Identification of radial drainage networks based on topographic and geometric features
The radial drainage network, which has a typical spatial distribution pattern, is a crucial component for reflecting regional geographical landscapes. Identification of radial drainage networks mainly involves the extraction of drainage segments that constitute the entire drainage network, which is...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
De Gruyter
2025-04-01
|
| Series: | Open Geosciences |
| Subjects: | |
| Online Access: | https://doi.org/10.1515/geo-2025-0792 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The radial drainage network, which has a typical spatial distribution pattern, is a crucial component for reflecting regional geographical landscapes. Identification of radial drainage networks mainly involves the extraction of drainage segments that constitute the entire drainage network, which is conducive to enhancing the existing data and obtaining map knowledge. Previous studies on the spatial distribution of drainage networks are primarily focused on pattern identification and generalization for single-connection drainage networks. However, radial drainage networks, which normally contain multiple local drainage networks, have typical regional and combinatorial features. Thus, identifying them using available methods is challenging. In this study, a method for identifying radial drainage networks is proposed by using dual constraints on topography and geometry. First, mountain peaks that satisfy the distribution criteria are identified from the window and contour features. Then, various parameters are designed to distinguish the features of drainage segments and basins, and the components of radial drainage networks are determined. Finally, through topological operation and the distance metric, the boundary segments of the radial drainage networks are divided into different groups, and the corresponding radial drainage networks are identified. The test results demonstrate that compared with manual identification, the proposed method has achieved over 90% identification accuracy, recall rate, and F1 value. |
|---|---|
| ISSN: | 2391-5447 |