Deep sub-arcsecond wide-field imaging of the Lockman Hole field at 144 MHz

Sweijen, F.; Weeren, R. J. van; Röttgering, H. J. A.; Morabito, L. K.; Jackson, Neal; Offringa, A. R.; Tol, S. van der; Veenboer, Bram; Oonk, J. B. R.; Best, P. N.; Bondì, M.; Shimwell, T. W.; Tasse, C.; Thomson, A. P.

doi:10.1038/s41550-021-01573-z

Cited by 27 publications

(29 citation statements)

References 27 publications

(23 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Overall, our results are encouraging for finding new galaxy-scale gravitational lenses with the ILT, and in the short term we will apply our best trained model to the first tranche of data from LoTSS that includes the ILT baselines (e.g. Morabito et al 2022;Sweĳen et al 2022a) and has detected over 2500 radio sources at an angular resolution of around 350 mas. Given the expected probability for galaxy-scale gravitational lensing, there should be between 2 and 4 systems within such a dataset, which we are now focusing on finding.…”

Section: Figure 17mentioning

confidence: 67%

“…This will likely produce more conservative, but less biased results. The sample of non-lensed events includes 15385 compact radio sources and 2000 double-lobed radio sources, which ensures that the ratio of compact-to-extended radio sources is similar to that observed with the ILT (Sweĳen et al 2022a). The total number of test samples was chosen so that around 15 to 25 FP events would be returned by each of the three network structures.…”

Section: Results For the Final Network Testmentioning

confidence: 99%

“…This distribution of source sizes is based in part on the typical sizes of radio emitting star-forming galaxies at 1.5 GHz (Muxlow et al 2020), but as most of our sources will likely be associated with AGN activity, which can have a variety of sizes, we have also added a distribution of sources with sizes < 0.5 arcsec. This upper-limit on the AGN source size was chosen to be consistent with recent ILT observations of the Lockman Hole region, where 88 per cent of the detected sources were found to be compact at 144 MHz (beam size 0.4 arcsec; rms 25 𝜇Jy beam −1 ; Sweĳen et al 2022a). As gravitational lensing has only rarely been observed when the object has large-scale radio lobes (Haarsma et al 2005), we initially chose not to densely sample source sizes larger than 2.5 arcsec (see Section 3.4, where we sample non-lensed sources that are extended between 0.5 and 6 arcsec).…”

Section: Simulating a Training Dataset For The Iltmentioning

confidence: 96%

“…The position relative to the lens was chosen so that at least one of the radio components of the background source would likely be strongly lensed, that is, produce multiple images. Note that the maximum distance between the radio components was chosen considering the properties of radio sources observed by the ILT at 144 MHz (Sweĳen et al 2022a). However, as stated above, we also tested larger separations between two radio components for the non-lensed samples to test the reliability of our approach to distinguish between double-lobed radio sources and gravitational lenses that produce two lensed images.…”

Section: Simulating a Training Dataset For The Iltmentioning

confidence: 99%

“…In particular, we concentrate on lens surveys with the Low Frequency Array (LOFAR; van Haarlem et al 2013), which mainly operates between 120 and 170 MHz, and has the sensitivity to detect the non-thermal emission from lensed radio sources (Stacey et al 2019) and the long baselines needed to resolve their structure (Badole et al 2022). Also, given that the International LOFAR Telescope (ILT) has now gone through the commissioning phase (Morabito et al 2022;Jackson et al 2022;Bonnassieux et al 2022;Harwood et al 2022;Sweĳen et al 2022b;Timmerman et al 2022), the routine imaging of the survey data with the international stations of LOFAR will soon start (Sweĳen et al 2022a). Therefore, developing methods for identifying lensed radio galaxies amongst the expected 15 million objects to be imaged with LOFAR is also timely.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

A machine learning based approach to gravitational lens identification with the International LOFAR Telescope

S.¹,

P.²,

Biehl³

et al. 2022

Preprint

View full text Add to dashboard Cite

We present a novel machine learning based approach for detecting galaxy-scale gravitational lenses from interferometric data, specifically those taken with the International LOFAR Telescope (ILT), which is observing the northern radio sky at a frequency of 150 MHz, an angular resolution of 350 mas and a sensitivity of 90 µJy beam −1 (1𝜎). We develop and test several Convolutional Neural Networks to determine the probability and uncertainty of a given sample being classified as a lensed or non-lensed event.By training and testing on a simulated interferometric imaging data set that includes realistic lensed and non-lensed radio sources, we find that it is possible to recover 95.3 per cent of the lensed samples (true positive rate), with a contamination of just 0.008 per cent from non-lensed samples (false positive rate). Taking the expected lensing probability into account results in a predicted sample purity for lensed events of 92.2 per cent. We find that the network structure is most robust when the maximum image separation between the lensed images is ≥ 3 times the synthesized beam size, and the lensed images have a total flux density that is equivalent to at least a 20𝜎 (point-source) detection. For the ILT, this corresponds to a lens sample with Einstein radii ≥ 0.5 arcsec and a radio source population with 150 MHz flux densities ≥ 2 mJy. By applying these criteria and our lens detection algorithm we expect to discover the vast majority of galaxy-scale gravitational lens systems contained within the LOFAR Two Metre Sky Survey.

show abstract

Section: Figure 17mentioning

confidence: 67%

Section: Results For the Final Network Testmentioning

confidence: 99%

Section: Simulating a Training Dataset For The Iltmentioning

confidence: 96%

Section: Simulating a Training Dataset For The Iltmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A machine learning based approach to gravitational lens identification with the International LOFAR Telescope

S.¹,

P.²,

Biehl³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

The energy consumption and carbon footprint of the LOFAR telescope

Kruithof,

Bassa,

Bonati

et al. 2023

Exp Astron

View full text Add to dashboard Cite

The LOw Frequency ARray (LOFAR) is a European radio telescope operating since 2010 in the frequency bands 10 - 80 MHz and 110 - 250 MHz. This article provides an analysis of the energy consumption and the carbon footprint of LOFAR. The approach used is a Life Cycle Analysis (LCA). We find that one year of LOFAR operations requires 3,627 MWh of electricity, 48,714 m3 gas and 135,497 liters of fuel. The associated carbon emission is 1,867 tCO2e/year. Results include the footprint stemming from operations of all LOFAR stations and central processing, but exclude scientific post-processing and activities. The electrical energy required for scientific processing is assessed separately. It ranges from 1% (standard imaging and time-domain), to 40% (wide field long baseline imaging) of the energy consumption for the observation. The outcome provides a transparent baseline in making LOFAR more sustainable and can serve as a blueprint for the analysis of other research infrastructures.

show abstract