This research was prompted by the discovery of 35 new or candidate symbiotic stars during a targeted search in the Local Group of Galaxies. A catalogue of a further 200 or so such objects has now been compiled. Many of them could be identified with counterparts in the POINT-AGAPE Catalogue. However, information in the Catalogue is limited to position, brightness and possible period, and light-curves are not available. The poster presented an example of a light-curve of a symbiotic star retrieved from original Point-Agape Catalogue data.

This talk presented a summary of our study of different types of long-term variability in the high-mass X-ray binary LMC X-4, by taking advantage of more than 43 years of measurements in the X-ray domain. In particular, we investigated the 30-day cycle of modulation of the X-ray emission from the source (super-orbital or precessional variability), and refined the orbital period and its first derivative. We showed that the precession period in the time-interval 1991–2015 is near its equilibrium value of Psup = 30.370 days, while the observed historical changes in the phase of this variability can be interpreted in terms of the ‘red noise’ model. We obtained an analytical law from which the precession phase can be determined to within 5% throughout the entire time-interval under consideration. Our analysis revealed for the first time that the source is displaying near-periodic variations of its spin period, on a time-scale of roughly 6.8 years, thus making LMC X-4 one of the (few) known binary systems that show remarkable long-term spin–torque reversals.

KIC 4851217 is a short period eclipsing binary (P = 2.47 days) in the field of the Kepler K1 mission. As well as variability caused by the eclipses, low-amplitude pulsations are also present in the data. A frequency analysis of the residual light-curve revealed δ Sct pulsations in the frequency range from 15–21 d−1 with amplitudes up to 3.5 mmag. Strong linear coupling (fi = fp + kforb) to orbital frequency was found, indicating tidally locked modes. From an analysis of 5 selected groups of frequencies we identified a radial mode on the secondary component, 3 dipole modes (l = |m| = 1), one of them present on the secondary component, and a quadrupole mode (l = |m| = 2), also located on the secondary component.

In the last decade Astronomy has been transformed by a deluge of data that will grow exponentially when near-future telescopes such as LSST and the SKA begin routine observing. Astroinformatics, a broad field encompassing many techniques in statistics, machine learning and data mining, is the key to extracting meaningful information from large amounts of data. This talk outlined Astroinformatics as a field, and gave a few examples of the use of machine learning and Bayesian statistics from my own work in survey Astronomy. The era of massive surveys in which we now find ourselves has the potential to revolutionise completely many fields, including time-domain Astronomy, but only if coupled with the powerful tools of Astroinformatics.

One of the great challenges in time-domain astronomy is the problem of combining data obtained at various epochs with very different instruments. These problems are mostly discussed from within a specific observational mode, for example photometry, spectroscopy or imaging. This Workshop explored by example diverse pitfalls of time-domain calibration by discussing calibration and standardisation problems across various types of variables.

The recent spectacular progress in machine learning and artificial intelligence opens up exciting opportunities for time-domain and transient astronomy. This talk outlined the fundamental challenges that we will need to overcome in order to realise that potential, from discovering completely new classes to fully-automated follow-up campaigns.

As shown by recent gravitational wave detections, galaxies harbour an unknown population of black holes at high masses. In our Galaxy such dark objects can be found and studied solely via gravitational microlensing methods. This paper described our search for black-hole lenses both in archived OGLE data and among on-going microlensing events found by OGLE and Gaia. That combination of superb time-domain astrometry and photometry will enable us to derive masses and distances to these dark lenses uniquely, and to describe the demographics of the unseen component of the Milky Way.

This talk presented and discussed some recent results obtained from a photometric and spectroscopic optical follow-up survey of bright classical novæ. The survey concerned the role of those objects in Galactic chemical evolution, with particular attention to the production of lithium.

We have identified a new population of luminous, optical, narrow-lined transients (FWHM ∼1000 km s−1) coincident with the nuclear region of Seyfert galaxies. According to extensive spectrophotometric follow-ups of the main event (PS1-10adi), we could exclude both normal active galactic nucleus activity and changing-look quasars as the origin. The integrated energy output and spectral evolution over a time-scale of several years point to two possible paths of origin: a tidal disruption of a star by a supermassive black hole, or an extremely energetic supernova occurring within the Seyfert galaxy’s narrow-line (or broad-line) region. The former model would require invoking a specific variant of a tidal disruption, while the latter would require an extremely efficient conversion of kinetic energy via shock interaction between the supernova ejecta and the dense ambient medium.

In March 2017, the Intermediate Palomar Transient Factory (iPTF) ceased operations. This Symposium was an appropriate opportunity to review the scientific returns from iPTF and its predecessor survey, the Palomar Transient Factory (PTF), and to summarise the lessons learned. Succeeding iPTF on the Palomar Observatory 48-inch Schmidt telescope is the Zwicky Transient Facility (ZTF), a new survey with an order of magnitude faster survey speed that is now being commissioned. I described the design and scientific rationale for ZTF. ZTF is prototyping new alert stream technologies being explored by the Large Synoptic Survey Telescope (LSST) to distribute millions of transient alerts per night to downstream science users. I described the design of the alert system and discussed it in the context of the wider LSST and community broker ecosystem.

Our poster presented a new analysis of the transit-time variations displayed by the extrasolar planet Kepler-410Ab. We assumed that the observed changes in the transit times are caused by the gravitational influence of another body in the system. To determine the mass of that perturbing body, we considered the light-time effect and an analytical approximation of the perturbation model. The solutions resulting from both methods gave comparable results, with an orbital period of 970 days and a slightly eccentric orbit for the third body. We proposed two possible models of a perturbing body orbiting a common barycentre with Kepler-410A: a single star with mass of at least 0.906 M⊙, or a binary star with a total component mass of at least 2.15 M⊙.

