Search

Service mode observations at ESO's Paranal Observatory relies on the ability to grade science observations. The grading criteria, as defined by the submitter of the program are based mainly on external conditions (e.g. sky transparency, atmospheric seeing) and the grading is done by the night crew immediately after the observations. One of the top-level requirements for operating the E-ELT is to improve observation grading scheme by including criteria based on pipeline-reduced science data. The implementation of a science data grading for a range of different instruments and data types requires the development of new versatile software tools. They would feature real time data visualization, the ability to measure a variety of data quality indicators and able to render the necessary information to the night operators in order to properly grade the science data. We will present the SCUBA software which provides a visualization interface for quality control to all data taken at Paranal.

Several challenges will have to be faced by the staff at Paranal Observatory in order to be well prepared for a seamless integration of the ELT in the current VLT operations scheme. The Telescopes and Instruments Operator group (TIO) is already undergoing changes connected with some of the identified technological and operational needs for the ELT. This paper will have detailed information about the current training needs, group structural changes, the current activities using the adopted engineering-TIO [2] (eTIO) scheme and the staffing plan that will have to be applied in order to keep the centralized support of the biggest world infrastructure in astronomy at the time of the ELT, to handle daily science operations for seven different telescopes, the VLT interferometer and twenty-one scientific instruments in parallel.

Paranal Observatory has a department called Science Operations (SciOps), which is in charge of operating the instruments within the global scheme established for the Very Large Telescope. This scheme was improved on what was called SciOps 2.0. The main operational goals of this new scheme were to strengthen the coordination of science operations activities within, and between, the department groups, by increasing the time allocated to “high-level” activities. It also improves the efficiency of the core science operations support to service mode (SM) and visitor mode (VM) observations, and the quality of the astronomical data delivered to the community of Paranal users. In this context of efficiency and quality improvement of operations within the SciOps department, we had identified a strong need to optimize the management of daily operation tasks, via the development of a daily activity monitoring integrated tool, so this paper details the findings of the Daily Activity Monitoring Integrated Tool (DAMIT), the proof of Concept phase and the first delivered phase. The technical proof of concept was the first phase in development of a daily operation-monitoring tool for the science operations department. The primary objective of this phase was to evaluate the viability and impact of such a tool to improve the quality and efficiency of SciOps at Paranal. This tool is running after overcoming the first phase of development, after followed an on-site technical analysis of the SciOps daily operation (day and night), the current procedures to certify the completeness and quality of the daily operations, and requirements for this new daily operation monitoring tool.

Observatories and operational strategies are evolving in connection with the facilities that will be built. For those new facilities, the strategy for dealing with the telescopes, instrumentation, data-flow, reduction process and relationship with the community is more or less handled from its conception. However, for those Observatories already in place, the challenge is to adapt the processes and prepare the existing people for these changes. This talk will show detailed information about current activities, the implemented training plan, the definition of the current operational model, the involvement of the group in projects towards improving operational processes and efficiency, and what new challenges will be involved during the definition of the strategies for the new generation instruments and facilities to be installed.

Since 2009, operations began at the Survey Telescopes at Paranal Observatory. The surveys aimed to observe using a large field of view targeting much fainter sources and covering wide areas of sky quickly. The first to enter operations was VISTA (Visible and Infrared Survey Telescope for Astronomy) and then the VST Telescope (VLT Survey Telescope). The survey telescopes introduced a change into the operational model of the time. The observations were wholly conducted by the telescope and instrument operator without the aid of a support astronomer. This prompted the gradual and steady improvement of tools for the operation of the observatory both generally and in particular for the Survey Telescopes. Examples of these enhancements include control systems for image quality, selection of OBs, logging of evening activities, among others. However, the new generation instruments at the Very Large Telescope (VLT) posed a new challenge to the observatory from a scientific and operational point of view. As these new systems were more demanding and complex, they would be more complicated to operate and require additional support. Hence, the focus of this study is to explore the possible development and optimization of the operations of the Survey telescopes, which would give greater operational flexibility in regards to the new generation instruments. Moreover, we aim to evaluate the feasibility of redistributing of telescope operators during periods of increased demand from other VLT systems.

