1. Remote Sensing of River Discharge From Medium‐Resolution Satellite Imagery Based on Deep Learning.
- Author
-
Hao, Zhen, Xiang, Naier, Cai, Xiaobin, Zhong, Ming, Jin, Jin, Du, Yun, and Ling, Feng
- Subjects
STREAM measurements ,REMOTE-sensing images ,REMOTE sensing ,DEEP learning ,BRAIDED rivers ,DROUGHT management - Abstract
Accurate monitoring of river discharge variations is essential for managing floods and droughts and understanding the response of global river systems to climate change. Remote sensing of discharge (RSQ) offers a timely and efficient alternative for widespread monitoring, particularly in ungauged areas. Current methods often struggle with accuracy, especially when estimating the width of narrow rivers from medium‐resolution images. We first observe that, although estimating the width variation of narrow rivers can be challenging from medium‐resolution satellite imagery, river discharge still correlates with river surface color or reflectance. However, existing methods can only correlate river surface reflectance with discharge in gauged rivers. Here, we introduce a novel method employing an advanced Transformer architecture to map river discharge variations directly from time‐series reflectance imagery. Our model, trained on quality‐checked data from 2,036 discharge gauges, outperforms existing methods in discharge estimation accuracy and is less affected by the need for precise river width estimation. The proposed model yields positive Kling‐Gupta Efficiency (KGE) in 68.6% of ungauged rivers, a substantial improvement over the BAM and geoBAM methods, which show positive KGEs in only 28.4% and 33.1% of rivers, respectively. Notably, this performance is achieved despite two‐thirds of the rivers being less than 100 m wide, a range where traditional RSQ methods typically struggle, and the RSQ performance does not show degradation for braided rivers. Our approach suggests a significant shift toward more efficient, extensive, and adaptable space‐based river discharge monitoring. Plain Language Summary: Monitoring river discharge is crucial for managing water‐related disasters like floods and droughts and for understanding how rivers respond to climate change. Traditional methods that use remote sensing to monitor discharge (RSQ) often face challenges, especially with accurately measuring the width of narrow rivers using satellite images. However, we've noticed that the color or reflectance of a river's surface, which can be seen from space, still tells us something about the river's discharge, even if the width is hard to measure accurately. Our study introduces a new method using a Transformer architecture that learns to estimate river discharge directly from how the river looks in time‐series satellite images. We trained our model using data from 2,036 river gauges and found that it performs better than existing methods. It is particularly effective in areas where we don't have direct river discharge measurements, achieving good results in 68.6% of these cases‐a big improvement over the 28.4% and 33.1% success rates of previous methods. This is impressive because two‐thirds of the rivers we studied are less than 100 m wide, where older methods usually fail. Key Points: Deep learning model maps river discharge from satellites in ungauged rivers, improving accuracy for rivers over 30 m wideThe model uses Transformer architecture, correlating river surface reflectance data with discharge when width changes are minimalValidated with positive Kling‐Gupta Efficiency in 68.6% of ungauged rivers, median width 67 m, significantly better than existing Mass Conserved Flow Law Inversion (McFLI) methods [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF