EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Deep Learning-Based Spectral Reconstruction Technology for Water Color Remote Sensing and Error Analysis

Rugang Tang, Li He, Biyun Guo, Cuishuo Ye

Remote Sensing · 17.08.2025 · 17 (16) · 2860

DOI: 10.3390/rs17162860 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Remote Sensing, online erschienen am 17.08.2025. Verfasst von Rugang Tang und 3 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 51 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Key R&D projects in Zhejiang Province, National Nature Science Foundation of China, Open Research Fund Program of Key Laboratory for Water Ecology Management and Protection in River Source Areas, Ministry of Water Resources.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Land observation multispectral satellites (e.g., Landsat-8/9 and Sentinel-2) offer high spatial resolution but have limited spectral bands for water color observation and insufficient spectral resolution. This study proposes a spectral reconstruction model based on a residual neural network (Deep Spectral Reconstruction Learning Network, DSR-Net) to provide additional spectral bands support for nearshore water observations. The model is trained on 60 million pairs of quasi-synchronous reflectance data, and achieves stable reconstruction of 15 water color channels of the surface level reflectance for water pixels (ρw) from visible to near-infrared bands, considering sensor noise and atmospheric correction errors. Validation results based on AERONET-OC data show that the root mean square error of reconstructed ρw by DSR-Net ranges from 4.09 to 5.18 × 10−3, representing a reduction of 25% to 43% compared to original atmospheric correction results. The reconstruction accuracy reaches the observation level of the Sentinel-3/OLCI water color sensor and is universally applicable to different water categories, effectively supporting nearshore water color observation tasks such as colored dissolved organic matter inversion and cyanobacteria monitoring. The errors in the multispectral reflectance-based ρw primarily arise from sensor noise and atmospheric correction errors. After DSR-Net reconstruction, approximately 59% of the uncertainty caused by sensor noise and 38% of that caused by atmospheric correction errors are reduced. In summary, the spectral reconstruction products generated by DSR-Net not only significantly enhance the water color observation capabilities of current satellite sensors but also provide critical technical support for marine environmental monitoring and the design of next-generation sensors.

Förderung

  • Key R&D projects in Zhejiang Province
  • National Nature Science Foundation of China
  • Open Research Fund Program of Key Laboratory for Water Ecology Management and Protection in River Source Areas, Ministry of Water Resources
  • Consultation and Evaluation Program of the Department of Chinese Academy of Sciences
  • University Students Science and Technology Innovation Program of Zhejiang Province

Zitieren

APA

Tang, R., He, L., Guo, B., & Ye, C. (2025). Deep Learning-Based Spectral Reconstruction Technology for Water Color Remote Sensing and Error Analysis. Remote Sensing, 17(16), 2860. https://doi.org/10.3390/rs17162860

Vancouver

Tang R, He L, Guo B, Ye C. Deep Learning-Based Spectral Reconstruction Technology for Water Color Remote Sensing and Error Analysis. Remote Sensing. 2025;17(16):2860. doi:10.3390/rs17162860

BibTeX
@article{Tang2025,
  author = {Tang, Rugang and He, Li and Guo, Biyun and Ye, Cuishuo},
  title = {Deep Learning-Based Spectral Reconstruction Technology for Water Color Remote Sensing and Error Analysis},
  journal = {Remote Sensing},
  year = {2025},
  volume = {17},
  number = {16},
  pages = {2860},
  doi = {10.3390/rs17162860}
}
Kostenlos starten →