We report the discovery of a microlensing planet OGLE-2012-BLG-0950Lb with the planet/host mass ratio of $q \sim 2 \times 10^{-4}$. A long term distortion detected in both MOA and OGLE light curve can be explained by the microlens parallax due to the Earth's orbital motion around the Sun. Although the finite source effect is not detected, we obtain the lens flux by the high resolution Keck AO observation. Combining the microlens parallax and the lens flux reveal the nature of the lens: a planet with mass of $M_{\rm p} = 35^{+17}_{-9} M_{\oplus}$ is orbiting around a M-dwarf with mass of $M_{\rm host} = 0.56^{+0.12}_{-0.16} M_{\odot}$ with a planet-host projected separation of $r_{\perp} =2.7^{+0.6}_{-0.7}$ AU located at $D_{\rm L} = 3.0^{+0.8}_{-1.1}$ kpc from us. This is the first mass measurement from only microlens parallax and the lens flux without the finite source effect. In the coming space observation-era with $Spitzer$, $K2$, $Euclid$, and $WFIRST$, we expect many such events for which we will not be able to measure any finite source effect. This work demonstrates an ability of mass measurements in such events., Comment: 40 pages, 10 figures, 4 Tables. Accepted to AJ

This paper presents the recent changes undergone by the Science Operations department of the ESO Paranal Observatory. This revised science operations model, named SciOps2, aims at improving operations efficiency and quality of the data delivered to our community of users. The changes regarding the new department structure, its staffing, and the distribution of tasks and responsibilities, are described in details, as well as the measured impact of these changes.

The ESO Very Large Telescope Interferometer (VLTI) using the Unit Telescope (UT) was strongly affected by vibrations since the first observations. Investigation by ESO on that subject had started in 2007, with a considerable effort since mid 2008. An important number of investigations on various sub-systems (On telescope: Guiding, Passive supports, Train Coude, insulation of electronics cabinets; On Instruments: dedicated campaign on each instruments with a special attention on the ones equipped with Close Cycle Cooler) were realized. Vibrations were not only recorded and analyzed using the usual accelerometers but also using on use sub-systems as InfRared Image Sensor (IRIS) and Multiple Applications Curvature Adaptive Optics (MACAO) and using a specific tool developed for vibrations measurements Mirror vibrAtion Metrology systeM for the Unit Telescope (MAMMUT). Those tools and systems have been used in order to improve the knowledge on telescope by finding sources. The sources whenever it was possible were damped. As known for years, instruments are still the principal sources of vibrations, for the majority of the UT. A special test in which 2 UTs instruments were completely shut down was realized to determine the minimum Optical Path Length (OPL) achievable. Vibrations is now a part of the instruments interface document and during the installation of any new instrument (KMOS) or system (AOF) a test campaign is realized. As a result some modifications (damping of CCC) can be asked in case of non-compliance. To ensure good operational conditions, levels of vibrations are regularly recorded to control any environmental change.

We analyze MOA-2010-BLG-311, a high magnification (A_max>600) microlensing event with complete data coverage over the peak, making it very sensitive to planetary signals. We fit this event with both a point lens and a 2-body lens model and find that the 2-body lens model is a better fit but with only Delta chi^2~80. The preferred mass ratio between the lens star and its companion is $q=10^(-3.7+/-0.1), placing the candidate companion in the planetary regime. Despite the formal significance of the planet, we show that because of systematics in the data the evidence for a planetary companion to the lens is too tenuous to claim a secure detection. When combined with analyses of other high-magnification events, this event helps empirically define the threshold for reliable planet detection in high-magnification events, which remains an open question., 29 pages, 6 Figures, 3 Tables. For a brief video presentation on this paper, please see http://www.youtube.com/user/OSUAstronomy 10/25/2012 - Updated author list. Replaced 10/10/13 to reflect the version published in ApJ

