Search

Organometallic complexes are ubiquitous in numerous technological applications, and in particular in homogeneous catalysis. Optimization of such complexes for specific applications is challenging due to the large variety of possible metal-ligand combinations and ligand-ligand interactions. Here we present OM-Diff, an inverse-design framework based on a diffusion generative model for in silico design of such complexes. Due to the importance of the spatial structure of a catalyst, the model operates on all-atom (including H) representations in 3D space. To handle the symmetries inherent to that data representation, OM-Diff combines an equivariant diffusion model with an equivariant property predictor. The diffusion model generates ligands conditioned on a specified metal-center, while the property predictor guides the generation towards novel complexes with desired properties. We demonstrate the potential of OM-Diff by designing optimized catalysts for a family of cross-coupling reactions, and validating a selection of novel proposed compounds with DFT calculations.

The processes of the Earth system drive interactions between energy, matter, and life, and a comprehensive understanding of their full evolutionary trajectory is critical for sustainable human development. Traditional modeling primarily relies on a set of theoretical equations to simulate dynamic process such as carbon-nitrogen cycle, solar radiation dynamics, and terrestrial ecosystem dynamics. Despite the extensive modeling experience of Earth scientists, the rapid advancement of Earth observation techniques has led to a significant increase in the volume of databases, with data accumulating daily, or even hourly. This has exacerbated the conflict between the capacity for data collection and utilization for big Earth data. Consequently, there is an urgent need to enhance the intelligent processing and analysis capabilities of big Earth data. At this critical juncture, the emergence of foundational models has revitalized the unique advantages of maximizing information retrieval and deriving insights from big Earth data. However, the mathematical principles underpinning their success are somewhat elusive, raising concerns about trustworthiness due to the lack of a clearly defined internal chain of reasoning and decision-making processes. Therefore, interpretable foundational models are crucial. They enhance our understanding and security of geoscientific applications, breakthrough performance limitations, and improve the controllability of their social impacts.

This thesis investigates the methods used by internship supervisors in social work to teach their students, as well as the emotional challenges these supervisors encounter during the internship semester. The study is based on recorded interviews with supervisors, which were analyzed in conjunction with previous research articles on the subject and two theoretical frameworks: Emotional Labor Theory and Kolb's Experiential Learning Cycle. The findings reveal that supervisors employ a variety of strategies to support student learning, including fostering independent decision-making, effective communication, and leadership skills. This study found that some supervisors don’t have a strategy for supervising their students, and instead tend to wing it. Supervisors also face significant emotional challenges, such as managing their own insecurities and balancing personal and professional boundaries. The results align with previous studies, confirming the complexity of the supervisory role and the need for empathy, patience, and adaptability. However, the study underscores the necessity for further research to deepen understanding and improve support for supervisors in social work education.

This thesis examines the effects of the UGL course (Development of Group and Leader), a popular leadership training program in Sweden. The study analyzes differences in leadership styles, particularly transformational and destructive leadership, between leaders who have completed the UGL course and those who have not. Using a quantitative method, participants self-evaluate their leadership behaviors with the Developmental Leadership Questionnaire (DLQ). The results indicate that UGL participants report an increase in transformational leadership and a decrease in destructive leadership, although these changes were not statistically significant. The study highlights how self-awareness, a major component of UGL, can influence leadership quality. It also points to the need for further research with larger and more diverse sample groups to validate the results and establish statistical significance.

A novel online clustering algorithm is presented where an Evolving Restricted Boltzmann Machine (ERBM) is embedded with a Kohonen Network called ERBM-KNet. The proposed ERBM-KNet efficiently handles streaming data in a single-pass mode using the ERBM, employing a bias-variance strategy for neuron growing and pruning, as well as online clustering based on a cluster update strategy for cluster prediction and cluster center update using KNet. Initially, ERBM evolves its architecture while processing unlabeled image data, effectively disentangling the data distribution in the latent space. Subsequently, the KNet utilizes the feature extracted from ERBM to predict the number of clusters and updates the cluster centers. By overcoming the common challenges associated with clustering algorithms, such as prior initialization of the number of clusters and subpar clustering accuracy, the proposed ERBM-KNet offers significant improvements. Extensive experimental evaluations on four benchmarks and one industry dataset demonstrate the superiority of ERBM-KNet compared to state-of-the-art approaches., Comment: 9 pages, 11 figures, 3 tables

Drumlins are important bedforms of former glaciated landscapes as they demonstrate past ice-flow directions and elucidate processes that operated at the ice/bed interface. Recently mapped drumlins and other streamlined subglacial bedforms in northeast Iceland reveal the flow-sets of cross-cutting palaeo-ice streams that were active within the Iceland Ice Sheet (IIS) during and following the Last Glacial Maximum. Here we study the Bustarfell drumlin field within the Vopnafjörður–Jökuldalsheiði flow-set. The internal architecture of two drumlins was investigated using sedimentological analysis and ground-penetrating radar (GPR, 50 and 100 MHz) to illuminate subglacial processes that contributed to drumlin formation, as well as the history and dynamics of the IIS. On the stoss side of one of the drumlins, two subglacial traction till units were identified, separated by a thick unit of deformed glaciofluvial sand and gravel. The core of glaciofluvial material suggests that the drumlin formed around well-drained patches (sticky spots) in the subglacial bed that retarded the ice flow locally through increased basal drag and encouraged till deposition. Furthermore, our GPR data indicate a combination of erosional and depositional processes. We suggest that the glaciofluvial sediments were deposited as small ice-marginal fans on the Bustarfell plateau, possibly during the Bølling–Allerød interstadial, and that the drumlins were formed around these fans during a subsequent readvance during the Younger Dryas.

