1,760 Hits in 2.2 sec

Cloud und IoT

Ulrich Hamann, Matthias Reinwarth
2018 Digitale Welt  
Hamann: Sicherheitsbedenken und Datenschutz- gründe halten insbesondere den deutschen Mittel- stand davon ab, Ulrich Hamann, Geschäftsführung, Bundesdruckerei Ulrich Hamann Von Schnittstellen zu geplanten  ...  Ulrich Hamann, Vorsitzender der Geschäftsführung der Bundesdruckerei, über Speichertechnologien, die Unternehmen einfachen und vor allem sicheren Datenzugriff versprechen. Cloud Computing zu nutzen.  ... 
doi:10.1007/s42354-018-0077-3 fatcat:7i43aiaabrdsbocycvduozk7oe

Seamless lightning nowcasting with recurrent-convolutional deep learning [article]

Jussi Leinonen, Ulrich Hamann, Urs Germann
2022 arXiv   pre-print
A deep learning model is presented to nowcast the occurrence of lightning at a five-minute time resolution 60 minutes into the future. The model is based on a recurrent-convolutional architecture that allows it to recognize and predict the spatiotemporal development of convection, including the motion, growth and decay of thunderstorm cells. The predictions are performed on a stationary grid, without the use of storm object detection and tracking. The input data, collected from an area in and
more » ... rrounding Switzerland, comprise ground-based radar data, visible/infrared satellite data and derived cloud products, lightning detection, numerical weather prediction and digital elevation model data. We analyze different alternative loss functions, class weighting strategies and model features, providing guidelines for future studies to select loss functions optimally and to properly calibrate the probabilistic predictions of their model. Based on these analyses, we use focal loss in this study, but conclude that it only provides a small benefit over cross entropy, which is a viable option if recalibration of the model is not practical. The model achieves a pixel-wise critical success index (CSI) of 0.45 to predict lightning occurrence within 8 km over the 60-min nowcast period, ranging from a CSI of 0.75 at a 5-min lead time to a CSI of 0.32 at a 60-min lead time.
arXiv:2203.10114v2 fatcat:vbuzetf2xbgnff64yy6fkzniem

Recent Advances in Scalable Network Generation [article]

Manuel Penschuck and Ulrik Brandes and Michael Hamann and Sebastian Lamm and Ulrich Meyer and Ilya Safro and Peter Sanders and Christian Schulz
2020 arXiv   pre-print
Hamann et al. [78, 79] propose an external memory algorithm that uses I/O-efficient ES and a streaming implementation of Havel-Hakimi to generate graphs.  ... 
arXiv:2003.00736v1 fatcat:pxy6g2qvs5hutbw4qyughe4dk4

Satellite-Based Rainfall Retrieval: From Generalized Linear Models to Artificial Neural Networks

