3HWC: The Third HAWC Catalog of Very-High-Energy Gamma-ray Sources
A. Albert, R. Alfaro, C. Alvarez, J.R. Angeles Camacho, J.C. Arteaga-Velázquez, K.P. Arunbabu, D. Avila Rojas, H.A. Ayala Solares, V. Baghmanyan, E. Belmont-Moreno, S.Y. BenZvi, C. Brisbois, K.S. Caballero-Mora, T. Capistrán, A. Carramiñana, S. Casanova, U. Cotti, S. Coutiño de León, E. De la Fuente, R. Diaz Hernandez, L. Diaz-Cruz, B.L. Dingus, M.A. DuVernois, M. Durocher, J.C. Díaz-Vélez, R.W. Ellsworth, K. Engel, C. Espinoza, K.L. Fan, K. Fang, M. Fernández Alonso, H. Fleischhack, N. Fraija, A. Galván-Gámez, D. Garcia, J.A. García-González, F. Garfias, G. Giacinti, M.M. González, J.A. Goodman, J.P. Harding, S. Hernandez, J. Hinton, B. Hona, D. Huang, F. Hueyotl-Zahuantitla, P. Hüntemeyer, A. Iriarte, A. Jardin-Blicq, V. Joshi, D. Kieda, A. Lara, W.H. Lee, H. León Vargas, J.T. Linnemann, A.L. Longinotti, G. Luis-Raya, J. Lundeen, R. López-Coto, K. Malone, V. Marandon, O. Martinez, I. Martinez-Castellanos, J. Martínez-Castro, J.A. Matthews, P. Miranda-Romagnoli, J.A. Morales-Soto, E. Moreno, M. Mostafá, A. Nayerhoda, L. Nellen, M. Newbold, M.U. Nisa, R. Noriega-Papaqui, L. Olivera-Nieto, N. Omodei, A. Peisker, Y. Pérez Araujo, E.G. Pérez-Pérez, Z. Ren, C.D. Rho, C. Rivière, D. Rosa-González, E. Ruiz-Velasco, H. Salazar, F. Salesa Greus, A. Sandoval, M. Schneider, H. Schoorlemmer, F. Serna, G. Sinnis, A.J. Smith, R.W. Springer, P. Surajbali, E. Tabachnick, K. Tollefson, I. Torres, R. Torres-Escobedo, T.N. Ukwatta, F. Ureña-Mena, et al. (7 additional authors not shown)
DDraft version July 20, 2020
Typeset using L A TEX default style in AASTeX63
A. Albert , R. Alfaro , C. Alvarez, J.R. Angeles Camacho, J.C. Arteaga-Velzquez, K.P. Arunbabu, D. Avila Rojas, H.A. Ayala Solares , V. Baghmanyan , E. Belmont-Moreno , S.Y. BenZvi , C. Brisbois , K.S. Caballero-Mora , T. Capistrn , A. Carramiana , S. Casanova , U. Cotti , S. Coutio de Len , E. De la Fuente , R. Diaz Hernandez, L. Diaz-Cruz, B.L. Dingus , M.A. DuVernois , M. Durocher, J.C. Daz-Vlez , R.W. Ellsworth , K. Engel , C. Espinoza , K.L. Fan, K. Fang, M. Fernndez Alonso, H. Fleischhack , N. Fraija, A. Galvn-Gmez, D. Garcia, J.A. Garca-Gonzlez , F. Garfias , G. Giacinti, M.M. Gonzlez , J.A. Goodman , J.P. Harding, S. Hernandez , J. Hinton , B. Hona, D. Huang , F. Hueyotl-Zahuantitla , P. Hntemeyer, A. Iriarte , A. Jardin-Blicq ,
17, 19, 20
V. Joshi , D. Kieda , A. Lara , W.H. Lee , H. Len Vargas , J.T. Linnemann, A.L. Longinotti , G. Luis-Raya , J. Lundeen, R. Lpez-Coto, K. Malone , V. Marandon, O. Martinez , I. Martinez-Castellanos , J. Martnez-Castro, J.A. Matthews , P. Miranda-Romagnoli , J.A. Morales-Soto, E. Moreno , M. Mostaf , A. Nayerhoda , L. Nellen , M. Newbold , M.U. Nisa , R. Noriega-Papaqui , L. Olivera-Nieto, N. Omodei , A. Peisker, Y. Prez Araujo , E.G. Prez-Prez , Z. Ren , C.D. Rho , C. Rivire , D. Rosa-Gonzlez , E. Ruiz-Velasco , H. Salazar, F. Salesa Greus ,
7, 30
A. Sandoval , M. Schneider , H. Schoorlemmer , F. Serna, G. Sinnis , A.J. Smith, R.W. Springer , P. Surajbali , E. Tabachnick, K. Tollefson , I. Torres , R. Torres-Escobedo, T.N. Ukwatta , F. Urea-Mena, T. Weisgarber, F. Werner , E. Willox, A. Zepeda , H. Zhou, C. de Len And J.D. lvarez (HAWC Collaboration) Physics Division, Los Alamos National Laboratory, Los Alamos, NM, USA Instituto de F´ısica, Universidad Nacional Autnoma de Mxico, Ciudad de Mexico, Mexico Universidad Autnoma de Chiapas, Tuxtla Gutirrez, Chiapas, Mxico Universidad Michoacana de San Nicols de Hidalgo, Morelia, Mexico Instituto