University of Bremen

About: University of Bremen is a education organization based out in Bremen, Germany. It is known for research contribution in the topics: Population & Context (language use). The organization has 14563 authors who have published 37279 publications receiving 970381 citations. The organization is also known as: Universität Bremen.

New chronology for the late Paleocene thermal maximum and its environmental implications

Abstract: The late Paleocene thermal maximum (LPTM) is associated with a brief, but intense, interval of global warming and a massive perturbation of the global carbon cycle. We have developed a new orbital chronology for Ocean Drilling Program (ODP) Site 690 (Weddell Sea, Southern Ocean) by using spectral analysis of high-resolution geochemical records. The LPTM interval spans 11 precessional cycles yielding a duration of 210 to 220 k.y. The δ 13 C anomaly associated with the LPTM has a magnitude of about −2.5‰ to −3‰; we show that about −2‰ of the excursion occurs within two steps that each were less than 1000 yr in duration. The remainder developed through a series of steps over ∼52 k.y. The timing of these steps is consistent with a series of nearly catastrophic releases of methane from gas hydrates, punctuated by intervals of relative equilibria between hydrate dissociation and carbon burial. Further, we are able to correlate the records between ODP Sites 690 and 1051 (western North Atlantic) on the scale of 21 k.y. cycles, which demonstrates that the details of the δ 13 C excursion are recognizable between distant sites. Comparison of cycle records at Sites 690 and 1051 suggests that sediment representing the interval ∼30 k.y. just prior to and at the onset of the LPTM are missing in the latter location. This unconformity probably resulted from slope failure accompanying methane hydrate dissociation within 10 k.y. of the start of the LPTM.

Real-Time Dynamics in Quantum-Impurity Systems: A Time-Dependent Numerical Renormalization-Group Approach

Abstract: We develop a general approach to the nonequilibrium dynamics of quantum-impurity systems for arbitrary coupling strength. The numerical renormalization group is used to generate a complete basis set necessary for the correct description of the time evolution. We benchmark our method with the exact analytical solution for the resonant-level model. As a first application, we investigate the equilibration of an ultrasmall quantum dot subject to a sudden change of gate voltage and external magnetic field. Two distinct relaxation times are identified for the spin and charge dynamics.

Global temperature calibration of the alkenone unsaturation index (UK′37) in surface waters and comparison with surface sediments

Abstract: In this paper, we compile the current surface seawater C37 alkenone unsaturation (UK′37) measurements (n = 629, −1 to 30°C temperature range) to derive a global, field-based calibration of UK′37 with alkenone production temperature. A single nonlinear “global” surface water calibration of UK′37 accurately predicts alkenone production temperatures over the diversity of modern-day oceanic environments and alkenone-synthesizing populations (T = −0.957 + 54.293(UK′37) − 52.894(UK′37)2 + 28.321(UK′37)3, r2 = 0.97, n = 567). The mean standard error of estimation is 1.2°C with insignificant bias in estimated production temperature among the different ocean regions sampled. An exception to these trends is regions characterized by strong lateral advection and extreme productivity and temperature gradients (e.g., the Brazil-Malvinas Confluence). In contrast to the surface water data, the calibration of UK′37 in surface sediments with overlying annual mean sea surface temperature (AnnO) is best fit by a linear model (AnnO = 29.876(UK′37) − 1.334, r2 = 0.97, n = 592). The standard error of estimation (1.1°C) is similar to that of the surface water production calibration, but a higher degree of bias is observed among the regional data sets. The sediment calibration differs significantly from the surface water calibration. UK′37 in surface sediments is consistently higher than that predicted from AnnO and the surface water production temperature calibration, and the magnitude of the offset increases as the surface water AnnO decreases. We apply the global production temperature calibration to the coretop UK′37 data to estimate the coretop alkenone integrated production temperature (coretop IPT) and compare this with the overlying annual mean sea surface temperature (AnnO). We use simple models to explore the possible causes of the deviation observed between the coretop temperature signal, as estimated by UK′37, and AnnO. Our results indicate that the deviation can best be explained if seasonality in production and/or thermocline production as well as differential degradation of 37:3 and 37:2 alkenones both affect the sedimentary alkenone signal.

Invertible Residual Networks

Abstract: We show that standard ResNet architectures can be made invertible, allowing the same model to be used for classification, density estimation, and generation. Typically, enforcing invertibility requires partitioning dimensions or restricting network architectures. In contrast, our approach only requires adding a simple normalization step during training, already available in standard frameworks. Invertible ResNets define a generative model which can be trained by maximum likelihood on unlabeled data. To compute likelihoods, we introduce a tractable approximation to the Jacobian log-determinant of a residual block. Our empirical evaluation shows that invertible ResNets perform competitively with both state-of-the-art image classifiers and flow-based generative models, something that has not been previously achieved with a single architecture.

Less Is More in Modeling Large Genetic Networks

Abstract: Will we ever be able to build predictive models of the complex genetic networks that determine function as well as malfunction of living cells? Studies of small cellular circuits that represent building blocks indicate that there might be a higher level of description that would reduce the vast complexity of genetic network models. In his Perspective, Bornholdt discusses the results reported in the same issue by Brandman et al . who characterize a frequent building block of cellular circuits. Such elements exhibit particularly simple dynamics and the hope is that models of large genetic networks become feasible when based on the dynamics of such simple building blocks.