Unveiling the nature of the Unassociated Gamma-ray Sources in the fourth Fermi catalog: a multi-wavelength approach

Fifteen years have passed since the launch of the Fermi Gamma-ray Space Telescope, a spaceobservatory designed to study celestial objects emitting in the 𝛾-ray energy range. Equippedwith two main instruments, the Large Area Telescope (LAT) and the Gamma-ray Burst Monitor(GBM), Fermi has transformed our understanding of the high-energy universe.Fermi’s all-sky monitoring capabilities have provided an unprecedented view of the universein 𝛾-rays, detecting a wide variety of sources such as active galactic nuclei (AGNs), pulsars,supernova remnants, and 𝛾-ray bursts with exceptional precision. Its observations have unveiledthe dynamic and energetic processes driving these objects, offering new insights into extremeastrophysical phenomena and shedding light on some of the most energetic events in the cosmos.By significantly improving sensitivity and spatial resolution compared to its predecessors, Fermihas uncovered thousands of previously unknown 𝛾-ray sources, enriching our understanding ofhigh-energy astrophysics.One of the key achievements of the Fermi mission has been the development of comprehensive𝛾-ray source catalogs, essential resources for the high-energy astrophysics community. Thesecatalogs list and characterize the 𝛾-ray sources detected by the LAT, providing critical data ontheir locations, energy spectra, and variability. Over the years, several iterations of these catalogshave been released, each incorporating new data and refined analysis methods.The most recent and extensive catalog is the Fourth Fermi-LAT Gamma-ray Source Catalog(4FGL), first released in 2019. In 2023, the Fermi collaboration published an incrementalupdate to this catalog, referred to as 4FGL-DR4. This version reports sources detected over14 years of LAT observations, cataloging a total of 7195 objects. Of these, 4765 have beenassociated or identified at other wavelengths through positional overlap, correlated variability, ormulti-wavelength spectral properties. Among the associated and identified sources, the majorityare blazars—a subclass of AGN with relativistic jets pointing at an angle of 𝜃 ≤ 10◦ with respectto the observer’s line of sight.However, a significant portion of the Fermi catalog, approximately 30% of sources, remainsunclassified and unassociated with lower-energy counterparts, collectively referred to as unassoci-ated gamma-ray sources (UGSs). UGSs play a key role in the high-energy sky, potentially hidingnew blazars or other types of AGNs. Identifying and classifying these sources can enhance our understanding of the most extreme environments in the universe, such as those involvingsupermassive black holes, and may reveal missing components in our current models of thehigh-energy universe. UGSs are typically faint, with lower 𝛾-ray fluxes. On average, UGSs havefluxes ∼ 5.3 × 10−12 erg cm−2 s−1, compared to ∼ 1.6 × 10−11 erg cm−2 s−1 for associated AGNin the 100 MeV to 100 GeV range), suggesting that they may represent a higher redshift AGNpopulation and/or lower luminosity sources. Identifying UGSs is crucial for population studies,developing physical models, understanding the cosmic evolution of 𝛾-ray sources and estimatingthe very high-energy (VHE) cosmic background.The goal of this thesis is to associate and classify extragalactic UGSs from the fourth Fermicatalog, with the aim of unveiling new blazars and, more broadly, new AGNs. To achieve this,the analysis focuses on the 1284 UGSs located outside the Galactic plane (|𝑏| ≥ 10◦), searchingfor X-ray, optical, and radio counterparts that spatially coincide with the 𝛾-ray positions. Giventhe large uncertainty regions in 𝛾-rays, the search begins in the X-ray band, where the potentialcounterpart confusion is limited compared to lower energy bands such as optical and radio, andwhere positional uncertainties are more constrained relative to 𝛾-rays. This strategy allows for amore focused search region.In the past decade, the Swift X-ray satellite has been conducting a dedicated observationalcampaign targeting Fermi’s unassociated sources, providing real-time access to its data. Thisthesis begins by analyzing all the available Swift X-ray images, which cover the 𝛾-ray positions ofthe UGSs. The number of UGSs with Swift observations is 714. From this analysis, I found that274 of these 𝛾-ray emitters have at least one X-ray source detected with a significance of ≥ 3𝜎within their Fermi error box. Among these, 193 UGSs have a single potential X-ray counterpart(referred to as UGS1), while 81 have multiple potential X-ray counterparts within the Fermi errorbox (referred to as UGS2). Of the UGS2, 54 have two X-ray counterparts, 11 have three, and theremaining 16 have more than three.Starting from the X-ray