de Geof´ısica, Universidad Nacional Autnoma de Mxico, Ciudad de Mexico, Mexico Department of Physics, Pennsylvania State University, University Park, PA, USA Institute of Nuclear Physics Polish Academy of Sciences, PL-31342 IFJ-PAN, Krakow, Poland Department of Physics & Astronomy, University of Rochester, Rochester, NY , USA Department of Physics, University of Maryland, College Park, MD, USA Instituto Nacional de Astrof´ısica, ptica y Electrnica, Puebla, Mexico Departamento de F´ısica, Centro Universitario de Ciencias Exactase Ingenierias, Universidad de Guadalajara, Guadalajara, Mexico Facultad de Ciencias F´ısico Matemticas, Benemrita Universidad Autnoma de Puebla, Puebla, Mexico Department of Physics, University of Wisconsin-Madison, Madison, WI, USA Department of Physics, Stanford University: Stanford, CA 943054060, USA Department of Physics, Michigan Technological University, Houghton, MI, USA Instituto de Astronom´ıa, Universidad Nacional Autnoma de Mxico, Ciudad de Mexico, Mexico Max-Planck Institute for Nuclear Physics, 69117 Heidelberg, Germany Department of Physics and Astronomy, University of Utah, Salt Lake City, UT, USA Department of Physics, Faculty of Science, Chulalongkorn University, 254 Phayathai Road,Pathumwan, Bangkok 10330, Thailand National Astronomical Research Institute of Thailand (Public Organization), Don Kaeo, MaeRim, Chiang Mai 50180, Thailand
Corresponding author: Henrike Fleischhackhfl[email protected] author: Mehr Un [email protected] author: Alison [email protected] a r X i v : . [ a s t r o - ph . H E ] J u l The HAWC Collaboration Erlangen Centre for Astroparticle Physics, Friedrich-Alexander-Universit¨at Erlangen-N¨urnberg, Erlangen, Germany Department of Physics and Astronomy, Michigan State University, East Lansing, MI, USA Universidad Politecnica de Pachuca, Pachuca, Hgo, Mexico INFN and Universita di Padova, via Marzolo 8, I-35131,Padova,Italy Centro de Investigaci´on en Computaci´on, Instituto Polit´ecnico Nacional, M´exico City, M´exico. Dept of Physics and Astronomy, University of New Mexico, Albuquerque, NM, USA Universidad Autnoma del Estado de Hidalgo, Pachuca, Mexico Instituto de Ciencias Nucleares, Universidad Nacional Autnoma de Mexico, Ciudad de Mexico, Mexico Natural Science Research Institute, University of Seoul, Seoul, Republic Of Korea Instituto de F´ısica Corpuscular, CSIC, Universitat de Val`encia, E-46980, Paterna, Valencia, Spain Department of Chemistry and Physics, California University of Pennsylvania, California, Pennsylvania, USA Physics Department, Centro de Investigacion y de Estudios Avanzados del IPN, Mexico City, DF, Mexico Tsung-Dao Lee Institute & School of Physics and Astronomy, Shanghai Jiao Tong University, Shanghai, China
Submitted to ApJSABSTRACTWe present a new catalog of TeV gamma-ray sources using 1523 days of data from the High Alti-tude Water Cherenkov (HAWC) observatory. The catalog represents the most sensitive survey of theNorthern gamma-ray sky at energies above several TeV, with three times the exposure compared tothe previous HAWC catalog, 2HWC. We report 65 sources detected at ≥ ◦ of previously detected TeV emitters, and twentysources that are more than 1 ◦ away from any previously detected TeV source. Of these twenty newsources, fourteen have a potential counterpart in the fourth Fermi
Large Area Telescope catalog ofgamma-ray sources. We also explore potential associations of 3HWC sources with pulsars in the ATNFpulsar catalog and supernova remnants in the Galactic supernova remnant catalog.