This study leverages a graph-based genetic algorithm (GB-GA) for the design ofefficient nitrogen-fixing catalysts as alternatives to the Schrock catalyst, with the aim toimprove the energetics of key reaction steps. Despite the abundance of nitrogen in theatmosphere, it remains largely inaccessible due to its inert nature. The Schrock catalyst,a molybdenum-based complex, offered a breakthrough but its practical application islimited due to low turnover numbers and energetic bottlenecks. The genetic algorithmin our study explores the chemical space for viable modifications of the Schrockcatalyst, evaluating each modified catalyst’s fitness based on reaction energies of keycatalytic steps and synthetic accessibility. Through a series of selection and optimizationprocesses, we obtained fully converged catalytic cycles for 20 molecules at the B3LYPlevel of theory. From these results, we identified three promising molecules, eachdemonstrating unique advantages in different aspects of the catalytic cycle. This studyoffers valuable insights into the potential of generative models for catalyst design. Ourresults can help guide future work on catalyst discovery for the challenging nitrogenfixation process.

Organometallic complexes are ubiquitous in homogenous catalysis, and their optimisation is of particular interest for many technologically relevant reactions. However, due to the large variety of possible metal-ligand and ligand-ligand interactions, finding the best combination of metal and ligands is an immensely challenging task. Here we present an inverse design framework based on a diffusion generative model for in-silico design of such complexes. Given the importance of the spatial structure of a catalyst, the model directly operates on all-atom (including explicit H) representations in 3D space. To handle the symmetries inherent to that data representation, it combines an equivariant diffusion model and an equivariant property predictor to drive sampling at inference time. We illustrate the potential of the proposed framework by optimising catalysts for the SuzukiMiyaura cross-coupling reaction, and validating a selection of novel proposed complexes with DFT.

Drumlins are important bedforms of former glaciated landscapes as they demonstrate past ice-flow directions and elucidate processes that operated at the ice/bed interface. Recently mapped drumlins and other streamlined subglacial bedforms in northeast Iceland reveal the flow-sets of cross-cutting palaeo-ice streams that were active within the Iceland Ice Sheet (IIS) during and following the Last Glacial Maximum. Here we study the Bustarfell drumlin field within the Vopnafjörður–Jökuldalsheiði flow-set. The internal architecture of two drumlins was investigated using sedimentological analysis and ground-penetrating radar (GPR, 50 and 100 MHz) to illuminate subglacial processes that contributed to drumlin formation, as well as the history and dynamics of the IIS. On the stoss side of one of the drumlins, two subglacial traction till units were identified, separated by a thick unit of deformed glaciofluvial sand and gravel. The core of glaciofluvial material suggests that the drumlin formed around well-drained patches (sticky spots) in the subglacial bed that retarded the ice flow locally through increased basal drag and encouraged till deposition. Furthermore, our GPR data indicate a combination of erosional and depositional processes. We suggest that the glaciofluvial sediments were deposited as small ice-marginal fans on the Bustarfell plateau, possibly during the Bølling–Allerød interstadial, and that the drumlins were formed around these fans during a subsequent readvance during the Younger Dryas.

Crevasse-squeeze ridges (CSRs) are landforms that have been unequivocally linked to surge-type glaciers. The formation of CSRs has been discussed since they were first defined in the mid-1980s. Here, we describe geometric CSR networks from the terrestrial glacier forefields of two glaciers in Trygghamna, Western Svalbard. No glacier surges have been observed in Trygghamna; however, the presence of the CSRs signifies past surge activity. Detailed geomorphological maps were constructed, and the spatial context of these landforms described. Cross-sections of several CSRs highlight ridge architecture, structure and relationships to surrounding landforms and sediments. Most CSRs are symmetrical in cross-profile, orientated perpendicular or oblique to the ice-flow direction. Like previous investigations, we observe these ridge networks on top of till and flutes. Additionally, we, for the first time, document CSRs deposited directly on non-glaciogenic subsurfaces, namely, beach gravels and bedrock. Our findings confirm previous CSR formation theories; basal sediments are squeezed into bottom-up crevasses during surges, which are subsequently transported englacially until surge termination and are finally released by melt out from stagnant ice. Consequently, a network of CSRs is the product of a significant reorganisation and down-glacier transport of basal sediment, exemplifying how single surges are agents of glacial sediment redistribution. These formation processes are illustrated in a refined schematic model. The results further contemporary understanding of CSRs in terrestrial surge-type glacier settings and may also apply to landforms and sediments in certain marine settings and palaeoglacial environments.

Both modern and palaeo-ice streams experience shutdown, which has critical implications for their mass balance and influence on relative sea level. Reconstructions of palaeo-ice streams have mainly focused on their phase of active flow and thus less is understood of their shutdown and style of deglaciation. Mapping of streamlined subglacial bedforms (SSBs), including drumlins and mega-scale glacial lineations, in northeast Iceland reveals cross-cutting flow-sets of palaeo-ice streams within the Iceland Ice Sheet (IIS) during and following the last glacial maximum (LGM). Here we map geometrical ridges (linear and reticular) within the Bakkaflói and Þistilfjörður flow-sets and combine the morphological data with sedimentological analyses to increase our understanding of the dynamics of the IIS during deglaciation in northeast Iceland. We interpret the ridges as crevasse-squeeze ridges (CSRs), based on their interconnected network, primary orientation transverse and/or oblique to former ice flow and internal composition of homogenous subglacial till. In both areas, the CSRs are superimposed on the SSBs, indicating that they post-date the SSBs and signify the waning stage of ice streaming associated with the readvance of the IIS during the Younger Dryas period. The preservation of the CSRs suggests ice stagnation following the readvance and ice-stream shutdown. The morphological difference of the CSRs between the flow-sets is taken to indicate different kinematic setting within the ice streams; the linear CSRs in Bakkaflói formed further upstream where extensional forces parallel to ice flow were dominant, whereas the reticular CSRs in Þistilfjörður are more indicative of transverse and longitudinal forces near the terminal zone. Future research reconstructing past ice-sheet behaviour and ice-stream dynamics would benefit from high-resolution bathymetric data from the adjoining shelf as well as enhanced geochronological constraints.

