The ability to generate physically plausible ensembles of variable sources is critical to the optimization of time-domain survey cadences and the training of classification models on datasets with few to no labels. Traditional data augmentation techniques expand training sets by reenvisioning observed exemplars, seeking to simulate observations of specific training sources under different (exogenous) conditions. Unlike fully theory-driven models, these approaches do not typically allow principled interpolation nor extrapolation. Moreover, the principal drawback of theory-driven models lies in the prohibitive computational cost of simulating source observables from {\it ab initio} parameters. In this work, we propose a computationally tractable machine learning approach to generate realistic light curves of periodic variables capable of integrating physical parameters and variability classes as inputs. Our deep generative model, inspired by the Transparent Latent Space Generative Adversarial Networks (TL-GANs), uses a Variational Autoencoder (VAE) architecture with Temporal Convolutional Network (TCN) layers, trained using the \hbox{OGLE-III} optical light curves and physical characteristics (e.g., effective temperature and absolute magnitude) from Gaia DR2. A test using the temperature-shape relationship of RR\,Lyrae demonstrates the efficacy of our generative "Physics-Enhanced Latent Space VAE" (PELS-VAE) model. Such deep generative models, serving as non-linear non-parametric emulators, present a novel tool for astronomers to create synthetic time series over arbitrary cadences.

We report the observations of solar system objects during the 2015 campaign of the High cadence Transient Survey (HiTS). We found 5740 bodies (mostly Main Belt asteroids), 1203 of which were detected in different nights and in $g'$ and $r'$. Objects were linked in the barycenter system and their orbital parameters were computed assuming Keplerian motion. We identified 6 near Earth objects, 1738 Main Belt asteroids and 4 Trans-Neptunian objects. We did not find a $g'-r'$ color-size correlation for $14<H_{g'}<18$ ($1<D<10$ km) asteroids. We show asteroids' colors are disturbed by HiTS' 1.6 hour cadence and estimate that observations should be separated by at most 14 minutes to avoid confusion in future wide-field surveys like LSST. The size distribution for the Main Belt objects can be characterized as a simple power law with slope $\sim0.9$, steeper than in any other survey, while data from HiTS 2014's campaign is consistent with previous ones (slopes $\sim0.68$ at the bright end and $\sim0.34$ at the faint end). This difference is likely due to the ecliptic distribution of the Main Belt since 2015's campaign surveyed farther from the ecliptic than did 2014's and most previous surveys.

Intermediate mass black holes (IMBHs) have masses between the $10^2\!-\!10^6$ M$_\odot$ and are key to our understanding of the formation of massive black holes. The known population of IMBH remains small, with a few hundred candidates and only a handful of them confirmed as bona-fide IMBHs. Until now, the most widely used selection method is based on spectral analysis. Here we present a methodology to select IMBH candidates via optical variability analysis of the nuclear region of local galaxies ($z \leqslant 0.35$). Active IMBH accreting at low rates show small amplitude variability with time scales of hours, as it is seen in one of the known IMBH NGC4395. We found a sample of $\sim \!500$ galaxies evidencing fast and small amplitude variation in their weekly based light curves. We estimate an average occupancy fraction of 4\% and a surface density of $\sim \!3$ deg$^{-2}$, which represent an increase by a factor of $\sim\!40$ compared to previous searches. A large fraction ($78\%$) of the candidates are in spiral galaxies. We preliminary confirm the AGN nature of 22 sources via BPT diagrams using SDSS legacy spectra. Further confirmation of these candidates will require multiwavelength observations, especially in X-ray and radio bands.

