Detection of marine oil-like features in Sentinel-1 SAR images by supplementary use of deep learning and empirical methods: Performance assessment for the Great Barrier Reef marine park

Automated marine oil spill detection algorithm based on single-image generative adversarial network and YOLO-v8 under small samples

As marine resources and transportation develop, oil spill incidents are increasing, endangering marine ecosystems and human lives. Rapidly and accurately identifying marine oil spill is of utmost importance in protecting marine ecosystems. Marine oil spill detection methods based on deep learning and computer vision have the great potential significantly enhance detection efficiency and accuracy, but their performance is often limited by the scarcity of real oil spill samples, posing a challenging to train a precise detection model. This study introduces a detection method specifically designed for scenarios with limited sample sizes. First, the small sample dataset of marine oil spill taken by Landsat-8 satellite is used as the training set. Then, a single image generative adversarial network (SinGAN) capable of training with a single oil spill image is constructed for expanding the dataset, generating diverse marine oil spill samples with different shapes. Second, a YOLO-v8 model is pretrained via the method of transfer learning and then trained with dataset before and after augmentation separately for real-time and efficient oil spill detection. Experimental results have demonstrated that the YOLO-v8 model, trained on an expanded dataset, exhibits notable enhancements in recall, precision, and average precision, with improvements of 12.3 %, 6.3 %, and 11.3 % respectively, compared to the unexpanded dataset. It reveals that our marine oil spill detection model based on YOLO-v8 exhibits leading or comparable performance in terms of recall, precision, and AP metrics. The data augmentation technique based on SinGAN contributes to the performance of other popular object detection algorithms as well.

Tracking the behavior of an accidental oil spill and its impacts on the marine environment in the Eastern Mediterranean

The eastern Mediterranean region is a vital hub for oil transportation and production because of its strategic location between Europe, Asia, and Africa. But its unique attributes, including narrow shipping routes, heavy marine traffic, and proximity to vulnerable ecosystems, render it particularly susceptible to accidental oil spills. In this research, an oil spill detection model, along with bathymetric and oceanographic parameters, was used to track oil spills that occurred at the Syrian Baniyas Station in the Eastern Mediterranean on August 23, 2021. Furthermore, the study employed a pairwise comparison matrix (PWCM) to assess the relative importance of wind speed and direction, water depth, and sea surface temperature (SST) in the dispersion of oil spills. Analysis of Sentinel-1 data obtained prior to, during, and after the incident revealed the accumulation of oil slicks along the Syrian coast from Baniyas to Latakia for up to twenty days. The spilled oil reached the coast of Cyprus six days after the incident. The study determined that wind speed and direction played a critical role in the dispersion of spilled oil, while water depth and SST were comparatively less significant factors in this process. The overall accuracy (OA) and Kappa coefficient (KC) for land, water, and oil slick classes derived from the random forest (RF) algorithm ranged from 90 % to 98 % and from 0.86 to 0.98, respectively. The spread of oil slicks at the incident location was revealed by the decorrelation stretch and band ratios of Sentinel-2 MultiSpectral Instrument (MSI) data. The accidental oil spill could have negative effects on the organic carbon cycle, chlorophyll production, and ecosystem productivity. It is essential to consider the vulnerability of specific regions in the Eastern Mediterranean to oil spills when developing adaptation strategies.

A Triplet Network Fusing Optical and SAR Images for Colored Steel Building Extraction

The identification of colored steel buildings in images is crucial for managing the construction sector, environmental protection, and sustainable urban development. Current deep learning methods for optical remote sensing images often encounter challenges such as confusion between the roof color or shape of regular buildings and colored steel structures. Additionally, common semantic segmentation networks exhibit poor generalization and inadequate boundary regularization when extracting colored steel buildings. To overcome these limitations, we utilized the metal detection and differentiation capabilities inherent in synthetic aperture radar (SAR) data to develop a network that integrates optical and SAR data. This network, employing a triple-input structure, effectively captures the unique features of colored steel buildings. We designed a multimodal hybrid attention module in the network that discerns the varying importance of each data source depending on the context. Additionally, a boundary refinement (BR) module was introduced to extract the boundaries of the colored steel buildings in a more regular manner, and a deep supervision strategy was implemented to improve the performance of the network in the colored steel building extraction task. A BR module and deep supervision strategy were also implemented to sharpen the extraction of building boundaries, thereby enhancing the network's accuracy and adaptability. The results indicate that, compared to mainstream semantic segmentation, this method effectively enhances the precision of colored steel building detection, achieving an accuracy rate of 83.19%. This improvement marks a significant advancement in monitoring illegal constructions and supporting the sustainable development of the Beijing-Tianjin-Hebei metropolitan region.

Tracking the early signals of crude oil in seawater and plankton after a major oil spill in the Red Sea

Understanding the immediate impacts of oil spills is essential to recognizing their long-term consequences on the marine environment. In this study, we traced the early (within one week) signals of crude oil in seawater and plankton after a major oil spill in October 2019 in the Red Sea. At the time of sampling, the plume had moved eastward, but we detected significant signs of incorporation of oil carbon into the dissolved organic carbon pool, resulting in a 10-20% increase in the ultraviolet (UV) absorption coefficient (a₂₅₄) of chromophoric dissolved organic matter (CDOM), elevated oil fluorescence emissions, and depletion of the carbon isotope composition (δ¹³C) of the seawater. The abundance of the picophytoplankton Synechococcus was not affected, but the proportion of low nucleic acid (LNA) bacteria was significantly higher. Moreover, specific bacterial genera (Alcanivorax, Salinisphaera, and Oleibacter) were enriched in the seawater microbiome. Metagenome-assembled genomes (MAGs) suggested that such bacteria presented pathways for growing on oil hydrocarbons. Traces of polycyclic aromatic hydrocarbons (PAHs) were also detected in zooplankton tissues, revealing the rapid entry of oil pollutants into the pelagic food web. Our study emphasizes the early signs of short-lived spills as an important aspect of the prediction of long-term impacts of marine oil spills.