The foundation model has recently garnered significant attention due to its potential to revolutionize the field of visual representation learning in a self-supervised manner. While most foundation models are tailored to effectively process RGB images for various visual tasks, there is a noticeable gap in research focused on spectral data, which offers valuable information for scene understanding, especially in remote sensing (RS) applications. To fill this gap, we created for the first time a universal RS foundation model, named SpectralGPT, which is purpose-built to handle spectral RS images using a novel 3D generative pretrained transformer (GPT). Compared to existing foundation models, SpectralGPT 1) accommodates input images with varying sizes, resolutions, time series, and regions in a progressive training fashion, enabling full utilization of extensive RS big data; 2) leverages 3D token generation for spatial-spectral coupling; 3) captures spectrally sequential patterns via multi-target reconstruction; 4) trains on one million spectral RS images, yielding models with over 600 million parameters. Our evaluation highlights significant performance improvements with pretrained SpectralGPT models, signifying substantial potential in advancing spectral RS big data applications within the field of geoscience across four downstream tasks: single/multi-label scene classification, semantic segmentation, and change detection., Comment: Accepted by IEEE TPAMI

Remote sensing image retrieval (RSIR), aiming at searching for a set of similar items to a given query image, is a very important task in remote sensing applications. Deep hashing learning as the current mainstream method has achieved satisfactory retrieval performance. On one hand, various deep neural networks are used to extract semantic features of remote sensing images. On the other hand, the hashing techniques are subsequently adopted to map the high-dimensional deep features to the low-dimensional binary codes. This kind of method attempts to learn one hash function for both the query and database samples in a symmetric way. However, with the number of database samples increasing, it is typically time-consuming to generate the hash codes of large-scale database images. In this article, we propose a novel deep hashing method, named asymmetric hash code learning (AHCL), for RSIR. The proposed AHCL generates the hash codes of query and database images in an asymmetric way. In more detail, the hash codes of query images are obtained by binarizing the output of the network, while the hash codes of database images are directly learned by solving the designed objective function. In addition, we combine the semantic information of each image and the similarity information of pairs of images as supervised information to train a deep hashing network, which improves the representation ability of deep features and hash codes. The experimental results on three public datasets demonstrate that the proposed method outperforms symmetric methods in terms of retrieval accuracy and efficiency. The source code is available at https://github.com/weiweisong415/Demo_AHCL_for_TGRS2022 .

Carbon-based systems of energy are rapidly unravelling. The imperative of climate emergency is reshaping energy landscapes, in some instances leading to the reappraisal of energy options hitherto sidelined. This paper deals with the emerging energy landscapes of the Netherlands and Iceland through historically informed tales that focus on islands. While vastly different in historical and geographical context and scale, these cases reveal the necessity of geographical nuance facilitated by the ways insular places offer insights into energy imaginaries of the Anthropocene. The former is a historicised narrative about the reinvention of wind energy as natural gas is being ousted. It focuses on the proposed Dogger Bank Power Link Islands, the first of which is scheduled to emerge in the coming years. The latter, also historically informed, identifies the context for current large wind energy proposals in Iceland, and then contrasts these with the authors’ empirical observations from the small peripheral island of Grímsey. There wind energy is also being reinvented for ousting the predominant oil infrastructure on the island. These cases represent experimental opportunities for envisioning Anthropocene futures intended to destabilize imaginaries of growth in ways that open spaces for negotiation and contestation. They problematize dominant narratives that render wind energy development visible and knowable as a necessary intervention. Emergent from this is wind’s decentred energy landscape in the Anthropocene; an epoch where energy is revealed in its importance to our societies, dispelling the human exceptionalism implicit in the nomenclature whilst at the same time showing how our actions come to matter. The collision of the Earth and ourselves under the terms of climate emergency begs the question whether our differences are the only ones that matter? But also, if it matters what we have done, then surely it matters what we have not done and chosen to ignore.

A Bayesian Deep Restricted Boltzmann-Kohonen architecture for data clustering termed as DRBM-ClustNet is proposed. This core-clustering engine consists of a Deep Restricted Boltzmann Machine (DRBM) for processing unlabeled data by creating new features that are uncorrelated and have large variance with each other. Next, the number of clusters are predicted using the Bayesian Information Criterion (BIC), followed by a Kohonen Network-based clustering layer. The processing of unlabeled data is done in three stages for efficient clustering of the non-linearly separable datasets. In the first stage, DRBM performs non-linear feature extraction by capturing the highly complex data representation by projecting the feature vectors of $d$ dimensions into $n$ dimensions. Most clustering algorithms require the number of clusters to be decided a priori, hence here to automate the number of clusters in the second stage we use BIC. In the third stage, the number of clusters derived from BIC forms the input for the Kohonen network, which performs clustering of the feature-extracted data obtained from the DRBM. This method overcomes the general disadvantages of clustering algorithms like the prior specification of the number of clusters, convergence to local optima and poor clustering accuracy on non-linear datasets. In this research we use two synthetic datasets, fifteen benchmark datasets from the UCI Machine Learning repository, and four image datasets to analyze the DRBM-ClustNet. The proposed framework is evaluated based on clustering accuracy and ranked against other state-of-the-art clustering methods. The obtained results demonstrate that the DRBM-ClustNet outperforms state-of-the-art clustering algorithms., Comment: 14 pages, 7 figures

Remote sensing image retrieval (RSIR), aiming at searching for a set of similar items to a given query image, is a very important task in remote sensing applications. Deep hashing learning as the current mainstream method has achieved satisfactory retrieval performance. On one hand, various deep neural networks are used to extract semantic features of remote sensing images. On the other hand, the hashing techniques are subsequently adopted to map the high-dimensional deep features to the low-dimensional binary codes. This kind of methods attempts to learn one hash function for both the query and database samples in a symmetric way. However, with the number of database samples increasing, it is typically time-consuming to generate the hash codes of large-scale database images. In this paper, we propose a novel deep hashing method, named asymmetric hash code learning (AHCL), for RSIR. The proposed AHCL generates the hash codes of query and database images in an asymmetric way. In more detail, the hash codes of query images are obtained by binarizing the output of the network, while the hash codes of database images are directly learned by solving the designed objective function. In addition, we combine the semantic information of each image and the similarity information of pairs of images as supervised information to train a deep hashing network, which improves the representation ability of deep features and hash codes. The experimental results on three public datasets demonstrate that the proposed method outperforms symmetric methods in terms of retrieval accuracy and efficiency. The source code is available at https://github.com/weiweisong415/Demo AHCL for TGRS2022., Comment: 14 pages, 12 figures, and 2 tables

