Gravitational waves from black-hole merging events have revealed a population of extra-galactic BHs residing in short-period binaries with masses that are higher than expected based on most stellar evolution models - and also higher than known stellar-origin black holes in our Galaxy. It has been proposed that those high-mass BHs are the remnants of massive metal-poor stars. Gaia astrometry is expected to uncover many Galactic wide-binary systems containing dormant BHs, which may not have been detected before. The study of this population will provide new information on the BH-mass distribution in binaries and shed light on their formation mechanisms and progenitors. As part of the validation efforts in preparation for the fourth Gaia data release (DR4), we analysed the preliminary astrometric binary solutions, obtained by the Gaia Non-Single Star pipeline, to verify their significance and to minimise false-detection rates in high-mass-function orbital solutions. The astrometric binary solution of one source, Gaia BH3, implies the presence of a 32.70 \pm 0.82 M\odot BH in a binary system with a period of 11.6 yr. Gaia radial velocities independently validate the astrometric orbit. Broad-band photometric and spectroscopic data show that the visible component is an old, very metal-poor giant of the Galactic halo, at a distance of 590 pc. The BH in the Gaia BH3 system is more massive than any other Galactic stellar-origin BH known thus far. The low metallicity of the star companion supports the scenario that metal-poor massive stars are progenitors of the high-mass BHs detected by gravitational-wave telescopes. The Galactic orbit of the system and its metallicity indicate that it might belong to the Sequoia halo substructure. Alternatively, and more plausibly, it could belong to the ED-2 stream, which likely originated from a globular cluster that had been disrupted by the Milky Way.

Machine Learning (ML) algorithms had successfully contributed in the creation of automated methods of recognizing patterns in high-dimensional data. Remote sensing data covers wide geographical areas and could be used to solve the problem of the demand of various in-situ data. Lithologicall mapping using remotely sensed data is one of the most challenging applications of ML algorithms. In the framework of the “AI for Geoapplications” project , ML and especially Deep Learning (DL) methodologies are investigated for the identification and characterization of the lithology based on remote sensing data in various pilot areas in Greece. In order to train and test the various ML algorithms, a dataset consisting of 30 ROIs selected mainly from low -vegetated areas, that cover 2% of the total area of Greece was created . For each ROI the corresponding shape file with the lithological units the corresponding Sentinel2 (10 bands) and/or Aster (14 bands) images are provided The dataset is being publicly available in the cloud along with the necessary code for visualization and processing.

High-resolution (HR) satellite images can provide detailed information about land usage/land cover. Often, it is necessary that the satellite sensor inherent spatial resolution is increased through algorithmic processing of the image data acquired. Machine-learning and in particular deep-learning based super-resolution (SR) techniques are an effective tool for increasing the spatial resolution of images. In the current work, Sentinel-2 images are super-resolved to spatial resolution equal to 2.5 m/pixel by means of deep-learning based SR techniques. The area of study is Zakynthos island in Greece. A novel index called Normalized Carotenoid Reflectance Index (NCRI) is proposed for the assessment of land cover by olive trees.

In the present work deep-learning based super-resolution (SR) is applied on Sentinel-2 images of the Zakynthos island, Greece, with the intention of detecting stress levels in supercentenarian olive trees due to water deficiency. The aim of this study is monitoring the stress in supercentenarian olive trees over time and over season. Specifically, the Carotenoid Reflectance Index 2 (CRI2) is calculated utilizing the Sentinel-2 bands B2 and B5. CRI2 maps at 10m and at 2.5mspatial resolutions are generated. In fact, the images of band B2 with original spatial resolution 10m are super-resolved to 2.5m. Regarding the images of band B5, these are SR resolved from 20m firstly to 10m and secondly to 2.5m. Deep-learning based SR techniques, namely DSen2 and RakSRGAN, are utilized for enhancing the spatial resolution to 10m and 2.5m. The following five seasons are considered autumn 2019, spring 2019, spring 2020, summer 2019 and summer 2020. In the future, comparisons with field measurements could better assess for the proposed methodology effectiveness regarding the recognition of stress levels in very old olive trees.

