Oral Session 11 | Friday, October 7, 10:20–10:40 | Abstract 474

ANN-based Atmospheric Correction of Satellites Images for Inland Waters

The atmospheric effects are generally significant and complex, which make atmospheric corrections (ACs) difficult to accurately derive the remote-sensing reflectance. With a revisit cycle of 16-days, free-of-charge, and a high resolution of 30 meters, Landsat 8 OLI imagery is widely utilized for water quality monitoring in inland waters. In this study, an atmospheric correction method, which is based on artificial neural networks (ANN) with the inputs of top-of-atmosphere reflectance, geometric angles, and aerosol optical thickness, is proposed to retrieve remote sensing reflectance of inland waters on Landsat-8 imagery. To acquire sufficient quantity of samples for training the proposed model, the synthetic remote sensing reflectance was firstly generated by means of an existing AC model. The proposed AC model was validated with in-situ remote sensing reflectance measured in the field campaigns, and the validated model was further tested by using in situ data in other lakes for feasible and effective assessments. The proposed model was compared with land- and water-based AC models. The results demonstrated that the proposed ANN model with the performance of RMSE=0.004, Bias=0.0005, and MAPE=4.19 outperforms the compared AC models. In addition, the results further reveal that the proposed model can avoid producing negative remote sensing reflectance, showing its effectiveness for atmospheric correction over inland waters. 

*Van Manh Nguyen, National Cheng-Kung University, and Vietnam Academy of Science and Technology, 0000-0001-6758-0626

Thi Oanh La, National Cheng-Kung University

Thi Thu Ha Nguyen, Vietnam National University, University of Science

Chao-Hung Lin, National Cheng-Kung University

Dewinta Heriza, National Cheng-Kung University

Quang Vinh Pham, Vietnam Academy of Science and Technology

Keep up to date

Sign up to receive email updates to be sure to catch all the meeting news.


Contact Jenny Ramarui,
Conference Coordinator,
at [email protected]
or (1) 301-251-7708

Translate »