We present an analysis of the anomalous microlensing event, MOA-2010-BLG-073, announced by the Microlensing Observations in Astrophysics survey on 2010-03-18. This event was remarkable because the source was previously known to be photometrically variable. Analyzing the pre-event source lightcurve, we demonstrate that it is an irregular variable over time scales >200d. Its dereddened color, $(V-I)_{S,0}$, is 1.221$\pm$0.051mag and from our lens model we derive a source radius of 14.7$\pm$1.3 $R_{\odot}$, suggesting that it is a red giant star. We initially explored a number of purely microlensing models for the event but found a residual gradient in the data taken prior to and after the event. This is likely to be due to the variability of the source rather than part of the lensing event, so we incorporated a slope parameter in our model in order to derive the true parameters of the lensing system. We find that the lensing system has a mass ratio of q=0.0654$\pm$0.0006. The Einstein crossing time of the event, $T_{\rm{E}}=44.3$\pm$0.1d, was sufficiently long that the lightcurve exhibited parallax effects. In addition, the source trajectory relative to the large caustic structure allowed the orbital motion of the lens system to be detected. Combining the parallax with the Einstein radius, we were able to derive the distance to the lens, $D_L$=2.8$\pm$0.4kpc, and the masses of the lensing objects. The primary of the lens is an M-dwarf with $M_{L,p}$=0.16$\pm0.03M_{\odot}$ while the companion has $M_{L,s}$=11.0$\pm2.0M_{\rm{J}}$ putting it in the boundary zone between planets and brown dwarfs., 24 pages, 8 figures, best viewed in colour, accepted by ApJ

The Galactic bulge source MOA-2010-BLG-523S exhibited short-term deviations from a standard microlensing lightcurve near the peak of an Amax ~ 265 high-magnification microlensing event. The deviations originally seemed consistent with expectations for a planetary companion to the principal lens. We combine long-term photometric monitoring with a previously published high-resolution spectrum taken near peak to demonstrate that this is an RS CVn variable, so that planetary microlensing is not required to explain the lightcurve deviations. This is the first spectroscopically confirmed RS CVn star discovered in the Galactic bulge., 29 pp, 6 figs, submitted to ApJ

We present the first example of binary microlensing for which the parameter measurements can be verified (or contradicted) by future Doppler observations. This test is made possible by a confluence of two relatively unusual circumstances. First, the binary lens is bright enough (I=15.6) to permit Doppler measurements. Second, we measure not only the usual 7 binary-lens parameters, but also the 'microlens parallax' (which yields the binary mass) and two components of the instantaneous orbital velocity. Thus we measure, effectively, 6 'Kepler+1' parameters (two instantaneous positions, two instantaneous velocities, the binary total mass, and the mass ratio). Since Doppler observations of the brighter binary component determine 5 Kepler parameters (period, velocity amplitude, eccentricity, phase, and position of periapsis), while the same spectroscopy yields the mass of the primary, the combined Doppler + microlensing observations would be overconstrained by 6 + (5 + 1) - (7 + 1) = 4 degrees of freedom. This makes possible an extremely strong test of the microlensing solution. We also introduce a uniform microlensing notation for single and binary lenses, we define conventions, summarize all known microlensing degeneracies and extend a set of parameters to describe full Keplerian motion of the binary lenses., Comment: 51 pages, 8 figures, 2 appendices. Submitted to ApJ. Fortran codes for Appendix B are attached to this astro-ph submission and are also available at http://www.astronomy.ohio-state.edu/~jskowron/OGLE-2009-BLG-020/

We present the analysis result of a gravitational binary-lensing event OGLE-2005-BLG-018. The light curve of the event is characterized by 2 adjacent strong features and a single weak feature separated from the strong features. The light curve exhibits noticeable deviations from the best-fit model based on standard binary parameters. To explain the deviation, we test models including various higher-order effects of the motions of the observer, source, and lens. From this, we find that it is necessary to account for the orbital motion of the lens in describing the light curve. From modeling of the light curve considering the parallax effect and Keplerian orbital motion, we are able to measure not only the physical parameters but also a complete orbital solution of the lens system. It is found that the event was produced by a binary lens located in the Galactic bulge with a distance $6.7\pm 0.3$ kpc from the Earth. The individual lens components with masses $0.9\pm 0.3\ M_\odot$ and $0.5\pm 0.1\ M_\odot$ are separated with a semi-major axis of $a=2.5 \pm 1.0$ AU and orbiting each other with a period $P=3.1 \pm 1.3$ yr. The event demonstrates that it is possible to extract detailed information about binary lens systems from well-resolved lensing light curves., 19 pages, 6 figures