Introduction: Diabetic nephropathy (DN) and hypertensive nephrosclerosis (HN) represent the most common causes of chronic kidney disease (CKD) and many patients progress to -end-stage renal disease. Patients are treated primarily through the management of cardiovas-cular risk factors and hypertension; however patients with HN have a more favorable outcome. A noninvasive clinical approach to separate these two entities, especially in hypertensive patients who also have diabetes, would allow for targeted treatment and more appropriate resource allocation to those patients at the highest risk of CKD progression. Meth-ods: In this preliminary study, high-spatial-resolution matrix-assisted laser desorption/ion-ization (MALDI) mass spectrometry imaging (MSI) was integrated with high-mass accuracy MALDI-FTICR-MS and nLC-ESI-MS/MS analysis in order to detect tissue proteins within kidney biopsies to discriminate cases of DN (n = 9) from cases of HN (n = 9). Results: Differences in the tryptic peptide profiles of the 2 groups could clearly be detected, with these becoming even more evident in the more severe histological classes, even if this was not evident with routine histology. In particular, 4 putative proteins were detected and had a higher signal intensity within regions of DN tissue with extensive sclerosis or fibrosis. Among these, 2 proteins (PGRMC1 and CO3) had a signal intensity that increased at the latter stages of the disease and may be associated with progression. Discussion/Conclusion: This preliminary study represents a valuable starting point for a future study employing a larger cohort of patients to develop sensitive and specific protein biomarkers that could reliably differentiate between diabetic and hypertensive causes of CKD to allow for improved diagnosis, fewer biopsy procedures, and refined treatment approaches for clinicians.

Understanding patterns of colonisation is important for explaining both the distribution of single species and anticipating how ecosystems may respond to global warming. Insular flora may be especially vulnerable because oceans represent severe dispersal barriers. Here we analyse two lake sediment cores from Iceland for ancient sedimentary DNA to infer patterns of colonisation and Holocene vegetation development. Our cores from lakes Torfdalsvatn and Nykurvatn span the last c. 12,000 cal yr BP and c. 8600 cal yr BP, respectively. With near-centennial resolution, we identified a total of 191 plant taxa, with 152 taxa identified in the sedimentary record of Torfdalsvatn and 172 plant taxa in the sedimentary record of Nykurvatn. The terrestrial vegetation at Torfdalsvatn was initially dominated by bryophytes, arctic herbs such as Saxifraga spp. and grasses. Around 10,100 cal yr BP, a massive immigration of new taxa was observed, and shrubs and dwarf shrubs became common whereas aquatic macrophytes became dominant. At Nykurvatn, the dominant taxa were all present in the earliest samples; shrubs and dwarf shrubs were more abundant at this site than at Torfdalsvatn. There was an overall steep increase both in the local accumulated richness and regional species pool until 8000 cal yr BP, by which time ¾ of all taxa identified had arrived. The period 4500-1000 cal yr BP witnessed the appearance of a a small number of bryophytes, graminoids and forbs that were not recorded in earlier samples. The last millennium, after human settlement of the island (Landnám), is characterised by a sudden disappearance of Juniperus communis, but also reappearance of some high arctic forbs and dwarf shrubs. Notable immigration during the Holocene coincides with periods of increased incidence of sea ice, and we hypothesise that this may have acted as a dispersal vector. Thus, although ongoing climate change might provide a suitable habitat in Iceland for a large range of species only found in

Differing environmental conditions can have profound effects on many behaviours in animals, especially where species have large geographic ranges. Seasonal changes or progression through life history stages impose differential constraints, leading to changes in behaviours. Furthermore, species which show flexibility in behaviours, may have a higher capacity to adapt to anthropogenic-induced changes to their environment. The red-throated diver (RTD) is an aquatic bird, that is able to forage in both freshwater and marine environments, though little else is known about its behaviours and its capacity to adapt to different environmental conditions. Here, we use time-depth recorders and saltwater immersion loggers to examine the foraging behaviour of RTDs from three regions across northwest Europe. We found that in the breeding season, birds from two regions (Iceland and Scotland) foraged in the marine environment, while birds from Finland, foraged predominantly in freshwater. Most of the differences in diving characteristics were at least partly explained by differences in foraging habitat. Additionally, while time spent foraging did not change through the breeding season, dives generally became more pelagic and less benthic over the season, suggesting RTDs either switched prey or followed vertical prey movements, rather than increasing foraging effort. There was a preference for foraging in daylight over crepuscular hours, with a stronger effect at two of the three sites. Overall, we provide the first investigation of RTD foraging and diving behaviour from multiple geographic regions and demonstrate variation in foraging strategies in this generalist aquatic predator, most likely due to differences in their local environment.

Bakgrund: Långvarig smärta är smärtproblematik som har funnits hos en person i minst tre månader. Smärtan kan ha olika orsaker och är en subjektiv upplevelse, därmed uttrycks den olika. Syfte: Syftet med studien var att beskriva tonåringars erfarenheter av att leva med långvarig smärta och dess konsekvenser i dagligt liv. Metod: Litteraturstudie genomförd genom analys av åtta valda studier, varav en med mixad metod och resterande med kvalitativ ansats. Resultat: Under analysen identifierades fem kategorier: Att ha svårt att kommunicera sin smärta, Att uppleva smärtans påverkan på dagligt liv, Att vara beroende av andras stöd, Att utveckla copingstrategier och Att känna osäkerhet inför framtiden. Slutsats: Långvarig smärta hos tonåringar påverkar, utöver det fysiska och psykiska välbefinnandet, dagliga livets aktiviteter. Utbildning, applikationer och informativa filmer kan bidra till bättre smärtterminologi som i sin tur kan hjälpa tonåringen beskriva smärtan och förbättra sin egenvård.

