17 results
Search Results
Now showing 1 - 10 of 17
- ItemAn AI-based open recommender system for personalized labor market driven education(Amsterdam [u.a.] : Elsevier Science, 2022) Tavakoli, Mohammadreza; Faraji, Abdolali; Vrolijk, Jarno; Molavi, Mohammadreza; Mol, Stefan T.; Kismihók, GáborAttaining those skills that match labor market demand is getting increasingly complicated, not in the last place in engineering education, as prerequisite knowledge, skills, and abilities are evolving dynamically through an uncontrollable and seemingly unpredictable process. Anticipating and addressing such dynamism is a fundamental challenge to twenty-first century education. The burgeoning availability of data, not only on the demand side but also on the supply side (in the form of open educational resources) coupled with smart technologies, may provide a fertile ground for addressing this challenge. In this paper, we propose a novel, Artificial Intelligence (AI) driven approach to the development of an open, personalized, and labor market oriented learning recommender system, called eDoer. We discuss the complete system development cycle starting with a systematic user requirements gathering, and followed by system design, implementation, and validation. Our recommender prototype (1) derives the skill requirements for particular occupations through an analysis of online job vacancy announcements
- ItemMitigating poverty: The patterns of multiple carbon tax and recycling regimes for Peru(Amsterdam [u.a.] : Elsevier Science, 2021) Malerba, Daniele; Gaentzsch, Anja; Ward, HaukeCarbon taxes are an economically effective and efficient policy measure to address climate change mitigation. However, they can have severe adverse distributional effects. Recycling parts of the fiscal revenues to vulnerable, lower income households through cash transfers (social assistance) is an option to also overcome associated political difficulties. This paper simulates the distributional impacts of such a combined policy reform in Peru. In a first step, we assess the distributional impacts of varying carbon tax rates. In a second step, we evaluate different scenarios of recycling revenues through existing or expanded transfer schemes towards vulnerable households. The results indicate that a national carbon tax, without compensation, would increase poverty but have no significant impact on inequality. When tax revenues are recycled through transfer schemes, however, poverty would actually decrease. Depending on the amount to be redistributed and the design of the cash transfer scheme, our simulations show a proportional reduction in the poverty headcount of up to around 17%. In addition, the paper underlines how crucial it is to go beyond aggregate measures of poverty to better identify losers from such reform; and assure that the “leave no one behind” principle of the Sustainable Development Goals (SDGs) is addressed.
- ItemS-wave seismic imaging of near-surface sediments using tailored processing strategies(Amsterdam [u.a.] : Elsevier Science, 2020) Burschil, Thomas; Buness, HermannReflection seismic imaging using horizontally-polarized S-waves (SH) can increase resolution and it could be cost-efficient compared to the common use of P-waves. However, since S-wave application often delivers varying data quality, appropriate processing schemes are required for particular imaging and interpretation purposes. In this paper, we present four tailored processing strategies that are applied to SH-wave data acquired in an overdeepened Quaternary basin in the Alpine foreland, the Tannwald Basin. The applied processing schemes consist of (1) processing using a short automatic gain control window that enhances structural details and highlights small-scale structures, (2) offset restriction indicating that relative small offsets are sufficient for adequate imaging, which offers reduced field operation costs, (3) coherency-enhancement that reveals large-scale structures for interpretation, and (4) adapted amplitude scaling that enables structural comparison of P-wave and S-wave seismic sections. With respect to P-wave data measured on the same profile, we demonstrate the benefits of the S-wave seismic reflection method. P-waves offer robust imaging results, but S-waves double the resolution, better depict shallow reflections, and may image reflectors in areas where the P-wave struggles. At least for the Tannwald Basin, S-wave imaging is also more cost-efficient than P-wave imaging. © 2020 The Authors
