Quantifying model uncertainty in terrestrial water cycle models due to channel hydraulics

Poster Number: 
145
Presenter/Primary Author: 
Ed Beighley
Co-Authors: 
Ram Ray
Co-Authors: 
Yiping He
Co-Authors: 
John Melack
Co-Authors: 
Blair Goodridge

Understanding the impacts of landuse and/or climate change on streamflow characteristics, such as peak discharge, sediment transport capacity, flow velocities and depths, within a given region or watershed requires knowledge of fine scale (0.01-10 m) hydraulic channel properties (i.e., detailed cross-sections, roughness, bed material). However, data for channel/reach properties are limited to primarily in-situ measurements. For watersheds greater than 10-100 sq km or regional to continental systems, hydrologic models that include hydraulic routing require representative channel cross-sections at 100’s to 1,000’s of locations. Although these models are capably of providing hydraulic realism (channel/floodplain flow depth, width and velocity) at the reach scale, channel data limitations are significant. This research addresses this limitation for watersheds in the Santa Barbara Coastal (SBC) LTER study region. Specifically, this study investigates the applicability of representing channels using: w=ad^b, where w is width, d is depth, a and b are coefficients determined from SBC-LTER data. The importance of a and b for estimating peak discharges and flow velocities are presented as a function of discharge, drainage area and slope for current land use and climate conditions.