Science Inventory

Riparian Sediment Delivery Ratio: Stiff Diagrams and Artifical Neural Networks

Citation:

Ssegane, H., E. W. Tollner, AND S. C. MCCUTCHEON. Riparian Sediment Delivery Ratio: Stiff Diagrams and Artifical Neural Networks. Transactions of the ASABE. AMERICAN SOCIETY OF AGRICULTURAL AND BIOLOGICAL ENGINEERS, ST. JOSEPH, MI, 52(6):1885-1893, (2009).

Impact/Purpose:

see description

Description:

Various methods are used to estimate sediment transport through riparian buffers and grass jilters with the sediment delivery ratio having been the most widely applied. The U.S. Forest Service developed a sediment delivery ratio using the stiff diagram and a logistic curve to integrate some of the factors influencing sediment delivery heuristically. This study independently tested the Forest Service sediment delivery ratio contrasted with artificial neural networks to represent the multiple nonlinearities between important factors and sediment delivery. The Forest Service sediment delivery ratio was not adequate when compared to published sediment yields from 30 small experimental buffers from three countries, including four forested buffers. However, artificial neural networks gave estimates of the delivery ratio that were highly correlated to the observations. The 30 buffer observations produced such good estimates of the sediment delivery ratio with both seven and five buffer parameters that this study suggests that as few as 30 sediment yield observations can be the basis for applying neural networks to interpolate the complex, multiple nonlinearities of hydrology and sediment transport on riparian buffers.

URLs/Downloads:

Riparian Sediment Delivery Ratio: Stiff Diagrams and Artifical Neural Networks   Exit

MCCUTCHEON 09 092 ABSTRACT ONLY.PDF   (PDF,NA pp, 11 KB,  about PDF)

Record Details:

Record Type: DOCUMENT (JOURNAL/PEER REVIEWED JOURNAL)
Product Published Date: 11/01/2009
Record Last Revised: 04/22/2010
OMB Category: Other
Record ID: 212089