Abstract
Accurate habitat prediction of Bigeye Tuna, the main fishing target of tuna pelagic fishery, is of great significance to the fishing operation. In response to the fact that most of the current studies use single-source data for habitat prediction, and the association between spatiotemporal features and habitat distribution is not fully explored and that this has limited the further improvement of prediction accuracy, this paper analyzes the spatiotemporal distribution of the characteristics of Bigeye Tuna’s highly migratory nature. Additionally, it puts forward a method of habitat prediction that utilizes heterosource remote-sensing data for the four-dimensional time–space–environment–spectrum (TSES) for deep-level feature extraction. First, a multi-source heterogeneous dataset was constructed by combining the spatiotemporal distribution characteristics of the product-level environmental remote-sensing data and the L1B-level original spectral remote-sensing data, and then a multi-branch, dynamic spatiotemporal feature extraction, Long Short-Term Memory Network (LSTM) time-series model was constructed to extract the characteristics of the heterogeneous data. This model was constructed to fully explore and utilize the multidimensional deep-level TSES distribution features affecting the habitat prediction. Finally, the two types of heterogeneous data were subjected to the weighted average-based decision-level fusion to obtain the final prediction results. The experimental results show that compared with other methods, the proposed method in this paper outperforms traditional machine-learning models and other single-source, data-based time-series models, with R2 reaching 0.96278 and RMSE decreasing to 0.031361 in the validation experiments of these models. In contrast, the method in this paper demonstrates good generalization ability and achieves accurate prediction of future fishery distribution.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.