December 2022

14/12/22 (Wednesday)
10:45, Library (ESO HQ, Garching) | ESO Garching
Informal Discussion
Talk — The Atacama Large Aperture Submillimeter Telescope (AtLAST)
Tony Mroczkowski (ESO)

Abstract

I will discuss, largely using the whiteboard and a few printout plots/graphics, the ideas behind a next-generation widefield 50-meter telescope design that could achieve mapping speeds roughly a million times faster than ALMA. Come and imagine what you could do with AtLAST in one year what you couldn't even begin to do with ALMA in our lifetimes. AtLAST has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 951815 (see https://cordis.europa.eu/project/id/951815), and ESO is one of the participants in this EU-funded design study.

 

07/12/22 (Wednesday)
10:45, Auditorium Telescopium (ESO HQE, Garching) | ESO Garching
Informal Discussion
Talk — JWST Commissioning - From First Light to First Science
Jens Kammerer (STScI)

November 2022

23/11/22 (Wednesday)
10:45, Auditorium Eridanus (ESO HQE, Garching) | ESO Garching
Informal Discussion
Talk — The next generation near-IR instrument for the VLT: ERIS
Harald Kuntschner & Marianne Heida (ESO)

Abstract

The Enhanced Resolution Imager and Spectrograph (ERIS) is mounted on UT4 on Paranal and recently the commissioning came to an end. Science Verification will start beginning of December and regular science observing in April 2023.

The instrument separates into the near-IR integral field unit SPIFFIER which is an enhanced version of the former SINFONI instrument and an imaging camera called NIX reaching out to M-band. NIX features also several coronographic modes and a long slit for L-band.

We will describe the different instrument modes and the enhanced use of the AO using the deformable mirror. We will then discuss some possible science cases and hopefully get you thinking about what you could do with ERIS.

16/11/22 (Wednesday)
10:45, Pavo (ESO room A.2.01) | ESO Garching
Informal Discussion
Talk — Towards a complete picture of early structure formation using the Thesan simulations
Enrico Garaldi (MPA)

Abstract

The formation of the first galaxies in the Universe is the new frontier of both galaxy formation and reionization studies. In fact, we will soon directly observe primeval galaxies thanks to the James Webb Space Telescope, and witness the reionization process through 21cm intensity mapping experiments. This unique moment in human history creates a fierce new challenge, i.e. to simultaneously understand in a unique and coherent picture the processes of galaxy formation and reionization, and – crucially – their connection. The latter, in particular, has escaped past numerical efforts. In this talk I will present the outcome of an years-long effort geared toward achieving such comprehensive picture, culminated in the Thesan suite of cosmological radiation-magneto-hydrodynamical simulations. They are unique for a number of reasons, namely they: (i) cover a very broad range of spatial and temporal scales, connecting cosmic reionization with galaxy formation; (ii) simulate an unprecedentedly-broad range of physical processes for simulation of such scales and resolution, including stellar and black hole feedback, individual metals transport, galactic winds, dust creation and destruction, self-consistent radiation transport from stars and black holes,and magnetic fields; (iii) exploit knowledge accumulated at low redshift to minimize the number of free parameters in the physical model; (iv) use a variance-suppression technique in the production of initial conditions to increase their statistical fidelity; (v) includes multiple runs exploring current unknowns in the physics of dark matter and ionizing sources. I will finally show an example of how Thesan is informing observations of the high-z IGM-galaxy connection.

09/11/22 (Wednesday)
10:45, Auditorium Eridanus (ESO HQE, Garching) | ESO Garching
Informal Discussion
Talk — Characterizing the filamentary structure of the ISM – The need for data combination
Francesca Bonanomi (VISESS, University of Vienna)

Abstract

The study of the filamentary nature of the Interstellar Medium  has been revolutionized by the use of ALMA’s high-resolution observations. However, the analysis of interferometric data is not straightforward, as the emission properties of the observed sources include large regions with diffuse, extended emission, strongly affected by the interferometric filtering effects on the true sky brightness distribution. We aim to quantify the impact of instrumental effects on the recovery of these fundamental filamentary structures at different spatial scales in the ISM by detailed simulations. Our results quantitatively illustrate the intrinsic limitations of interferometers for the recovery of sources such as cores and filaments with extended emission, both not recovering the total emission of the source and critically affecting the radius and the slope of their true radial profile.  In our work we investigate different data-combination methods, comparing the results obtained from the combination of the ALMA 12m array data with the ALMA short-spacing contribution (ALMA 7m array and Total Power) and a with a large Single Dish image.