Lea Beusch, Loris Foresti, Marco Gabella, Ulrich Hamann
2018 Remote Sensing  
In this study, we develop and compare satellite rainfall retrievals based on generalized linear models and artificial neural networks. Both approaches are used in classification mode in a first step to identify the precipitating areas (precipitation detection) and in regression mode in a second step to estimate the rainfall intensity at the ground (rain rate). The input predictors are geostationary satellite infrared (IR) brightness temperatures and Satellite Application Facility (SAF)
more » ... g products which consist of cloud properties, such as cloud top height and cloud type. Additionally, a set of auxiliary location-describing input variables is employed. The output predictand is the ground-based instantaneous rain rate provided by the European-scale radar composite OPERA, that was additionally quality-controlled. We compare our results to a precipitation product which uses a single infrared (IR) channel for the rainfall retrieval. Specifically, we choose the operational PR-OBS-3 hydrology SAF product as a representative example for this type of approach. With generalized linear models, we show that we are able to substantially improve in terms of hits by considering more IR channels and cloud property predictors. Furthermore, we demonstrate the added value of using artificial neural networks to further improve prediction skill by additionally reducing false alarms. In the rain rate estimation, the indirect relationship between surface rain rates and the cloud properties measurable with geostationary satellites limit the skill of all models, which leads to smooth predictions close to the mean rainfall intensity. Probability matching is explored as a tool to recover higher order statistics to obtain a more realistic rain rate distribution. Rain gauges directly measure surface precipitation, but, since they are point measurements, they lack representativity for regional-scale QPE [17, 18] . This problem is further increased by the poor coverage of gauges over oceans and certain land regions [19] . Additional challenges include measurement errors, most importantly due to wind effects, but also caused by evaporation and wetting processes [20] , with the largest uncertainty being observed for solid precipitation [21] . Weather radars are active remote sensing instruments that allow for monitoring the spatio-temporal evolution of precipitation systems over large areas. Radar-based QPE involves a complex data processing chain and is subject to several sources of uncertainty, among which there are the conversion of the measured reflectivity into rainfall rates, the attenuation of the signal, anomalous propagation, range degradation, spatial variability of the vertical profile of reflectivity, and residual clutter [22, 23] . Satellite infrared (IR) and microwave (MW) radiometers, on the other hand, are passive remote sensing devices and QPE can be carried out based on the spectral information they provide. However, compared to active remote sensing instruments, the relationship between the spectral information and the rainfall is weaker. Nevertheless, satellite radiometer-based rainfall retrievals can deliver global-scale, spatially and temporally continuous, high-resolution precipitation products. Thus, they are most valuable in covering regions outside of radar domains or during radar-downtime [24] . They are especially helpful over oceans and in developing countries with no radar infrastructure, poor coverage, or limited resources to maintain the radar networks (e.g., [25, 26] ). Several comprehensive overviews treating different aspects of satellite-based QPE can be found in the literature [27] [28] [29] [30] [31] [32] [33] . In the following, "satellite-based precipitation" refers to rainfall that is derived solely from radiometers. Such precipitation products can be derived from low-level earth-orbiting satellites (LEO, typically at 350-850 km altitude; e.g., [7]), geostationary satellites (GEO, at about 36,000 km altitude; e.g., [34, 35] ), or a combination thereof (e.g., [8, 36] ). Since MW emission and scattering depend on the hydrometer size distribution, they are more directly related to rainfall intensity than cloud top temperature derived from IR measurements, making instantaneous rainfall retrievals with passive MW sensors on-board of LEOs attractive [37] . However, GEO satellites provide a much better temporal resolution rendering them more suitable for real-time precipitation monitoring [38] . Additionally, they give better estimations of daily and monthly precipitation accumulations. Adler et al. [39] were the first to combine the advantages of the two data sources by matching microwave and IR data to improve the algorithm of Arkin and Meisner [34] . The rainfall retrievals presented above are generally split into precipitation detection and rainfall rate estimation. Frequently, spectral information is taken as proxy for cloud top properties and conceptual understanding of rainfall processes is employed to determine the parametric relationships between satellite-derived information and precipitation patterns [40] . Precipitating areas are often distinguished from the non-precipitating ones using threshold tests for selected satellite channels and/or derived properties (e.g., [35, [41] [42] [43] ). Rainfall rates are then estimated by relating the satellite information to modelled or observed rain rates (e.g., [38, 43, 44] ). If the goal is to obtain the best-possible performance in the rainfall retrieval rather than improving the conceptual understanding of the physical processes, machine learning approaches exploiting large numbers of input predictors can be helpful [40] . Also when using machine learning approaches, precipitation detection (e.g., [45, 46] ) and rain rate estimation (e.g., [47] ) are often treated separately. Many studies can be found on such rainfall retrievals and we present a few in the following. The probably most widely known machine learning precipitation retrieval algorithm is PERSIANN [36] . PERSIANN, however, groups pixels according to cloud surface characteristics derived from GEO IR information with the use of an artificial neural network instead of explicitly identifying precipitating areas. For each group, a separate multivariate linear function mapping is employed to relate the input predictors to rain rates. LEO instantaneous rain rates are furthermore used to update the linear mapping network parameters. PERSIANN has been continuously developed e.g., by introducing a cloud classification system and thereby switching from pixel-by-pixel fitting of rain rates to cloud patches (PERSIANN-CCS, [48] ). Recently, Tao et al. [49]
doi:10.3390/rs10060939 fatcat:gjhcaabsg5bn7deicwdznvqkdu

