Bathymetric mapping using quantitative remote sensing techniques is a crucial research domain for accurately retrieving oceanic depths.This study uses GF5-AHSI hyperspectral remote sensing data to evaluate the accurac...Bathymetric mapping using quantitative remote sensing techniques is a crucial research domain for accurately retrieving oceanic depths.This study uses GF5-AHSI hyperspectral remote sensing data to evaluate the accuracy of three semi-empirical models for shallow water depth retrieval:single-band,multi-band,and band-ratio models.The methodology involved parameter extraction,optimal band selection,and combining bands to create the models.A Pearson correlation analysis was conducted to assess parameter sensitivity,optimizing the models for water depth retrieval.The models’precision was evaluated by comparing their outputs with actual underwater topography measurements from Meizhou Bay,Fujian Province.Error margins in estimated water depths ranged from 10%to 50%across the three models,with accuracy generally improving at greater depths.Among the models,the band-ratio model showed the highest reliability,followed by the multi-band model,and the single-band model was the least reliable.However,in depths greater than 30 m,the single-band model’s error margin could be reduced to within 10%,surpassing the performance of the multi-band and band-ratio models.A spectral reflectance sensitivity test revealed variations in reflectance across different water depths,with a slight increase in the nearinfrared band due to water turbidity.To further improve model accuracy,strategies must be implemented to mitigate the interference of suspended sediments and reduce noise,thereby enhancing the reliability of water depth retrieval.展开更多
基金supported by the National Natural Science Foundation of China(Grant No.42171282)the Open Project of Middle Yarlung Zangbo River Natural Resources Observation and Research Station,China(No.2024YJZKF005)+3 种基金Key Laboratory of Spatial Data Mining&Information Sharing of Ministry of Education(No.2023LSDMIS04)Spatial Information Acquisition and Application Joint Laboratory of Anhui Province(No.2024tlxykjxx002)the Zhejiang Provincial Natural Science Foundation of China(No.LY22D010002)the Tibet Autonomous Region Science and Technology Plan Projects(No.XZ202401JD0024).
文摘Bathymetric mapping using quantitative remote sensing techniques is a crucial research domain for accurately retrieving oceanic depths.This study uses GF5-AHSI hyperspectral remote sensing data to evaluate the accuracy of three semi-empirical models for shallow water depth retrieval:single-band,multi-band,and band-ratio models.The methodology involved parameter extraction,optimal band selection,and combining bands to create the models.A Pearson correlation analysis was conducted to assess parameter sensitivity,optimizing the models for water depth retrieval.The models’precision was evaluated by comparing their outputs with actual underwater topography measurements from Meizhou Bay,Fujian Province.Error margins in estimated water depths ranged from 10%to 50%across the three models,with accuracy generally improving at greater depths.Among the models,the band-ratio model showed the highest reliability,followed by the multi-band model,and the single-band model was the least reliable.However,in depths greater than 30 m,the single-band model’s error margin could be reduced to within 10%,surpassing the performance of the multi-band and band-ratio models.A spectral reflectance sensitivity test revealed variations in reflectance across different water depths,with a slight increase in the nearinfrared band due to water turbidity.To further improve model accuracy,strategies must be implemented to mitigate the interference of suspended sediments and reduce noise,thereby enhancing the reliability of water depth retrieval.
