Search

Introduction: This research explores the application of generative artificial intelligence, specifically the novel ARISGAN framework, for generating highresolution synthetic satellite imagery in the challenging arctic environment. Realistic and high-resolution surface imagery in the Arctic is crucial for applications ranging from satellite retrieval systems to the wellbeing and safety of Inuit populations relying on detailed surface observations. Methods: The ARISGAN framework was designed by combining dense block, multireceptive field, and Pix2Pix architecture. This innovative combination aims to address the need for high-quality imagery and improve upon existing state-of-the-art models. Various tasks and metrics were employed to evaluate the performance of ARISGAN, with particular attention to land-based and sea icebased imagery. Results: The results demonstrate that the ARISGAN framework surpasses existing state-of-the-art models across diverse tasks and metrics. Specifically, land-based imagery super-resolution exhibits superior metrics compared to sea ice-based imagery when evaluated across multiple models. These findings confirm the ARISGAN framework's effectiveness in generating perceptually valid highresolution arctic surface imagery. Discussion: This study contributes to the advancement of Earth Observation in polar regions by introducing a framework that combines advanced image processing techniques with a well-designed architecture. The ARISGAN framework's ability to outperform existing models underscores its potential. Identified limitations include challenges in temporal synchronicity, multispectral image analysis, preprocessing, and quality metrics. The discussion also highlights potential avenues for future research, encouraging further refinement of the ARISGAN framework to enhance the quality and availability of high-resolution satellite imagery in the Arctic. [ABSTRACT FROM AUTHOR]

The European Space Agency Sentinel-1 satellites provide good resolution all weather SAR images. We describe algorithms for detection and classification of ships, icebergs and other objects at sea. Sidelobes from strongly reflecting objects as large ships are suppressed for better determination of ship parameters. The resulting improved ship lengths and breadths are larger than the ground truth values known from Automatic Identification System (AIS) data due to the limited resolution in the processing of the SAR images as compared to previous analyses of Sentinel-2 optical images. The limited resolution in SAR imagery degrades spatial classification algorithms but it is found that the backscatter horizontal and vertical polarizations can be exploited to distinguish icebergs in the Arctic from large ships but not small boats or wakes. [ABSTRACT FROM AUTHOR]

Snow, ice, and melt ponds cover the surface of the Arctic Ocean in fractions that change throughout the seasons. These surfaces control albedo and exert tremendous influence over the energy balance in the Arctic. Increasingly available meter- to decimeter-scale resolution optical imagery captures the evolution of the ice and ocean surface state visually, but methods for quantifying coverage of key surface types from raw imagery are not yet well established. Here we present an open-source system designed to provide a standardized, automated, and reproducible technique for processing optical imagery of sea ice. The method classifies surface coverage into three main categories: snow and bare ice, melt ponds and submerged ice, and open water. The method is demonstrated on imagery from four sensor platforms and on imagery spanning from spring thaw to fall freeze-up. Tests show the classification accuracy of this method typically exceeds 96%. To facilitate scientific use, we evaluate the minimum observation area required for reporting a representative sample of surface coverage. We provide an open-source distribution of this algorithm and associated training datasets and suggest the community consider this a step towards standardizing optical sea ice imagery processing. We hope to encourage future collaborative efforts to improve the code base and to analyze large datasets of optical sea ice imagery. [ABSTRACT FROM AUTHOR]

Despite accounting for a small fraction of the surface area in the Arctic, long and narrow sea ice fractures, known as "leads", play a critical role in the energy flux between the ocean and atmosphere. As the volume of sea ice in the Arctic has declined over the past few decades, it is increasingly important to monitor the corresponding changes in sea ice leads. A novel approach has been developed using artificial intelligence (AI) to detect sea ice leads using satellite thermal infrared window data from the Moderate Resolution Imaging Spectroradiometer (MODIS) and the Visible Infrared Imaging Radiometer Suite (VIIRS). In this new approach, a particular type of convolutional neural network, a U-Net, replaces a series of conventional image processing tests from our legacy algorithm. Results show the new approach has a high detection accuracy with F1 Scores on the order of 0.7. Compared to the legacy algorithm, the new algorithm shows improvement, with more true positives, fewer false positives, fewer false negatives, and better agreement between satellite instruments. [ABSTRACT FROM AUTHOR]