- ItemGeophysical investigation of a freshwater lens on the island of Langeoog, Germany – Insights from combined HEM, TEM and MRS data(Amsterdam [u.a.] : Elsevier Science, 2017) Costabel, Stephan; Siemon, Bernhard; Houben, Georg; Günther, ThomasA multi-method geophysical survey, including helicopter-borne electromagnetics (HEM), transient electromagnetics (TEM), and magnetic resonance sounding (MRS), was conducted to investigate a freshwater lens on the North Sea island of Langeoog, Germany. The HEM survey covers the entire island and gives an overview of the extent of three freshwater lenses that reach depths of up to 45 m. Ground-based TEM and MRS were conducted particularly on the managed western lens to verify the HEM results and to complement the lithological information from existing boreholes. The results of HEM and TEM are in good agreement. Salt- and freshwater-bearing sediments can, as expected, clearly be distinguished due to their individual resistivity ranges. In the resistivity data, a large transition zone between fresh- and saltwater with a thickness of up to 20 m is identified, the existence of which is verified by borehole logging and sampling. Regarding lithological characterisation of the subsurface, the MRS method provides more accurate and reliable results than HEM and TEM. Using a lithological index derived from MRS water content and relaxation time, thin aquitard structures as well as fine and coarse sand aquifers can be distinguished. Complementing the existing borehole data with the lithology information estimated from MRS, we generate a map showing the occurrence of aquitard structures, which significantly improves the hydrogeological model of the island. Moreover, we demonstrate that the estimates of groundwater conductivity in the sand aquifers from geophysical data are in agreement with the fluid conductivity measured in the boreholes.
- ItemNon-invasive prospection techniques and direct push sensing as high-resolution validation tools in wetland geoarchaeology – Artificial water supply at a Carolingian canal in South Germany?(Amsterdam [u.a.] : Elsevier Science, 2020) Rabiger-Völlmer, Johannes; Schmidt, Johannes; Linzen, Sven; Schneider, Michael; Werban, Ulrike; Dietrich, Peter; Wilken, Dennis; Wunderlich, Tina; Fediuk, Annika; Berg, Stefanie; Werther, Lukas; Zielhofer, ChristophThe prospection of (geo-)archaeological sites yield important knowledge about the concept and the utilisation of pre-historical and historical infrastructure. The satisfactory conduction of classical prospection methods like archaeological excavations or geoarchaeological vibra-coring might be challenging in the case of large sites or difficult underground conditions. This is particularly problematic in wetlands featuring a high groundwater table and high compaction rates of organic layers. In this study, we provide an alternative and non- to minimal-invasive exploration approach to discover hydro-engineering structures for artificial water supply in the surrounding of a Carolingian summit canal in South Germany. The Early Medieval Fossa Carolina was intended 792/793 CE to bridge the Central European watershed between Rhine-Main and Danube catchments. As the canal was constructed as a summit canal, an artificial water supply at the highest levels seemed very likely or even obligatory. In order to explore these obligatory hydro-engineering features, we use a wide range of on-site and off-site tools in a spatial hierarchical way. Our approach includes the large-scale SQUID magnetic survey and the sighting of historical maps. Furthermore, we integrate high-resolution direct push colour logs, and subsequent vibra-coring for small-scale stratigraphical verification and sedimentological analyses. The SQUID magnetic survey and related depth models discover two pronounced linear anomalies that might represent potential artificial water inlets in the North-Eastern and Northern Sections of the canal. I) In the North-Eastern Section, direct push colour logs, vibra-coring and 14C dating provide no evidence for a Carolingian hydro-engineering feature but reveal a natural lenticular structure of Early Holocene age. II) The linear magnetic anomaly in the Northern Section can be excluded with high probability as a hydro-engineering structure as well. Here, direct push colour logs, vibra-coring, 14C dating and the comparison with a historic map reveal evidence for a historic gravel road. Thus, we have nicely verified the magnetic information but have no prove for an artificial Carolingian water inlet from the Swabian Rezat River that contradicts with assumptions of former studies. © 2020 The Authors
