Remote Sensing Applications in Ocean Observation (Second Edition)

Abstract

The articles presented in this Special Issue epitomize the convergence of cutting-edge sensor technologies, innovative data processing techniques, and advanced algorithmic approaches in ocean remote sensing. Through studies ranging from sensor calibration and data fusion to the application of deep learning and transformer models, the research showcased here pushes the boundaries of what can be achieved in ocean observation. A recurring theme among these contributions is the importance of integrating data from multiple sources and employing state-of-the-art computational methods. Deep learning and the transformer architecture highlight a paradigm shift in remote sensing data analysis. These advanced techniques help extract complex features from high-dimensional datasets and can process large amounts of data quickly and automatically. Furthermore, research focusing on spatiotemporal dynamics and environmental monitoring highlights the critical role of remote sensing in addressing global challenges. By capturing the dynamic interactions between atmospheric, oceanic, and terrestrial processes, these studies provide important insights into the drivers of climate and environmental change. This information is valuable for developing predictive models and informing policy decisions related to climate change mitigation and adaptation.

1. Introduction

Oceans cover more than 70% of the Earth’s surface and are vital to the continuation of life on Earth [1,2]. It is where life on Earth began. The ocean also provides humans with food, transportation, recreation, minerals, and electricity. Moreover, because of its large specific heat, the ocean stores more energy than the atmosphere and land, so it has a great impact on global weather and even climate [3,4]. However, our knowledge of the ocean is limited. While ships, autonomous vehicles, coasts, and islands provide opportunities to observe, sample, and study the ocean. These methods only allow us to observe a small part of the global ocean and cannot conduct comprehensive observations in a short time. Therefore, a better location to observe the ocean is necessary. Space provides this place. Satellites circling the Earth can survey an entire ocean in a short time. These satellites can observe sea surface phenomena during the day or night, cloud-free or even through clouds, using different remote sensor sensors. Satellite remote sensing has long been a cornerstone of ocean observation, enabling scientists and practitioners to monitor the ocean dynamic processes and environmental changes over vast spatial and temporal scales. In recent years, rapid advances in sensor technologies, data processing algorithms, and computational capabilities have transformed the field, expanding its applications from traditional mapping to intricate environmental and geophysical analyses.

Since the advent of various ocean remote sensing sensors carried by satellites, they have been widely used to observe various phenomena or characteristics of the ocean. Using different remote sensing sensors installed on satellites, the oceans can be observed during the day, in cloudless conditions, and even in all weather conditions [5,6]. For example, ocean color sensors can be used to explore the composition and concentration of various substances in seawater [7,8,9,10]. Thermal infrared sensors can provide insights into the distribution of sea surface temperature and, thus, the movement of water masses [11,12,13]. In addition to observing changes in sea surface water levels [14,15,16,17], active microwaves can also observe changes in sea surface wind fields [18,19,20]. Passive microwaves can be applied to observe sea surface wind, temperature, and salinity [21,22,23].

The first edition of “Remote Sensing Applications in Ocean Observation” received an enthusiastic response, so this second edition was produced. Like the first edition, the second edition also received a wide response. Therefore, the call for the third edition is now underway. This second edition, featuring 16 rigorously peer-reviewed articles, highlights state-of-the-art research in remote sensing, spanning topics such as multi-source data fusion, advanced machine learning applications, and spatiotemporal dynamics in oceanic, atmospheric, and terrestrial systems. The contributions presented here illustrate innovative methodologies and technological breakthroughs and pave the way for future developments in the discipline.

In this preface, an integrated overview of the work included in this Special Issue was provided. For clarity, the articles have been grouped into three thematic sections: (I) Multi-Source Data Fusion, Sensor Calibration, and Information Extraction [Contributions 1–6]; (II) Advanced Algorithms and Machine Learning Applications [Contributions 7–10]; and (III) Spatiotemporal Dynamics and Environmental Monitoring [Contributions 11–16]. In each section, the unique contributions of the individual studies and discussions were highlighted on how they collectively advance our understanding of ocean remote sensing applications in various domains.

2. Multi-Source Data Fusion, Sensor Calibration, and Information Extraction

One of the central challenges in ocean remote sensing is the effective integration of data from diverse sources and sensors. With the increasing availability of high-resolution satellite imagery, hyperspectral data, and in situ measurements, researchers can develop more sophisticated techniques to extract meaningful information from these heterogeneous datasets. Several articles in this Special Issue tackle these challenges head-on, proposing innovative methods that enhance data accuracy and enable improved feature extraction.