In this work, very deep super-resolution (VDSR) method is presented for improving the spatial resolution of remotely sensed (RS) images for scale factor 4. The VDSR net is re-trained with Sentinel-2 images and with drone aero orthophoto images, thus becomes RS-VDSR and Aero-VDSR, respectively. A novel loss function, the Var-norm estimator, is proposed in the regression layer of the convolutional neural network during re-training and prediction. According to numerical and optical comparisons, the proposed nets RS-VDSR and Aero-VDSR can outperform VDSR during prediction with RS images. RS-VDSR outperforms VDSR up to 3.16 dB in terms of PSNR in Sentinel-2 images.

Deep learning techniques are applied so as to increase the spatial resolution of Sentinel2 satellite imagery, depicting the Amynteo lignite mine in Ptolemaida, Greece. Resolution enhancement by factors 2 and 4 as well as by factors 2 and 6 using Very-Deep SuperResolution (VDSR) and DSen2 networks, respectively, provides fairly well results on Amynteo lignite mine images.

Nowadays, governmental programs like ESA’s Copernicus provide freely available data that can be easily utilized for earth observation. In the present work, the problem of detecting agricultural and non-agricultural land cover is addressed. The methodology is based on classification with convolutional neural networks (CNNs) and transfer learning using AlexNet. The study area is located at the Ionian Islands, which include several land cover classes according to Copernicus CORINE Land Cover 2018 (CLC 2018). Furthermore, the dataset consists of natural color images acquired by Sentinel-2A multi-spectral instrument. Experimentation proves that extra addition of training data from foreign grounds, unfamiliar to the Greek data, serves much as a confusing agent regarding network performance.

Artificial Neural Networks (ANN) are mathematical computing paradigms imitating the operations of biological neural systems. Their nonlinear nature and ability to learn from the environment make them highly suited to solve real-world problems from those that are still under development. In the field of Physics there are many problems which cannot be adequately solved with the physics–based methods and the use of ANN may yield better results. In the present work ANNs have been tested in predicting nuclear radii considering as input the atomic and mass numbers, exclusively. The performance of different supervised ANNs is evaluated. The dataset used for the training and testing was based on evaluated data of nuclear radii available in IAEA tables.

A methodology for monitoring and mapping lignite mining areas using Sentinel-1 and Sentinel-2 ESA Copernicus satellite systems is presented. A stochastic regularised super-resolution reconstruction (SRSR) for the enhancement of the Sentinel-2 optical data is developed, and a land monitoring/change analysis based on the enhanced Sentinel-2 images is performed. Additionally, the ground motion is monitored using the Sentinel-1 radar data via the Rheticus service. The proposed methodology is tested on the Amyntaio lignite mine in Ptolemais basin, Greece, for Sentinel images obtained from 2014 to 2018. The Amyntaio area has been of particular interest, as a landslide event occurred on June 10th, 2017, causing major operational disruption and a severe economic loss to the Public Power Plant Cooperation of Greece SA. The methodology proves to be useful for facilitating mapping and monitoring mining and post-mining areas facing similar problems with the Amyntaio lignite site.

This work super-resolves the lowest-resolution 60m/pixel Sentinel-2 B1 and B9 to the highest-resolution 10m/pixel. Two different categories of super-resolution (SR) techniques are utilized, in specific a SR technique which performs information transfer among different bands and the stochastic regularized SR technique Var-norm+BTV. The study area is the Lysimachia Lake, Western Greece. The Sentinel-2 image of 10th November 2018 has been selected to test the different techniques.

We investigate Titan's low-latitude and midlatitude surface using spectro-imaging near-infrared data from Cassini/Visual and Infrared Mapping Spectrometer. We use a radiative transfer code to first evaluate atmospheric contributions and then extract the haze and the surface albedo values of major geomorphological units identified in Cassini Synthetic Aperture Radar data, which exhibit quite similar spectral response to the Visual and Infrared Mapping Spectrometer data. We have identified three main categories of albedo values and spectral shapes, indicating significant differences in the composition among the various areas. We compare with linear mixtures of three components (water ice, tholin-like, and a dark material) at different grain sizes. Our fits of the data are overall successful, except in some cases at 0.94, 2.03, and 2.79 μm, indicative of the limitations of our simplistic compositional model and the need for additional components to reproduce Titan's complex surface. Our results show a latitudinal dependence of Titan's surface composition, with water ice being the major constituent at latitudes beyond 30°N and 30°S, while Titan's equatorial region appears to be dominated partly by a tholin-like or by a very dark unknown material. The albedo differences and similarities among the various geomorphological units give insights on the geological processes affecting Titan's surface and, by implication, its interior. We discuss our results in terms of origin and evolution theories.