Keywords:
High-energy astrophysics, Gamma-ray astronomy — catalogs — surveys INTRODUCTIONHigh-sensitivity, unbiased surveys of the gamma-ray sky are important to finding new astrophysical objects – bothto understand their bulk properties, and to constrain new physics beyond the standard model. Discovering the sourcesof cosmic rays and determining the underlying acceleration mechanisms requires precise measurements of gamma-rayspectra of objects above several tens of TeV. In addition, the indirect search for dark matter particles in the GeV–TeVregime also hinges on detecting a steady flux of photons from several galactic and extra-galactic targets of interest. Thecurrent generation of ground-based gamma-ray telescopes, in particular Imaging Atmospheric Cherenkov Telescopes(Holder et al. 2009; Balzer et al. 2014; Ahnen et al. 2017; Anderhub et al. 2013), are capable of resolving gamma-raysources to ≤ . ◦ precision. The highly complementary survey instruments, such as Tibet-AS γ (Amenomori et al.2019), LHAASO (Zhen 2014), ARGO-YBJ (Bartoli et al. 2013), and HAWC, have further extended the reach of TeVastronomy with their high up-time and unprecedented sensitivity to spatially extended sources.The High Altitude Water Cherenkov (HAWC) observatory has been continuously monitoring the Northern sky inTeV cosmic rays and gamma rays since commencing full operations in 2015, and has achieved a sensitivity down to afew percent of the Crab flux in five years. This work presents the results of an all-sky time-integrated search for point-like and extended sources using 1523 days of HAWC data. As a follow-up to the 2HWC catalog of TeV gamma-raysources (Abeysekara et al. 2017b), we introduce an updated catalog using more data and improved analysis methods.This paper is structured as follows. Section 2 provides a brief description of the HAWC detector, as well as thedata and the analysis method we use in the construction of the catalog. Section 3 presents the results of the catalogsearch and provides preliminary spatial and spectral information on all sources. A broad discussion of the results isalso presented. Section 4 discusses the systematic uncertainties and methodological limitations of this work. Section5 concludes the paper. INSTRUMENT AND DATA ANALYSIS2.1.
The HAWC Gamma-ray Observatory
HAWC consists of 300 water tanks, each filled with ∼ ◦ N and its wide field-of-view, HAWC can observe about two thirds of the gamma-ray sky (from about-26 ◦ to +64 ◦ in declination) every day, with an instantaneous field-of-view of > ◦ and 1.0 ◦ depending on the energy and zenith angle of thesignal. More details about the HAWC detector can be found in Abeysekara et al. (2017b,c).2.2. Data Selection and Reduction
The temporal and spatial distribution of charge deposited in HAWC’s PMTs is used to reconstruct the propertiesof the primary particle producing the air shower. The difference in timing between the signals recorded in differenttanks allows us to reconstruct the direction of the primary particle. The spatial distribution and magnitude of thecharges can be used to reconstruct the primary energy and to separate gamma-ray induced showers from cosmic-rayinduced ones.We distribute the reconstructed events into nine analysis bins according to the fraction of the operating PMTs thatrecorded a signal for a given event. The fraction of PMTs hit is correlated with the primary energy, allowing us toextract the energy spectrum of gamma-ray sources. We apply gamma-hadron separation cuts to reduce the backgroundof cosmic-ray induced showers. A detailed description of air shower reconstructions and quality cuts applied to thedata can be found in Abeysekara et al. (2017b,c).We further bin the gamma-ray candidate events according to the direction of the primary particle in celestialcoordinates (Right Ascension and Declination, J2000 epoch). We use the HEALPix binning scheme (Gorski et al.2005), with an NSIDE parameter of 1024.The dominant background is given by hadronic showers that pass the gamma-hadron cuts. For each pixel, weestimate the expected number of remaining background events after cuts using the method of Direct Integration asdescribed in Atkins et al. (2003). 2.3.