October 2022

26/10/22 (Wednesday)
10:45, Auditorium Telescopium (ESO HQE, Garching) | ESO Garching
Informal Discussion
Talk — Wild as Dust: the origin of dust in early-type galaxies
Tom Richtler (Universidad de Concepcion)

Abstract

Dust is found frequently in early-type galaxies. The traditional view that dust is mainly produced by young stellar populations, can only be rescued if dust in old stellar populations would have an "external origin" (whatever that means). I shall show a gallery of dust manifestations in various galaxies. A few galaxies like NGC 1316, NGC 4696, and NGC 5102 will be presented in more detail with regard to their dust and gas content. The lesson to be learned is that apparently a significant part of the dust is produced and distributed internally, by dusty winds from nuclei and/or by dust formation in magnetic field structures.

19/10/22 (Wednesday)
10:45, Pavo (ESO room A.2.01) | ESO Garching
Informal Discussion
Talk — A Fast Radio Burst from a galaxy at z=1.02
Elaine Sadler (Sydney University and CSIRO)

Abstract

I will report on the very recent discovery of a Fast Radio Burst (FRB) localised to a galaxy at redshift z=1.02. This FRB host galaxy is twice as distant as any previously identified FRB host galaxy (all of which are at z < 0.5). In my talk, I will give a brief introduction to the FRB phenomenon and explain why these enigmatic objects are of interest in areas as diverse as stellar astrophysics, galaxy evolution and cosmology.

I will also describe the recent ESO observations that allowed our team to identify and study the host galaxy of this very distant FRB and will outline some of the key astrophysical questions (both observational and theoretical) that remain unanswered.

12/10/22 (Wednesday)
10:45, Webinar | ESO Garching
Informal Discussion
Talk — The small-scale structure of the circum-galactic medium
Ramona Augustin (Space Telescope, US)

Abstract

Gas flows in and out of galaxies are typically probed by quasar absorption lines which are usually limited to a single sightline through the halo, giving no information on the structure of the gas probed in absorption. First studies using lensed, multiple or extended background objects have shown that there can be significant variation of absorber strengths on relatively small (~kpc and less) scales, hinting at an inhomogeneous clumpy circum-galactic medium. I am going to present our ongoing efforts of combining observational IFU data of lensed quasars with high spatial resolution cosmological simulations in order to study the small-scale structure of the CGM. I will show the results of our non-targeted "blind" search for intermediate absorbers and their metal line variations over kpc scales using MUSE data of a lensed quasar field. To learn more about the underlying physical structure of the gas probed in the CGM, I am using FOGGIE's highly resolved cosmological simulations to extract and analyse the clumps and filaments in galactic halos and I will report on the current state of our efforts to mock and interpret observables from these simulations.  

September 2022

28/09/22 (Wednesday)
10:45, Webinar | ESO Garching
Informal Discussion
Talk — Tracing the Magellanic periphery with Mira variables
Camila Navarrete (ESO)

Abstract

The Large and Small Magellanic Clouds provide the closest example of a galactic “three-body train wreck” in the local Universe. In fact, several stellar substructures, based on photometry and/or proper motions, have been uncovered in the outskirts of the Clouds, evidence of past interactions between the two dwarfs, and disturbances due to the Milky Way's tidal field. Mira variables are long-period pulsating red giant stars, bright enough to trace the Magellanic outskirts with precise Gaia eDR3 proper motions. In this talk, I will present a spectroscopic campaign to derive radial velocities for tens of Mira stars in the Magellanic periphery. On-sky position, distances from period-luminosity relations, and 3D velocities were used to elucidate the origin of these tracers and their observables. Using a suite of dynamical models for the past interactions of the Clouds, a recent disk-crossing of the SMC, largely disturbing the east side of the LMC, successfully reproduces the observed properties of the Mira stars. The implications of our findings will be discussed.