The third observing run by LVC has brought the discovery of many compact binary coalescences. Following the detection of the first binary neutron star merger in this run (LIGO/Virgo S190425z), we performed a dedicated follow-up campaign with the Zwicky Transient Facility (ZTF) and Palomar Gattini-IR telescopes. The initial skymap of this single-detector gravitational wave (GW) trigger spanned most of the sky observable from Palomar Observatory. Covering 8000 deg$^2$ of the initial skymap over the next two nights, corresponding to 46\% integrated probability, ZTF system achieved a depth of $\approx$\,21 $m_\textrm{AB}$ in $g$- and $r$-bands. Palomar Gattini-IR covered 2200 square degrees in $J$-band to a depth of 15.5\,mag, including 32\% integrated probability based on the initial sky map. The revised skymap issued the following day reduced these numbers to 21\% for the Zwicky Transient Facility and 19\% for Palomar Gattini-IR. We narrowed 338,646 ZTF transient "alerts" over the first two nights of observations to 15 candidate counterparts. Two candidates, ZTF19aarykkb and ZTF19aarzaod, were particularly compelling given that their location, distance, and age were consistent with the GW event, and their early optical lightcurves were photometrically consistent with that of kilonovae. These two candidates were spectroscopically classified as young core-collapse supernovae. The remaining candidates were ruled-out as supernovae. Palomar Gattini-IR did not identify any viable candidates with multiple detections only after merger time. We demonstrate that even with single-detector GW events localized to thousands of square degrees, systematic kilonova discovery is feasible.

The first two months of the third Advanced LIGO and Virgo observing run (2019 April-May) showed that distant gravitational wave (GW) events can now be readily detected. Three candidate mergers containing neutron stars (NS) were reported in a span of 15 days, all likely located more than 100 Mpc away. However, distant events such as the three new NS mergers are likely to be coarsely localized, which highlights the importance of facilities and scheduling systems that enable deep observations over hundreds to thousands of square degrees to detect the electromagnetic counterparts. On 2019-05-10 02:59:39.292 UT the GW candidate S190510g was discovered and initially classified as a BNS merger with 98% probability. The GW event was localized within an area of 3462 deg2, later refined to 1166 deg2 (90%) at a distance of 227 +- 92 Mpc. We triggered Target of Opportunity observations with the Dark Energy Camera (DECam), a wide-field optical imager mounted at the prime focus of the 4m Blanco Telescope at CTIO in Chile. This Letter describes our DECam observations and our real-time analysis results, focusing in particular on the design and implementation of the observing strategy. Within 24 hours of the merger time, we observed 65% of the total enclosed probability of the final skymap with an observing efficiency of 94%. We identified and publicly announced 13 candidate counterparts. S190510g was re-classified 1.7 days after the merger, after our observations were completed, with a "binary neutron star merger" probability reduced from 98% to 42% in favor of a "terrestrial" classification.

The discovery of a transient kilonova following the gravitational-wave event GW170817 highlighted the critical need for coordinated rapid and wide-field observations, inference, and follow-up across the electromagnetic spectrum. In the Southern hemisphere, the Dark Energy Camera (DECam) on the Blanco 4-m telescope is well-suited to this task, as it is able to cover wide-fields quickly while still achieving the depths required to find kilonovae like the one accompanying GW170817 to $\sim$500 Mpc, the binary neutron star horizon distance for current generation of LIGO/Virgo collaboration (LVC) interferometers. Here, as part of the multi-facility followup by the Global Relay of Observatories Watching Transients Happen (GROWTH) collaboration, we describe the observations and automated data movement, data reduction, candidate discovery, and vetting pipeline of our target-of-opportunity DECam observations of S190426c, the first possible neutron star--black hole merger detected via gravitational waves. Starting 7.5hr after S190426c, over 11.28\,hr of observations, we imaged an area of 525\,deg$^2$ ($r$-band) and 437\,deg$^2$ ($z$-band); this was 16.3\% of the total original localization probability and nearly all of the probability density visible from the Southern hemisphere. The machine-learning based pipeline was optimized for fast turnaround, delivering transient candidates for human vetting within 17 minutes, on average, of shutter closure. We reported nine promising counterpart candidates 2.5 hours before the end of our observations. Our observations yielded no detection of a bona fide counterpart to $m_z = 22.5$ and $m_r = 22.9$ at the 5$\sigma$ level of significance, consistent with the refined LVC positioning. We view these observations and rapid inferencing as an important real-world test for this novel end-to-end wide-field pipeline.