Construction of the Catalog
The 3HWC catalog is based on data collected by the HAWC observatory between November 2014 and June 2019,corresponding to a livetime of 1523 days – about three times the livetime of the 2HWC catalog data set. For themost part, the construction of the catalog follows the same method as the previous 2HWC catalog, which is describedin Abeysekara et al. (2017b). We summarize the algorithm below, with particular focus on the differences from theprevious catalog search. 2.3.1.
Source Search
We perform a blind search for sources across HAWC’s field-of-view using the likelihood framework discussed inYounk et al. (2015). The likelihood calculation assumes that the number of counts in each bin/pixel is distributedaccording to a Poisson distribution, with the mean given by the estimated background counts plus (if applicable) thepredicted number of gamma-ray counts from the convolution of the source model with the detector response.For each HEALPix pixel, we calculate a likelihood ratio λ = ˆ L s + b / L b , comparing the likelihood, ˆ L s + b , of thebest-fit model with a gamma-ray source centered on that pixel to that of a background-only model, L b . We definea test statistic, T S = 2 log (Λ). Assuming that the null hypothesis is true, the
T S is distributed according to a χ distribution with one degree of freedom (Wilks 1938), which can be approximated by a gaussian distribution. Then, ±√ T S corresponds to the (“pre-trials”) significance. The negative sign is used for pixels in which the best-fit fluxnormalization is negative.The signal hypothesis considers a fixed source morphology and an E − . power-law energy spectrum. The only freeparameter of the likelihood fit is the flux normalization. We repeat source searches for four different hypotheticalmorphologies: point sources, and extended disk-like sources with radii of 0.5 ◦ , 1.0 ◦ , and 2.0 ◦ . This procedure is very The HAWC Collaboration similar to what was used in the previous 2HWC catalog, with the only change being the spectral index hypothesis(the prior catalog used a spectral index of -2.7 for point sources, -2.0 for extended sources). The resulting all-skysignificance map for a point-source assumption can be seen in Figure 1.For each significance map, we compile a list of candidate sources comprising the local maxima with √ T S >
5. Dueto Poisson fluctuations in the number of detected gamma rays, a single source may produce multiple local maxima.To avoid double-counting such sources, candidate sources are promoted to sources if they pass the following “TSvalley” criterion: the significance profile connecting the source candidate with any other source within 5 ◦ of the sourcecandidate in question has to “dip” by ∆ T S > < ∆ T S <
2. We mark secondary sources with adagger ( † ) in Table 1. The ∆ T S criterion used for 3HWC is a little stricter than the one used in the 2HWC catalog,in which a source only had to pass the ∆
T S with its closest neighboring source.We then combine the four source lists (for the four different assumptions of the source morphology) to yield the3HWC catalog. We include all sources found in the point source search in 3HWC. We only include sources found inthe extended source searches if they are more than 1 ◦ away from any point source or smaller extended source alreadyin the catalog. Table 1 shows the resulting list of sources comprising the 3HWC catalog. For each source, we also showthe closest known TeV source listed in the TeVCat (Wakely & Horan 2008) if it is within 1 ◦ of the HAWC source.2.3.2. Spectral Fits
After identifying the primary and secondary sources, we perform likelihood fits to obtain each source’s energyspectrum. We assume a simple power-law spectrum for each source, fitting for the spectral index and flux normalizationas done for the 2HWC catalog (Abeysekara et al. 2017b). There are two changes to the spectral fitting procedure usedin this work compared to 2HWC. First, for 3HWC, we only report the spectral fit for the same morphology for whichthe source was first found in the source-search stage. Second, we dynamically treat the instrument response duringthe fit. 2HWC fits relied on a method where the angular resolution was pre-calculated (and fit with a double-Gaussianfunction in each analysis bin) before the spectral fit, for a fixed spectral assumption. 3HWC spectral fits use a newmethod in which we recalculate the angular resolution for each tested spectral assumption during the fit. Additionally, TS Figure 1.
