Work in time-domain astronomy necessitates robust, automated data processing pipelines that operate in real time. We present the BANZAI pipeline which processes the thousands of science images produced across the Las Cumbres Observatory Global Telescope (LCOGT) network of robotic telescopes each night. BANZAI is designed to perform near real-time preview and end-of-night final processing for four types of optical CCD imagers on the three LCOGT telescope classes. It performs instrumental signature removal (bad pixel masking, bias and dark removal, flat-field correction), astrometric fitting and source catalog extraction. We discuss the design considerations for BANZAI, including testing, performance, and extensibility. BANZAI is integrated into the observatory infrastructure and fulfills two critical functions: (1) real-time data processing that delivers data to users quickly and (2) derive metrics from those data products to monitor the health of the telescope network. In the era of time-domain astronomy, to get from these observations to scientific results, we must be able to automatically reduce data with minimal human interaction, but still have insight into the data stream for quality control.

Modern astronomical surveys such as the Large Synoptic Sky Survey (LSST) promise an unprecedented wealth of discoveries, delivered in the form of ~10 million alerts of time-variable events per night. Astronomers are faced with the daunting challenge of identifying the most scientifically important events from this flood of data in order to conduct effective and timely follow-up observations. Several ongoing observing programs have proven databases to be extremely valuable in conducting efficient follow-up, particularly when combined with tools to select targets, submit observation requests directly to ground- and space-based facilities (manual, remotely-operated and robotic), handle the resulting data, interface with analysis software and share information with collaborators. We draw on experience from a number of follow-up programs running at LCOGT, all of which have independently developed systems to provide these capabilities, including the Microlensing Key Project (RoboNet, PI: Tsapras, Co-I: Street), the Global Supernovae Project (SNEx, PI: Howell), and the Near-Earth Object Project (NEOExchange, PI: Lister). We refer to these systems in general as Target and Observation Managers, or TOMs. Future projects, facing a much greater and rapidly-growing list of potential targets, will find such tools to be indispensable, but the systems developed to date are highly specialized to the projects they serve and are not designed to scale to the LSST alert rate. We present a project to develop a general-purpose software toolkit that will enable astronomers to easily build TOM systems that they can customize to suit their needs, while a professionally-developed codebase will ensure that the systems are capable of scaling to future programs.

Telescope networks are gaining traction due to their promise of higher resource utilization than single telescopes and as enablers of novel astronomical observation modes. However, as telescope network sizes increase, the possibility of scheduling them completely or even semi-manually disappears. In an earlier paper, a step towards software telescope scheduling was made with the specification of the Reservation formalism, through the use of which astronomers can express their complex observation needs and preferences. In this paper we build on that work. We present a solution to the discretized version of the problem of scheduling a telescope network. We derive a solvable integer linear programming (ILP) model based on the Reservation formalism. We show computational results verifying its correctness, and confirm that our Gurobi-based implementation can address problems of realistic size. Finally, we extend the ILP model to also handle the novel observation requests that can be specified using the more advanced Compound Reservation formalism.

We describe the operational capabilities of the Las Cumbres Observatory Global Telescope Network. We summarize our hardware and software for maintaining and monitoring network health. We focus on methodologies to utilize the automated system to monitor availability of sites, instruments and telescopes, to monitor performance, permit automatic recovery, and provide automatic error reporting. The same jTCS control system is used on telescopes of apertures 0.4m, 0.8m, 1m and 2m, and for multiple instruments on each. We describe our network operational model, including workloads, and illustrate our current tools, and operational performance indicators, including telemetry and metrics reporting from on-site reductions. The system was conceived and designed to establish effective, reliable autonomous operations, with automatic monitoring and recovery - minimizing human intervention while maintaining quality. We illustrate how far we have been able to achieve that.

Las Cumbres Observatory Global Telescope (LCOGT) is a young organization dedicated to time-domain observations at optical and (potentially) near-IR wavelengths. To this end, LCOGT is constructing a world-wide network of telescopes, including the two 2m Faulkes telescopes, as many as 17 x 1m telescopes, and as many as 23 x 40cm telescopes. These telescopes initially will be outfitted for imaging and (excepting the 40cm telescopes) spectroscopy at wavelengths between the atmospheric UV cutoff and the roughly 1-micron limit of silicon detectors. Since the first of LCOGT's 1m telescopes are now being deployed, we lay out here LCOGT's scientific goals and the requirements that these goals place on network architecture and performance, we summarize the network's present and projected level of development, and we describe our expected schedule for completing it. In the bulk of the paper, we describe in detail the technical approaches that we have adopted to attain the desired performance. In particular, we discuss our choices for the number and location of network sites, for the number and sizes of telescopes, for the specifications of the first generation of instruments, for the software that will schedule and control the network's telescopes and reduce and archive its data, and for the structure of the scientific and educational programs for which the network will provide observations.

