Premium
TerEx Toolbox for semi‐automated selection of fluvial terrace and floodplain features from lidar
Author(s) -
Stout Justin C.,
Belmont Patrick
Publication year - 2014
Publication title -
earth surface processes and landforms
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.294
H-Index - 127
eISSN - 1096-9837
pISSN - 0197-9337
DOI - 10.1002/esp.3464
Subject(s) - floodplain , terrace (agriculture) , channel (broadcasting) , digital elevation model , riparian zone , geology , fluvial , hydrology (agriculture) , watershed , remote sensing , geomorphology , cartography , geography , computer science , structural basin , computer network , ecology , geotechnical engineering , machine learning , habitat , biology , archaeology
ABSTRACT Terraces and floodplains are important indicators of near‐channel sediment dynamics, serving as potential sediment sources and sinks. Increasing availability of high resolution topography data over large areas calls for development of semi‐automated techniques for identification and measurement of these features. In this study we introduce a novel tool that accommodates user‐defined parameters including, a local‐relief threshold selected by a variable‐size moving window, minimum area threshold, and maximum distance from the channel to identify and map discrete terrace and floodplain surfaces. Each of the parameters can easily be calibrated for a given watershed or reach. Subsequently, the tool automatically measures planform area, absolute elevation, and height relative to the local river channel for each terrace polygon. We validate the tool in two locations where terrace maps were previously developed via manual digitization from lidar and extensive field mapping campaigns. The tool is also tested on six different types of rivers to provide examples of starting selection parameters, and to test effectiveness of the tool across a wide range of landscapes. Generally, the tool provides a high quality draft map of terrace and floodplain surfaces across the wide range of environmental conditions for which it has been tested. We find that the tool functions best in catchments where the terraces are spatially extensive, with distinct differences between the terrace and floodplain. The most challenging environments for semi‐automated terrace and floodplain mapping include steep catchments with dense riparian vegetation, and very small terraces (~10 m 2 in areal extent). We then apply the tool to map terraces and floodplains in the Root River watershed, southeastern Minnesota and generate exceedance plots for terrace heights. These plots provide a first pass analysis to indicate the tributaries and reaches of the river where terraces constitute a significant source of sediment. Copyright © 2013 John Wiley & Sons, Ltd.