149544 research outputs found
Sort by
Automatic grounding line delineation of DInSAR interferograms using deep learning
No full textThe regular and robust mapping of grounding lines is essential for various applications related to the mass balance of marine ice sheets and glaciers in Antarctica and Greenland. Differential Interferometric Synthetic Aperture Radar (DInSAR) enables precise detection of tide-induced ice shelf flexure at a continent-wide scale with temporal resolutions of just a few days. While automated pipelines for generating differential interferograms are well established, grounding line delineation remains largely a manual process, which is labor-intensive and increasingly impractical given the growing data streams from current and upcoming synthetic aperture radar (SAR) missions. To address this limitation, we developed an automated pipeline employing the holistically nested edge detection (HED) neural network to delineate grounding lines from DInSAR interferograms. The network was trained in a supervised manner using 421 manually annotated grounding lines of outlet glaciers and ice shelves of the Antarctic Ice Sheet. We also evaluated the utility of non-interferometric features such as surface elevation, ice velocity, and differential tide levels for enhancing delineation performance. Our recommended neural network, trained on the real and imaginary interferometric features, achieved a median offset of 265 m and a mean offset of 421 m from manual grounding line delineations, as well as a predictive uncertainty of 401 m. Furthermore, we demonstrated this network's capacity to generalize by generating grounding lines for previously undelineated interferograms, highlighting its potential for large-scale, high-resolution spatiotemporal mappings
Hierarchical Modeling and Architecture Optimization: Review and Unified Framework
No full textSimulation-based problems involving mixed-variable inputs frequently feature domains that are hierarchical, conditional, heterogeneous, or tree-structured. These characteristics pose challenges for data representation, modeling, and optimization. This paper reviews extensive literature on these structured input spaces and proposes a unified framework that generalizes existing approaches. In this framework, input variables may be continuous, integer, or categorical. A variable is described as meta if its value governs the presence of other decreed variables, enabling the modeling of conditional and hierarchical structures.
We further introduce the concept of partially-decreed variables, whose activation depends on contextual conditions. To capture these inter-variable hierarchical relationships, we introduce design space graphs, combining principles from feature modeling and graph theory. This allows the definition of general hierarchical domains suitable for describing complex system architectures. The framework supports the use of surrogate models over such domains and integrates hierarchical kernels and distances for efficient modeling and optimization. The proposed methods are implemented in the open-source Surrogate Modeling Toolbox (SMT 2.0), and their capabilities are demonstrated through applications in Bayesian optimization for complex system design, including a case study in green aircraft architecture
Almost 2500 days in space – the RAMIS radiation detector on the DLR Eu:CROPIS mission
No full textRAMIS (size: 140x140x35mm³, mass: 608 gram, power consumption 1.8 Watt) is a radiation detector developed for the DLR Eu:CROPIS satellite mission. RAMIS uses an arrangement of two silicon detectors in telescope geometry and maps the radiation environment in the course of the mission. For this RAMIS stores count- and dose rate data with a 1-minute and energy deposition spectra with a 5-minute cadence. Eu:CROPIS was launched on December 03rd 2018 into a polar orbit circling around Earth at an average altitude of ~600 km. RAMIS is located on the outside of the satellite and was activated on December 05th 2018 and has continuously provided data during the course of the mission. Due to the polar orbit of the satellite RAMIS measures a) the variation of galactic cosmic radiation (GCR) in dependence on the orbit, b) the contributions of protons in the inner Earth radiation (Van Allen), c) variations of the trapped electron intensity during crossings of the outer radiation belt and finally (d) the changes in the radiation environment due to the changes in the solar cycle as seen in a high number of detected Solar Particle Events (SPE). Due to the high complexity of the radiation environment and the various particle contributions a separation of GCR, SAA, Electrons and SPEs becomes a challenge. We tackled this problem by using for the first time a machine learning approach classifying data points based on a random forest algorithm. For this a trainings data set was defined and used as input for the radiation environment classification. This enabled the correct attribution/mapping of the various particle populations. First results applying this machine learning tool will be presented focusing on SPEs as measured with RAMIS in the course of the mission with a focus on the year 2024
Seagrass Mapping in Cyprus Using Earth Observation Advances
Get PDFSeagrass meadows are vital for biodiversity and provide a plethora of ecosystem services, but significant losses due to human activity and climate change have been observed in recent decades. This study aims to evaluate whether the integration of Sentinel-2 composites, cloud computing (Google Earth Engine, GEE), and machine learning (ML) classifiers can produce accurate, scalable maps of seagrass habitats, enabling reliable estimates of associated carbon stocks. In this case study, we developed a methodological workflow for local-scale seagrass mapping in Cyprus, covering a total area of 310 km2. ML techniques, specifically the Random Forest (RF) classifier and Classification And Regression Tree (CART), were employed in the main processing stage. The RF classifier achieved an overall accuracy of 73.5%, with a seagrass-specific F1-score of 69.4%. Class-specific F1-scores ranged from 63.2% for hard bottoms to 98.2% for deep water, accounting for variability in habitat separability. The workflow is designed to be scalable across Cyprus and potentially the broader EMMENA region (Eastern Mediterranean, Middle East, and North Africa). Based on the mapped extent of Posidonia oceanica meadows, preliminary estimates suggest a carbon stock of approximately 19,000 Mg C in Cyprus
Reduced infection risks: A tailored ventilation concept for passenger trains decreasing the dispersion of exhaled aerosol particles
No full textWe report on an experimental study of state-of-the-art “classical” ventilation systems and a novel headrest-integrated ventilation system for passenger compartments. The main objective of the novel ventilation system is to reduce the spread of exhaled particles, often referred to simply as aerosols, in densely occupied spaces such as passenger train compartments or aircraft cabins.
