A significant portion of plastic floating on the ocean surface originates from land, often transported by rivers. A large share of this debris leaks from dumpsites located near rivers, especially during storms or similar events, and eventually reaches the ocean. In this paper, Machine Learning (ML) methods are used to track waste in dumpsites near waterways and classify floating debris accumulations on the ocean surface using data from spatio-temporal fusion of Sentinel-2 and Planet Scope. The study involves two case studies: the first focuses on monitoring dumpsite evolution using both ML standard pixel-based classification approaches and Deep Learning (DL) methods, while the second addresses the classification of floating matter on the ocean surface using a pre-trained algorithm on the Sentinel-2 data and adapted to Sentinel-2/Planet fused image. The goal of this study is to improve the temporal accuracy of debris detection and contribute to environmental conservation through the integration of AI and satellite data fusion techniques.

Coastal erosion and sea-level rise are included in the seven climate-change related processes listed by the United Nations that have a negative impact on Cultural Heritage (CH). Coastal archaeological sites are threatened by coastal retreat, accelerated erosion and anthropogenic use. Left unchecked, the increasing sea levels, intensified erosion, and more frequent and severe mega storms pose a significant threat to many of the world's important archaeological sites, leading to their potential destruction.This paper proposes a methodology for coastline extraction in Landsat 5 and Sentinel-2 images, utilizing Artificial Intelligence (AI) models and image processing techniques. Focusing on shoreline alterations near Mykonos Town, Greece, our approach demonstrates satisfactory coastline extraction, with observed erosion within acceptable margins. Leveraging deep learning networks, particularly U-Net CNN, alongside Super Resolution (SR) techniques, enhances accuracy in coastline detection. Additionally, Mean Absolute Difference (MAD) values serve as metrics for evaluating coastline accuracy. The MAD value for the whole area was about 1.4 pixels i.e. 14m. This study highlights the potential of free of charge high resolution satellite images and RS methods to provide valuable insights into the dynamic nature of coastal environments and safeguarding coastal heritage.

Snow, a crucial component of the cryosphere, significantly impacts global climate monitoring and has an important role in freshwater supply, hydropower energy and tourism. Advances in remote sensing technologies, particularly Interferometric Synthetic Aperture Radar (InSAR), enhance our understanding of snow processes. This study investigates the potential use of interferometric observables (phase, coherence, and phase closure) from a ground-based L-band SAR sensor to analyze snow dynamics. Time series data expanding one winter and one summer season are assessed with in-situ snow depth, snow water equivalent and air temperature observations. We found that by exploiting all three interferometric observables the identification of the start and the end of each snow season (wet snow, dry snow, no snow) can be performed more accurately. This study highlights the potential of L-band interferometric observables that are relevant for future L-band SAR missions.

The THETIDA project addresses climate change threats, utilizing a holistic approach to safeguard Europe’s coastal and underwater cultural heritage. Employing environmental modeling for specific climate change scenarios, the project conducts quantitative and qualitative impact assessments, combining environmental and pollution analyses at seven pilot sites across Europe. The developed and deployed technology includes sensors, satellite image processing, smart buoys, AUVs, wearables and crowdsourcing applications supported by ocean forecasting and hazard mapping services. The obtained datasets inform adaptive strategies through a Decision Support System with enhanced visualization capabilities and aim to support sustainable plans for cultural heritage preservation and resilient, climate-neutral policies.

This study presents a dynamic phenology stage estimation methodology for cotton towards early warning and mitigation advice against natural disasters. First, a time-series comparison algorithm, based on Earth Observation (EO) data, is used to assign pseudo-labels to approximately 1,000 parcels. For this, we employ only a limited number of ground truth samples. The pseudo-labels are then used to train Random Forest (RF) regression models for phenology stage estimation. The pseudo-labeling process is used to augment the annotated dataset and allow for modelling the growth of cotton. The models are applied and evaluated on two different test sites in Greece; for which field campaigns were carried out to collect the labels. The results are satisfactory and showcase the successful generalization of the models to other areas. The dynamic predictions for cotton growth and extreme weather events, from numerical weather prediction (NWP) models, are invaluable information for decision-making relevant to agricultural insurance schemes and farm management.

Sentinel-2 (S2) satellite images present directional reflectance effects caused by the configuration of the 12 detectors of the Multispectral Instrument (MSI). These effects are more distinct over water bodies due to water optical properties and they are often described by the bidirectional reflectance distribution function (BRDF). Kernel based models that simulate the BRDF can lead to their elimination. In this study, the c-factor approach along with the Ross-Li BRDF model are employed in order to calculate the Nadir BRDF Adjusted Reflectance (NBAR) for all the pixels of S2 images. NBAR allows prediction of reflectance for nadir observations for all the detectors of the MSI. Results prove the effectiveness of the methodology over ocean scenes.

