Logo image
Back
EVALUATING AND DEVELOPING STREAMBANK EROSION MODELS IN THE GULF OF MEXICO COASTAL PLAIN
Thesis   Open access

EVALUATING AND DEVELOPING STREAMBANK EROSION MODELS IN THE GULF OF MEXICO COASTAL PLAIN

Mitchell McMillan
University of West Florida
Master of Science (MS), University of West Florida
2016

Metrics

4 File views/ downloads
49 Record Views

Abstract

The BANCS method has become a standard for predicting streambank erosion rates within a given region, but relies on subjective estimates of root density and has not been calibrated in the Gulf of Mexico coastal plain. This thesis presents three approaches to modeling streambank erosion rates. First, standard BANCS models were calibrated. Statistical analyses show that they are not significantly correlated to erosion rates due to inadequate estimates near-bank shear stress. Modifications to this estimate improve the models significantly (R2 = 0:31). Second, a statistical modeling process was developed using physical measurements, biomass density surveys, and soil analyses. The best statistical model requires similar data collection to BANCS, but offers higher accuracy (R2 = 0:54). Finally, a unique spatially-distributed model was developed. The model was designed to utilize global remote-sensing data and a physics-based hydrodynamic model and include the major sources of spatial and temporal variability. The model reproduces measured erosion rates with a Pearson's coefficient of R = 0:55. The spatially-distributed model allows streambank erosion rate to be estimated at any location on earth where channel geometry data are available and requires little to no field data collection. The model fills a long-standing gap in the prediction of regional sediment yield.
pdf
uwf:61217DownloadView
Open Access

Details

Logo image