For example, Huang et al. (contribution 1) studied shallow water depth, focusing on the complex relationship between water depth and spectral response. They found that careful consideration of the one-to-many relationship can significantly improve the inversion accuracy of shallow water depths. By exploiting high-resolution Sentinel-2 imagery, the authors demonstrate that optimized inversion techniques can produce more reliable bathymetric maps—an advance with important implications for coastal management and marine ecology. Similarly, Zeng et al. (contribution 2) explored the internal wave generation mechanism in the northern South China Sea through numerical simulation, synthetic aperture radar, and field measurements. Their multi-source approach provides a comprehensive understanding of internal wave processes in the region, highlighting the importance of combining observational data with model simulations. The study demonstrates that complex physical phenomena can be elucidated by combining different data sources, providing a solid framework for future studies in ocean dynamics. Accurate retrieval of ocean color and radiometric data is another critical aspect of remote sensing. Li et al. (contribution 3) addressed this challenge in their paper. By focusing on the limitations imposed by high solar zenith angles, the authors propose correction techniques that enhance the reliability of ocean color products. These improvements are critical for research related to biogeochemical cycles and marine resource management. The study by Lipinskaya et al. (contribution 4) expanded the topic of ocean data analysis. Through detailed analysis of ocean color images obtained by satellite, they described the depth and spatial variability of the formation of submesoscale eddies. This work not only deepens our understanding of mesoscale and submesoscale processes but also highlights the potential of remote sensing techniques to reveal fine-scale ocean features that were previously difficult to detect. Complementing these studies are two articles focusing on sensor calibration and optimization of data collection methods. Zheng et al. (contribution 5) proposed a comprehensive method for calibrating the HY-2C microwave radiometer. Their approach, which includes advanced wet tropospheric correction techniques, ensures the reliability of radiometric measurements, a prerequisite for precise atmospheric and oceanic observations. Nekrasov et al. (contribution 6) proposed an improved sampling strategy for scatterometer data, which improved the accuracy of sea breeze measurements. Both studies provided important calibration and processing strategies, thereby improving the overall quality of remote sensing datasets.

Taken together, these articles illustrate how multi-source data fusion and advanced calibration techniques are essential to improving the accuracy and interpretability of remote sensing observations. The integration of diverse data types not only augments the spatial and temporal resolution of the obtained information but also improves the robustness of subsequent analyses. Such progress is critical for applications ranging from coastal management and marine ecology to atmospheric science and climate research.

3. Advanced Algorithms and Machine Learning Applications in Remote Sensing

The rapid growth of computing power and the emergence of machine learning have ushered in a new era of remote sensing research. Advanced algorithms, especially deep learning and transformation models, are revolutionizing the way researchers process and interpret large amounts of data. This Special Issue contains several exemplary studies that utilize these cutting-edge techniques to address long-standing challenges in feature detection, classification, and prediction.

Yang et al. (contribution 7) leveraged transformer architecture to accurately delineate coastal boundaries from satellite imagery. They demonstrated that the transformer model outperforms traditional convolutional neural network approaches in terms of both accuracy and computational efficiency. Their work highlights the transformative potential of attention-based models for extracting complex spatial features and provides a promising new tool for coastal zone management. Chowdhury et al. (contribution 8) proposed an innovative method to successfully detect macroalgal blooms with high precision using deep learning technology to process Sentinel-1 radar images. This research is particularly significant given the ecological and economic impacts of macroalgal blooms on coastal ecosystems. The deep learning framework developed in this study can not only automate the detection process but also provide timely information that is critical for environmental monitoring and response.

Also using deep learning, Shang et al. (contribution 9) combined high-resolution optical imagery with state-of-the-art deep learning algorithms to monitor the spatiotemporal evolution of green tides. Their method demonstrated exceptional accuracy in identifying affected areas, allowing for more effective management of these harmful algal blooms. By harnessing the power of deep learning, the study highlights how modern computational methods can derive actionable insights from complex, high-dimensional datasets. Yu et al. (contribution 10) proposed an iterative algorithm to predict seafloor topography based on gravity anomalies. While this research is not based on deep neural networks, it represents a major algorithmic advance in ocean remote sensing. The iterative algorithm they proposed uses gravity anomaly data to accurately predict seafloor topography. This approach provides an efficient and reliable alternative to traditional inversion techniques and has important applications in marine geology and geophysics.

