ORCID Profile
0000-0002-5826-0548
Current Organisations
University of Birmingham
,
Georg-August-Universität Göttingen
Does something not look right? The information on this page has been harvested from data sources that may not be up to date. We continue to work with information providers to improve coverage and quality. To report an issue, use the Feedback Form.
Publisher: American Astronomical Society
Date: 09-01-2019
Publisher: Oxford University Press (OUP)
Date: 25-07-2014
Publisher: Oxford University Press (OUP)
Date: 02-11-2022
Abstract: Developing an effective automatic classifier to separate genuine sources from artifacts is essential for transient follow-ups in wide-field optical surveys. The identification of transient detections from the subtraction artifacts after the image differencing process is a key step in such classifiers, known as real-bogus classification problem. We apply a self-supervised machine learning model, the deep-embedded self-organizing map (DESOM) to this ‘real-bogus’ classification problem. DESOM combines an autoencoder and a self-organizing map to perform clustering in order to distinguish between real and bogus detections, based on their dimensionality-reduced representations. We use 32 × 32 normalized detection thumbnails as the input of DESOM. We demonstrate different model training approaches, and find that our best DESOM classifier shows a missed detection rate of $6.6{{\\ \\rm per\\,cent}}$ with a false-positive rate of $1.5{{\\ \\rm per\\,cent}}$. DESOM offers a more nuanced way to fine-tune the decision boundary identifying likely real detections when used in combination with other types of classifiers, e.g. built on neural networks or decision trees. We also discuss other potential usages of DESOM and its limitations.
Publisher: Oxford University Press (OUP)
Date: 07-03-2013
DOI: 10.1093/MNRAS/STT293
Publisher: American Astronomical Society
Date: 03-2023
Abstract: We present James Webb Space Telescope (JWST) and Hubble Space Telescope (HST) observations of the afterglow of GRB 221009A, the brightest gamma-ray burst (GRB) ever observed. This includes the first mid-IR spectra of any GRB, obtained with JWST/Near Infrared Spectrograph (0.6–5.5 micron) and Mid-Infrared Instrument (5–12 micron), 12 days after the burst. Assuming that the intrinsic spectral slope is a single power law, with F ν ∝ ν − β , we obtain β ≈ 0.35, modified by substantial dust extinction with A V = 4.9. This suggests extinction above the notional Galactic value, possibly due to patchy extinction within the Milky Way or dust in the GRB host galaxy. It further implies that the X-ray and optical/IR regimes are not on the same segment of the synchrotron spectrum of the afterglow. If the cooling break lies between the X-ray and optical/IR, then the temporal decay rates would only match a post-jet-break model, with electron index p 2, and with the jet expanding into a uniform ISM medium. The shape of the JWST spectrum is near-identical in the optical/near-IR to X-SHOOTER spectroscopy obtained at 0.5 days and to later time observations with HST. The lack of spectral evolution suggests that any accompanying supernova (SN) is either substantially fainter or bluer than SN 1998bw, the proto-type GRB-SN. Our HST observations also reveal a disk-like host galaxy, viewed close to edge-on, that further complicates the isolation of any SN component. The host galaxy appears rather typical among long-GRB hosts and suggests that the extreme properties of GRB 221009A are not directly tied to its galaxy-scale environment.
Publisher: Springer Science and Business Media LLC
Date: 22-06-2023
Publisher: Oxford University Press (OUP)
Date: 09-10-2020
Abstract: The amount of observational data produced by time-domain astronomy is exponentially increasing. Human inspection alone is not an effective way to identify genuine transients from the data. An automatic real-bogus classifier is needed and machine learning techniques are commonly used to achieve this goal. Building a training set with a sufficiently large number of verified transients is challenging, due to the requirement of human verification. We present an approach for creating a training set by using all detections in the science images to be the s le of real detections and all detections in the difference images, which are generated by the process of difference imaging to detect transients, to be the s les of bogus detections. This strategy effectively minimizes the labour involved in the data labelling for supervised machine learning methods. We demonstrate the utility of the training set by using it to train several classifiers utilizing as the feature representation the normalized pixel values in 21 × 21 pixel st s centred at the detection position, observed with the Gravitational-wave Optical Transient Observer (GOTO) prototype. The real-bogus classifier trained with this strategy can provide up to $95{{\\ \\rm per\\ cent}}$ prediction accuracy on the real detections at a false alarm rate of $1{{\\ \\rm per\\ cent}}$.