Aims: We present a detailed analysis of OGLE 2004-BLG-482, a relatively high-magnification single-lens microlensing event which exhibits clear extended-source effects. These events are relatively rare, but they potentially contain unique information on the stellar atmosphere properties of their source star, as shown in this study. Methods: Our dense photometric coverage of the overall light curve and a proper microlensing modelling allow us to derive measurements of the OGLE 2004-BLG-482 source star's linear limb-darkening coefficients in three bands, including standard Johnson-Cousins I and R, as well as in a broad clear filter. In particular, we discuss in detail the problems of multi-band and multi-site modelling on the expected precision of our results. We also obtained high-resolution UVES spectra as part of a ToO programme at ESO VLT from which we derive the source star's precise fundamental parameters. Results: From the high-resolution UVES spectra, we find that OGLE 2004-BLG-482's source star is a red giant of MK type a bit later than M3, with Teff = 3667 +/- 150 K, log g = 2.1 +/- 1.0 and an assumed solar metallicity. This is confirmed by an OGLE calibrated colour-magnitude diagram. We then obtain from a detailed microlensing modelling of the light curve linear limb-darkening coefficients that we compare to model-atmosphere predictions available in the literature, and find a very good agreement for the I and R bands. In addition, we perform a similar analysis using an alternative description of limb darkening based on a principal component analysis of ATLAS limb-darkening profiles, and also find a very good agreement between measurements and model predictions., Accepted in A&A

The Astronomical Multi-BEam Recombiner (AMBER), has been operational at the Very Large Telescope Interferometer (VLTI) for many years. We present here some of the constant improvements we have been providing while still operating the instrument, with a heavy load of visitor and service observing programs, most of the nights of the year. In particular, we present here improvements regarding the spectral calibration and correction of the atmospheric loss in squared visibility due to path difference jitter, allowing the instrument to achieve greater precision.

The ESO Very Large Telescope Interferometer (VLTI) offers access to the four 8-m Unit Telescopes (UT) and the four 1.8-m Auxiliary Telescopes (AT) of the Paranal Observatory located in the Atacama Desert in northern Chile. The two VLTI instruments, MIDI and AMBER deliver regular scientific results. In parallel to the operation, the instruments developments are pursued, and new modes are studied and commissioned to offer a wider range of scientific possibilities to the community. New configurations of the ATs array are discussed with the science users of the VLTI and implemented to optimize the scientific return. The monitoring and improvement of the different systems of the VLTI is a continuous work. The PRIMA instrument, bringing astrometry capability to the VLTI and phase referencing to the instruments has been successfully installed and the commissioning is ongoing. The possibility for visiting instruments has been opened to the VLTI facility.

The ESO Very Large Telescope Interferometer (VLTI) offers the unique access to the combination of the four 8-meter Unit Telescopes (UT) of Cerro Paranal. The quality of the scientific observations in interferometric mode is strongly related to the stability of the optical path difference (OPD) between the telescopes. Vibrations at the level of the telescopes and affecting the mirrors were shown to be an important source of perturbation for the OPD. ESO has thus started an important effort on the UTs and VLTI to tackle this effect. Active controls based on accelerometers and phase measurements have been developed to provide real-time correction of the variation of OPD introduced by vibrations. Systematic studies and measurement of the sources of vibration (instruments, wind, telescope altitude, ...) have been performed. Solutions to reduce the vibrations via design modification and/or new operation configurations are studied and implemented. To ensure good operational conditions, the levels of vibrations are regularly monitored to control any environmental change. This document will describe the modifications implemented and foreseen and give a status of the VLTI-UT vibrations evolution.

Multiple Application Curvature Adaptive Optics (MACAO) systems are used at the coud´e focus of the unit telescopes (UTs) at the La-Silla Paranal Observatory, Paranal, to correct for the wave-front aberrations induced by the atmosphere. These systems are in operation since 2005 and are designed to provide beams with 10 mas residual rms tip-tilt error to the VLTI laboratory. We have initiated several technical studies such as measuring the Strehl ratio of the images recorded at the guiding camera of the VLTI, establishing the optimum setup of the MACAO to get collimated and focused beam down to the VLTI laboratory and to the instruments, and ascertaining the data generated by the real time computer, all aimed at characterizing and improving the overall performance of these systems. In this paper we report the current status of these studies.