High-Resolution (HR) satellite images are a prerequisite in many applications such as astronomy, remote sensing, geoscience and geographical information systems, not only for providing better visualization but also for extracting extra information details. In the present work a comparative study of different single image resolution enhancement techniques is carried out on Sentinel-2 images of bands B2, B3, B4 and B8. The authors describe the stochastic regularized super-resolution (SR) reconstruction technique and compare with others. The techniques under comparison are stochastic regularized SR reconstruction (SRSR), spatial-wavelet SR reconstruction (SWSR) and the conventional interpolation techniques nearest neighbor (NN), bilinear (BL), bicubic (BC) and spline (SP). These techniques are tested against each other in terms of Root Mean Square Error (RMSE), Xydeas and Petrovich (XP), and Correlation Coefficient (CC). Simulated experiments of single image resolution increase take place.

We infer surface properties, such as surface albedo and atmospheric contributions in the form of haze content, of the mid-latitude region of Titan. In previous studies [1;2] we reported results on two areas presenting indications for possible changes in surface albedo with time [2]. We also investigate the endogenic or exogenic processes linked to the formation of the various mid-latitude geomorphological units. These could be aeolian, fluvial, sedimentary, cryovolcanic, lacustrine, and more. Furthermore, deposition of organics through the atmosphere seems to be predominantly present [1]. We now focus on constraining the chemical composition of the various geomorphological units [5;6] by investigating the lower atmosphere of Titan from Visual and Infrared Mapping Spectrometer (VIMS) spectro-imaging data by use of a recently updated radiative transfer code in the near-IR range. For the distinction of geomorphological units we use RADAR/SAR data [4]. We study the units of interest identified in [1;3] and [4]: mountains, plains, labyrinths, dune fields, and possible cryovolcanic and/or evaporitic features (the latter two are albedo features, [4;5]). Our findings indicate that many of the regions from the same geomorphological unit show compositional variations depending on location, while units of significant geomorphological differences seem to consist of very similar material mixtures. Preliminary results on the chemical composition of the regions that have shown temporal changes (i.e. Tui Regio and Sotra Patera; [6]) are also presented. The albedo differences and similarities among the various geomorphological terrains set constraints on the possible geological processes that govern Titan's surface. References: [1] Lopes, R.M.C., et al.: Icarus, 270, 162-182, 2016; [2] Solomonidou, A., et al.: Icarus, 270, 85-99, 2016; [3] Lopes, R.M.C., et al.: Icarus, 205, 540-558, 2010; [4] Malaska, M., et al.: Icarus, 270, 130-161, 2016; [4] Barnes, J., et al.: Pl. Scie., 2:1, 2013; [5] Solomonidou, A., et al.: JGR, 119, 1729-1747, 2014; [6] Schmitt, B., et al.: GhoSST database (ghosst.osug.fr).