14/09/22 (Wednesday)
10:45, Auditorium Telescopium (ESO HQE, Garching) | ESO Garching
Informal Discussion
Talk — Asymmetric wind-driving in supergiant high-mass X-ray binaries
Ryosuke Hirai (Monash University)

Abstract

High-mass X-ray binaries (HMXBs) are systems consisting of a compact object (neutron star or black hole) with a relatively massive star. About a third of them are considered to be emitting X-rays by accreting part of the wind emitted from the donor star, and only a handful of them are known to host black holes. The amount of accretion is usually estimated by using the Bondi-Hoyle accretion model with an assumed velocity distribution and mass-loss rate for the wind. It is usually also assumed that the wind is spherically symmetric, which is true as long as the wind is faster than the orbital velocity. However, if this assumption holds, the wind flows symmetrically around the black hole and therefore there is insufficient angular momentum in the accreted material to form an accretion disk and emit X-rays. In this study, we model the wind acceleration process in close binary stars where the wind velocity can be slower or be comparable to the orbital velocity. We find that due to the tidal force from the companion and the associated deformation of the donor star, the wind can have a highly asymmetric velocity distribution. This strongly influences the amount of angular momentum that accretes onto the black hole, enabling accretion disk formation. There is a strong orbital separation dependence for this effect, and we predict that accretion disks cannot be formed when the Roche lobe filling factor of the donor is less than ~80%.

07/09/22 (Wednesday)
10:45, Fornax (ESO HQ, Garching) | ESO Garching
Informal Discussion
Talk — A new planet hunter awakens: NIRPS instrument is offered to the community
Celine Peroux & Louise Nielsen (ESO)

Abstract

The Near InfraRed Planet Searcher (NIRPS) instrument is offered to the community for the first time at the upcoming ESO Call for Proposals (27th of Sep 2022). NIRPS is a near-infrared, high-resolution spectrograph equipped with an adaptive optic system and optimised to reach a long-term stability of < 1 m/s. The mission of NIRPS is to search for new exoplanets around the Milky Way’s coolest stars. The instrument will focus its search on rocky worlds, which are key targets for understanding how planets form and evolve, and are the most likely planets where life may develop. NIRPS will search for exoplanets using the radial velocity method and characterise exoplanet atmospheres with transmission spectroscopy. This informal discussion aims at presenting the instrument capabilities, detailing its science goals and answering any questions you might have to help you shape your NIRPS science projects.

July 2022

27/07/22 (Wednesday)
10:45, Auditorium Eridanus (ESO HQE, Garching) | ESO Garching
Informal Discussion
Talk — "Protoplanet Express”, a video game based on hydrodynamical simulations
Jorge Cuadra (Universidad Adolfo Ibáñez)

Abstract

High-resolution observations from the last decade show that protoplanetary discs are very complex objects. Rather than uniform discs, we see that they have large cavities, spiral arms, rings, gaps, and asymmetries. To understand these features, astrophysicists many times have to develop numerical simulations, in which planets or unseen stellar companions are added to reproduce the observed morphologies. While the spectacular observed images are popular in mainstream media, the simulation side is rarely known by the general public.
We have created "Protoplanet express", a video game in which the user visits different known protoplanetary discs, which are rendered in 3D from actual hydrodynamical simulations. The user will encounter in the models the same features observed with telescopes, and their goal will be to find the reason for each of them. For example, the user completes a level once they find the planet causing an observable gap in the disc.
In this presentation I will show excerpts from the game, discuss its development, and upcoming release.

06/07/22 (Wednesday)
10:45, Auditorium Eridanus (ESO HQE, Garching) | ESO Garching
Informal Discussion
Talk — Comparing codes: a numerical (and human) adventure
Stephane Blondin (LAM & ESO Visitor)

Abstract

Code validation and verification is - or at least should be! - part of our daily routine as astronomers. Comparing our results to those of our colleagues is both valuable and necessary. But how often do we go the extra mile and attempt to not only replicate each other's results, but also using the same tools (simulation/analysis software) to obtain them?
In this informal discussion I will present the results of a supernova radiative-transfer code-comparison initiative (StaNdaRT - this won the acronym contest) bringing together 30 researchers and 10 numerical codes, starting with how (and where) it all began and ending with the submission of our first paper one month ago... and all the excitement and drama that happened in between! An enriching scientific and human adventure that I'd like to share with you, hoping to receive feedback/advice and hear from your own similar experiences.
 

June 2022

15/06/22 (Wednesday)
10:45, Auditorium Telescopium (ESO HQE, Garching) | ESO Garching
Informal Discussion
Talk — Evolution of massive stars with new hydrodynamic wind models
Alex Gormaz-Matamala (Universidad de Valparaiso, Chile)

Abstract

We present evolutionary models for a set of massive stars, introducing a new prescription for the mass-loss rate obtained from hydrodynamical calculations in which the wind velocity profile, $\varv(r)$, and the line-acceleration, $g_\text{​​​​​​​​line}​​​​​​​​$, are obtained in a self consistently way.

