The Impact of Snow Cover on River Discharge Simulation: Insights from the Barandozchay River Basin
This study presents a comprehensive analysis aimed at predicting the discharge of the Barandozchay River using machine learning algorithms and meteorological data from both satellite and ground sources over the period from 2002 to 2022. The research highlights the significance of incorporating snow...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
University of Birjand
2025-03-01
|
| Series: | Water Harvesting Research |
| Subjects: | |
| Online Access: | https://jwhr.birjand.ac.ir/article_3487_1285f9b7b3581273203d5cb9954047fc.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This study presents a comprehensive analysis aimed at predicting the discharge of the Barandozchay River using machine learning algorithms and meteorological data from both satellite and ground sources over the period from 2002 to 2022. The research highlights the significance of incorporating snow cover data in enhancing predictive accuracy, particularly during the spring and summer seasons. Utilizing Artificial Neural Networks (ANN), Support Vector Machines (SVM), and Random Forest (RF), the study evaluates various parameters affecting river discharge, including temperature, precipitation, and solar radiation. The results indicate that the Random Forest model outperforms the others in accuracy and generalization, while SVM demonstrates improved predictive capabilities with the inclusion of snow cover data. Specifically, the integration of snow cover data significantly enhanced the simulation accuracy of river discharge. The SVM model showed notable improvements in evaluation metrics, with R2 increasing from 0.64 to 0.72, MAE decreasing from 0.4 to 0.61, and RMSE reducing from 0.81 to 0.29 in the test data. Conversely, the RF model experienced an increase in error for the test data, but the correlation coefficient R2 improved from 0.85 to 0.88. The findings underscore the necessity of employing advanced machine learning techniques for water resource management, especially in regions facing water crises due to climate change. |
|---|---|
| ISSN: | 2476-6976 2476-7603 |