We investigate the surface of Titan using spectro-imaging near-infrared data from the Cassini Visual and Infrared Mapping Spectrometer (VIMS). We apply a radiative transfer code to first determine the contributions of atmospheric haze to the Titan spectrum and then derive the surface albedo (Solomonidou et al. 2014; 2016). We focus here on the geological major units identified in Lopes et al. (2010, 2016), Malaska et al. (2016) and Radebaugh et al. (2016) from Synthetic Aperture Radar (SAR), data including mountains, different types of plains, labyrinths, impact craters, dune fields, and alluvial fans. We find that all regions classified as being the same geomorphological unit in SAR exhibit a coherent spectral response after the VIMS data analysis, thus suggesting a good correlation in the classification between SAR and VIMS. The Huygens landing site appears to be compositionally similar to one type of plains unit (variable plains), suggesting similar plain formation mechanisms. We have sub-categorized the VIMS data into three albedo categories (high, medium, low). By matching the extracted albedos with candidate materials for Titan’s surface (GhoSST database), we find that all regions of interest fall into one out of three main types of major candidate constituents: water ice, tholin-like material, or an unknown, very dark material. This suggests that Titan’s surface is possibly dominated by tholin-like material and a very dark unknown (most likely organic) material, and that most of the surface is covered by atmospheric/organic deposits. Water ice is also present at a number of regions as major constituent at latitudes higher than 30N and 30S. The surface albedo differences and similarities among the various geomorphological units constrain the implications for the geological processes that govern Titan’s surface and interior (e.g. aeolian, fluvial, sedimentary, lacustrine, cryovolcanic, tectonic).References: Lopes et al.: Icarus, 205, 540-558, 2010; Lopes et al.: Icarus, 270, 162-182, 2016; Malaska et al.: Icarus, 270, 130-161, 2016; [4] Solomonidou et al.: JGR, 119, 1729-1747, 2014; [6] Solomonidou et al.: Icarus, 270, 85-99, 2016; [7] Schmitt et al.: GhoSST database (ghosst.osug.fr).

The detailed three-dimensional modeling of buildings utilizing elevation data, such as those provided by light detection and ranging (LiDAR) airborne scanners, is increasingly demanded today. There are certain application requirements and available datasets to which any research effort has to be adapted. Our dataset includes aerial orthophotos, with a spatial resolution 20 cm, and a digital surface model generated from LiDAR, with a spatial resolution 1 m and an elevation resolution 20 cm, from an area of Athens, Greece. The aerial images are fused with LiDAR, and we classify these data with a multilayer feedforward neural network for building block extraction. The innovation of our approach lies in the preprocessing step in which the original LiDAR data are super-resolution (SR) reconstructed by means of a stochastic regularized technique before their fusion with the aerial images takes place. The Lorentzian estimator combined with the bilateral total variation regularization performs the SR reconstruction. We evaluate the performance of our approach against that of fusing unprocessed LiDAR data with aerial images. We present the classified images and the statistical measures confusion matrix, kappa coefficient, and overall accuracy. The results demonstrate that our approach predominates over that of fusing unprocessed LiDAR data with aerial images. (C) 2017 Society of Photo-Optical Instrumentation Engineers (SPIE)

This paper examines the utility of high-resolution airborne RGB orthophotos and LiDAR data for mapping residential land uses within the spatial limits of suburb of Athens, Greece.Modern remote sensors deliver ample information from the AOI (area of interest) for the estimation of 2D indicators or with the inclusion of elevation data 3D indicators for the classification of urban land.In this research, two of these indicators, BCR (building coverage ratio) and FAR (floor area ratio) are automatically evaluated.In the pre-processing step, the low resolution elevation data are fused with the high resolution optical data through a mean-shift based discontinuity preserving smoothing algorithm.The outcome is an nDSM (normalized digital surface model) comprised of upsampled elevation data with considerable improvement regarding region filling and "straightness" of elevation discontinuities.Following this step, a MFNN (multilayer feedforward neural network) is used to classify all pixels of the AOI into building or non-building categories.The information derived from the BCR and FAR building indicators, adapted to landscape characteristics of the test area is used to propose two new indices and an automatic post-classification based on the density of buildings.

Building detection has been a prominent area in the area of image classification. Most of the research effort is adapted to the specific application requirements and available datasets. Our dataset includes aerial orthophotos (with spatial resolution 20cm), a DSM generated from LiDAR (with spatial resolution 1m and elevation resolution 20 cm) and DTM (spatial resolution 2m) from an area of Athens, Greece. Our aim is to classify these data by means of Markov Random Fields (MRFs) in a Bayesian framework for building block extraction and perform a comparative analysis with other supervised classification techniques namely Feed Forward Neural Net (FFNN), Cascade-Correlation Neural Network (CCNN), Learning Vector Quantization (LVQ) and Support Vector Machines (SVM). We evaluated the performance of each method using a subset of the test area. We present the classified images, and statistical measures (confusion matrix, kappa coefficient and overall accuracy). Our results demonstrate that the MRFs and FFNN perform better than the other methods.