Hyperspectral images (HSIs) provide detailed spectral information through hundreds of (narrow) spectral channels (also known as dimensionality or bands), which can be used to accurately classify diverse materials of interest. The increased dimensionality of such data makes it possible to significantly improve data information content but provides a challenge to conventional techniques (the so-called curse of dimensionality) for accurate analysis of HSIs.

Pressurised meltwater has a major impact on ice dynamics, as well as on sedimentary and deformational processes occurring below/in front of glaciers and ice sheets, but its role in glaciotectonic processes is yet to be fully understood. This study explores micro- and macroscale structures developed within décollements in two thrust-block moraines of Late Weichselian age in Melasveit, western Iceland. The aim is to investigate how pressurised subglacial meltwater can aid the dislocation and transport of large, unfrozen and unlithified sediment blocks by glaciers. A detailed model is constructed for the development of the thrust-block moraines and the microscale processes occurring along their detachments during thrusting. The detachments are characterized by relatively thin zones of crosscutting hydrofractures, which reflect fluctuating water pressures during glaciotectonism. Little evidence of shearing is observed along the leading edges of the thrusts in both moraines. This is supported by high water pressures along the detachments and indicates that the thrust blocks were initially decoupled from the underlying deposits. As the thrust moraines evolved, an increased amount of shear occurred in between events of sediment liquefaction, hydrofracturing, and fluid escape. This was followed by progressive locking up of the detachments and eventual cessation in the accretion of the thrust blocks.

Hyperspectral images provide detailed spectral information through hundreds of (narrow) spectral channels (also known as dimensionality or bands) with continuous spectral information that can accurately classify diverse materials of interest. The increased dimensionality of such data makes it possible to significantly improve data information content but provides a challenge to the conventional techniques (the so-called curse of dimensionality) for accurate analysis of hyperspectral images. Feature extraction, as a vibrant field of research in the hyperspectral community, evolved through decades of research to address this issue and extract informative features suitable for data representation and classification. The advances in feature extraction have been inspired by two fields of research, including the popularization of image and signal processing as well as machine (deep) learning, leading to two types of feature extraction approaches named shallow and deep techniques. This article outlines the advances in feature extraction approaches for hyperspectral imagery by providing a technical overview of the state-of-the-art techniques, providing useful entry points for researchers at different levels, including students, researchers, and senior researchers, willing to explore novel investigations on this challenging topic. In more detail, this paper provides a bird's eye view over shallow (both supervised and unsupervised) and deep feature extraction approaches specifically dedicated to the topic of hyperspectral feature extraction and its application on hyperspectral image classification. Additionally, this paper compares 15 advanced techniques with an emphasis on their methodological foundations in terms of classification accuracies. Furthermore, the codes and libraries are shared at https://github.com/BehnoodRasti/HyFTech-Hyperspectral-Shallow-Deep-Feature-Extraction-Toolbox.

Impact craters, as "lunar fossils", are the most dominant lunar surface features and occupy most of the Moon's surface. Their formation and evolution record the history of the Solar System. Sixty years of triumphs in the lunar exploration projects accumulated a large amount of lunar data. Currently, there are 9137 existing recognized craters. However, only 1675 of them have been determined age, which is obviously not satisfactory to reveal the evolution of the Moon. Identifying craters is a challenging task due to their enormous difference in size, large variations in shape and vast presence. Furthermore, estimating the age of craters is extraordinarily difficult due to their complex and different morphologies. Here, in order to effectively identify craters and estimate their age, we convert the crater identification problem into a target detection task and crater age estimation into a taxonomy structure. From an initial small number of available craters, we progressively identify craters and estimate their age from Chang'E data by transfer learning (TL) using deep neural networks. For comprehensive identification of multi-scale craters, a two-stage craters detection approach is developed. Thus 117240 unrecognized lunar craters that range in diameter from 532 km to 1 km are identified. Then, a two-stage classification approach is developed to estimate the age of craters by simultaneously extracting their morphological features and stratigraphic information. The age of 79243 craters larger than 3 km in diameter is estimated. These identified and aged craters throughout the mid and low-latitude regions of the Moon are crucial for reconstructing the dynamic evolution process of the Solar System., Comment: 6 pages, 3 figures

Long Term Evolution (LTE) is the most widespread cellular technology around the globe. An LTE device must communicate with base stations without encryption and authentication before a connection to a network is established and security steps executed. During that period the device is vulnerable to attacks from rogue base stations; that is, transmitters set up by adversaries to imitate legitimate base stations. The received LTE signals have physical layer properties,such as Channel State Information (CSI) and Carrier Frequency Oset (CFO), which can be utilized to distinguish between legitimate transmissions and illegitimate ones. A method to detect intrusions based on CSI using a Kalman lterand a hypothesis test is studied in this thesis and its performance evaluated.Downlink LTE signals are collected from genuine LTE base stations in an2x2 Multiple-Input Multiple-Output (MIMO) system using Universal SoftwareRadio Peripherals (USRPs) and Software Defined Radio (SDR). Measurements performed at dierent frequency bands (i.e., dierent operators) are used to represent signals from a legitimate transmitter and attackers. The performance of the proposed scheme is evaluated by comparing the acceptance rate of symbols from the legitimate transmitter and the attacker. Additionally, the eects of errors due to false alarms and missed detection are explored. The filtering based intrusion detection shows improved performance compared to a non-filtering CSI based approach., Long Term Evolution (LTE) är den mest utbredda teknologin för mobil kom-munikation runt om världen. Innan anslutning till ett nätverk upprättas och följande säkerhetssteg utförs, måste en LTE-enhet kommunicera med basstationen utan kryptering och autentisering. Under den perioden är enheten sårbar för attacker från så kallade falska basstationer, det vill säga sändare som sätts upp med syftet att imitera legitima basstationer. Mottagna LTE-signaler besitter vissa fysikaliska egenskaper relaterade till den trådlösa kanalen och sändarens hårdvara, till exempel Channel State Information (CSI) samt Carrier Frequency Offset (CFO), som kan utnyttjas för att skilja mellan legitima och illegitima meddelanden. En metod för att upptäcka intrång baserat på CSI med hjälp av ett Kalman-filter och ett hypotestest studeras i denna uppsats och dess prestanda utvärderas.Downlink LTE-signaler samlas in från äkta LTE-basstationer i ett 2x2 MultipleInput Multiple-Output (MIMO) system med Universal Software Radio Peripherals (USRPs) och Software Defined Radio (SDR). Mätningar som utförs vid olika frekvensband (dvs olika operatörar) används för att representera signaler från en legitim respektive falsk sändare. Utförandet av den föreslagna metoden utvärderas genom att jämföra acceptansgraden för symboler från den legitima sändaren respektive från den falska basstationen. Dessutom utvärderas effekterna av felaktiga beslut i detektionsmetoden (av typ-1 och typ-2). Den filterbaserade metoden för detektion av falska basstationer visar förbättrad prestanda jämför med en icke-filtrerande CSI-baserad metod.