We found important differences between standard and our new self-consistent tracks.

Models with the new recipe for $\dot M$ retain more mass during their evolution, which is expressed in larger radii and consequently more luminous tracks over the Hertzsprung-Russell diagram.

Later increments in the mass-loss rate for tracks when self-consistency is no longer used, attributed to the LBV stage, produce different final stellar radii and masses before the end of H-burning stage, which are analysed case to case.

Moreover, we observed remarkable differences for the evolution of the radionuclide isotope $^{​​​​​​​​26}​​​​​​​​$Al in the core and the surface of the star.

Since $\dot M_\text{​​​​​​​​sc}​​​​​​​​$ are weaker than the commonly adopted values for evolutionary tracks, self-consistent tracks predict a later modification in the abundance number of $^{​​​​​​​​26}​​​​​​​​$Al in the stellar winds.

This new behaviour could provide useful information about the real contribution of this isotope from massive stars to the Galactic interstellar medium.

08/06/22 (Wednesday)
10:45, Auditorium Eridanus (ESO HQE, Garching) | ESO Garching
Informal Discussion
Talk — How stellar rotation shapes the main sequences of star clusters
Sebastian Kamann (LJMU, Liverpool)

Abstract

Star clusters with ages up to about 2 Gyr show peculiar features in their colour magnitude diagrams, such as split main sequences and extended turn-offs. These features appear to be fundamentally different from those observed in ancient globular clusters, as young star clusters lack the light-element abundance variations that are characteristic for globulars. Instead, stellar rotation has been suggested as the underlying cause, altering the colours of the stars via changes to their hydrostatic equilibria. While an observational confirmation of this idea had been lacking for many years, MUSE proved to be a game changer thanks to its unique capabilities of performing spectroscopy in crowded stellar fields. In my talk, I will present our efforts to measure the stellar rotation rates in a sample of massive star clusters in the Magellanic Clouds using MUSE. I will also discuss the implications for the formation of star clusters that arise from the presence of groups of stars with very different rotational velocities.

01/06/22 (Wednesday)
10:45, Auditorium Eridanus (ESO HQE, Garching) | ESO Garching
Informal Discussion
Talk — Emission-line diagnostics of HII regions using conditional Invertible Neural Networks
Da Eun Kang (ITA, Heidelberg)

Abstract

The stellar feedback is an energetic interaction between star clusters and the surrounding interstellar medium (ISM) that self-regulates the star formation in giant molecular clouds. Studying stellar feedback generally relies on observations of star-forming regions but inferring the physical properties from photometric and spectroscopic measurements is difficult because observational data are highly degenerate due to the complexity and non-linearity of stellar feedback. In this talk, I will introduce a novel method that couples a conditional invertible neural network (cINN) with the WARPFIELD-emission predictor (WARPFIELD-EMP) to estimate the physical properties of star-forming regions from spectral observations. In our first paper (Kang et al. 2022), we presented a cINN that predicts seven physical parameters (cloud mass, star formation efficiency, etc) from the luminosity of 12 optical emission lines ranging from 3700 to 9600Å. Testing and validating our network with synthetic models that are not used for training, we confirmed that our network is a time-efficient and powerful tool that provides accurate and precise posterior distributions for each parameter. However, degeneracy sometimes remains depending on the characteristics of the object to be analyzed. Since we have to take into account observational uncertainty in analyzing the real observation data, I will present two methods of considering observation errors in using the cINN: treating the error after training or training the network together with the error.

April 2022

13/04/22 (Wednesday)
10:45, Webinar | ESO Garching
Informal Discussion
Talk — The SKA Precursors and opening up of the low column density HI Universe
Sambit Roychowdhury (ICRAR, UWA & ESO Visitor)

Abstract

The low column density atomic hydrogen (HI) needs to be observationally traced in order to fully understand the accretion of gas onto galaxies, and the evolution of galaxies in general. And ideally this gas needs to be traced at ~kpc resolutions. Pushing the observed HI column density limit down has demanded considerable investment of time on existing radio interferometers, but with the Square Kilometre Array (SKA) Precursors coming online, we can soon expect to have a large samples of galaxies observed with high spatial resolution down to very low column densities. I would like to discuss how 'low' a HI column density can be realistically observed by these Precursors and the SKA in future, and what science we can expect from such observations. I will introduce the two relevant SKA Precursors, the relevant deep HI surveys on them, and hopefully also discuss some of the exciting new observations these surveys are starting to produce.