We analyze Cassini VIMS data of several areas on Titan's surface looking for variations with time. Three of these locations are near the equator (10-30 degrees S), namely Hotei Regio, Tui Regio and Sotra Patera; in some cases changes in brightness and/or in appearance were reported therein. We also investigate a portion of the undifferentiated plains, areas relatively homogeneous and dark in radar observations, located near 20-25 degrees N. This is a follow-up on a previous paper in which we had inferred surface albedos for some distinct regions of interest (Rols) identified within the Hotei, Tui and Sotra areas through a Principal Component Analysis (PCA) and radiative transfer (RT) modeling (Solomonidou [2014]. J. Geophys. Res. 119, 1729-1747). We apply the same methods here to a larger dataset looking for variations of the surface albedo with time and using the Huygens landing site as the 'ground truth' for calibration purposes. As expected, the undifferentiated plains remain unchanged from January 2010 to June 2012. Our analysis of Hotei Regio data from March 2005 to March 2009 also does not show any significant surface albedo variations within uncertainties. We note however that our RT retrievals are not optimal in this case because of the use of a plane-parallel code and the unfavorable geometry of the associated datasets. Conversely, Tui Regio and Sotra Patera show surface albedo fluctuations with time with pronounced trends for darkening and for brightening respectively. The Tui Regio spectrum exhibits a surface albedo decrease from March 2005 to February 2009, at 0.94, 1.08, 2.03, and 5 mu m wavelengths, while the spectrum shape remains the same over that time. On the contrary, the Sotra Patera area became at least two times brighter within a year (April 2005-February 2006), at 1.58 mu m, 2.03 mu m, and 5 mu m. We also retrieved surface albedo spectra for three reference regions surrounding Hotei, Tui and Sotra and for three additional regions we use as 'test cases' that correspond to dune fields. During the time periods explored here we find that, as expected and contrary to Tui Regio and Sotra Patera, the test cases did not show any significant changes in surface albedo. We therefore suggest that temporal variations of surface albedo exist for some areas on Titan, but that their origin may differ from one region to the other. They could be due to diverse, past and/or ongoing formation processes (endogenic and/or exogenic, possibly cryovolcanic), as discussed here. (C) 2015 Elsevier Inc. All rights reserved.

We investigate the lower atmosphere of Titan from Visual and Infrared Mapping Spectrometer (VIMS) spectro-imaging data by use of a recently updated radiative transfer code in the near-IR range and RADAR/SAR data for the distinction of geomorphological units. We focus here on the geological major units identified in [1;2] and [3]: mountains, plains, labyrinths, dune fields, and possible cryovolcanic and/or evaporitic features (the latter two are albedo features, [4;5;6]). We infer surface properties (like absolute surface albedo and morphology) and atmospheric contributions, in particular the haze content. We find that the Huygens landing site and the candidate evaporitic regions pair compositionally with the variable plains, thus indicating that units of significant geomorphological differences seem to consist of very similar materials. Similarly for the labyrinth terrains and the undifferentiated plains. On the contrary, many regions from the same geomorphological unit show compositional variations depending on location (i.e. undifferentiated plains). These differences provide implications on the endogenic or exogenic origin of the various units. In previous studies we showed that the processes most likely linked to the formation of the various geomorphological units are aeolian, fluvial, sedimentary, and lacustrine, in addition to the deposition of organics through the atmosphere. Currently, we are working on deriving information on the chemical composition of the aforementioned regions from the extracted surface albedos using an extensive library of ices and tholins [e.g. 7]. This will shed light on the potential formation processes (Solomonidou et al. in prep.). Preliminary results on the chemical composition of the regions that have shown temporal changes (i.e. Tui Regio and Sotra Patera; [6]) are also presented.References: [1] Lopes, R.M.C., et al.: Icarus, 205, 540-558, 2010; [2] Lopes, R.M.C., et al.: Icarus, 270, 162-182, 2016; [3] Malaska, M., et al.: Icarus, 270, 130-161, 2016; [4] Barnes, J., et al.: Pl. Scie., 2:1, 2013; [5] Solomonidou, A., et al.: JGR, 119, 1729-1747, 2014; [6] Solomonidou, A., et al.: Icarus, 270, 85-99, 2016; [7] Schmitt, B., et al.: GhoSST database (ghosst.osug.fr).