دورية أكاديمية
A novel reflectance transformation and convolutional neural network framework for generating bathymetric data for long rivers: A case study on the Bei River in South China
| العنوان: | A novel reflectance transformation and convolutional neural network framework for generating bathymetric data for long rivers: A case study on the Bei River in South China |
|---|---|
| المؤلفون: | Ting On Chan, Simin Zhang, Linyuan Xia, Ming Luo, Jinhua Wu, Joseph Awange |
| المصدر: | International Journal of Applied Earth Observations and Geoinformation, Vol 127, Iss , Pp 103682- (2024) |
| بيانات النشر: | Elsevier, 2024. |
| سنة النشر: | 2024 |
| المجموعة: | LCC:Physical geography LCC:Environmental sciences |
| مصطلحات موضوعية: | Satellite derived bathymetry, Reflectance transformation, Convolutional neural network, Spatial-temporal analysis, Physical geography, GB3-5030, Environmental sciences, GE1-350 |
| الوصف: | Satellite-derived bathymetry plays a significant role in characterizing river systems, furnishing invaluable insights for applications such as flood risk management. In this paper, we introduce a bathymetry inversion framework, namely reflectance transformation - convolutional neural network (RT-CNN), geared towards long river segments, exemplified by a 210-kilometer stretch of the Bei River's in South China. The framework hinges on a neural network driven by a reflectance transformation model, an optical construct that harnesses radiance data from visible, near-infrared, and shortwave infrared bands. This modified input undergoes spatial–temporal fusion and convolutional neural network processing, culminating in high-resolution (10-meter) bathymetric estimations. The RT-CNN approach adeptly blends Landsat and Sentinel imagery with a substantial dataset of over 18,000 SONAR measurements, integrating convolutional neural network regressions for bathymetric inference. Comparative analysis against established machine learning methodologies, including random forest, gradient boosting decision tree, support vector machines, and multiple linear regression, underscores the supremacy of the proposed RT-CNN framework. Despite the Bei River's changing hydrodynamic attributes along its length, the RT-CNN delivers accurate bathymetric estimates. Specifically, across 2012, 2016, and 2019, the convolutional neural network consistently outperforms other algorithms, achieving an accuracy range of 0.43 to 1.63 m (with a 18% accuracy improvement compared to the existing methods) within an approximate elevation difference of 60 m from upstream to downstream. Temporal trend analyses indicate that heightened precipitation volumes and vegetation loss correlate with downstream shallowing in the Bei River during recent years. |
| نوع الوثيقة: | article |
| وصف الملف: | electronic resource |
| اللغة: | English |
| تدمد: | 1569-8432 |
| Relation: | http://www.sciencedirect.com/science/article/pii/S1569843224000360; https://doaj.org/toc/1569-8432 |
| DOI: | 10.1016/j.jag.2024.103682 |
| URL الوصول: | https://doaj.org/article/2ecceb2df1ba40bdbc3168e30e9c7995 |
| رقم الأكسشن: | edsdoj.2ecceb2df1ba40bdbc3168e30e9c7995 |
| قاعدة البيانات: | Directory of Open Access Journals |
Full Text Finder | تدمد: | 15698432 |
|---|---|
| DOI: | 10.1016/j.jag.2024.103682 |