Ice Pathfinder (Code: BNU-1), launched on 12 September 2019, is the first Chinese polar observation microsatellite. Its main payload is a wide-view camera with a ground resolution of 74 m at the subsatellite point and a scanning width of 744 km. BNU-1 takes into account the balance between spatial resolution and revisit frequency, providing observations with finer spatial resolution than Terra/Aqua MODIS data and more frequent revisits than Landsat-8 OLI and Sentinel-2 MSI. It is a valuable supplement for polar observations. Geolocation is an essential step in satellite image processing. This study aims to geolocate BNU-1 images; this includes two steps. For the first step, a geometric calibration model is applied to transform the image coordinates to geographic coordinates. The images calibrated by the geometric model are the Level1A (L1A) product. Due to the inaccuracy of satellite attitude and orbit parameters, the geometric calibration model also exhibits errors, resulting in geolocation errors in the BNU-1 L1A product. Then, a geometric correction method is applied as the second step to find the control points (CPs) extracted from the BNU-1 L1A product and the corresponding MODIS images. These CPs are used to estimate and correct geolocation errors. The BNU-1 L1A product corrected by the geometric correction method is processed to the Level1B (L1B) product. Although the geometric correction method based on CPs has been widely used to correct the geolocation errors of visible remote sensing images, it is difficult to extract enough CPs from polar images due to the high reflectance of snow and ice. In this study, the geometric correction employs an image division and an image enhancement method to extract more CPs from the BNU-1 L1A products. The results indicate that the number of CPs extracted by the division and image enhancements increases by about 30% to 182%. Twenty-eight images of Antarctica and fifteen images of Arctic regions were evaluated to assess the performance of the geometric correction. The average geolocation error was reduced from 10 km to ~300 m. In general, this study presents the geolocation method, which could serve as a reference for the geolocation of other visible remote sensing images for polar observations. [ABSTRACT FROM AUTHOR]

The Northwest Passage (NWP) in the Arctic is usually covered with hazardous multi-year ice (MYI) and seasonal first-year ice (FYI) in winter, with possible thin ice and open-water areas during transition seasons. Ice classification is important for both marine navigation and climate change studies. Satellite-based Synthetic Aperture Radar (SAR) systems have shown advantages of retrieving this information. Operational ice mapping relies on visual analysis of SAR images along with ancillary data. However, these maps estimate ice types and concentrations within large-size polygons of a few tens or hundreds of kilometers, which are subjectively identified and selected by analysts. This study aims at developing an automated algorithm to identify individual MYI floes from SAR images then classify the rest of the image as FYI and other ice types. The algorithm identifies the MYI floes using extended-maximum operator, morphological image processing, and a few geometrical features. Classifying the rest of the image uses texture and neural network model. The input data is a set of Sentinel-1 A/B Extended Wide (EW) mode images, acquired between September and March 2016–2019. Although the overall accuracy (for all type classification) from the new method scored 93.26%, the accuracy from using the texture classifier only was 75.81%. The kappa coefficient from the former was higher than the latter by 0.25. Compared with the operational ice charts from the Canadian Ice Service, ice type maps from the new method show better distribution of MYI at the fine scale of individual floes. Comparison against MYI concentration from two automated algorithms that use a combination of coarse-resolution passive and active microwave data also confirms the advantage of resolving MYI floes from the fine-resolution SAR. [ABSTRACT FROM AUTHOR]

The presence of sea ice leads in the sea ice cover represents a key feature in polar regions by controlling the heat exchange between the relatively warm ocean and cold atmosphere due to increased fluxes of turbulent sensible and latent heat. Sea ice leads contribute to the sea ice production and are sources for the formation of dense water which affects the ocean circulation. Atmospheric and ocean models strongly rely on observational data to describe the respective state of the sea ice since numerical models are not able to produce sea ice leads explicitly. For the Arctic, some lead datasets are available, but for the Antarctic, no such data yet exist. Our study presents a new algorithm with which leads are automatically identified in satellite thermal infrared images. A variety of lead metrics is used to distinguish between true leads and detection artefacts with the use of fuzzy logic. We evaluate the outputs and provide pixel-wise uncertainties. Our data yield daily sea ice lead maps at a resolution of 1 km2 for the winter months November– April 2002/03–2018/19 (Arctic) and April–September 2003–2019 (Antarctic), respectively. The long-term average of the lead frequency distributions show distinct features related to bathymetric structures in both hemispheres. Dataset: 10.1594/PANGAEA.917588 Dataset License: CC-BY [ABSTRACT FROM AUTHOR]