All-sky significance map in celestial coordinates, assuming a point-source hypothesis. The bright band on the left ispart of the Galactic plane (c.f. Figures 4-7), and the bright region on the right is the Galactic anti-center region containing theCrab Nebula and the Geminga halo (c.f. Figure 3). The two off-plane hotspots are the two TeV-bright blazars Mrk 421 (right)and Mrk 501 (left). http://tevcat.uchicago.edu/ we do not assume that the angular resolution follows a specific analytical shape. This allows for a more completecharacterization of HAWC’s PSF and of the systematic uncertainties on the fit parameters. See Martinez-Castellanos(2019) for more details on the new spectral fit method. Table 2 shows the results of the spectral fits. RESULTS3.1.
The 3HWC catalog contains 65 sources, 17 of which are considered secondary sources (not well separated fromneighboring sources according to the ∆
T S criterion). The source positions can be found in Table 1 and the results ofthe spectral fits can be found in Table 2. Twenty-eight of these sources do not lie within 1 ◦ of any 2HWC source. Wediscuss some of these sources in more detail in Section 3.4.We compare the flux measurements with the sensitivity for the underlying dataset. The flux sensitivity is definedas the flux normalization required to have a 50% probability of detecting a source at the 5 σ level. Figure 2 shows theHAWC 1523-days sensitivity and the flux measurements from Table 2 as a function of declination. HAWC is moresensitive to sources transiting directly overhead, corresponding to a declination of 19.0 ◦ , than to sources transiting atlarger zenith angles. HAWC is also more sensitive to hard-spectrum sources. For the optimal case (an E − sourcetransiting directly overhead), HAWC’s sensitivity approaches ∼
2% of the flux of the Crab Nebula.Most of the sources were found in the point source search. With about three times the livetime compared to the2HWC catalog, many extended sources are now also significantly detected in the point source map. For example, Figure3 shows five 3HWC sources ( , , , ,and , all found in the point source search) clustering near the Geminga pulsar. We believe thatthese five sources are all part of the extended halo around Geminga, described in Abeysekara et al. (2017a). Similarly,both and are part of the extended source announcedin the aforementioned publication. It is not clear if these sources correspond to real features in the morphology of thetwo pulsar halos, or if they are just due to statistical fluctuations in the number of photons recorded by HAWC.As seen in the the all-sky significance map (Figure 1), the majority of the sources in the 3HWC catalog are locatedalong the Galactic plane. Figures 4, 5, 6, and 7 show the significance maps of the Galactic plane from the Cygnusregion ( l = 85 ◦ ) to the inner Galaxy ( l = 2 ◦ ). The Galactic center itself falls outside of the part of the sky visible toHAWC. Figure 3 shows a region near the Galactic anti-center containing the Crab Nebula, Geminga, and other sources.For this region, both the point-source significance map and the significance map from the 1 ◦ extended source searchare shown. For convenience, the locations of 3HWC sources and TeVCat sources have been marked in these images.Figures 8 and 9 show the distribution of 3HWC sources as a function of galactic latitude and longitude respectively. The HAWC Collaboration
Table 1 . Source list and nearest TeVCat sources (within 1 ◦ of each 3HWC source). Secondary sources (i.e., sources that are not separatedfrom their neighbor(s) by a large TS gap) are marked with a dagger ( † ). The position uncertainty reported here is statistical only. Thesystematic uncertainty on the position is discussed in Section 4.3. TeVCat source names within 0.5 ◦ of a 3HWC source are printed inbold. For sources without a TeVCat counterpart within 1 ◦ , the angular distance to the nearest TeVCat source is printed for reference.Nearest TeVCat sourceName Radius TS RA Dec l b σ stat. unc. Dist. Name[ ◦ ] [ ◦ ] [ ◦ ] [ ◦ ] [ ◦ ] [ ◦ ] [ ◦ ]3HWC J0534+220 0.0 35736.5 83.63 22.02 184.55 -5.78 0.06 0.01 Crab † HAWC J0543+233
IC 443 † Geminga
HAWC J0635+070 † Geminga Pulsar † Markarian 421
Markarian 501 † HESS J1809-193
HESS J1813-126 † HESS J1831-098 † HESS J1848-018
IGR J18490-0000 † HESS J1857+026 † Table 1 continued on next page
Table 1 (continued)
Nearest TeVCat sourceName Radius TS RA Dec l b σ stat. unc. Dist. Name[ ◦ ] [ ◦ ] [ ◦ ] [ ◦ ] [ ◦ ] [ ◦ ] [ ◦ ]3HWC J1907+085 0.0 75.5 286.79 8.57 42.35 0.44 0.09 0.07 MGRO J1908+06
HESS J1912+101 † SS 433 e1 † W 51 † SNR G054.1+00.3 † † † † VER J2019+368
VER J2019+407
TeV J2032+4130 † Boomerang
The HAWC Collaboration
Table 2 . Source radius, fitted spectrum, and nearest TeVCat source if within 1 ◦ . TeVCat sources within 0.5 ◦ of a particular 3HWC source are printed in bold face. The flux F is the differential flux at 7 TeV. The twosets of reported uncertainties correspond to statistical and systematic, respectively. The spectral fit for did not converge.Name Radius Index F Nearest TeVCat source[ ◦ ] [10 − TeV − cm − s − ]3HWC J0534+220 0.0 − . ± . +0 . − . ± . +55 . − . Crab − . ± . +0 . − . +0 . − . . − . HAWC J0543+2333HWC J0543+231 0.0 − . +0 . − .