We present an analysis of the rotation of young stars in the associations Cepheus OB3b, NGC 2264, NGC 2362 and the Orion Nebula Cluster (ONC). We discover a correlation between rotation rate and position in a colour-magnitude diagram (CMD) such that stars which lie above an empirically determined median pre-main sequence rotate more rapidly than stars which lie below this sequence. The same correlation is seen, with a high degree of statistical significance, in each association studied here. If position within the CMD is interpreted as being due to genuine age spreads within a cluster, then the stars above the median pre-main sequence would be the youngest stars. This would in turn imply that the most rapidly rotating stars in an association are the youngest, and hence those with the largest moments of inertia and highest likelihood of ongoing accretion. Such a result does not fit naturally into the existing picture of angular momentum evolution in young stars, where the stars are braked effectively by their accretion discs until the disc disperses. Instead, we argue that, for a given association of young stars, position within the CMD is not primarily a function of age, but of accretion history. We show that this hypothesis could explain the correlation we observe between rotation rate and position within the CMD.

Gravitational microlensing is not only a successful tool for discovering distant exoplanets, but it also enables characterization of the lens and source stars involved in the lensing event. In high magnification events, the lens caustic may cross over the source disk, which allows a determination of the angular size of the source and additionally a measurement of its limb darkening. When such extended-source effects appear close to maximum magnification, the resulting light curve differs from the characteristic Paczynski point-source curve. The exact shape of the light curve close to the peak depends on the limb darkening of the source. Dense photometric coverage permits measurement of the respective limb-darkening coefficients. In the case of microlensing event OGLE 2008-BLG-290, the K giant source star reached a peak magnification of about 100. Thirteen different telescopes have covered this event in eight different photometric bands. Subsequent light-curve analysis yielded measurements of linear limb-darkening coefficients of the source in six photometric bands. The best-measured coefficients lead to an estimate of the source effective temperature of about 4700 +100-200 K. However, the photometric estimate from colour-magnitude diagrams favours a cooler temperature of 4200 +-100 K. As the limb-darkening measurements, at least in the CTIO/SMARTS2 V and I bands, are among the most accurate obtained, the above disagreement needs to be understood. A solution is proposed, which may apply to previous events where such a discrepancy also appeared.

We present a photometric study of I-band variability in the young association Cepheus OB3b. The study is sensitive to periodic variability on timescales of less than a day, to more than 20 days. After rejection of contaminating objects using V, I, R and narrowband H-alpha photometry, we find 475 objects with measured rotation periods, which are very likely pre-main-sequence members of the Cep OB3b star forming region. We revise the distance and age to Cep OB3b, putting it on the self-consistent age and distance ladder of Mayne & Naylor (2008). This yields a distance modulus of 8.8 +/- 0.2 mags, corresponding to a distance of 580 +/- 60 pc, and an age of 4-5Myrs. The rotation period distribution confirms the general picture of rotational evolution in young stars, exhibiting both the correlation between accretion (determined in this case through narrowband H-alpha photometry) and rotation expected from disc locking, and the dependence of rotation upon mass that is seen in other star forming regions. However, this mass dependence is much weaker in our data than found in other studies. Comparison to the similarly aged NGC 2362 shows that the low-mass stars in Cep OB3b are rotating much more slowly. This points to a possible link between star forming environment and rotation properties. Such a link would call into question models of stellar angular momentum evolution, which assume that the rotational period distributions of young clusters and associations can be assembled into an evolutionary sequence, thus ignoring environmental effects.

We present three new methods for determining the age of groups of pre-main-sequence stars. The first, creating empirical isochrones allows us to create a robust age ordering, but not to derive actual ages. The second, using the width of the gap in colour-magnitude space between the pre-main-sequence and main-sequence (the radiative convective gap) has promise as a distance and extinction independent measure of age, but is as yet uncalibrated. Finally we discuss tau-squared fitting of the main sequence as the stars approach the terminus of the main sequence. This method suggests that there is a factor two difference between these "nuclear" ages, and more conventional pre-main-sequence contraction ages.