March 2022

23/03/22 (Wednesday)
15:00, Webinar | ESO Garching
Informal Discussion
Talk — What we talk about when we talk about refereeing
Megan Reiter & Alice Concas (ESO)

Abstract

The peer review process is a central part of publishing in journals, but a seldom discussed aspect of our work. In this discussion, we direct a curious eye toward all parts of the process. We have invited a panel of editors to provide their perspective. We ask that you, the audience, provide the perspective of authors and referees. To do so, please fill out our survey. We will present the survey results and hope to stimulate a discussion of where we as a field need to put our attention to improve the process. The survey is available here: https://forms.office.com/r/eDYfzsJeFG

16/03/22 (Wednesday)
10:45, Webinar | ESO Garching
Informal Discussion
Talk — RECA - The Network of Colombian Astronomy Students
Sofia Rojas (MPIA Heidelberg)

Abstract

I want to introduce RECA (https://www.astroreca.org) and the many pioneer programs we are building for astronomy education in Colombia. We are a group of astronomy students and junior researchers who want to connect the new generations of students from primary school to undergraduates with global opportunities in astronomy. I would like to discuss our RECA programs and how many of you at ESO could help us grow the network with your support and advice.

 

09/03/22 (Wednesday)
10:45, Webinar | ESO Garching
Informal Discussion
Talk — Putting brakes on star formation: evidence for the integrated role of historic AGN feedback
Joanna Piotrowska (University of Cambridge)

Abstract

Understanding the physical processes responsible for ceasing star formation in galaxies is one of the most important unresolved questions in the field of galaxy evolution. Over the past two decades multiple mechanisms were suggested as potential drivers of the transition between the star-forming and quiescent galaxy categories, referred to as galaxy ‘quenching’. In this talk I will present the results of our recent study, in which we combine machine learning with partial correlation analysis to determine which among the three potential quenching mechanisms: supernova feedback, halo shock heating or AGN feedback are most likely responsible for bringing star formation to a halt in massive, central galaxies. To this end we bridge the gap between theory and observation by extracting theoretical predictions from three state-of-the-art cosmological simulations – EAGLE, Illustris and IllustrisTNG and comparing them with the Sloan Digital Sky Survey (SDSS) observations. We find that the supermassive black hole mass (MBH) is the most powerful parameter in determining whether a galaxy is star-forming or quenched across all datasets. Remarkably, this result is true for all different implementations of AGN feedback in the simulations and is met overwhelmingly well in the SDSS, where we infer MBH from a variety of calibrations for ~230 000 local galaxies. In my brief talk I will share our results together with our methodology to make a convincing case for star formation being quenched by AGN in massive, central galaxies. If you would like to learn more about this study, you can now read it on arXiv/MNRAS at the following link: https://arxiv.org/abs/2112.07672

02/03/22 (Wednesday)
10:45, Webinar | ESO Garching
Informal Discussion
Talk — Sonification and Sound Design for Research, Communication and Education of Astronomy
Chris Harrison (University of Newcastle)

Abstract

The idea of turning data into sound is not new; however, over the last decade there has been a dramatic increase in the number of projects using sound for astronomy research, communication and education. I will present a brief overview of the 98 such projects we identified (as of December 2021) that form the basis of our recently submitted review article. Through some examples, I will discuss the large potential benefits of using sound to both enhance scientific discovery and to make astronomy more accessible and engaging for a wide variety of audiences. On the flip-side, I will present some of the challenges and limitations that have limited the progress in using sound within the mainstream astronomy research and communication communities. At the present time North America and Australia are leading the way in resolving these issues and are embracing these new sound-based approaches. Therefore, during this session I welcome a discussion on how ESO and the European astronomy community might make more progress in this arena and take the lead in unleashing the huge potential of sonification and sound design.

February 2022

23/02/22 (Wednesday)
10:45, Webinar | ESO Garching
Informal Discussion
Talk — What I would like to know...can interferometry help me find out?
Christian Hummel (ESO)

Abstract

Over the last few periods, VLTI operations have adapted better to science cases requiring imaging or monitoring of variable targets, and the European network of expertise centers offers help with proposal preparation and even full data reduction and calibration.  The astrometric mode of GRAVITY and mid-infrared imaging with MATISSE have attracted more astronomers who recognize the unique capabilities these instruments have for their science projects. However, preparing a proposal including the description of astrophysical constraints to be delivered by the interferometric observations is challenging for newcomers, and I my presentation I will highlight the use of interferometry for some of these science cases.

