Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userCLASS B0631+519: Last of the CLASS lenses
176
0
0.0
(
0
)
Authors
T. York
Ask ChatGPT about the research
No Arabic abstract
We report the discovery of a new gravitational lens system from the CLASS survey, CLASS B0631+519. VLA, MERLIN and VLBA observations show a doubly-imaged radio core, a doubly-imaged lobe and a second lobe that is probably quadruply-imaged. The maximum image separation is 1.16 arcseconds. The VLBA resolves the most magnified image of the flat-spectrum radio core into a number of sub-components spread across approximately 20 milli-arcseconds. Optical and near-infrared imaging with the ACS and NICMOS cameras on the HST show that there are two galaxies along the line of sight to the lensed source, as was previously discovered by optical spectroscopy. The nearer galaxy at z=0.0896 is a small blue irregular, while the more distant galaxy at z=0.6196 is an elliptical type and appears to contribute most of the lensing effect. The host galaxy of the lensed source is visible in NICMOS imaging as a set of arcs that form an almost complete Einstein ring. Mass modelling using non-parametric techniques can reproduce the ring and indicates that the irregular galaxy has a (localised) effect on the flux density distribution in the Einstein ring at the 5-10% level.
rate research
Read More
We present the optical spectra of four newly discovered gravitational lenses from the Cosmic Lens All-Sky Survey (CLASS). These observations were carried out using the Low Resolution Imaging Spectrograph on the W. M. Keck-I Telescope as part of a program to study galaxy-scale gravitational lenses. From our spectra we found the redshift of the background source in CLASS B0128+437 (z_s=3.1240+-0.0042) and the lensing galaxy redshifts in CLASS B0445+123 (z_l=0.5583+-0.0003) and CLASS B0850+054 (z_l=0.5883+-0.0006). Intriguingly, we also discovered that CLASS B0631+519 may have two lensing galaxies (z_l,1=0.0896+-0.0001, z_l,2=0.6196+-0.0004). We also found a single unidentified emission line from the lensing galaxy in CLASS B0128+437 and the lensed source in CLASS B0850+054. We find the lensing galaxies in CLASS B0445+123 and CLASS B0631+519 (l,2) to be early-type galaxies with Einstein Radii of 2.8-3.0 h^{-1} kpc. The deflector in CLASS B0850+054 is a late-type galaxy with an Einstein Radius of 1.6 h^{-1} kpc.
We present flux-ratio curves of the fold and cusp (i.e. close multiple) images of six JVAS/CLASS gravitational lens systems. The data were obtained over a period of 8.5 months in 2001 with the Multi-Element Radio-Linked Interferometer Network (MERLIN) at 5-GHz with 50 mas resolution, as part of a MERLIN Key-Project. Even though the time delays between the fold and cusp images are small (<~1 day) compared to the time-scale of intrinsic source variability, all six lens systems show evidence that suggests the presence of extrinsic variability. In particular, the cusp images of B2045+265 -- regarded as the strongest case of the violation of the cusp relation (i.e. the sum of the magnifications of the three cusp images add to zero) -- show extrinsic variations in their flux-ratios up to ~40 percent peak-to-peak on time scales of several months. Its low Galactic latitude of b=-10 degree and a line-of-sight toward the Cygnus superbubble region suggest that Galactic scintillation is the most likely cause. The cusp images of B1422+231 at b=+69 degree do not show strong extrinsic variability. Galactic scintillation can therefore cause significant scatter in the cusp and fold relations of some radio lens systems (up to 10 percent rms), even though these relations remain violated when averaged over a <~1 year time baseline.
We develop the concept of ABC-Boost (Adaptive Base Class Boost) for multi-class classification and present ABC-MART, a concrete implementation of ABC-Boost. The original MART (Multiple Additive Regression Trees) algorithm has been very successful in large-scale applications. For binary classification, ABC-MART recovers MART. For multi-class classification, ABC-MART considerably improves MART, as evaluated on several public data sets.
Numerous researchers recently applied empirical spectral analysis to the study of modern deep learning classifiers. We identify and discuss an important formal class/cross-class structure and show how it lies at the origin of the many visually striking features observed in deepnet spectra, some of which were reported in recent articles, others are unveiled here for the first time. These include spectral outliers, spikes, and small but distinct continuous distributions, bumps, often seen beyond the edge of a main bulk. The significance of the cross-class structure is illustrated in three ways: (i) we prove the ratio of outliers to bulk in the spectrum of the Fisher information matrix is predictive of misclassification, in the context of multinomial logistic regression; (ii) we demonstrate how, gradually with depth, a network is able to separate class-distinctive information from class variability, all while orthogonalizing the class-distinctive information; and (iii) we propose a correction to KFAC, a well-known second-order optimization algorithm for training deepnets.
The discovery of a small cosmological constant has stimulated interest in the measure problem. One should expect to be a typical observer, but defining such a thing is difficult in the vastness of an eternally inflating universe. We propose that a crucial prerequisite is understanding why one should exist as an observer at all. We assume that the Physical Church Turing Thesis is correct and therefore all observers (and everything else that exists) can be described as different types of information. We then argue that the observers collectively form the largest class of information (where, in analogy with the Faddeev Popov procedure, we only count over gauge invariant forms of information). The statistical predominance of the observers is due to their ability to selectively absorb other forms of information from many different sources. In particular, it is the combinatorics that arise from this selection process which leads us to equate the observer class $mathcal{O}$ with the nontrivial power set $hat{mathcal{P}}(mathcal{U})$ of the set of all information $mathcal{U}$. Observers themselves are thus the typical form of information. If correct, this proposal simplifies the measure problem, and leads to dramatic long term predictions.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