positions and error boxes, optical and radio counterparts are searchedwithin several catalogs or by dedicated observations. I found that each UGS1 has a potentialoptical counterpart, and 113 also could be associated to a radio counterpart. Regarding theUGS2 sub-sample, I found a large degeneracy of potential counterparts, which complicates theassociation process. I separated the sources in the sample in radio-loud and radio-quiet sources(accordingly to the historical threshold of the radio-loudness ratio, placed to R=10).To assess the reliability of the associations, I compared the absolute and relative fluxes ofthe potential counterparts with those of blazars from the fourth Fermi catalog. The comparisonrevealed that radio-loud sources exhibit emission profiles similar to those of blazars, makingthem strong candidates for counterparts. In contrast, radio-quiet sources tend to deviate fromtypical blazar behavior. Given that only radio-loud AGNs are listed in the Fermi catalog, a morein-depth analysis is necessary to confirm these associations.I found that 33 UGS1 have optical spectra already available in the literature. The analysis ofoptical spectra is a fundamental step in determining the extragalactic origin of UGSs, estimatingtheir distance, and classifying them among the various AGN sub-classes. From the analysis ofthese 33 UGSs, I found that the X-ray emitter located within their Fermi error box is a BL Lac (a subclass of blazars with an optical spectrum described by a featureless power law or weakemission lines) for 21 of them, a Flat Spectrum Radio Quasar (FSRQ, a subclass of blazarsdistinguished from BL Lacs by its strong and broad optical emission lines) for one source, aradio galaxy (a subclass of AGN characterized by extended radio emission) for one source, and aSeyfert or Quasi Stellar Object (subclasses of AGN with strong emission lines in the optical butweak or absent radio emission) for 10 of them. The presence of radio-quiet AGNs in the Fermicatalog is unexpected and potentially interesting, as all AGNs listed in the fourth Fermi catalogso far are radio-loud.The 33 UGSs associated and classified in this thesis, along with a sample of 44 UGSs fromthe second and third Fermi catalogs, which were associated and classified by Paiano et al. (2017a,2019), are characterized through their multi-wavelength emission. Specifically, I analyzed thespectral energy distributions (SEDs) of these 77 𝛾-ray emitters to study their properties, while alsoexamining the emission intensities of the BL Lacs in the sample to identify new masquerading BLLacs (FSRQs with featureless optical spectra typical of BL Lacs, due to their strong continuummasking the spectral lines).The SED analysis leads to the conclusion that radio-loud AGNs in the sample (64 objectson a total of 77) are well-associated, as they exhibit continuity between the emission of thecounterpart and the 𝛾-ray emitter. These objects are then classified into the blazar subclasses lowsynchrotron peak (LSP), intermediate synchrotron peak (ISP), and high synchrotron peak (HSP),based on the position of the synchrotron emission peak in their SEDs. The majority are classifiedas HSP (46 objects), with 11 as ISP and 7 as LSP.Furthermore the multi-wavelength analysis led to the identification of 9 candidates mas-querading BL Lacs, representing approximately 15% (possibly ∼ 30%) of the radio-loud sample,which is somewhat smaller than the value of ≳ 34% found in the previous studies.Regarding the 13 radio-quiet sources in the sample, I found that they exhibit behavior thatdiverges from that of jetted-AGNs. Many of them do not show radio emission, and their SEDsdeviate from the typical double peak trend, indicating the absence of a jet. This makes reconcilingthe 𝛾-ray emission with lower-energy emissions challenging. Since 𝛾-ray emission requiresnon-thermal processes, and having excluded the jet as the source of this emission, I wonderedabout the possibility of having a 𝛾-ray emission from star formation processes. This analysisyields negative results. In the absence of a jet and star formation processes, linking the 𝛾-rayemission to lower-energy emission is complicated.For this reason I searched for alternative counterparts, looking for radio sources that arepositionally coincident with the 𝛾-ray emitters. From this search, I find a possible alternativecounterpart for 7 out of 13 UGSs, which shows a jet and continuity across the entire SED,described by a typical double peak trend of jetted AGNs. These sources will be the subject of afuture observational campaign, which will allow for a definitive classification and verificationof the associations. No valid alternative counterparts are found for the remaining 6 sources.Therefore, these sources require a more in-depth analysis regarding the origin of the 𝛾-rayemission.