We propose a new sequential classification model for astronomical objects based on a recurrent convolutional neural network (RCNN) which uses sequences of images as inputs. This approach avoids the computation of light curves or difference images. This is the first time that sequences of images are used directly for the classification of variable objects in astronomy. The second contribution of this work is the image simulation process. We generate synthetic image sequences that take into account the instrumental and observing conditions, obtaining a realistic, set of movies for each astronomical object. The simulated dataset is used to train our RCNN classifier. This approach allows us to generate datasets to train and test our RCNN model for different astronomical surveys and telescopes. We aim at building a simulated dataset whose distribution is close enough to the real dataset, so that a fine tuning could match the distributions between real and simulated dataset. To test the RCNN classifier trained with the synthetic dataset, we used real-world data from the High cadence Transient Survey (HiTS) obtaining an average recall of 85%, improved to 94% after performing fine tuning with 10 real samples per class. We compare the results of our model with those of a light curve random forest classifier. The proposed RCNN with fine tuning has a similar performance on the HiTS dataset compared to the light curve classifier, trained on an augmented training set with 10 real samples per class. The RCNN approach presents several advantages in an alert stream classification scenario, such as a reduction of the data pre-processing, faster online evaluation and easier performance improvement using a few real data samples. These results encourage us to use this method for alert brokers systems that will process alert streams generated by new telescopes such as the Large Synoptic Survey Telescope.

Type II supernovae (SNe) originate from the explosion of hydrogen-rich supergiant massive stars. Their first electromagnetic signature is the shock breakout, a short-lived phenomenon which can last from hours to days depending on the density at shock emergence. We present 26 rising optical light curves of SN II candidates discovered shortly after explosion by the High cadence Transient Survey (HiTS) and derive physical parameters based on hydrodynamical models using a Bayesian approach. We observe a steep rise of a few days in 24 out of 26 SN II candidates, indicating the systematic detection of shock breakouts in a dense circumstellar matter consistent with a mass loss rate $\dot{M} > 10^{-4} M_\odot yr^{-1}$ or a dense atmosphere. This implies that the characteristic hour timescale signature of stellar envelope SBOs may be rare in nature and could be delayed into longer-lived circumstellar material shock breakouts in most Type II SNe.

The High Cadence Transient Survey (HiTS) aims to discover and study transient objects with characteristic timescales between hours and days, such as pulsating, eclipsing and exploding stars. This survey represents a unique laboratory to explore large etendue observations from cadences of about 0.1 days and to test new computational tools for the analysis of large data. This work follows a fully \textit{Data Science} approach: from the raw data to the analysis and classification of variable sources. We compile a catalog of ${\sim}15$ million object detections and a catalog of ${\sim}2.5$ million light-curves classified by variability. The typical depth of the survey is $24.2$, $24.3$, $24.1$ and $23.8$ in $u$, $g$, $r$ and $i$ bands, respectively. We classified all point-like non-moving sources by first extracting features from their light-curves and then applying a Random Forest classifier. For the classification, we used a training set constructed using a combination of cross-matched catalogs, visual inspection, transfer/active learning and data augmentation. The classification model consists of several Random Forest classifiers organized in a hierarchical scheme. The classifier accuracy estimated on a test set is approximately $97\%$. In the unlabeled data, $3\,485$ sources were classified as variables, of which $1\,321$ were classified as periodic. Among the periodic classes we discovered with high confidence, 1 $\delta$-scutti, 39 eclipsing binaries, 48 rotational variables and 90 RR-Lyrae and for the non-periodic classes we discovered 1 cataclysmic variables, 630 QSO, and 1 supernova candidates. The first data release can be accessed in the project archive of HiTS.