Using high-resolution topography for advancing the understanding of mass and energy transfer across landscapes

Channel networks are a critical means of mass and energy transfer across the landscape. High resolution topographic imagery derived from lidar scans, provides new opportunities in the observation and analysis of these processes, especially as the resolution of these data is proportionate to channel and hillslope process scales. Channel feature extraction algorithms supply a method with which to analyze hydrologic and geomorphic processes; automatic, open-source frameworks such as GeoNet aim to provide a reliable platform for this task to the academic, public service, and professional communities. In this thesis, the GeoNet algorithm is tested across different types of landscapes and data resolutions. Innovative analysis methods are also assessed within the framework in order to advance the method in parallel with advancing understanding of channel processes and improving technologies. The goal of this research is to assess GeoNet performance across landscape type in terms of relief, vegetation, and anthropogenic influence and make recommendations for future development of the algorithm and the GeoNet user community. An alternate spectral analysis-based filtering method is tested, as well as curvature-based filtering of erroneously identified channel heads. Results from alternate filtering testing indicate that the nonlinear Perona-Malik filter is superior to a spectral-based filtering approach. The use of a contour curvature threshold is not wholly successful at removing spuriously identified channel heads in natural landscapes. This research also introduces the analysis of urban landscapes into the GeoNet repertoire, and examines the effects of data resolution on feature extraction therein. An urban dataset from the Austin, TX area is tested and the optimal settings for GeoNet are identified; field work at the site is used to validate the results. The recommendations resulting from this work aim to improve in the functionality and versatility of GeoNet, and enhance accessibility for the user-community.

Name thesis:degree_name

Master of Science in Engineering

Level thesis:degree_level

Masters

Discipline thesis:degree_discipline

Environmental and Water Resources Engineering

Grantor

The University of Texas at Austin

Year dc:date.issued

2015

Author dc:creator

• Kladzyk, Anna Gabrielle

Advisor dc:contributor.advisor

• Passalacqua, Paola

Committee member dc:contributor.committeemember

• Maidment, David

• High resolution topography
• Channel feature extraction

Language dc:language.iso

en

Identifier

doi:10.15781/T2NS4D

OAI identifier oai:identifier

oai:repositories.lib.utexas.edu:2152/31773