The interaction between drumlins and overriding glacier ice is not well studied, largely due to the difficulty of identifying and accessing suitable active subglacial environments. The surge‐type glacier Múlajökull, in central Iceland, overlies a known field of actively forming drumlins and therefore provides a rare opportunity to investigate the englacial structures that have developed in association with ice flow over the subglacial drumlins. In this study detailed ground‐penetrating radar surveys are combined with field observations to identify clear sets of up‐glacier and down‐glacier dipping fractures at Múlajökull's margin. These are interpreted as conjugate shear planes or P‐ and R‐type Reidel shears that developed and filled with saturated sediment derived from the glacier bed, during a previous surge. The fracture sets exhibit focused spatial distributions that are influenced by the subglacial topography. In particular, down‐glacier dipping fractures are strongly focused over drumlin stoss slopes. These fractures, although well developed at depth, were mostly unable to transmit basal water and sediment up to the glacier surface during the surge cycle. In contrast, up‐glacier dipping fractures formed over drumlin lee sides and in more gently sloping swales, and more frequently connected to the glacier surface, providing a pathway for the evacuation of basal water and water‐saturated sediment. The study suggests that the subglacial drumlins under Múlajökull's margin have influenced the nature and distribution of englacial fractures, which could potentially contribute to spatial variations in basal water pressure during a surge

The recent, sharp increase in the availability of data captured by different sensors, combined with their considerable heterogeneity, poses a serious challenge for the effective and efficient processing of remotely sensed data. Such an increase in remote sensing and ancillary data sets, however, opens up the possibility of utilizing multimodal data sets in a joint manner to further improve the performance of the processing approaches with respect to applications at hand. Multisource data fusion has, therefore, received enormous attention from researchers worldwide for a wide variety of applications. Moreover, thanks to the revisit capability of several.

The recent, sharp increase in the availability of data captured by different sensors, combined with their considerable heterogeneity, poses a serious challenge for the effective and efficient processing of remotely sensed data. Such an increase in remote sensing and ancillary data sets, however, opens up the possibility of utilizing multimodal data sets in a joint manner to further improve the performance of the processing approaches with respect to applications at hand. Multisource data fusion has, therefore, received enormous attention from researchers worldwide for a wide variety of applications. Moreover, thanks to the revisit capability of several.

Hyperspectral image (HSI) classification has become a hot topic in the field of remote sensing. In general, the complex characteristics of hyperspectral data make the accurate classification of such data challenging for traditional machine learning methods. In addition, hyperspectral imaging often deals with an inherently nonlinear relation between the captured spectral information and the corresponding materials. In recent years, deep learning has been recognized as a powerful feature-extraction tool to effectively address nonlinear problems and widely used in a number of image processing tasks. Motivated by those successful applications, deep learning has also been introduced to classify HSIs and demonstrated good performance. This survey paper presents a systematic review of deep learning-based HSI classification literature and compares several strategies for this topic. Specifically, we first summarize the main challenges of HSI classification which cannot be effectively overcome by traditional machine learning methods, and also introduce the advantages of deep learning to handle these problems. Then, we build a framework that divides the corresponding works into spectral-feature networks, spatial-feature networks, and spectral-spatial-feature networks to systematically review the recent achievements in deep learning-based HSI classification. In addition, considering the fact that available training samples in the remote sensing field are usually very limited and training deep networks require a large number of samples, we include some strategies to improve classification performance, which can provide some guidelines for future studies on this topic. Finally, several representative deep learning-based classification methods are conducted on real HSIs in our experiments.

Hyperspectral image (HSI) classification is a core task in the remote sensing community, and recently, deep learning-based methods have shown their capability of accurate classification of HSIs. Among the deep learning-based methods, deep convolutional neural networks (CNNs) have been widely used for the HSI classification. In order to obtain a good classification performance, substantial efforts are required to design a proper deep learning architecture. Furthermore, the manually designed architecture may not fit a specific data set very well. In this paper, the idea of automatic CNN for the HSI classification is proposed for the first time. First, a number of operations, including convolution, pooling, identity, and batch normalization, are selected. Then, a gradient descent-based search algorithm is used to effectively find the optimal deep architecture that is evaluated on the validation data set. After that, the best CNN architecture is selected as the model for the HSI classification. Specifically, the automatic 1-D Auto-CNN and 3-D Auto-CNN are used as spectral and spectral-spatial HSI classifiers, respectively. Furthermore, the cutout is introduced as a regularization technique for the HSI spectral-spatial classification to further improve the classification accuracy. The experiments on four widely used hyperspectral data sets (i.e., Salinas, Pavia University, Kennedy Space Center, and Indiana Pines) show that the automatically designed data-dependent CNNs obtain competitive classification accuracy compared with the state-of-the-art methods. In addition, the automatic design of the deep learning architecture opens a new window for future research, showing the huge potential of using neural architectures' optimization capabilities for the accurate HSI classification.