A new unsupervised classification approach for polarimetric synthetic aperture radar (POLSAR) data is proposed in this paper. The Wishart-Chernoff distance is calculated and used in an agglomerative hierarchical clustering approach. Initial segmentation of POLSAR data into clusters is obtained based on the total backscattering power (SPAN) combined with the entropy, alpha angle, and anisotropy. The complex Wishart clustering is performed to optimize the initialization. Optimized clusters with minimum Wishart-Chernoff distance are merged hierarchically into an appropriate number of classes. The appropriate number of classes is estimated based on the data log-likelihood algorithm. Classification results show that the use of Wishart-Chernoff distance is superior to that of the Wishart test statistic distance. The effectiveness of the proposed Wishart-Chernoff distance is demonstrated using Advanced Land Observing Satellite POLSAR data.

The majority of the existing unmixing methods use a unique set of endmembers for spectral mixture analysis of the entire image, failing to account that each pixel may be comprised of a different combination of endmembers. Multiple endmember spectral mixture analysis (MESMA) allows the number and types of endmembers to vary on a per-pixel basis. The existing MESMA algorithms have high computational cost. In this paper, a novel MESMA is introduced, called MESMA-SAD, which aims to minimize the time-processing by combining the Spectral Angle Distance (SAD) values and the mean absolute errors (MAE). In order to evaluate the proposed method, an AVIRIS hyperspectral data has been used.

This paper presents a new methodology for automatically labeling extracted endmembers from hyperspectral images, using a reference Spectral Library (SL). The SL is structured with three main layers: main consisting material, semantic meaning, and illumination conditions in order to classify the endmembers. The labeling process is based on two similarity metrics: the Spectral Angle Mapper and the Cross Correlation. These two metrics are initially introduced in spectral inequalities which characterize each endmember of the reference SL. During the labeling procedure, inequalities should be respected for labeling the endmembers, otherwise the values of metrics are used. Updating of the SL refers to incorporation of the not labeled endmembers in the SL. Photo-interpretation methodology and SAM calculation for checking the shifted values due to illumination variations are required. The proposed methodology is evaluated using CASI-550 radiance images. The methodology produced satisfactory results.

This paper presents a new nonlinear unmixing method. Based on relative distances which imply nonlinearity, the method introduces the “fractional distance” as a key variable that quantifies interactions between pixels and endmembers. Relationships between fractional distances and abundance fractions are built through networks. Because an equal spectral mixture of ground spectral classes present on the surface sensed is likely impossible, the proposed method, due to its mathematical concept, reveals unknown endmembers. Three versions of the method have been developed: the nonconstrained, the sum-to-one, and the fully constrained versions. Evaluation of the method using synthetic and real data showed that the method is robust with clear and interpretable results and provides reliable abundance fractions, particularly the sum-to-one and the fully constrained versions of the method. The new unmixing method has also been compared with the fully constrained least squares method.

Image segmentation is a basic step of any segment-based classification method. Various segmentation approaches of polarimetric SAR data, such as region growing and splitmerge to name a few, have been proposed recently. This paper describes the development of a new segmentation approach that improves the polarimetric SAR data analysis by including information from the backscattering behavior of objects in the Freeman-Durden analysis images. This method is based on the main scattering mechanism that appears in each image pixel and the second most important scattering mechanism that might have been contributed significantly in the scattering process. Further segmentation is performed based on the calculated histograms of sub-regions. The state-of-art ALOS polarimetric SAR data are used in this study. The study area is located in the south of the United Kingdom and includes the city of Minehead.

This paper develops mathematical techniques and makes use of interferometric synthetic aperture radar (InSAR) for improving the quality of digital elevation models derived from SAR images and providing accurate estimations of atmospheric path and absolute phase delays and water vapor content estimation. The problem of solving parameters such as atmospheric path delay, height, and unwrapping errors is to be described in a mathematical form that uses QR factorization techniques for solving rank-deficient systems. A new approach for the solution of rank-deficient systems with few independent equations is proposed. In the new approach, the bounds are considered known and provided by the eigenvalues. New eigenvalues are added inside the bounds. During the implementation, attention is given not to exceed the bounds of the reorganized matrix. This approach is the first time that is used for phase parameterization in terms of height, tropospheric delay, and phase unwrapping errors. Expanding the approach was also studied for absolute phase delay and water vapor content estimation. The investigation and the implementation of the approach make use of ENVISAT images. The validation of results was implemented through GPS and meteorological measurements.

In the current survey, the performance of the shift-invariant discrete wavelet transform and dual-tree complex wavelet transform (DT-CWT) for Quickbird image fusion is investigated. For this purpose, a DT-CWT fusion algorithm is developed and implemented on high-resolution multispectral and panchromatic Quickbird images of Heraclion, Crete, Greece. In order to point out the effectiveness of the aforementioned transforms, the resulting imagery is visually (through photointerpretation) and computationally (through index computations) compared to fusion products derived by other commonly used methods, such as the intensity hue saturation transform (IHS), the discrete wavelet transform, and the crossbred wavelet and IHS transform. The DT-CWT has been proved to provide a complete and effective tool for Quickbird image fusion