Kiloparsec-scale jets in three radio-loud narrow-line Seyfert 1 galaxies

We have discovered kiloparsec-scale extended radio emission in three narrow-line Seyfert 1 galaxies (NLS1s) in sub-arcsecond resolution 9 GHz images from the Karl G. Jansky Very Large Array (VLA). We find all sources show two-sided, mildly core-dominated jet structures with diffuse lobes dominated by termination hotspots. These span 20-70 kpc with morphologies reminiscent of FR II radio galaxies, while the extended radio luminosities are intermediate between FR I and FR II sources. In two cases the structure is linear, while a $45^{circ}$ bend is apparent in the third. Very Long Baseline Array images at 7.6 GHz reveal parsec-scale jet structures, in two cases with extended structure aligned with the inner regions of the kiloparsec-scale jets. Based on this alignment, the ratio of the radio core luminosity to the optical luminosity, the jet/counter-jet intensity and extension length ratios, and moderate core brightness temperatures ($lesssim10^{10}$ K), we conclude these jets are mildly relativistic ($betalesssim0.3$, $deltasim1$-$1.5$) and aligned at moderately small angles to the line of sight (10-15$^{circ}$). The derived kinematic ages of $sim10^6$-$10^7$ y are much younger than radio galaxies but comparable to other NLS1s. Our results increase the number of radio-loud NLS1s with known kiloparsec-scale extensions from seven to ten and suggest that such extended emission may be common, at least among the brightest of these sources.

The Parsec-scale Structure, Kinematics, and Polarization of Radio-Loud Narrow-Line Seyfert 1 Galaxies

Several narrow-line Seyfert 1 galaxies (NLS1s) have now been detected in gamma rays, providing firm evidence that at least some of this class of active galactic nuclei (AGN) produce relativistic jets. The presence of jets in NLS1s is surprising, as these sources are typified by comparatively small black hole masses and near- or super-Eddington accretion rates. This challenges the current understanding of the conditions necessary for jet production. Comparing the properties of the jets in NLS1s with those in more familiar jetted systems is thus essential to improve jet production models. We present early results from our campaign to monitor the kinematics and polarization of the parsec-scale jets in a sample of 15 NLS1s through multifrequency observations with the Very Long Baseline Array. These observations are complemented by fast-cadence 15 GHz monitoring with the Owens Valley Radio Observatory 40m telescope and optical spectroscopic monitoring with with the 2m class telescope at the Guillermo Haro Astrophysics Observatory in Cananea, Mexico.

An Exceptional Radio Flare in Markarian 421

In September 2012, the high-synchrotron-peaked (HSP) blazar Markarian 421 underwent a rapid wideband radio flare, reaching nearly twice the brightest level observed in the centimeter band in over three decades of monitoring. In response to this event we carried out a five epoch centimeter- to millimeter-band multifrequency Very Long Baseline Array (VLBA) campaign to investigate the aftermath of this emission event. Rapid radio variations are unprecedented in this object and are surprising in an HSP BL Lac object. In this flare, the 15 GHz flux density increased with an exponential doubling time of about 9 days, then faded to its prior level at a similar rate. This is comparable with the fastest large-amplitude centimeter-band radio variability observed in any blazar. Similar flux density increases were detected up to millimeter bands. This radio flare followed about two months after a similarly unprecedented GeV gamma-ray flare (reaching a daily E>100 MeV flux of (1.2 +/- 0.7)x10^(-6) ph cm^(-2) s^(-1)) reported by the Fermi Large Area Telescope (LAT) collaboration, with a simultaneous tentative TeV detection by ARGO-YBJ. A cross-correlation analysis of long-term 15 GHz and LAT gamma-ray light curves finds a statistically significant correlation with the radio lagging ~40 days behind, suggesting that the gamma-ray emission originates upstream of the radio emission. Preliminary results from our VLBA observations show brightening in the unresolved core region and no evidence for apparent superluminal motions or substantial flux variations downstream.

Construction of a Calibrated Probabilistic Classification Catalog: Application to 50k Variable Sources in the All-Sky Automated Survey

With growing data volumes from synoptic surveys, astronomers must become more abstracted from the discovery and introspection processes. Given the scarcity of follow-up resources, there is a particularly sharp onus on the frameworks that replace these human roles to provide accurate and well-calibrated probabilistic classification catalogs. Such catalogs inform the subsequent follow-up, allowing consumers to optimize the selection of specific sources for further study and permitting rigorous treatment of purities and efficiencies for population studies. Here, we describe a process to produce a probabilistic classification catalog of variability with machine learning from a multi-epoch photometric survey. In addition to producing accurate classifications, we show how to estimate calibrated class probabilities, and motivate the importance of probability calibration. We also introduce a methodology for feature-based anomaly detection, which allows discovery of objects in the survey that do not fit within the predefined class taxonomy. Finally, we apply these methods to sources observed by the All Sky Automated Survey (ASAS), and unveil the Machine-learned ASAS Classification Catalog (MACC), which is a 28-class probabilistic classification catalog of 50,124 ASAS sources. We estimate that MACC achieves a sub-20% classification error rate, and demonstrate that the class posterior probabilities are reasonably calibrated. MACC classifications compare favorably to the classifications of several previous domain-specific ASAS papers and to the ASAS Catalog of Variable Stars, which had classified only 24% of those sources into one of 12 science classes. The MACC is publicly available at http://www.bigmacc.info.