This Data in Brief paper contains data (including images) from Quaternary sedimentary successions investigated along the Bol'shaya Balakhnya River and the Luktakh-Upper Taimyra-Logata river system on southern Taimyr Peninsula, NW Siberia (Russia). Marine foraminifera and mollusc fauna composition, extracted from sediment samples, is presented. The chronology (time of deposition) of the sediment successions is reconstructed from three dating methods; (i) radiocarbon dating of organic detritus (from lacustrine/fluvial sediment) and molluscs (marine sediment) as finite ages (usually <42 000 years) or as non-finite ages (>42 000-48 000 years) on samples/sediments beyond the radiocarbon dating limit; (ii) Electron Spin Resonance (ESR) dating on marine molluscs (up to ages >400 000 years); (iii) Optically Stimulated Luminescence (OSL) dating, usually effective up to 100-150 0000 years. Terrestrial Cosmogenic Nuclide (TCN) exposure dating has been applied to boulders resting on top of moraine ridges (Ice Marginal Zones). See (Moller et al., 2019) (doi.org/10.1016/j.earscirev.2019.04.004) for interpretation and discussion of all data. (c) 2019 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Hyperspectral image (HSI) classification has become a hot topic in the field of remote sensing. In general, the complex characteristics of hyperspectral data make the accurate classification of such data challenging for traditional machine learning methods. In addition, hyperspectral imaging often deals with an inherently nonlinear relation between the captured spectral information and the corresponding materials. In recent years, deep learning has been recognized as a powerful feature-extraction tool to effectively address nonlinear problems and widely used in a number of image processing tasks. Motivated by those successful applications, deep learning has also been introduced to classify HSIs and demonstrated good performance. This survey paper presents a systematic review of deep learning-based HSI classification literatures and compares several strategies for this topic. Specifically, we first summarize the main challenges of HSI classification which cannot be effectively overcome by traditional machine learning methods, and also introduce the advantages of deep learning to handle these problems. Then, we build a framework which divides the corresponding works into spectral-feature networks, spatial-feature networks, and spectral-spatial-feature networks to systematically review the recent achievements in deep learning-based HSI classification. In addition, considering the fact that available training samples in the remote sensing field are usually very limited and training deep networks require a large number of samples, we include some strategies to improve classification performance, which can provide some guidelines for future studies on this topic. Finally, several representative deep learning-based classification methods are conducted on real HSIs in our experiments.

Driven by the urgent demand for managing remote sensing big data, large-scale remote sensing image retrieval (RSIR) attracts increasing attention in the remote sensing field. In general, existing retrieval methods can be regarded as visual-based retrieval approaches which search and return a set of similar images from a database to a given query image. Although retrieval methods have achieved great success, there is still a question that needs to be responded to: Can we obtain the accurate semantic labels of the returned similar images to further help analyzing and processing imagery? Inspired by the above question, in this paper, we redefine the image retrieval problem as visual and semantic retrieval of images. Specifically, we propose a novel deep hashing convolutional neural network (DHCNN) to simultaneously retrieve the similar images and classify their semantic labels in a unified framework. In more detail, a convolutional neural network (CNN) is used to extract high-dimensional deep features. Then, a hash layer is perfectly inserted into the network to transfer the deep features into compact hash codes. In addition, a fully connected layer with a softmax function is performed on hash layer to generate class distribution. Finally, a loss function is elaborately designed to simultaneously consider the label loss of each image and similarity loss of pairs of images. Experimental results on two remote sensing datasets demonstrate that the proposed method achieves the state-of-art retrieval and classification performance.

© 2019, The Author(s). The original HTML version of this Article was updated shortly after publication to add links to the Peer Review file. In addition, affiliations 16 and 17 incorrectly read ‘School of Medicine Sydney, University of Notre Dame Australia, Sydney, WA, 6160, Australia’ and ‘St Vincent’s Clinical School, University of New South Wales Medicine, University of New South Wales, Sydney, NSW, 2052, Australia.’ This has now been corrected in both the PDF and HTML versions of the Article.

Patients with chemotherapy-induced bone marrow aplasia often develop severe thrombocytopenia which requires platelet transfusion. However, up to 25% of the patients show a less than adequate response to the transfusion. This phenomenon is called platelet refractoriness and is associated with increased morbidity and mortality as well as longer hospital stays and increased costs.It is well documented in the critical care setting that both a low platelet count (PLC) and a decline in PLC is associated with increased mortality. However, it is unknown whether the results of routine coagulation tests on admission to ICU may predict mortality and morbidity.Study I, performed on hematological patients receiving platelet transfusions, showed that platelet increments, measured as corrected count increments (CCI), declined in a linear manner from hours 1 to 24 post-transfusion. In patients with bone marrow aplasia due to chemotherapy for either acute leukemia or stem cell transplantation, the CCI declined 2.0%  0.6% (mean  95% confidence interval) per hour. In patients who were transfused prior to an intervention, the CCI declined 2.8%  1.2% per hour. Study II, a sub-study of Study I, showed that platelet increments did not correlate with endothelial damage measured with the endothelial cell markers syndecan-1, soluble thrombomodulin and vascular endothelial growth factor. Study III, a retrospective cohort study, showed that prolonged APTT and increased PT-INR on admission to the ICU correlated with increased mortality in patients with severe sepsis or septic shock.Study IV, a retrospective study on an unselected ICU cohort, showed that prolonged APTT on admission to the ICU correlated with increased mortality and the need for vasopressors and renal replacement therapy. Increased PT-INR was, on the other hand, not associated with mortality, but was nevertheless associated with the need for vasopressors and invasive ventilation.The findings in Studies III and IV were adjusted