We use I band imaging to perform a variability survey of the 13 Myr-old cluster h Per. We find a significant fraction of the cluster members to be variable. Most importantly, we find that variable members lie almost entirely on the convective side of the gap in the cluster sequence between fully convective stars and those which have a radiative core. This result is consistent with a scenario in which the magnetic field changes topology when the star changes from being fully convective, to one containing a radiative core. When the star is convective the magnetic field appears dominated by large-scale structures, resulting in global-size spots that drive the observed variability. For those stars with radiative cores we observe a marked absence of variability due to spots, which suggests a switch to a magnetic field dominated by smaller-scale structures, resulting in many smaller spots and thus less apparent variability. This implies that wide field variability surveys may only be sensitive to fully convective stars. On the one hand this reduces the chances of picking out young groups (since the convective stars are the lower mass and therefore fainter objects), but conversely the absolute magnitude of the head of the convective sequence provides a straightforward measure of age for those groups which are discovered.

RoboNet-II uses a global network of robotic telescopes to perform follow-up observations of microlensing events in the Galactic Bulge. The current network consists of three 2m telescopes located in Hawaii and Australia (owned by Las Cumbres Observatory) and the Canary Islands (owned by Liverpool John Moores University). In future years the network will be expanded by deploying clusters of 1m telescopes in other suitable locations. A principal scientific aim of the RoboNet-II project is the detection of cool extra-solar planets by the method of gravitational microlensing. These detections will provide crucial constraints to models of planetary formation and orbital migration. RoboNet-II acts in coordination with the PLANET microlensing follow-up network and uses an optimization algorithm ("web-PLOP") to select the targets and a distributed scheduling paradigm (eSTAR) to execute the observations. Continuous automated assessment of the observations and anomaly detection is provided by the ARTEMiS system.

We discuss work by the eSTAR project which demonstrates a fully closed loop autonomous system for the follow up of possible micro-lensing anomalies. Not only are the initial micro-lensing detections followed up in real time, but ongoing events are prioritised and continually monitored, with the returned data being analysed automatically. If the ``smart software'' running the observing campaign detects a planet-like anomaly, further follow-up will be scheduled autonomously and other telescopes and telescope networks alerted to the possible planetary detection. We further discuss the implications of this, and how such projects can be used to build more general autonomous observing and control systems.

We describe the design and implementation of an autonomous adaptive software agent that addresses the practical problem of observing undersampled, periodic, time-varying phenomena using a network of HTN-compliant robotic telescopes. The algorithm governing the behaviour of the agent uses an optimal geometric sampling technique to cover the period range of interest, but additionally implements proactive behaviour that maximises the optimality of the dataset in the face of an uncertain and changing operating environment.

We have selected pre-main-sequence stars in 12 groups of notional ages ranging from 1 Myr to 35 Myrs, using heterogeneous membership criteria. Using these members we have constructed empirical isochrones in V, V-I colour magnitude diagrams (CMDs). This allows us to identify clearly the gap between the radiative main sequence and the convective pre-main-sequence (the R-C gap). We follow the evolution of this gap with age and show that it can be a useful age indicator for groups less than ~15 Myrs old. We also observe a reduction in absolute spreads about the sequences with age. Finally the empirical isochrones allow us to place the groups in order of age, independently of theory. The youngest groups can be collated into three sets of similar ages. The youngest set is the ONC, NGC6530 and IC5146 (nominally 1 Myrs); next Cep OB3b, NGC2362, Lambda Ori and NGC2264 (nominally 3 Myrs); and finally Sigma Ori and IC348 (nominally 4-5 Myrs). This suggests Cep OB3b is younger than previously thought, and IC348 older. For IC348 the stellar rotation rate distribution and fraction of stars with discs imply a younger age than we derive. We suggest this is because of the absence of O-stars in this cluster, whose winds and/or ionising radiation may be an important factor in the removal of discs in other clusters.

Robotic telescopes present the opportunity for the sparse temporal placement of observations when period searching. We address the best way to place a limited number of observations to cover the dynamic range of frequencies required by an observer. We show that an observation distribution geometrically spaced in time can minimise aliasing effects arising from sparse sampling, substantially improving signal detection quality. The base of the geometric series is however a critical factor in the overall success of this strategy. Further, we show that for such an optimal distribution observations may be reordered, as long as the distribution of spacings is preserved, with almost no loss of quality. This implies that optimal observing strategies can retain significant flexibility in the face of scheduling constraints, by providing scope for on-the-fly adaptation. Finally, we present optimal geometric samplings for a wide range of common observing scenarios, with an emphasis on practical application by the observer at the telescope. Such a sampling represents the best practical empirical solution to the undersampling problem that we are aware of. The technique has applications to robotic telescope and satellite observing strategies, where target acquisition overheads mean that a greater total target exposure time (and hence signal-to-noise) can often in practice be achieved by limiting the number of observations.