Repository logo
 

Identification and characterization of dendritic, parallel, pinnate, rectangular and trellis networks based on deviations from planform self-similarity

Date

2006

Authors

Mejía, Alfonso I., author
Niemann, Jeffrey D., advisor
Wohl, Ellen, committee member
Ramírez, Jorge A., committee member

Journal Title

Journal ISSN

Volume Title

Abstract

Geomorphologists have long recognized that the geometry of channel network planforms can vary significantly between regions depending on the local lithologic and tectonic conditions. This tendency has led to the classification of channel networks using terms such as dendritic, parallel, pinnate, rectangular, and trellis. Unfortunately, available classification methods are scale dependent and have no connection to an underlying quantitative theory of drainage network geometry or evolution. In this study, a new method is developed to classify drainage networks based on their deviations from self-similarity. The planform geometry of dendritic networks is known to be self-similar. It is our hypothesis that parallel, pinnate, rectangular, and trellis networks correspond to distinct deviations from this self-similarity. To identify such deviations, three measures of channel networks are applied to ten networks from each classification. These measures are the incremental accumulation of drainage area along channels, the irregularity of channel courses, and the angles formed by merging channels. The results confirm and characterize the self-similarity of dendritic networks. Parallel and pinnate networks are found to be self-affine with Hurst exponents around 0.8 and 0.7, respectively. Rectangular and trellis networks are approximately self-similar although deviations from self-similarity are observed. Rectangular networks have more sinuous channels than dendritic networks across all scales, and trellis networks have a slower rate of area accumulation than dendritic networks across all scales. Such observations are used to build and test classification trees, which are found to perform well in classifying networks.

Description

Covers not scanned.
Print version deaccessioned 2022.

Rights Access

Subject

Geomorphology
Drainage
Hydrologic models

Citation

Associated Publications