1704.02322
Automated lensing learner - I: an automated strong lensing identification pipeline
Avestruz, Li, Lightman, Collett, Luo
GL directly probes the underlying mass distribution of lensing systems, the high redshift universe, and cosmo models. The advent of large scale surveys such as LSST and Euclid has promoted a need for automatic and efficient identification of SL systems. Present (1) a SL identification pipeline and (2) a mock LSST dataset with strong gg-lenses. In this first application, employ a fast feature extraction method, Histogram of Oriented Gradients (HOG), to capture edge patterns that are characteristic of strong gravitational arcs in gg strong lensing. Use logistic regression to train a supervised classifier model on the HOG of HST- and LSST-like images. Use the area under the curve (AUC) of a Receiver Operating CHaractierstic (ROC) curve to assess model performance; AUC=1.0 is an ideal classifier, and AUC=0.5 is no better than randomly guessing. The best performing models on a training set of 10,000 lens containing images and 10,000 non-lens containing images exhibit an AUC of 0.975 for an HST-like sample. However, for one exposure of LSST, the model only reaches an AUC of 0.625. For 10-year mock LSST observations, the AUC improved to 0.809. Model performance appears to continually improve with the size of the training set. Models trained on fewer images perform better in absence of the light from the lens galaxy. However, with larger training data sets, information from the lens galaxy actually improves model performance. The results demonstrate an efficient and effective method for automatically identifying strong lenses that captures much of the complexity of the arc finding problem. The linear classifier both runs on a personal laptop and can easily scale to large data sets on a computing cluster, all while using existing open source tools.
1704.02451
Revealing the cosmic web dependent halo bias
Yang et al
Halo bias is one of the key ingredients of the halo models. It was shown at a given redshift to be only dependent, to the first order, on the halo mass. In this study, four types of cosmic web environments: clusters, filaments, sheets and voids are defined within a state of the art high resolution N-body simulation Within those environments, use both halo-dark matter cross-correlation and halo-halo auto correlation functions to probe the clustering properties of haloes. The nature of the halo bias differs strongly among the four different cosmic web environments described. With respect to the overall population, haloes in clusters have significantly lower biases in the {1e11.0-1e13.5} Msun/h mass range. In other environments however, halos show extremely enhanced bases up to a factor 10 in voids for halos of mass 1e12 Msun/h. Demonstrate for the first time that the cosmic web environment is another first order term tat should be rightfully implemented along with mass in halo bias models. In addition, age dependence is found to be only significant in clusters and filaments for relatively small halos 1e12.5 Msun/h.
1704.02744
Finding strong lenses in CFHTLS using convolutional neural networks
Jacobs, et al
Train and apply convolutional neural networks, a machine learning technique developed to learn from and classify image data, to CFHTLS imaging for the identification of potential strong lensing systems. An ensemble of four convolutional neural networks was trained on ims=ages of simulated gg lenses. The training sets consisted of a total of 62k simulated lenses and 64k non-lens negative examples generated with 2 different methodologies. The networks were able to learn the features of simulated lenses with accuracy of up to 99.8% and a purity and completeness of 84-100% on a test set of 2000 sims. An ensemble of trained networks was applied to all of the 171 sq deg of CFHTLS wide field image data, identifying 18k candidates including 63 known and 139 other potential lens candidates. A second search of 1.4 M early type galaxies selected from the survey catalog as potential deflectors, identified 2465 candidates including 117 previously known lens candidates, 29 confirmed lenses/ high-quality lens candidates, 266 novel probably or potential lenses and 2097 candidates classified as false positives. For the catalog-based search, estimate a completeness of 21-28% with respect to detectable lenses and a purity of 15%, with a false-positive rate of 1 in 671 images tested. Predict a human astronomer reviewing candidates produced by the system would identify ~20 probable lenses and 100 possible lenses per hour in a sample selected by the robot. Convolutional neural networks are therefore a promising tool for use in the search for lenses in current and forthcoming surveys such as DES and LSST.
1704.02970
Spectroscopic confirmation of an ultra-faint galaxy at the epoch of reionization
Hoag, Bradac, Trenti, Treu, et al
Within one billion years of the BB, intergalactic H was ionized by sources emitting UV and higher energy photons . This was the final phenomenon to globally affect all baryons (visible matter) in the Universe. It is referred to as cosmic reionization and is an integral component of cosmology. It is broadly expected that intrinsically faint galaxies were the primary ionizing sources due to their abundance in this epoch. However, at the highest redshifts (z>7.5; loopback time 13.1 Gyr), all galaxies with spectroscopic confirmations to date are intrinsically bright and, therefore, not necessary representative of the general population. Report the unequivocal spectroscopic detection of a low luminosity galaxy at z>7.5. Detected the Ly-alpha emission line at ~10504 AA in two separate observations with MOFIRE on Keck I and independently with the HST slit-less grism spectrograph, implying a source redshift of z=7.640±0.001. The galaxy is gravitationally magnified by the massive galaxy clusters MACS J1423.8+2404 (z=0.545), with an estimated intrinsic luminosity of M_AB=-19.6±0.2 mag and a stellar mass of M*=3.0+1.5-0.8e8 Msun. Both are an order of magnitude lower than the 4 other Ly-alpha emitters currently known at z>7.5, making it probably the most distant representative source of reionization found to date.
No comments:
Post a Comment