Collectively, these articles demonstrate that advanced algorithms and machine learning techniques are not only enhancing the precision of remote sensing analyses but also expanding the range of applications. Whether through the adoption of transformer models for feature extraction or deep learning frameworks for environmental monitoring, these studies provide compelling evidence that modern computational methods are redefining the frontiers of remote sensing. As data volumes continue to increase, these innovations are critical to transforming raw sensor data into actionable insights for decision-makers.

4. Spatiotemporal Dynamics and Environmental Monitoring

Understanding the spatial and temporal dynamics of natural systems is crucial to addressing environmental challenges such as climate change, resource management, and disaster response. In this context, remote sensing plays a vital role by providing continuous, large-scale observations to capture the dynamic changes in the atmosphere, ocean, and land surface. The third thematic section of this Special Issue brings together articles that focus on these dynamic processes and their environmental impacts.

Li et al. (contribution 11) presented a concept study that outlines an innovative sensor design capable of simultaneously measuring wind and wave characteristics. The dual-function sensor concept represents a significant advancement in ocean remote sensing, as it promises to enhance our ability to monitor ocean surface conditions with unprecedented accuracy and efficiency. Such measurements are critical for improving weather forecasts, maritime safety, and our understanding of air–sea interactions. Hu et al. (contribution 12) analyzed long-term satellite data and revealed how intraseasonal oscillations regulate ocean circulation patterns in the South China Sea. Their findings provide valuable insights into the complex interplay between atmospheric forcing and ocean dynamics, thereby helping to improve climate modeling and regional oceanography. Mitra et al. (contribution 13) further emphasized the importance of remote sensing for environmental monitoring. This comprehensive work integrated multivariate statistical analysis with satellite-based observations to map air quality variations in the Red Sea region. This study not only identified key pollutants and their temporal trends but also explored the meteorological and anthropogenic factors that affect air quality. The integrated approach adopted in this study provides a demonstrative model for environmental monitoring in regions with complex atmospheric dynamics.

Andreev (contribution 14) shifted the focus to sea level change. By examining high-resolution sea level data, the intraseasonal variability of the Bering Sea Shelf sea level and its impact on downstream regional ocean currents are emphasized. This research is critical for understanding the regional impacts of climate change and improving predictions of coastal processes and ocean circulation patterns. Two other studies explored the dynamics of ocean eddies, a key feature that influences the transport of heat and mass in the marine environment. Shi and Hu (contribution 15) studied the changes in eddies and their potential impacts on regional ocean circulation from a long-term perspective. Similarly, Yuan and Hu (contribution 16) provided insights into the transport mechanism of these eddies.

Together, these articles advance our understanding of mesoscale and submesoscale dynamics, which is crucial for accurately modeling ocean circulation and predicting the impacts of climate change.

5. Conclusions and Future Perspectives

The contributions in this Special Issue advance the current state of remote sensing research and provide a vision for its future development. The innovative methodologies and findings presented here will undoubtedly serve as a foundation for further exploration and application, fostering greater integration between scientific research and real-world problem-solving. As we continue to confront global challenges such as climate change, resource depletion, and environmental degradation, the role of remote sensing will become ever more critical in ensuring sustainable development and informed decision-making.

The insights presented in these 16 articles are expected to stimulate further research, promote international collaboration, and encourage the continued development of new remote sensing methods. It is hoped that the works compiled in this Special Issue will have a lasting impact on the academic community and practitioners working in various fields related to ocean observation.

These studies collectively illustrate how remote sensing data can be harnessed to capture the dynamic behavior of natural systems over multiple scales. By integrating observations from various satellite platforms with advanced analytical techniques, the research presented in this section contributes significantly to our ability to monitor and predict ocean environmental phenomena. Such work is particularly relevant in global climate change, as accurate, timely information on oceanic and atmospheric dynamics is essential for informed policy-making and effective resource management.

Looking ahead, the future of ocean remote sensing lies in the continued integration of emerging technologies such as big data analytics, cloud computing, and edge computing. As sensor networks expand and data volumes increase, the development of scalable, robust algorithms will be imperative for processing and interpreting this wealth of information. Moreover, interdisciplinary collaboration—bringing together expertise from oceanography, computer science, engineering, and environmental studies—will be key to addressing the multifaceted challenges of ocean observation.