Evaluating and Improving Cloud Parameter Retrievals

Rob Roebeling, Bryan Baum, Ralf Bennartz, Ulrich Hamann, Andy Heidinger, Anke Thoss, Andi Walther
2013 Bulletin of The American Meteorological Society - (BAMS)  
doi:10.1175/bams-d-12-00041.1 fatcat:kr7ecv3r7vby7cwhvrgvibrl54

Nowcasting thunderstorm hazards using machine learning: the impact of data sources on performance

Jussi Leinonen, Ulrich Hamann, Urs Germann, John R. Mecikalski
2022 Natural Hazards and Earth System Sciences  
Abstract. In order to aid feature selection in thunderstorm nowcasting, we present an analysis of the utility of various sources of data for machine-learning-based nowcasting of hazards related to thunderstorms. We considered ground-based radar data, satellite-based imagery and lightning observations, forecast data from numerical weather prediction (NWP) and the topography from a digital elevation model (DEM), ending up with 106 different predictive variables. We evaluated machine-learning
more » ... s to nowcast storm track radar reflectivity (representing precipitation), lightning occurrence, and the 45 dBZ radar echo top height that can be used as an indicator of hail, producing predictions for lead times of up to 60 min. The study was carried out in an area in the Northeastern United States for which observations from the Geostationary Operational Environmental Satellite-16 are available and can be used as a proxy for the upcoming Meteosat Third Generation capabilities in Europe. The benefits of the data sources were evaluated using two complementary approaches: using feature importance reported by the machine learning model based on gradient-boosted trees, and by repeating the analysis using all possible combinations of the data sources. The two approaches sometimes yielded seemingly contradictory results, as the feature importance reported by the gradient-boosting algorithm sometimes disregards certain features that are still useful in the absence of more powerful predictors, while, at times, it overstates the importance of other features. We found that the radar data is the most important predictor overall. The satellite imagery is beneficial for all of the studied predictands, and therefore offers a viable alternative in regions where radar data are unavailable, such as over the oceans and in less-developed ares. The lightning data are very useful for nowcasting lightning but are of limited use for the other hazards. While the feature importance ranks NWP data as an important input, the omission of NWP data can be well compensated for by using information in the observational data over the nowcast period. Finally, we did not find evidence that the nowcast benefits from the DEM data.
doi:10.5194/nhess-22-577-2022 fatcat:xy2yj4gcinhmzdnmswzbae4nsm

The libRadtran software package for radiative transfer calculations (version 2.0.1)

Claudia Emde, Robert Buras-Schnell, Arve Kylling, Bernhard Mayer, Josef Gasteiger, Ulrich Hamann, Jonas Kylling, Bettina Richter, Christian Pause, Timothy Dowling, Luca Bugliaro
2016 Geoscientific Model Development  
<p><strong>Abstract.</strong> libRadtran is a widely used software package for radiative transfer calculations. It allows one to compute (polarized) radiances, irradiance, and actinic fluxes in the solar and thermal spectral regions. libRadtran has been used for various applications, including remote sensing of clouds, aerosols and trace gases in the Earth's atmosphere, climate studies, e.g., for the calculation of radiative forcing due to different atmospheric components, for UV forecasting,
more » ... e calculation of photolysis frequencies, and for remote sensing of other planets in our solar system. The package has been described in Mayer and Kylling (2005). Since then several new features have been included, for example polarization, Raman scattering, a new molecular gas absorption parameterization, and several new parameterizations of cloud and aerosol optical properties. Furthermore, a graphical user interface is now available, which greatly simplifies the usage of the model, especially for new users. This paper gives an overview of libRadtran version 2.0.1 with a focus on new features. Applications including these new features are provided as examples of use. A complete description of libRadtran and all its input options is given in the user manual included in the libRadtran software package, which is freely available at <a href="" target="_blank"></a>.</p>
doi:10.5194/gmd-9-1647-2016 fatcat:gwforymamfaezhjzevbqmf6qtm

Parallel and I/O-efficient Randomisation of Massive Networks using Global Curveball Trades [article]

