Quantitative Inversion Method of Surface Suspended Sand Concentration in Yangtze Estuary Based on Selected Hyperspectral Remote Sensing Bands

Kuifeng Luan, Hui Li (), Jie Wang, Chunmei Gao, Yujia Pan, Weidong Zhu, Hang Xu, Zhenge Qiu and Cheng Qiu
Additional contact information
Kuifeng Luan: College of Marine Sciences, Shanghai Ocean University, Shanghai 201306, China
Hui Li: College of Marine Sciences, Shanghai Ocean University, Shanghai 201306, China
Jie Wang: College of Marine Sciences, Shanghai Ocean University, Shanghai 201306, China
Chunmei Gao: College of Marine Ecology and Environment, Shanghai Ocean University, Shanghai 201306, China
Yujia Pan: Shanghai Marine Monitoring and Forecasting Center, Shanghai 200062, China
Weidong Zhu: College of Marine Sciences, Shanghai Ocean University, Shanghai 201306, China
Hang Xu: College of Marine Sciences, Shanghai Ocean University, Shanghai 201306, China
Zhenge Qiu: College of Marine Sciences, Shanghai Ocean University, Shanghai 201306, China
Cheng Qiu: Shanghai Marine Monitoring and Forecasting Center, Shanghai 200062, China

Sustainability, 2022, vol. 14, issue 20, 1-22

Abstract: The distribution of the surface suspended sand concentration (SSSC) in the Yangtze River estuary is extremely complex. Therefore, effective methods are needed to improve the efficiency and accuracy of SSSC inversion. Hyperspectral remote sensing technology provides an effective technical means of accurately monitoring and quantitatively inverting SSSC. In this study, a new framework for the accurate inversion of the SSSC in the Yangtze River estuary using hyperspectral remote sensing is proposed. First, we quantitatively simulated water bodies with different SSSCs using sediment samples from the Yangtze River estuary, and analyzed the spectral characteristics of water bodies with different SSSCs. On this basis, we compared six spectral transformation forms, and selected the first derivative (FD) transformation as the optimal spectral transformation form. Subsequently, we compared two feature band extraction methods: the successive projections algorithm (SPA) and the competitive adaptive reweighted sampling (CARS) method. Then, the partial least squares regression (PLSR) model and back propagation (BP) neural network model were constructed. The BP neural network model was determined as the best inversion model. The new FD-CARS-BP framework was applied to the airborne hyperspectral data of the Yangtze estuary, with R 2 of 0.9203, RPD of 4.5697, RMSE of 0.0339 kg/m 3 , and RMSE% of 8.55%, which are markedly higher than those of other framework combination forms, further verifying the effectiveness of the FD-CARS-BP framework in the quantitative inversion process of SSSC in the Yangtze estuary.

Keywords: surface suspended sand concentration; first derivative; competitive adaptive reweighted sampling; neural network; feature band extraction; hyperspectral remote sensing (search for similar items in EconPapers)
JEL-codes: O13 Q Q0 Q2 Q3 Q5 Q56 (search for similar items in EconPapers)
Date: 2022
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2071-1050/14/20/13076/pdf (application/pdf)
https://www.mdpi.com/2071-1050/14/20/13076/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jsusta:v:14:y:2022:i:20:p:13076-:d:940361

Access Statistics for this article

Sustainability is currently edited by Ms. Alexandra Wu

More articles in Sustainability from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().