The broad-lined Type Ic Supernovæ (SNe) associated with Gamma-Ray Bursts (GRBs) were long considered the most luminous class as a whole among core-collapse SNe, until the discovery of Superluminous SNe (SLSNe) during the last decade. There are many differences between the two classes, but in recent years events have been discovered which form a link between them. Associated with GRBs but more luminous and hotter, and in one case clearly dissimilar spectrally from any other known GRB-SNe, these sources point the way towards common mechanisms underlying GRBs and SLSNe.

The Berkeley Visible Image Tube (BVIT) has been a user instrument on the SALT 10-m telescope for the past six years. It can observe transient astrophysical phenomena occurring on time-scales of micro-seconds. This overview presented some recent observations of a dMe flare star, and discussed the recent results of our optical Search for Extraterrestrial Intelligence (OSETI) around nearby exoplanet-hosting stars.

Many distinct classes of high-energy variability have been observed in astrophysical sources, and over a range of time-scales. The widest range, spanning microseconds to decades, is found in accreting, compact, stellar-mass objects, including neutron stars and black holes. Neutron stars are of particular observational interest as they exhibit surface effects giving rise to phenomena – such as thermonuclear bursts and pulsations – not seen in black holes.

This talk reviewed briefly the present understanding of thermonuclear (Type-I) X-ray bursts – events that are powered by an extensive chain of nuclear reactions which in many cases are unique to the environments. Thermonuclear bursts have been exploited over the last few years as an avenue to measure a neutron star’s mass and radius, although the contribution of systematic errors to the measurements remains contentious. We described recent efforts to match burst models to observations better, with a view to resolving some of the astrophysical uncertainties relating to those events. Our efforts have good prospects for providing information that is complementary to nuclear experiments.

The bright nearby binary α Centauri constitutes an excellent laboratory for testing stellar evolution models. The mass, radius, and luminosity of α Cen A and B are known to better than 1% accuracy thanks to recent interferometric and adaptive optical observations, and p-mode oscillations have been observed in both stars. We present new stellar models which fit simultaneously the classical and seismic observations, with particular emphasis on the convective mixing length parameter MLT – the adaptivity of which is necessary to fit the models to observations. The oscillation data provide an important constraint on the models, as the small frequency separation is sensitive to the composition gradient in the core of the stars, while the large frequency separation constrains the mean density of the stars, providing an independent check on the mass and radius.

MeerLICHT is a 65-cm optical telescope situated in Sutherland. The MeerLICHT project, as a part of the ThunderKAT project, will be linked to MeerKAT in real time, providing simultaneous optical and radio observations of the transient sky. Since transient events often require multi-wavelength observations within a short time-lapse, the combination of MeerLICHT and MeerKAT will provide invaluable information on emission at both wavelengths, including the relation between, and evolution of, emission from both wavelength windows. With the installation and commissioning of MeerLICHT now completed, early science with MeerLICHT has begun. The talk offered an overview of the MeerLICHT and ThunderKAT projects as a whole, presented early science results obtained with MeerLICHT, looked at future work with these projects, and discussed other related planned projects.

Astronomy’s heritage, comprised mainly of several million photographic plates (both images and spectra), extends backwards in time for many decades. It constitutes a unique and irreplaceable resource for research into long-term variations. Unfortunately, to date rather few of the plates can be accessed in digital formats and cannot therefore be used in modern research. That lack of use encourages the attitude that ‘old’ data have no value and might as well be destroyed. This Workshop discussed ways and means to avert such threats, and prioritised a list of actions that need to be taken as soon as possible.

The next-generation radio telescopes such as LOFAR and SKA will give access to high time-resolution and high instantaneous sensitivity that can be exploited to study slow and fast transients over the whole radio window. The search for radio transients in large datasets also represents a new signal-processing challenge requiring efficient and robust signal reconstruction algorithms. Using sparse representations and the general ‘compressed sensing’ framework, we developed a 2D–1D algorithm based on the primal-dual splitting method. We have performed our sparse 2D–1D reconstruction on three-dimensional data sets containing either simulated or real radio transients, at various levels of SNR and integration times. This report presents a summary of the current level of performance of our method.

14 Workshops were held during the Symposium, and all but one submitted a report of the discussions that took place. The reports are arranged in the following order:

1. Radio Transients in the Era of Multi-Messenger Astrophysics………. 207

2. Stellar Variability: From Citizen Science to Citizen Astronomy………. 215

3. Getting Ready for TESS: an On-Hand Software Tutorial………. 220

4. 25 Years of the Southern Skies Monitoring by OGLE………. 226

5. A-Type Stars as a Unique Challenge in Time-Domain Studies………. 230

6. X-ray Binary Transients in the Magellanic Clouds and the Milky Way………. 236

7. Towards Science with LSST: Data Products and Communications………. 241

8. Supernovæ………. 245

9. Exploring Data Sonification to Enable, Enhance, and Accelerate the Analysis of Big, Noisy, and Multi-Dimensional Data ………. 251

10. New Instrumentation for Transient Follow-Up ………. 257

11. Nuclear Transients ………. 263

12. Accessing Data for Long-Term Variability ………. 269

13. Machine Learning for Transient Classification ………. 274

14. Calibration and Standardisation ………. 275