Corrie Jacobien Carstens, Michael Hamann, Ulrich Meyer, Manuel Penschuck, Hung Tran, Dorothea Wagner
2018 arXiv   pre-print
Graph randomisation is a crucial task in the analysis and synthesis of networks. It is typically implemented as an edge switching process (ESMC) repeatedly swapping the nodes of random edge pairs while maintaining the degrees involved. Curveball is a novel approach that instead considers the whole neighbourhoods of randomly drawn node pairs. Its Markov chain converges to a uniform distribution, and experiments suggest that it requires less steps than the established ESMC. Since trades however
more » ... e more expensive, we study Curveball's practical runtime by introducing the first efficient Curveball algorithms: the I/O-efficient EM-CB for simple undirected graphs and its internal memory pendant IM-CB. Further, we investigate global trades processing every node in a graph during a single super step, and show that undirected global trades converge to a uniform distribution and perform superior in practice. We then discuss EM-GCB and EM-PGCB for global trades and give experimental evidence that EM-PGCB achieves the quality of the state-of-the-art ESMC algorithm EM-ES nearly one order of magnitude faster.
arXiv:1804.08487v2 fatcat:ujwf5wzqzjejbjpki6nmpm6vne

Summary of the Fourth Cloud Retrieval Evaluation Workshop

Rob Roebeling, Bryan Baum, Ralf Bennartz, Ulrich Hamann, Andrew Heidinger, Jan Fokke Meirink, Martin Stengel, Anke Thoss, Andi Walther, Phil Watts
2015 Bulletin of The American Meteorological Society - (BAMS)  
Since CREW-3 in 2012, more groups now perform these types of assessments (Hamann et al. 2014; Stengel et al. 2015) .  ... 
doi:10.1175/bams-d-14-00184.1 fatcat:bfretsnuhjgilhuzu7yxdslpkq

The acquisition of pronouns by French children: A parallel study of production and comprehension

2010 Applied Psycholinguistics  
Moreover, this difficulty with object clitics is found not only in L1 acquisition but also in young L2 learners of French (White, 1996) and in children with SLI (Chillier et al., 2001; Hamann et al.  ...  With respect to reflexive clitics, studies of spontaneous production reported a profile similar to the one observed for object clitics (Hamann, Rizzi, & Frauenfelder, 1996; Jakubowicz & Rigaut, 2000)  ... 
doi:10.1017/s0142716410000147 fatcat:rakphya3crha7pja7cs2woujli

Single-distance phase retrieval algorithm for Bragg Magnifier microscope

Stanislav Hrivňak, Jozef Uličný, Ladislav Mikeš, Angelica Cecilia, Elias Hamann, Tilo Baumbach, Libor Švéda, Zdenko Zápražný, Dušan Korytár, Eva Gimenez-Navarro, Ulrich H. Wagner, Christoph Rau (+2 others)
2016 Optics Express  
We present an improved, single-distance phase retrieval algorithm applicable for holographic X-ray imaging of biological objects for an in-line germanium Bragg Magnifier Microscope (BMM). The proposed algorithm takes advantage of a modified shrink-wrap algorithm for phase objects, robust unwrapping algorithm as well as other reasonable constraints applied to the wavefield at the object and the detector plane. The performance of the algorithm is analyzed on phantom objects and the results are
more » ... wn and discussed. We demonstrated the suitability of the algorithm for the phase retrieval on a more complex biological specimen Tardigrade, where we achieved successful phase retrieval from only a single hologram. The spatial resolution obtained by Fourier spectral power method for biological objects is ∼ 300 nm, the same value as obtained from the reconstructed test pattern. Our results achieved using the new algorithm confirmed the potential of BMM for in-vivo, dose-efficient single-shot imaging of biological objects. "X-ray phase-contrast imaging with submicron resolution by using extremely asymmetric bragg diffractions," Appl. Phys. Lett. 78, 132-134 (2001). 6. P. Modregger, D. Lübbert, P. Schäfer, and R. Köhler, "Magnified x-ray phase imaging using asymmetric bragg reflection: Experiment and theory," Phys. Rev. B 74, 054107 (2006). 7. M. Stampanoni, G. Borchert, R. Abela, and P. Ruegsegger, "Nanotomography based on double asymmetrical bragg diffraction," Appl.
doi:10.1364/oe.24.027753 pmid:27906343 fatcat:wthywuif7zc5zhoxej2kdcx5tq

