A5044344662- Oceanographic and Atmospheric Processes
- Ocean Waves and Remote Sensing
- Marine and fisheries research
- Marine and coastal ecosystems
- Remote-Sensing Image Classification
- Underwater Acoustics Research
- Climate variability and models
- IoT and Edge/Fog Computing
- Remote Sensing and Land Use
- Reservoir Engineering and Simulation Methods
- Advanced Image Fusion Techniques
- Machine Learning and ELM
- Text and Document Classification Technologies
- Meteorological Phenomena and Simulations
Ocean University of China
2021-2024
The eddy identification is an important part of human cognition the ocean. Significant achievements have been made by using sea level anomaly (SLA) data observed altimeter. However, abundant eddies, which do not cause surface characteristic anomalies, cannot be identified. In this study, subsurface vertical structure-oriented 3-D neural network developed to classify oceanic eddies. This study among first that explores ability deep learning in with structure. First, purified profiles dataset...
Oceanic eddies have a non-negligible impact on ocean energy transfer, nutrient distribution, and biological migration in global oceans. The fine detection of oceanic is significant for the development marine science. Remarkable achievements eddy recognition were achieved by mining satellite altimeter data its derived data. However, due to limited spatial resolution altimeters, it difficult detect submesoscale with radial dimensions less than 10 km. Different from previous works, context edge...
Abstract The forthcoming wave of progress in oceanographic technology is the digital twin ocean, a concept that integrates marine big data and artificial intelligence (AI). This development logical consequence combining science considered superior to previous models, such as transparent smart ocean. Amid swift advancement next-generation information technology, conditions are favorable for developing prototype which will integrate various functionalities—data fusion, situation presentation,...
Predicting the trajectory of ocean eddies can promote understanding transport matter and energy in ocean. However, accurately rapidly predicting poses a significant challenge due to their intricate nonlinear motion within physical environment. Regrettably, existing data-driven methods primarily focus on migration combination models, as well fusion processing diverse observational data oceanic eddies. These ways often overlook crucial aspect modeling underlying mechanism We believe that...
Efficient eddy trajectory prediction driven by multi-information fusion can facilitate the scientific research of oceanography, while complicated dynamics mechanism makes this issue challenging. Benefiting from ocean observing technology, dataset be qualified for data-intensive paradigms. In paper, is used to inspire design idea neural network (termed EddyTPNet) and also transformed into prior knowledge guide learning process. This study among first implement with physics informed network....
The inadequate spatial resolution of altimeter results in low identification efficiency oceanic eddies, especially for small-scale eddies. It is well known that eddies can not only induce sea surface signal but more importantly have typical vertical structure characteristics. However, although the characteristics are usually used statistical analysis, they seldom considered process eddy recognition. This study devoted to identifying from perspective their derived 18-year Argo data. Due...
The capability of continuous eddy tracking has been significantly advanced over the past decades, thanks to availability sea surface height measurements routinely collected by satellite altimeters. However, due complexity kinematic mechanisms and underlying oceanic processes, a medium-range forecast individual trajectories is challenging task. This study develops an algorithm effectively achieve precise prediction propagation with ten-day lead time building physics-embedded temporal...
Oceans at a depth ranging from ~100 to ~1000-m (defined as the intermediate water here), though poorly understood compared sea surface, is critical layer of Earth system where many important oceanographic processes take place. Advances in ocean observation and computer technology have allowed science enter era big data (to be precise, for surface layer, small bottom sits between) greatly promoted our understanding near-surface phenomena. During past few decades, however, also undergoing...
Oceanic mesoscale eddies are relatively small, short-lived circulation patterns that approximately in geostrophic balance. Meanwhile, omnipresent and can be characterized by dynamic sea level anomalies temperature anomalies. This makes the eddy identification mainstream with Sea Level Anomaly (SLA). Unfortunately, nearly 90% of caused oceanic cannot observed due to insufficient resolution satellite altimeters. Combining in-situ Expendable Bathythermograph (XBT) profiles data, surface data...
Abstract. Exceptional precision and excellent resolution reconstruction of sea surface currents are beneficial for exploring complex oceanic dynamic processes. Normally, this required physical inversion models global or regional oceans constructed to reconstruct currents. These based on the analysis geostrophic Ekman derived from satellite observations level wind stress fields. Nevertheless, presence various typical processes in marine environments, such as mesoscale eddies small-scale...
Oceanic eddies are an omnipresent phenomenon of seawater flow and critical in transporting oceanic energy material. Consequently, mastering comprehending the characteristics ocean through detecting recognizing contributes to understanding oceanography. In traditional oceanography, a series methods identify with physical or geometric have been developed. Deep learning frameworks recently applied eddy detection field. this paper, Dual-Pyramid UNet architecture that combines pyramid split...
The classification of Hyperspectral Image (HSI) plays an important role in various fields. To achieve more precise multi-target a short time, method for combining discrete non-local theory with traditional variable fraction Potts models is presented this paper. nonlocal operator makes better use the information certain region centered on that pixel. Meanwhile, adding constraint model can ensure every pixel HSI has only one class. proposed characteristics non-convex, nonlinear, and non-smooth...