Active Learning to Overcome Sample Selection Bias: Application to Photometric Variable Star Classification

Despite the great promise of machine-learning algorithms to classify and predict astrophysical parameters for the vast numbers of astrophysical sources and transients observed in large-scale surveys, the peculiarities of the training data often manifest as strongly biased predictions on the data of interest. Typically, training sets are derived from historical surveys of brighter, more nearby objects than those from more extensive, deeper surveys (testing data). This sample selection bias can cause catastrophic errors in predictions on the testing data because a) standard assumptions for machine-learned model selection procedures break down and b) dense regions of testing space might be completely devoid of training data. We explore possible remedies to sample selection bias, including importance weighting (IW), co-training (CT), and active learning (AL). We argue that AL---where the data whose inclusion in the training set would most improve predictions on the testing set are queried for manual follow-up---is an effective approach and is appropriate for many astronomical applications. For a variable star classification problem on a well-studied set of stars from Hipparcos and OGLE, AL is the optimal method in terms of error rate on the testing data, beating the off-the-shelf classifier by 3.4% and the other proposed methods by at least 3.0%. To aid with manual labeling of variable stars, we developed a web interface which allows for easy light curve visualization and querying of external databases. Finally, we apply active learning to classify variable stars in the ASAS survey, finding dramatic improvement in our agreement with the ACVS catalog, from 65.5% to 79.5%, and a significant increase in the classifiers average confidence for the testing set, from 14.6% to 42.9%, after a few AL iterations.

On Machine-Learned Classification of Variable Stars with Sparse and Noisy Time-Series Data

With the coming data deluge from synoptic surveys, there is a growing need for frameworks that can quickly and automatically produce calibrated classification probabilities for newly-observed variables based on a small number of time-series measurements. In this paper, we introduce a methodology for variable-star classification, drawing from modern machine-learning techniques. We describe how to homogenize the information gleaned from light curves by selection and computation of real-numbered metrics (feature), detail methods to robustly estimate periodic light-curve features, introduce tree-ensemble methods for accurate variable star classification, and show how to rigorously evaluate the classification results using cross validation. On a 25-class data set of 1542 well-studied variable stars, we achieve a 22.8% overall classification error using the random forest classifier; this represents a 24% improvement over the best previous classifier on these data. This methodology is effective for identifying samples of specific science classes: for pulsational variables used in Milky Way tomography we obtain a discovery efficiency of 98.2% and for eclipsing systems we find an efficiency of 99.1%, both at 95% purity. We show that the random forest (RF) classifier is superior to other machine-learned methods in terms of accuracy, speed, and relative immunity to features with no useful class information; the RF classifier can also be used to estimate the importance of each feature in classification. Additionally, we present the first astronomical use of hierarchical classification methods to incorporate a known class taxonomy in the classifier, which further reduces the catastrophic error rate to 7.8%. Excluding low-amplitude sources, our overall error rate improves to 14%, with a catastrophic error rate of 3.5%.

Blazars in the Fermi Era: The OVRO 40-m Telescope Monitoring Program

The Large Area Telescope (LAT) aboard the Fermi Gamma-ray Space Telescope provides an unprecedented opportunity to study gamma-ray blazars. To capitalize on this opportunity, beginning in late 2007, about a year before the start of LAT science operations, we began a large-scale, fast-cadence 15 GHz radio monitoring program with the 40-m telescope at the Owens Valley Radio Observatory (OVRO). This program began with the 1158 northern (declination>-20 deg) sources from the Candidate Gamma-ray Blazar Survey (CGRaBS) and now encompasses over 1500 sources, each observed twice per week with a ~4 mJy (minimum) and 3% (typical) uncertainty. Here, we describe this monitoring program and our methods, and present radio light curves from the first two years (2008 and 2009). As a first application, we combine these data with a novel measure of light curve variability amplitude, the intrinsic modulation index, through a likelihood analysis to examine the variability properties of subpopulations of our sample. We demonstrate that, with high significance (7-sigma), gamma-ray-loud blazars detected by the LAT during its first 11 months of operation vary with about a factor of two greater amplitude than do the gamma-ray-quiet blazars in our sample. We also find a significant (3-sigma) difference between variability amplitude in BL Lacertae objects and flat-spectrum radio quasars (FSRQs), with the former exhibiting larger variability amplitudes. Finally, low-redshift (z<1) FSRQs are found to vary more strongly than high-redshift FSRQs, with 3-sigma significance. These findings represent an important step toward understanding why some blazars emit gamma-rays while others, with apparently similar properties, remain silent.

15 GHz Monitoring of Gamma-ray Blazars with the OVRO 40 Meter Telescope in Support of Fermi

We present results from the first two years of our fast-cadence 15 GHz gamma-ray blazar monitoring program, part of the F-GAMMA radio monitoring project. Our sample includes the 1158 blazars north of -20 degrees declination from the Candidate Gamma-Ray Blazar Survey (CGRaBS), which encompasses a significant fraction of the extragalactic sources detected by the Fermi Gamma-ray Space Telescope. We introduce a novel likelihood analysis for computing a time series variability amplitude statistic that separates intrinsic variability from measurement noise and produces a quantitative error estimate. We use this method to characterize our radio light curves. We also present results indicating a statistically significant correlation between simultaneous average 15 GHz radio flux density and gamma-ray photon flux.