A method for extracting water from barrier lake in high mountain areas based on decision tree classification: A case study of Attabad barrier lake on the Karakoram Highway
The water dynamics of barrier lakes in high mountain areas are crucial for risk assessment, disaster prediction, safety management, and decision-making. Objective and Methods To accurately and efficiently extract the water boundaries of mountainous barrier lakes, this paper focuses on the Attabad ba...
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Editorial Department of Bulletin of Geological Science and Technology
2024-11-01
|
| Series: | 地质科技通报 |
| Subjects: | |
| Online Access: | https://dzkjqb.cug.edu.cn/en/article/doi/10.19509/j.cnki.dzkq.tb20240125 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The water dynamics of barrier lakes in high mountain areas are crucial for risk assessment, disaster prediction, safety management, and decision-making. Objective and Methods To accurately and efficiently extract the water boundaries of mountainous barrier lakes, this paper focuses on the Attabad barrier lake along the Karakoram Highway, proposing a water extraction method based on decision tree classification. This method incorporates slope information into six conventional water extraction methods for decision tree classification. The effectiveness of these six methods was compared for extracting water from barrier lakes in the experimental area. The best-performing methods suitable for barrier lakes in high-altitude areas were applied to extract the water body range of the Attabad barrier lake.Accuracy was assessed using a confusion matrix, and classification post-processing was performed to refine the water boundary extraction. Results The research results indicate that (1) among the six models, the CWI model demonstrates the best performance, effectively distinguishing between slope, water, and shadow water, leading to a highly accurate outline of the barrier lake. However, a limitation of this model is the presence of a few mountain shadows in the middle of the slope. (2) The decision tree classification method based on slope achieved an overall accuracy of 89.31% and a kappa coefficient of 0.84. It effectively extracts the actual water range, excluding slope shoreline and mountain shadows, and provides a clearer lake boundary. Nevertheless, black fragments observed in the lower area of the barrier lake, likely due to landslides and mountain shadows, remained challenging to classify. Overall, the decision tree classification-based method proved effective in identifying water bodies, particularly in areas with rugged terrain and numerous shadows. Conclusion This paper proposes a method for extracting water bodies from barrier lakes in high mountain areas using decision tree classification. By incorporating slope information into conventional water body extraction methods, this approach accurately extracts the water boundary, effectively eliminates shadows from steep slopes, retain shadowed water on gentler slopes, and improves extraction efficiently. The simplicity and high extraction efficiency of this method make it a practical solution for widespread application. |
|---|---|
| ISSN: | 2096-8523 |