I/O-Efficient Generation of Massive Graphs Following the LFR Benchmark [article]

Michael Hamann, Ulrich Meyer, Manuel Penschuck, Hung Tran, Dorothea Wagner
2017 arXiv   pre-print
LFR is a popular benchmark graph generator used to evaluate community detection algorithms. We present EM-LFR, the first external memory algorithm able to generate massive complex networks following the LFR benchmark. Its most expensive component is the generation of random graphs with prescribed degree sequences which can be divided into two steps: the graphs are first materialized deterministically using the Havel-Hakimi algorithm, and then randomized. Our main contributions are EM-HH and
more » ... S, two I/O-efficient external memory algorithms for these two steps. We also propose EM-CM/ES, an alternative sampling scheme using the Configuration Model and rewiring steps to obtain a random simple graph. In an experimental evaluation we demonstrate their performance; our implementation is able to handle graphs with more than 37 billion edges on a single machine, is competitive with a massive parallel distributed algorithm, and is faster than a state-of-the-art internal memory implementation even on instances fitting in main memory. EM-LFR's implementation is capable of generating large graph instances orders of magnitude faster than the original implementation. We give evidence that both implementations yield graphs with matching properties by applying clustering algorithms to generated instances. Similarly, we analyse the evolution of graph properties as EM-ES is executed on networks obtained with EM-CM/ES and find that the alternative approach can accelerate the sampling process.
arXiv:1604.08738v3 fatcat:l6ewm2btwfburmk22bofl7le4m

Similar contributions of BRCA1 and BRCA2 germline mutations to early-onset breast cancer in Germany

Ute Hamann, Xuan Liu, Nikola Bungardt, Hans Ulrich Ulmer, Gunther Bastert, Hans-Peter Sinn
2003 European Journal of Human Genetics  
BRCA 1/2 germline mutations in breast cancer patients U Hamann et al Table 2 2 BRCA1 and BRCA2 germline mutations in German early-onset breast cancer patients FS: frameshift mutation; MS: missense  ...  database (BIC) 13 , affects the tyrosine residue at codon 179 that is conserved among mouse, dog and rat homologues. 14, 15 This mutation was also identified in the German breast cancer family P 328 (U Hamann  ... 
doi:10.1038/sj.ejhg.5200988 pmid:12774040 fatcat:xvffdy34sjhxleabdbi7odl23a

Normal and pathological development of subject–verb agreement in speech production: a study on French children

Julie Franck, Stephany Cronel-Ohayon, Laurence Chillier, Ulrich H. Frauenfelder, Cornelia Hamann, Luigi Rizzi, Pascal Zesiger
2004 Journal of Neurolinguistics  
et al., 2002 Hamann et al., , 2001 Hamann et al., , 2003 Zesiger et al., 2001) .  ...  Similar findings were reported for young French-speaking SLI children (Hamann et al., 2003; Paradis & Crago, 2000) .  ... 
doi:10.1016/s0911-6044(03)00057-5 fatcat:rteldebvqbgn5bfmfazxlwarnu

Improved targeting of human CD4+ T cells by nanobody-modified AAV2 gene therapy vectors

Martin V. Hamann, Niklas Beschorner, Xuan-Khang Vu, Ilona Hauber, Ulrike C. Lange, Bjoern Traenkle, Philipp D. Kaiser, Daniel Foth, Carola Schneider, Hildegard Büning, Ulrich Rothbauer, Joachim Hauber (+1 others)
2021 PLoS ONE  
Hamann, Niklas Beschorner, Ulrike C. Lange, Hildegard Bün- ing, Ulrich Rothbauer, Joachim Hauber.  ...  These novel vector variants dem- Copyright: © 2021 Hamann et al.  ... 
doi:10.1371/journal.pone.0261269 pmid:34928979 pmcid:PMC8687595 fatcat:rcqci4ny5napjdhfnxu2x3yzzi
« Previous Showing results 1 — 15 out of 1,760 results