Lava flows pose a hazard in volcanic environments and reset ecosystem development. A succession of dated lava flows provides the possibility to estimate the direction and rates of ecosystem development and can be used to predict future development. We examine plant succession, soil development and soil carbon (C) accretion on the historical (post 874 AD) lava flows formed by the Hekla volcano in south Iceland. Vegetation and soil measurements were conducted all around the volcano reflecting the diverse vegetation communities on the lavas, climatic conditions around Hekla mountain and various intensities in deposition of loose material. Multivariate analysis was used to identify groups with similar vegetation composition and patterns in the vegetation. The association of vegetation and soil parameters with lava age, mean annual temperature, mean annual precipitation and soil accumulation rate (SAR) was analysed. Soil carbon concentration increased with increasing lava age becoming comparable to concentrations found on the prehistoric lavas. The combination of a sub-Arctic climate, gradual soil thickening due to input of loose material and the specific properties of volcanic soils allow for continuing accumulation of soil carbon in the soil profile. Four successional stages were identified: initial colonization and cover coalescence (ICC) of Racomitrium lanuginosum and Stereocaulon spp. (lavas <70 years of age); secondary colonization (SC) – R. lanuginosum dominance (170−700 years); vascular plant dominance (VPD) (>600 years); and highland conditions/retrogression (H/R) by tephra deposition (70−860 years). The long time span of the SC stage indicates arrested development by the thick R. lanuginosum moss mat. The progression from SC into VPD was linked to age of the lava flows and soil depth, which was significantly deeper within the VPD stage. Birch was growing on lavas over 600 years old indicating the development towards birch woodland, the climax ecosystem in

The empirical line (EL) calibration method is commonly used for atmospheric correction of remotely sensed spectral images and recovery of surface reflectance. The current EL-based methods are applicable to calibrate only single images. Therefore, the use of the EL calibration is impractical for imaging campaigns, where many (partially overlapped) images are acquired to cover a large area. In addition, the EL results are unconstrained and an undesired reflectance with negative values or larger than 100% can be obtained. In this paper, we use the standard EL model to formulate a new generalized empirical line (GEL) model. Based on the GEL, we present a novel method for simultaneous and constrained calibration of multiple images. This new method allows for calibration through multiple image constrained empirical line (MIcEL) and three additional calibration modes. Given a set of images, we use the available ground targets and automatically extracted tie points between overlapping images to calibrate all the images in the set simultaneously. Quantitative and visual assessments of the proposed method were carried out relatively to the off-the-shelf method quick atmospheric correction (QUAC), using real hyperspectral images and field measurements. The results clearly show the superiority of MIcEL with respect to the minimization of the difference between the reflectance values of the same object in different overlapping images. An assessment of the absolute accuracy, with respect to 11 field measurement points, shows that the accuracy of MIcEL, with an average mean absolute error (MAE) of ∼11%, is comparable with respect to the QUAC.

Geomorphological mapping is a well-established method for examining earth surface processes and landscape evolution in a range of environmental contexts. In glacial research, it provides crucial data for a wide range of process-oriented studies and palaeoglaciological reconstructions; in the latter case providing an essential geomorphological framework for establishing glacial chronologies. In recent decades, there have been significant developments in remote sensing and Geographical Information Systems (GIS), with a plethora of high quality remotely-sensed datasets now (often freely) available. Most recently, the emergence of unmanned aerial vehicle (UAV) technology has allowed sub-decimetre scale aerial images and Digital Elevation Models (DEMs) to be obtained. Traditional field mapping methods still have an important role in glacial geomorphology, particularly in cirque glacier, valley glacier and icefield/ice-cap outlet settings. Field mapping is also used in ice sheet settings, but often takes the form of necessarily highly-selective ground-truthing of remote mapping. Given the increasing abundance of datasets and methods available for mapping, effective approaches are necessary to enable assimilation of data and ensure robustness. This paper provides a review and assessment of the various glacial geomorphological methods and datasets currently available, with a focus on their applicability in particular glacial settings. We distinguish two overarching 'work streams' that recognise the different approaches typically used in mapping landforms produced by ice masses of different sizes: (i) mapping of ice sheet geomorphological imprints using a combined remote sensing approach, with some field checking (where feasible); and (ii) mapping of alpine and plateau-style ice mass (cirque glacier, valley glacier, icefield and ice-cap) geomorphological imprints using remote sensing and considerable field mapping. Key challenges to accurate and robust geomorphological mappi

Radiographic contrast agents cause acute kidney injury (AKI), yet the underlying pathogenesis is poorly understood. Nod-like receptor pyrin containing 3-deficient (Nlrp3-deficient) mice displayed reduced epithelial cell injury and inflammation in the kidney in a model of contrast-induced AKI (CI-AKI). Unexpectedly, contrast agents directly induced tubular epithelial cell death in vitro that was not dependent on Nlrp3. Rather, contrast agents activated the canonical Nlrp3 inflammasome in macrophages. Intravital microscopy revealed diatrizoate (DTA) uptake within minutes in perivascular CX3CR1+ resident phagocytes in the kidney. Following rapid filtration into the tubular luminal space, DTA was reabsorbed and concentrated in tubular epithelial cells via the brush border enzyme dipeptidase-1 in volume-depleted but not euvolemic mice. LysM-GFP+ macrophages recruited to the kidney interstitial space ingested contrast material transported from the urine via direct interactions with tubules. CI-AKI was dependent on resident renal phagocytes, IL-1, leukocyte recruitment, and dipeptidase-1. Levels of the inflammasome-related urinary biomarkers IL-18 and caspase-1 were increased immediately following contrast administration in patients undergoing coronary angiography, consistent with the acute renal effects observed in mice. Taken together, these data show that CI-AKI is a multistep process that involves immune surveillance by resident and infiltrating renal phagocytes, Nlrp3-dependent inflammation, and the tubular reabsorption of contrast via dipeptidase-1.Copyright © 2018 American Society for Clinical Investigation. All rights reserved.