16/02/22 (Wednesday)
10:45, Webinar | ESO Garching
Informal Discussion
Talk — ALMA Cycle 8 proposal selection using double-anonymous distributed peer review.
Enrique Macias, Martin Zwaan (ESO)

Abstract

For its Cycle 8, ALMA implemented two major changes in the proposal selection procedure. Dual-anonymous review in which the identity of the proposers was not revealed to the reviewers was introduced, and at the same time distributed peer review was adopted to award the majority of the observing time. Nearly 1500 proposals were assessed using this method, making it the largest implementation of distributed peer review in astronomy to date. In this Informal Discussion, we will give an overview of the ALMA proposal selection process, discuss some effects seen in e.g. the change of the amount of time requested per proposal, describe how the distributed peer review works in practice and highlight some interesting trends, correlations, and feedback from PIs.

09/02/22 (Wednesday)
10:45, Webinar | ESO Garching
Informal Discussion
Talk — Infrared Interferometric imaging with the VLTI
Joel Sánchez Bermúdez (Instituto de Astronomía, Universidad Nacional Autónoma de México)

Abstract

In Astronomy, interferometry is an observational technique that delivers us the major resolution possible to study physical processes at the smallest spatial scales that we can probe with our instruments. It is used extensively at radio wavelengths and, since more than a decade, it has been converted into an important technique for infrared Astronomy. However, in the infrared interferometry is restricted to the use of arrays with only a few telescopes. Meanwhile interferometric imaging is the most intuitive way to understand the data, recovering images from sparse arrays is an “ill-posed” problem, with less data (equations) than pixels (unknowns) to recover. In this talk, we will present the basic concepts for image infrared image reconstruction. In particular, we will present some examples of imaging cases observed with the near-infrared instrument GRAVITY at the Very Large Telescope Interferometer. Finally, we will describe the areas of opportunity to improve interferometric imaging.

02/02/22 (Wednesday)
15:00, Webinar | ESO Garching
Informal Discussion
Talk — The International Staff Association and its Committee – what is it all about?
Tania Johnston & Domingo Álvarez Méndez (ESO)

Abstract

The International Staff Association exists to defend the interests and rights of its members, and those members are you! Come along to hear how the ISC aims to fulfil the aims of the association and for a chance to meet some of the current committee representatives. We welcome your questions and look forward to an open discussion about how the ISA can help you.

January 2022

26/01/22 (Wednesday)
10:45, Webinar | ESO Garching
Informal Discussion
Talk — TULIPS: a Tool for Understanding the Lives, Interiors, and Physics of Stars
Eva Laplace (HITS)

Abstract

Understanding the lives and interior structures of stellar objects is a fundamental objective of astrophysics. Research in this domain often relies on the visualization of astrophysical data, for instance the results of theoretical simulations. However, the diagrams commonly employed to this effect are usually static, complex, and can sometimes be non-intuitive or even counter-intuitive to newcomers in the field.

In this informal talk I will introduce the python package TULIPS, which enables interactive visualization of stellar models computed with the MESA code. The audience will have the opportunity to try out TULIPS themselves through a hands-on interactive tutorial that makes use of jupyter notebooks. I will discuss applications of TULIPS for research, higher education, and outreach.

To try out TULIPS, please make sure to download the following:

19/01/22 (Wednesday)
10:45, Webinar | ESO Garching
Informal Discussion
Talk — A Novel Approach for Fast Streaks Detection in Astronomical Images
Vojtech Cvrcek (ESO)

Abstract

Streaks in astronomical images can either be an undesired obstacle for the analysis or the target of some observations. An example of the former are satellite tails in deep sky images, an example of the latter are near solar system objects. It is crucial to find them automatically in large surveys in both cases. We present a novel streak pattern representation called steerable representation, allowing noticeably faster processing time at a higher memory cost. Steerable representation compresses a template into a vector of size k, where k is negligible wrt the size of the represented template. Such representation allows to (1) compute an arbitrarily oriented streak response in O(k), (2) compute an arbitrarily oriented streak response for a set of k lengths in amortized O(1). Moreover, the steerable representation allows computing additional (steerable) features together with streak (ridge) responses (such as ridge-ends, rotation-invariant upper-bounds, edges, and ridge derivations by angle). Results show that we can compute response approximately 30 folds faster than without steerable representation, with a detection quality comparable to the state-of-the-art approach.