- ItemNon-remote reference noise cancellation - using reference data in the presence of surface-NMR signals(Amsterdam [u.a.] : Elsevier Science, 2020) Müller-Petke, MikeSurface-NMR measurements commonly suffer from low signal-to-noise ratios. In recent years, with the introduction of multi-channel surface-NMR instruments, the technique of remote-reference noise cancellation (RNC) was developed and significantly improved the applicability of surface-NMR. The current formulation of RNC requires a reference loop to be placed a remote distance from the transmitter loop such that no NMR signal is recorded. Reference loops placed at non-remote distances have been envisaged to provide both improved noise cancellation performance and field efficiency; however, the concept has not been previously applied because the theoretical framework was missing. In this paper, the theoretical framework is presented. It is demonstrated that reference loops placed at non-remote distances provide superior noise cancellation performance. Considerations for placing the reference loop relative to the transmitter loop are provided, and the theoretical framework is evaluated based on a semi-synthetic example using real field noise and synthetic surface-NMR data. © 2020
- ItemEvaluation of spectral induced polarization field measurements in time and frequency domain(Amsterdam [u.a.] : Elsevier Science, 2020) Martin, Tina; Günther, Thomas; Orozco, Adrian Flores; Dahlin, TorleifSpectral induced polarization (SIP) measurements have been demonstrated to correlate with important parameters in hydrogeological and environmental investigations. Although SIP measurements were often collected in the frequency domain (FDIP), recent developments have demonstrated the capabilities to solve for the frequency-dependence of the complex conductivity through measurements collected in the time domain (TDIP). Therefore, the aim of our field investigations is a comparison of the measured frequency-dependence at a broad frequency range resolved through FDIP and TDIP. In contrast to previous studies, we conducted measurements with different instruments and measuring technologies for both FDIP and TDIP. This allows for investigating the robustness of different measurements and assessing various sources of errors, for the assessment of the advantages and drawbacks from different measuring techniques. Our results demonstrate that data collected through different instruments are consistent. Apparent resistivity measurements as well as the inversion results revealed quantitatively the same values for all instruments. The measurements of the IP effect are also comparable, particularly FDIP readings in the low frequencies (< 10 Hz) revealed to be quantitatively the same for different instruments. TDIP measurements are consistent for data collected with both devices. As expected, the spatial distribution of the values is also consistent for low frequency data (in FDIP) and late times measurements in TDIP (> 0.1 s). However, data quality for higher frequencies in FDIP (i.e., early times in TDIP) show larger variations, which reflects the differences between the instruments to deal with the electromagnetic contamination of the IP data. Concluded in general, the different instruments and measuring techniques can provide consistent responses for varying signal-to-noise ratio and measuring configurations. © 2020 The Authors
- ItemComparison of novel semi-airborne electromagnetic data with multi-scale geophysical, petrophysical and geological data from Schleiz, Germany(Amsterdam [u.a.] : Elsevier Science, 2020) Steuer, Annika; Smirnova, Maria; Becken, Michael; Schiffler, Markus; Günther, Thomas; Rochlitz, Raphael; Yogeshwar, Pritam; Mörbe, Wiebke; Siemon, Bernhard; Costabel, Stephan; Preugschat, Benedikt; Ibs-von Seht, Malte; Zampa, Luigi Sante; Müller, FranzIn the framework of the Deep Electromagnetic Sounding for Mineral EXploration (DESMEX) project, we carried out multiple geophysical surveys from regional to local scales in a former mining area in the state of Thuringia, Germany. We prove the applicability of newly developed semi-airborne electromagnetic (EM) systems for mineral exploration by cross-validating inversion results with those of established airborne and ground-based investigation techniques. In addition, supporting petrophysical and geological information to our geophysical measurements allowed the synthesis of all datasets over multiple scales. An initial regional-scale reconnaissance survey was performed with BGR's standard helicopter-borne geophysical system deployed with frequency-domain electromagnetic (HEM), magnetic and radiometric sensors. In addition to geological considerations, the HEM results served as base-line information for the selection of an optimal location for the intermediate-scale semi-airborne EM