Modeling Bankfull Channel Geometry Based on Watershed and Precipitation Characteristics Using Dimensionless Parameters

Regression models have consistently identified discharge as a suitable variable for capturing the systematic trends in channel geometry. Using a discharge-based relation, however, does not reflect the more fundamental dependence of channel geometry on basic watershed properties and precipitation characteristics. To explore relations between bankfull channel geometry and watershed properties, we consider two sets of dimensionless regression models: (a) with bankfull discharge as an independent variable, and (b) with bankfull discharge as a dependent variable. While bankfull discharge is used as an input variable in the first set of regression analyses, average annual precipitation has been used in the second set. Other watershed and channel variables considered include drainage area, median sediment grain size, and valley slope. A data set of 538 sandy and gravel channel reaches from UK, and various physiographic provinces of the USA is considered. By representing bankfull discharge, width, depth and watershed variables in the form of dimensionless parameters, we specify the relationship between channel morphology and watershed and precipitation characteristics using an approach that is applicable to sandy and gravel streams from a range of climatic and physiographic regions. The formulation presented here is an effort toward the development of models that will be capable of predicting bankfull channel dimensions and discharge under modified average annual precipitation due to change in climate.

Automated summary

Regression models have identified discharge as a suitable variable for capturing channel geometry trends. However, using discharge alone does not reflect the fundamental dependence of channel geometry on watershed properties and precipitation characteristics.