradient of main channels, the areas of watershed, and the ratios of watershed lengths to widths were key factors which affect runoff. Dimensionless parameters and step-wise regression were used to quantify the parameters of the event-based runoff model. The established model used few parameters with simple structure. The featured with clearly defined physical meanings with 70% of prediction accuracy, which could supply a basis for further research of models for watershed runoff prediction.