experiments. The semi-airborne surveys utilized long grounded transmitters and two independent airborne receiver instruments: induction coil magnetometers and SQUID sensors. Due to the limited investigation depth of the HEM method, local-scale electrical resistivity tomography (ERT) and long-offset transient electromagnetic (LOTEM) measurements were carried out on a reference profile, enabling the validation of inversion results at greater depths. The comparison of all inversion results provided a consistent overall resistivity distribution. It further confirmed that both semi-airborne receiver instruments achieve the bandwidth and sensitivity required for the investigation of the resistivity structure down to 1 km depth and therewith the detection of deeply seated earth resources. A 3D geological model, lithological and geophysical borehole logs as well as petrophysical investigations were integrated to interpret of the geophysical results. Distinct highly-conductive anomalies with resistivities of less than 10 Om were identified as alum shales over all scales. Apart from that, the petrophysical investigations exhibited that correlating geophysical and geological information using only one single parameter, such as the electrical resistivity, is hardly possible. Therefore, we developed a first approach based on clustering methods and self-organizing maps (SOMs) that allowed us to assign geological units at the surface to a given combination of geophysical and petrophysical parameters, obtained on different scales. © 2020 The Authors
- ItemClassification of slag material by spectral induced polarization laboratory and field measurements(Amsterdam [u.a.] : Elsevier Science, 2021) Martin, Tina; Günther, Thomas; Weller, Andreas; Kuhn, KerstinHistorical slag dumps are of increasing interest due to economic, environmental or archaeological reasons. Geophysical investigations can help accessing the potential reuse of slag material to recover metallic raw material or for the estimation of the hazard potential of the buried slag material due to dissolution occurrence. In our study, we have investigated various slag material in the laboratory with the spectral induced polarization (SIP) method, obtained from different historical slag dumps, located in the Harz Mountains, Germany. We also present SIP results from field measurements at a historical slag dump where most of the slag samples reveal high amounts of iron, zinc, silica, and barium. Our results reveal a discrimination between three different slag grades (low, medium, high) by using the imaginary conductivity σ″ at a medium frequency (1–10 Hz) in both laboratory and field. Furthermore, additional information is obtained by a classification based on the spectral polarization behaviour and considering the field frequency range (0.1 Hz – 100 Hz). Five different types of spectra (ascending, descending, constant, maximum and minimum type) can be discriminated and recognized in the laboratory and in distinct areas of the slag dump. Even though a direct comparison between the laboratory and field results still needs to be proven, the buried slag material can be differentiated from the surrounding material by the polarization magnitude.
- ItemHigh-resolution net load forecasting for micro-neighbourhoods with high penetration of renewable energy sources(Amsterdam [u.a.] : Elsevier Science, 2020) Kobylinski, P.; Wierzbowski, M.; Piotrowski, K.Though extensive, the literature on electrical load forecasting lacks reports on studies focused on existing residential micro-neighbourhoods comprising small numbers of single-family houses equipped with solar panels. This paper provides a full description of an ANN-based model designed to predict short-term high-resolution (15-min intervals) micro-scale residential net load profiles. Since it seems especially relevant due to the specificity of local autocorrelations in load signal, in this paper we put stress on the systematic approach to feature selection in the context of lagged signal. We performed a case study of a real micro-neighbourhood comprising only 75 single-family houses. The obtained average prediction error was equivalent to 5.4 per cent of the maximal measured net load. The issues, i.e.: (1) the feasibility of micro-scale residential load forecasting taking into account renewable energy penetration, (2) the feasibility to predict net load with dense temporal resolution of 15 min, (3) the feature selection problem, (4) the proposed prosumption- and comparison-oriented prediction model key performance measure, could be of interest to engineers designing energy balancing systems for local smart grids. © 2019 The Authors