Publisher: Research Square Platform LLC
Date: 22-11-2022
DOI: 10.21203/RS.3.RS-2298504/V1
Abstract: The majority of long-duration ( s) gamma-ray bursts (GRBs) are believed to arise from the collapse of massive stars (Hjorth et al. 2003) with a small proportion created from the merger of compact objects (Rastinejad et al. 2022, Yang et al. 2022, Troja et al. 2022). Most of these systems are likely formed via standard stellar evolution pathways. However, it has long been thought that a fraction of GRBs may instead be an outcome of dynamical interactions in dense environments (Grindlay et al. 2006, Fragione et al. 2019, McKernan et al. 2020), channels which could also contribute significantly to the s les of compact object mergers detected as gravitational wave sources (O’Leary et al. 2016). Here we report the case of GRB 191019A, a long GRB (T_90 = 64.4 +/- 4.5 s) which we pinpoint close ( 100 pc projected) to the nucleus of an ancient ( Gyr old) host galaxy at z=0.248. The lack of evidence for star formation and deep limits on any supernova emission makes a massive star origin difficult to reconcile with observations, while the timescales of the emission rule out direct interaction with the supermassive black hole in the nucleus of the galaxy, We suggest that the most likely route for progenitor formation is via dynamical interactions in the dense nucleus of the host, consistent with the centres of such galaxies exhibiting interaction rates up to two orders of magnitude larger than typical field galaxies (French et al. 2014, Stone et al. 2016). The burst properties could naturally be explained via compact object mergers involving white dwarfs (WD), neutron stars (NS), or black holes (BH). These may form dynamically in dense stellar clusters, or originate in a gaseous disc around the supermassive black hole (Perna et al. 2021, Lazzati et al. 2022). Future electromagnetic and gravitational-wave observations in tandem thus offer a route to probe the dynamical fraction and the details of dynamical interactions in galactic nuclei and other high-density stellar systems.
Publisher: American Astronomical Society
Date: 16-10-2017
Publisher: Springer Science and Business Media LLC
Date: 30-11-2022
Publisher: Research Square Platform LLC
Date: 24-07-2023
DOI: 10.21203/RS.3.RS-3135743/V1
Abstract: The mergers of binary compact objects such as neutron stars and black holes are of central interest to several areas of astrophysics, including as the progenitors of gamma-ray bursts (GRBs), sources of high-frequency gravitational waves and likely production sites for heavy element nucleosynthesis via rapid neutron capture (the r-process). These heavy elements include some of great geophysical, biological and cultural importance, such as thorium, iodine, and gold. Here we present observations of the exceptionally bright gamma-ray burst GRB230307A. We show that GRB 230307A belongs to the class of long-duration gamma-ray bursts associated with compact object mergers, and contains a kilonova similar to AT2017gfo, associated with the gravitational-wave merger GW170817. We obtained James Webb Space Telescope mid-infrared (mid-IR) imaging and spectroscopy 29 & 61 days after the burst. The spectroscopy shows an emission line at 2.1 microns which we interpret as tellurium (atomic mass A=130), and a very red source, emitting most of its light in the mid-IR due to the production of lanthanides. These observations demonstrate that nucleosynthesis in GRBs can create r-process elements across a broad atomic mass range and play a central role in heavy element nucleosynthesis across the Universe.
Publisher: Springer Science and Business Media LLC
Date: 02-07-2018
Publisher: Springer Science and Business Media LLC
Date: 25-10-2023
Location: United Kingdom of Great Britain and Northern Ireland
Location: United Kingdom of Great Britain and Northern Ireland
Location: United Kingdom of Great Britain and Northern Ireland
Start Date: 2017
End Date: End date not available
Funder: Landwirtschaftliche Rentenbank
View Funded Activity