Here, we present a proof-of-concept on remote sensing of ocean plastics using airborne shortwave infrared (SWIR) imagery. We captured red, green, and blue (RGB) and hyperspectral SWIR imagery with equipment mounted on a C-130 aircraft surveying the “Great Pacific Garbage Patch” at a height of 400 m and a speed of 140 knots. We recorded the position, size, color, and type (container, float, ghost net, rope, and unknown) of every plastic piece identified in the RGB mosaics. We then selected the top 30 largest items within each of our plastic type categories (0.6–6.8 m in length) to investigate SWIR spectral information obtained with a SASI-600 imager (950–2450 nm). Our analyses revealed unique SWIR spectral features common to plastics. The SWIR spectra obtained (N = 118 items) were quite similar both in magnitude and shape. Nonetheless, some spectral variability was observed, likely influenced by differences in the object optical properties, the level of water submersion, and an intervening atmosphere. Our simulations confirmed that the ∼1215 and ∼1732 nm absorption features have potential applications in detecting ocean plastics from spectral information. We explored the potential of SWIR remote sensing technology for detecting and quantifying ocean plastics, thus provide relevant information to those developing better monitoring solutions for ocean plastic pollution.

Large quantities of mismanaged plastic waste are polluting and threatening the health of the blue planet. As such, vast amounts of this plastic waste found in the oceans originates from land. It finds its way to the open ocean through rivers, waterways and estuarine systems. Here we present a novel machine learning algorithm based on convolutional neural networks (CNNs) that is capable of detecting and quantifying floating and washed ashore plastic litter. The aquatic plastic litter detection, classification and quantification system (APLASTIC-Q) was developed and trained using very high geo-spatial resolution imagery (∼5 pixels cm⁻¹ = 0.002 m pixel⁻¹) captured from aerial surveys in Cambodia. APLASTIC-Q was made up of two machine learning components (i) plastic litter detector (PLD-CNN) and (ii) plastic litter quantifier (PLQ-CNN). PLD-CNN managed to categorize targets as water, sand, vegetation and plastic litter with an 83% accuracy. It also provided a qualitative count of litter as low or high based on a thresholding approach. PLQ-CNN further distinguished and enumerated the litter items in each of the classes defined as water bottles, Styrofoam, canisters, cartons, bowls, shoes, polystyrene packaging, cups, textile, carry bags small or large. The types and amounts of plastic litter provide benchmark information that is urgently needed for decision-making by policymakers, citizens and other public and private stakeholders. Quasi-quantification was based on automated counts of items present in the imagery with caveats of underlying object in case of aggregated litter. Our scientific evidence-based machine learning algorithm has the prospects of complementing net trawl surveys, field campaigns and clean-up activities for improved quantification of plastic litter. APLASTIC-Q is a smart algorithm that is easy to adapt for fast and automated detection as well as quantification of floating or washed ashore plastic litter from aerial, high-altitude pseudo satellites and space missions.

Marine environments are influenced by a wide diversity of anthropogenic and natural substances and organisms that may have adverse effects on human health and ecosystems. Real-time measurements of pollutants, toxins, and pathogens across a range of spatial scales are required to adequately monitor these hazards, manage the consequences, and to understand the processes governing their magnitude and distribution. Significant technological advancements have been made in recent years for the detection and analysis of such marine hazards. In particular, sensors deployed on a variety of mobile and fixed-point observing platforms provide a valuable means to assess hazards. In this review, we present state-of-the-art of sensor technology for the detection of harmful substances and organisms in the ocean. Sensors are classified by their adaptability to various platforms, addressing large, intermediate, or small areal scales. Current gaps and future demands are identified with an indication of the urgent need for new sensors to detect marine hazards at all scales in autonomous real-time mode. Progress in sensor technology is expected to depend on the development of small-scale sensor technologies with a high sensitivity and specificity towards target analytes or organisms. However, deployable systems must comply with platform requirements as these interconnect the three areal scales. Future developments will include the integration of existing methods into complex and operational sensing systems for a comprehensive strategy for long-term monitoring. The combination of sensor techniques on all scales will remain crucial for the demand of large spatial and temporal coverage.

The Coastal Observing System for Northern and Arctic Seas (COSYNA) was established in order to better understand the complex interdisciplinary processes of northern seas and the Arctic coasts in a changing environment. Particular focus is given to the German Bight in the North Sea as a prime example of a heavily used coastal area, and Svalbard as an example of an Arctic coast that is under strong pressure due to global change.The COSYNA automated observing and modelling system is designed to monitor real-time conditions and provide short-term forecasts, data, and data products to help assess the impact of anthropogenically induced change. Observations are carried out by combining satellite and radar remote sensing with various in situ platforms. Novel sensors, instruments, and algorithms are developed to further improve the understanding of the interdisciplinary interactions between physics, biogeochemistry, and the ecology of coastal seas. New modelling and data assimilation techniques are used to integrate observations and models in a quasi-operational system providing descriptions and forecasts of key hydrographic variables. Data and data products are publicly available free of charge and in real time. They are used by multiple interest groups in science, agencies, politics, industry, and the public.