16 +0 . − . ± . +1 . − . HAWC J0543+233 − . ± . +0 . − . ± . +1 . − . IC 443 − . +0 . − .
13 +0 . − . +1 . − . . − . ...3HWC J0630+186 0.0 − . +0 . − .
14 +0 . − . +0 . − . . − . ...3HWC J0631+107 0.0 − . +0 . − .
19 +0 . − . +0 . − . . − . ...3HWC J0631+169 0.0 − . +0 . − .
14 +0 . − . ± . +2 . − . Geminga − . +0 . − .
15 +0 . − . ± . +1 . − . ...3HWC J0634+067 0.5 − . ± . +0 . − . +1 . − . . − . HAWC J0635+070 − . +0 . − .
22 +0 . − . +0 . − . . − . Geminga3HWC J0634+180 0.0 − . +0 . − .
11 +0 . − . ± . +2 . − . Geminga Pulsar − . +0 . − .
24 +0 . − . +1 . − . . − . − . ± . +0 . − . ± . +16 . − . Markarian 421 − . ± . +0 . − . ± . +5 . − . Markarian 501 − . +0 . − .
14 +0 . − . ± . +1 . − . ...3HWC J1743+149 0.0 − . +0 . − .
16 +0 . − . +0 . − . . − . ...3HWC J1757-240 1.0 − . ± . +0 . − . +10 . − . . − . HESS J1800-240B3HWC J1803-211 0.0 − . +0 . − .
14 +0 . − . +5 . − . . − . HESS J1804-2163HWC J1809-190 0.0 − . ± . +0 . − . +4 . − . . − . HESS J1809-193 − . ± . +0 . − . ± . +5 . − . HESS J1813-126 − . ± . +0 . − . ± . +26 . − . − . +0 . − .
07 +0 . − . ± . +10 . − . − . ± . +0 . − . ± . +53 . − . − . ± . +0 . − . +1 . − . . − . HESS J1831-098 − . ± . +0 . − . ± . +25 . − . − . ± . +0 . − . ± . +19 . − . − . ± . +0 . − . ± . +2 . − . ...3HWC J1847-017 0.0 − . ± . +0 . − . ± . +8 . − . HESS J1848-018 − . ± . +0 . − . ± . +11 . − . IGR J18490-0000 − . ± . +0 . − . +1 . − . . − . − . ± . +0 . − . ± . +10 . − . HESS J1857+026 − . ± . +0 . − . ± . +1 . − . ...3HWC J1907+085 0.0 − . ± . +0 . − . +0 . − . . − . continued on next page Table 2 (continued)
Name Radius Index F Nearest TeVCat source[ ◦ ] [10 − TeV − cm − s − ]3HWC J1908+063 0.0 − . ± . +0 . − . ± . +20 . − . MGRO J1908+06 − . ± . +0 . − . ± . +4 . − . HESS J1912+101 − . ± . +0 . − . +0 . − . . − . SS 433 e1 − . ± . ± . +0 . − . . − . − . ± . +0 . − . ± . +1 . − . ...3HWC J1918+159 0.0 − . +0 . − .
