X-Ray Spectra and Multiwavelength Machine Learning Classification for Likely Counterparts to Fermi 3FGL Unassociated Sources

Abstract

We conduct X-ray spectral fits on 184 likely counterparts to Fermi-LAT 3FGL unassociated sources. Characterization and classification of these sources allows for more complete population studies of the high-energy sky. Most of these X-ray spectra are well fit by an absorbed power-law model, as expected for a population dominated by blazars and pulsars. A small subset of seven X-ray sources have spectra unlike the power law expected from a blazar or pulsar and may be linked to coincident stars or background emission. We develop a multiwavelength machine learning classifier to categorize unassociated sources into pulsars and blazars using gamma-ray and X-ray observations. Training a random forest (RF) procedure with known pulsars and blazars, we achieve a cross-validated classification accuracy of 98.6%. Applying the RF routine to the unassociated sources returned 126 likely blazar candidates (defined as P bzr ≥ 90%) and five likely pulsar candidates (P bzr ≤ 10%). Our new X-ray spectral analysis does not drastically alter the RF classifications of these sources compared to previous works, but it builds a more robust classification scheme and highlights the importance of X-ray spectral fitting. Our procedure can be further expanded with UV, visual, or radio spectral parameters or by measuring flux variability.

4FGL J1120.0-2204: A Unique Gamma-ray Bright Neutron Star Binary with an Extremely Low Mass Proto-White Dwarf

“…Three of the X-ray sources they analyzed had plausible optical counterparts, and two of these are present within the 4FGL 95% confidence ellipse (see Figure 1; we note these are the same two sources identified by Hui et al (2015) from the Swift data). As pointed out by Salvetti et al (2017), the fainter of these two X-ray sources is a known galaxy, and is unlikely to be the source of the γ-ray emission, especially given the highly curved γ-ray spectrum; Kerby et al (2021) estimate the probability that this 4FGL source is a blazar is only ∼3%, supporting the inference that the fainter X-ray source is unrelated to the γ-ray emission.…”

“…No spectral or temporal analysis of these data was carried out by this group. Kerby et al (2021) performed a more detailed spectral analysis of this source, finding a best fit power law index of Γ = 1.89 +0.18 −0.17 and unabsorbed 0.3-10 keV flux of (7.9 +1.2 −1.0 ) × 10 −14 erg cm -2 s -1 . It is notable that despite using the same datasets, these two flux estimates do not agree within the uncertainties.…”

“…The γ-ray spectrum is highly curved, with a curvature significance of 16.0σ for a LogParabola model (16.3σ for a cut-off power law), a common feature among Fermi -detected MSPs (Abdo et al 2010). In fact, a number of independent groups have classified this Fermi source as a highly-likely MSP based on machine learning classifications using the γ-ray data (Saz Parkinson et al 2016;Kaur et al 2019;Finke et al 2020;Kerby et al 2021;Bhat & Malyshev 2021).…”

Probing the origin of 4FGL J0822.8-4207: cosmic ray illumination from the SNR Puppis A and the Herbig-Haro object HH219

“…To investigate HH219 in more detail as a potential source of the gamma-ray emission from 4FGL J0822.8-4207, we submitted a Target of Opportunity request to the Niel Gehrels Swift Observatory (henceforth Swift). Observations with Swift are a powerful tool for localizing high-energy emission from Fermi-LAT sources by using the Swift X-ray Telescope (Swift-XRT, Burrows et al 2005) to observe 0.3−10.0 keV X-rays (for example, the ongoing Swift follow-up program described in Kerby et al 2021). As many sources of gamma rays observed with Fermi also produce X-rays via synchrotron emission, if HH219 is producing gamma-ray emission by leptonic processes, it may produce characteristic X-rays as well.…”

Multiwavelength Spectral Analysis and Neural Network Classification of Counterparts to 4FGL Unassociated Sources

“…From their initial list, 31 sources from the 3FGL unassociated list defied categorization and were labeled 'ambiguous'. Recent work by this collaboration (Kerby et al 2021) continued that analysis by including more detailed X-ray spectral fitting and systematically adding spectral parameters to machine learning. Adding UV-visual observations from the Swift-UVOT telescope is particularly useful, as pulsars are usually extremely dim in the UV-visual range Saz Parkinson et al 2016) while blazars emit at all wavelengths and at low redshift can be observed across the electromagnetic spectrum (Ghisellini & Tavecchio 2008).…”