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., Astronomy & Astrophysics in press

The mass function and statistics of binaries provide important diagnostics of the star formation process. Despite this importance, the mass function at low masses remains poorly known due to observational difficulties caused by the faintness of the objects. Here we report the microlensing discovery and characterization of a binary lens composed of very low-mass stars just above the hydrogen-burning limit. From the combined measurements of the Einstein radius and microlens parallax, we measure the masses of the binary components of $0.10\pm 0.01\ M_\odot$ and $0.09\pm 0.01\ M_\odot$. This discovery demonstrates that microlensing will provide a method to measure the mass function of all Galactic populations of very low mass binaries that is independent of the biases caused by the luminosity of the population., Comment: 6 pages, 3 figures, 1 table

We outline a method for fitting binary-lens caustic-crossing microlensing events based on the alternative model parameterisation proposed and detailed in Cassan (2008). As an illustration of our methodology, we present an analysis of OGLE-2007-BLG-472, a double-peaked Galactic microlensing event with a source crossing the whole caustic structure in less than three days. In order to identify all possible models we conduct an extensive search of the parameter space, followed by a refinement of the parameters with a Markov Chain-Monte Carlo algorithm. We find a number of low-chi2 regions in the parameter space, which lead to several distinct competitive best models. We examine the parameters for each of them, and estimate their physical properties. We find that our fitting strategy locates several minima that are difficult to find with other modelling strategies and is therefore a more appropriate method to fit this type of events., Comment: 12 pages, 11 figures

To improve the planet detection efficiency, current planetary microlensing experiments are focused on high-magnification events searching for planetary signals near the peak of lensing light curves. However, it is known that central perturbations can also be produced by binary companions and thus it is important to distinguish planetary signals from those induced by binary companions. In this paper, we analyze the light curves of microlensing events OGLE-2007-BLG-137/MOA-2007-BLG-091, OGLE-2007-BLG-355/MOA-2007-BLG-278, and MOA-2007-BLG-199/OGLE-2007-BLG-419, for all of which exhibit short-term perturbations near the peaks of the light curves. From detailed modeling of the light curves, we find that the perturbations of the events are caused by binary companions rather than planets. From close examination of the light curves combined with the underlying physical geometry of the lens system obtained from modeling, we find that the short time-scale caustic-crossing feature occurring at a low or a moderate base magnification with an additional secondary perturbation is a typical feature of binary-lens events and thus can be used for the discrimination between the binary and planetary interpretations., 17 pages, 4 figures, 1 table

The ESO Very Large Telescope Interferometer (VLTI) offers access to the four 8 m Unit Telescopes (UT) and the four 1.8 m Auxiliary Telescopes (AT) of the Paranal Observatory located in the Atacama Desert in northern Chile. The fourth AT has been delivered to operation in December 2006, increasing the flexibility and simultaneous baselines access of the VLTI. Regular science operations are now carried on with the two VLTI instruments, AMBER and MIDI. The FINITO fringe tracker is now used for both visitor and service observations with ATs and will be offered on UTs in October 2008, bringing thus the fringe tracking facility to VLTI instruments. In parallel to science observations, technical periods are also dedicated to the characterization of the VLTI environment, upgrades of the existing systems, and development of new facilities. We will describe the current status of the VLTI and prospects on future evolution.

We present here a short review of the calibration processes that are currently applied to the instruments AMBER and MIDI of the VLTI (Very Large Telescope Interferometer) at Paranal. We first introduce the general principles to calibrate the raw data (the “visibilities”) that have been measured by long-baseline optical interferometry. Then, we focus on the specific case of the scientific operation of the VLTI instruments. We explain the criteria that have been used to select calibrator stars for the observations with the VLTI instruments, as well as the routine internal calibration techniques. Among these techniques, the “P2VM” (Pixel-to-Visibility Matrix) in the case of AMBER is explained. Also, the daily monitoring of AMBER and MIDI, that has recently been implemented, is shortly introduced.

The ESO Very Large Telescope Interferometer (VLTI) is the first general-user interferometer that offers near- and mid-infrared long-baseline interferometric observations in service and visitor mode to the whole astronomical community. Over the last two years, the VLTI has moved into its regular science operation mode with the two science instruments, MIDI and AMBER, both on all four 8m Unit Telescopes and the first three 1.8m Auxiliary Telescopes. We are currently devoting up to half of the available time for science, the rest is used for characterization and improvement of the existing system, plus additional installations. Since the first fringes with the VLTI on a star were obtained on March 17, 2001, there have been five years of scientific observations, with the different instruments, different telescopes and baselines. These observations have led so far to more than 40 refereed publications. We describe the current status of the VLTI and give an outlook for its near future.