The concepts were evaluated in terms of many different parameters such as contaminant removal efficiency, the number of seats above a certain threshold or the mean particle concentration in the breathing zone of the passengers. It was found that the different state-of-the-art concepts are of similar quality in terms of aerosol dispersion. While some concepts are marginally better for single parameter, they are slightly worse in others. Overall, the variance of the evaluation parameter between the different state-of-the-art concepts is rather small. A generic prototype was designed, 3D-printed and tested for different operating configurations in a generic train laboratory. Regarding the new concept, the results demonstrated the positive effect of the active system in reducing the average particle load throughout the compartment as well as the peak concentration on the neighbouring seat. The best results were found for the combined blowing-suction configuration at high flow rates, reducing the peak load by more than 40% and the mean load by approx. 50%, compared to the standard ceiling-based ventilation case without in-seat ventilation system
Fokker-Planck Simulations of the SHEFEX~II Vehicle
No full textThe numerical prediction of aerodynamic and aerothermal loads is an essential part in the design and the optimisation of modern launch and re-entry vehicles. One of the goals of DLR is to be capable to predict the aerodynamic and aerothermal loads for the entire trajectory of these vehicles with high-fidelity simulation tools. The DLR in-house CFD solver TAU already provides very good capabilities for the prediction in the continuum region of flows. To extend our capabilities to the rarefied and transition regions we have started development on the kinetic Fokker-Planck (FP) method.
We will do full 3D simulations of the SHEFEX II vehicle during decent at select trajectory points for altitudes between 70 km and 110 km with both FP and CFD. The mach number ranges from 9.2 to 9.5. We analyse and compare the flow fields of the simulations. We calculate the aerodynamic coefficients and the surface heat fluxes and compare them to the measured data from the flight experiment. We investigate the differences and provide a first estimate for the applicability and accuracy of our simulation methods in this flow regime
Pods4Rail - D4.1 Description of Use Cases
Get PDFThe aim of the Deliverable 4.1 Description of use cases is to identify and describe potential use cases (UCs) for Pod systems. The definition of the UCs considers technological feasibility, environmental impact, economic viability as well as user and society centred design and operation. The main advantage of Pod systems is the user friendly transport of people and goods. Hence, the significance of UCs lies in their ability to pinpoint user requirements, emphasizing the essential fulfilment of system functionalities aligned with user needs.
The approach of this task is to use diverging and converging methods to iterate from a wide spectrum of possible UCs to a more specific description of the most valuable UCs. As diverging method morphological charts were used to create a list of various UCs that were further detailed by mobility journeys. Through classifying and prioritising several parameters, each UC could be characterized. This converging method was used to manage the large number of UCs. Additionally, an analysis of synergies between identified UCs and derivation of technical parameters for transport unit (TU) and carrier was carried out.
Task 4.1 shows that mainly three kinds of transport solutions have to be considered for Pod systems. Passenger transport UCs such as premium and individual transport services are highly relevant, since they can offer new services and better comfort for several user groups. For freight transport and combined (passenger and freight) transport, Pod systems offer new solutions for more effective, flexible freight transport and also for specific event-driven UCs. The identified UCs represent a range of possible and feasible UCs, even though further UCs are can still be realised with such a flexible system.
Based on the technical overview of Pod systems conducted for D2.2, D4.1 determines potential UCs and their specific characteristics, which gives valuable input for the subsequent Task 4.2 (SWOT analysis), Task 4.4 (Functional Requirement Specification) as well as several follow-up work packages (WPs) such as WP5 (Business Case Development), WP7 (Pod Technical Concept), WP8 (Design Variants) or WP11 (traffic coordination)
Validation of the Fokker-Planck chemistry implementation with the RFZ-ST2 upper stage
Get PDFWe test the chemistry implementation of our kinetic Fokker-Planck (FP) method for an industrial scale space transport problem. We use the FP method as it is able to simulate flows in the rarefied regime above the continuum limit, as found at high altitudes. The FP method provides this physical modelling advantage of a particle method with the computational efficiency needed for the simulations of industrial scale applications. We compare FP simulations of the generic RFZ-ST2 upper stage to computational fluid dynamics (CFD) Navier-Stokes simulations with the DLR TAU code. We use an artificial flight Mach number of M a = 15 to facilitate chemical reactions. We limit ourselves to dissociation and exchange reactions. We compare the flow fields and the surface distributions to validate the implementation. This is done as a step towards the simulation of re-usable launch vehicles at high altitudes in re-entry and re-entry burn situations