17 +0 . − . ± . +1 . − . ...3HWC J1920+147 0.0 − . ± . +0 . − . +0 . − . . − . W 513HWC J1922+140 0.0 − . ± . +0 . − . ± . +3 . − . W 51 − . ± . +0 . − . ± . +1 . − . ...3HWC J1928+178 0.0 − . ± . +0 . − . ± . +5 . − . − . ± . +0 . − . ± . +2 . − . SNR G054.1+00.3 − . ± . +0 . − . ± . +1 . − . ...3HWC J1936+223 0.0 − . ± . +0 . − . +0 . − . . − . ...3HWC J1937+193 0.0 − . ± . +0 . − . +0 . − . . − . ...3HWC J1940+237 0.0 − . +0 . − .
11 +0 . − . ± . +1 . − . − . +0 . − .
19 +0 . − . ± . +1 . − . − . +0 . − .
13 +0 . − . +1 . − . . − . ...3HWC J1951+293 0.0 − . +0 . − .
10 +0 . − . ± . +2 . − . − . ± . +0 . − . ± . +2 . − . − . +0 . − .
14 +0 . − . ± . +2 . − . − . ± . +0 . − . ± . +1 . − . ...3HWC J2006+340 0.0 − . +0 . − .
13 +0 . − . +0 . − . . − . − . ± . +0 . − . ± . +1 . − . ...3HWC J2019+367 0.0 − . ± . +0 . − . ± . +17 . − . VER J2019+368 − . ± . +0 . − . ± . +3 . − . VER J2019+407 − . ± . +0 . − . ± . +2 . − . ...3HWC J2023+324 1.0 − . +0 . − .
12 +0 . − . +1 . − . . − . ...3HWC J2031+415 0.0 − . ± . +0 . − . +1 . − . . − . TeV J2032+4130 − . +0 . − .
15 +0 . − . +2 . − . . − . ...3HWC J2227+610 0.0 − . ± . +0 . − . +5 . − . . − . Boomerang The HAWC Collaboration
20 0 20 40 60 [ ] d N d E [ T e V c m s ] E E E Figure 2. E − . , E − . and E − . . Also shown is the best-fit flux normalization at 7 TeVfor all sources in the 3HWC catalog. -173-167-161-155-149-143 l [ ]933915 b [] Galactic plane V; 0.0 ; 1523 days TS -173-167-161-155-149-143 l [ ]933915 b [] Galactic plane V; 1.0 ; 1523 days TS Figure 3.
Significance maps of the Galactic anti-center region for − ◦ ≤ l ≤ − ◦ , showing the Crab Nebula and theGeminga halo among other sources. Left: Point-source hypothesis. Right: 1 ◦ extended-source hypothesis. The grey lines showsignificance contours starting at √ T S = 40, increasing in steps of ∆ √ T S = 20. Top labels indicate positions of known TeVsources (from TeVCat), bottom labels indicate positions of 3HWC sources. b [] Galactic plane I; 0.0 ; 1523 days TS Figure 4.
Significance map of part of the Galactic plane for 62 ◦ ≤ l ≤ ◦ ; point-source hypothesis. The green lines showsignificance contours starting at √ T S = 26, increasing in steps of ∆ √ T S = 2. Top labels indicate positions of known TeVsources (from TeVCat), bottom labels indicate positions of 3HWC sources. b [] Galactic plane II; 0.0 ; 1523 days TS Figure 5.
Significance map of part of the Galactic plane for 42 ◦ ≤ l ≤ ◦ ; point-source hypothesis. Top labels indicatepositions of known TeV sources (from TeVCat), bottom labels indicate positions of 3HWC sources. The HAWC Collaboration b [] Galactic plane III; 0.0 ; 1523 days TS Figure 6.
Significance map of part of the Galactic plane for 22 ◦ ≤ l ≤ ◦ ; point-source hypothesis. The green lines showsignificance contours starting at √ T S = 26, increasing in steps of ∆ √ T S = 2. Top labels indicate positions of known TeVsources (from TeVCat), bottom labels indicate positions of 3HWC sources. b [] Galactic plane IV; 0.0 ; 1523 days TS Figure 7.