The VLTI has been operating for about 5 years using the VINCI instrument first, and later MIDI. In October 2005 (Period 76) the first Science Operations with the AMBER instrument started, with 14 Open Time proposals in the observing queues submitted by the astronomical community. AMBER, the near-infrared/red focal instrument of the VLTI, operates in the bands J, H, and, K (i.e. 1.0 to 2.5 micrometers) with three beams, thus enabling the use of closure phase techniques. Light was fed from the 8m Unit Telescopes (UT). The Instrument was offered with the Low Resolution Mode (JHK) and the Medium Resolution Mode in K-band on the UTs. We will present how the AMBER VLTI Science Operations currently are performed and integrated into the general Paranal Science Operations, using the extensive experience of Service Mode operations performed by the Paranal Science operations and in particular applying the know-how learned from the two years of MIDI Science Operations. We will also be presenting the operational statistics from these first ever Open Time observations with AMBER.

We present an analysis of OGLE 2004-BLG-254, a high-magnification and relatively short duration microlensing event in which the source star, a Bulge K3-giant, has been spatially resolved by a point-like lens. We have obtained dense photometric coverage of the event light curve with OGLE and PLANET telescopes, as well as a high signal-to-noise ratio spectrum taken while the source was still magnified by 20, using the UVES/VLT spectrograph. Our dense coverage of this event allows us to measure limb darkening of the source star in the I and R bands. We also compare previous measurements of linear limb-darkening coefficients involving GK-giant stars with predictions from ATLAS atmosphere models. We discuss the case of K-giants and find a disagreement between limb-darkening measurements and model predictions, which may be caused by the inadequacy of the linear limb-darkening law., Comment: 14 pages, 8 figures, accepted for publication in A&A (04/09/2006)

VIMOS is the Visible Multi-Object Spectrograph mounted at the Nasmyth focus of the 8.2m Melipal (UT3) telescope of the ESO Very Large Telescope. VIMOS operates with four channels in three observing modes: imaging, multi-object spectroscopy (MOS), and integral field spectroscopy. VIMOS data are pipeline-processed and quality-checked by the Data Flow Operation group in Garching. The quality check is performed in two steps. The first one is a visual check of each pipeline product that allows the identification of any potential major data problem, such as, for example, a failure in the MOS mask insertion or an over/under exposure. The second step is performed in terms of Quality Control (QC) parameters, which are derived from both raw and processed data to monitor the instrument performance. The evolution in time of the QC parameters is recorded in a publically available database (http://www.eso.org/qc/). The VIMOS QC parameters include, for each of the four VIMOS channels, the bias level, read-out-noise, dark current, gain factor, flat-field and arc-lamps efficiencies, resolution and rms of dispersion, sky flat-field structure, image quality and photometric zeropoints. We describe here some examples of quality checks of VIMOS data.

We discuss high-resolution, time-resolved spectra of the caustic exit of the binary microlensing event OGLE 2002-BUL-69 obtained with UVES on the VLT. The source star is a G5III giant in the Galactic Bulge. During such events, the source star is highly magnified, and a strong differential magnification around the caustic resolves its surface. Using an appropriate model stellar atmosphere generated by the NextGEN code we obtained a model light curve for the caustic exit and compared it with a dense set of photometric observations obtained by the PLANET microlensing follow up network. We further compared predicted variations in the H alpha equivalent width with those measured from our spectra. While the model and observations agree in the gross features, there are discrepancies suggesting shortcomings in the model, particularly for the H alpha line core, where we have detected amplified emission from the stellar chromosphere as the source star's trailing limb exited the caustic. This achievement became possible by the provision of the OGLE-III Early Warning System, a network of small telescopes capable of nearly-continuous round-the-clock photometric monitoring, on-line data reduction, daily near-real-time modelling in order to predict caustic crossing parameters, and a fast and efficient response of a 8m-class telescope to a ``Target-Of-Opportunity'' observation request., 4 pages Latex, 3 figures, accepted for publication to astronomy and astrophysics letters