Shallow-Water Benthic Identification Using Multispectral Satellite Imagery: Investigation on the Effects of Improving Noise Correction Method and Spectral Cover

Remote Sensing

Masita Dwi Mandini Manessa¹  Ariyo Kanno¹  Masahiko Sekine¹  Eghbert Elvan Ampou²  Nuryani Widagti² 

[1] Graduate School of Science and Engineering, Yamaguchi University, Yamaguchi 755-8611, Japan; E-Mails:;Ministries of Marine Affairs and Fisheries Republic of Indonesia—Institute for Marine Research and Observation, Bali 82251, Indonesia; E-Mails:

关键词: coral reef;    multispectral;    Lyzenga;    WorldView-2;    noise correction;   

DOI  :  10.3390/rs6054454

来源: mdpi

【 摘 要 】

Lyzenga’s method is used widely for radiative transfer analysis because of its simplicity of application to cases of shallow-water coral reef ecosystems with limited information of water properties. WorldView-2 imagery has been used previously to study bottom-type identification in shallow-water coral reef habitats. However, this is the first time WorldView-2 imagery has been applied to bottom-type identification using Lyzenga’s method. This research applied both of Lyzenga’s methods: the original from 1981 and the one from 2006 with improved noise correction that uses the near-infrared (NIR) band. The objectives of this study are to examine whether the utilization of NIR bands in the correction of atmospheric and sea-surface scattering improves the accuracy of bottom classification, and whether increasing the number of visible bands also improves accuracy. Firstly, it has been determined that the improved 2006 correction method, which uses NIR bands, is only more accurate than the original 1981 correction method in the case of three visible bands. When applying six bands, the accuracy of the 1981 correction method is better than that of the 2006 correction method. Secondly, the increased number of visible bands, when applied to Lyzenga’s empirical radiative transfer model, improves the accuracy of bottom classification significantly.

【 授权许可】

CC BY   
© 2014 by the authors; licensee MDPI, Basel, Switzerland