Significance map of the inner Galactic plane for 2 ◦ ≤ l ≤ ◦ ; point-source hypothesis. The green lines showsignificance contours starting at √ T S = 26, increasing in steps of ∆ √ T S = 2. Top labels indicate positions of known TeVsources (from TeVCat), bottom labels indicate positions of 3HWC sources. ◦ . Two of these sources ( and ) still show significant emission ( T S >
25) in the new data set, but do notpass the TS dip test. is now considered part of the complex, and is part of the complex.The remaining five sources do not pass the
T S >
25 criterion with the new data set. Out of these, lies in the Galactic plane near the W51 supernova remnant. In the 3HWC data set (point-source search), has a TS of 21.7, which is below the threshold for significant detection. It is possible that the2HWC detection was due to an upward fluctuation of a combination of diffuse emission and emission from the nearbyW51 complex. The other four sources ( , , , and ) were off-plane sources without known counterparts. So far, searches for variability of these sources havenot yielded conclusive results. It is unclear whether the previous detections were false positives due to randombackground fluctuations ro upward fluctuations of real, but weak, sources, or indicative of flaring activity/temporalvariability. We estimate the expected number of false positives, i.e. random background fluctuations that pass thedetection threshold, to be 0.75 for the 3HWC catalog (see Section 4), compared to 0.4 for 2HWC.3.3. New 3HWC Sources Associated with Known TeV Sources
Eight of the 3HWC sources have no counterpart (within 1 ◦ ) in 2HWC, but are potentially associated with knownTeV emitters. These sources are listed below. We define positional coincidence within 1 ◦ as the criterion for twosources to be considered as associated. ( T S = 28 .
8) and ( T S = 34 .
2) are part of the extended source that hadbeen previously announced as
HAWC J0543+233 (Riviere et al. 2017) – a potential TeV halo around the pulsar
PSR B0540+23 . ( T S = 32 .
3) is associated with the shell-type supernova remnant (SNR)
IC 443 ( SNR G189.1+03.0 ), which has been detected at TeV energies by MAGIC (Albert et al. 2007) and VERITAS(Acciari et al. 2009a). HAWC previously announced the detection of gamma-ray emission from
IC 443 withoutnaming the source (Fleischhack 2020). ( T S = 36 .
2) had been previously announced as
HAWC J0635+070 (Brisbois et al. 2018) –a potential TeV halo around the pulsar
PSR J0633+0632 . ( T S = 52 .
5) is associated with the known TeV gamma-ray sources
VER J2227+608 (Acciariet al. 2009b) and
MGRO J2228+61 (Abdo et al. 2009a,b). The most likely source of the emission is the shell-typesupernova remnant
SNR G106.3+2.7 . was recently announced as HAWC J2227+610 in adedicated publication (Albert et al. 2020). ( T S = 44 .
7) is associated with the eastern lobe of the micro-quasar
SS 433 . Detection ofTeV gamma-ray emission from this source (as well as the western lobe of
SS 433 ) had previously been announced byHAWC (Abeysekara et al. 2018). ( T S = 28 .
6) was found in the 1 ◦ extended source search. It overlaps with the W28 region,which contains at least four known TeV sources ( HESS J1801-233 , and
HESS J1800-240A/B/C ) (Aharonianet al. 2008). The emission seen by H.E.S.S. has been attributed to an SNR interacting with nearby molecular clouds.Due to HAWC’s limited sensitivity and relatively poor angular resolution at these declinations, we are currently unableto resolve the individual H.E.S.S. sources. ( T S = 38 .
4) is located near the known, but unidentified, TeV source
HESS J1804-216 (Aharonian et al. 2005, 2006). 3.4.
New TeV Sources
For each source in 3HWC, we scan several catalogs of known or potential gamma-ray sources for potential associationswithin 1 ◦ including the TeVCat (Wakely & Horan 2008), the fourth Fermi -LAT source catalog (4FGL) (Abdollahiet al. 2020), the ATNF Pulsar Catalog (v 1.62) (Manchester et al. 2005), and the Galactic supernova remnant catalogSNRCat (Ferrand & Safi-Harb 2012)..We report 20 new sources that do not have a potential counterpart in the TeVCat (see Table 3). Fourteen of thesenew sources are within 1 ◦ of a previously observed GeV source. Table 3 lists the GeV associations and their source The HAWC Collaboration
Table 3.
New HAWC sources with no TeV counterpart. For each source we list the following information in the various columns:Galactic longitude; Galactic latitude; the nearest GeV source in 4FGL (Abdollahi et al. 2020) and its separation from the 3HWCsource; the source class as listed in 4FGL where available (bcu: active galaxy of uncertain type, PSR: pulsar, identified by pulsations,unk: unknown); the nearest pulsar and corresponding separation from the ATNF pulsar catalog (Manchester et al. 2005); and thenearest SNR, separation distance, and type from the SNRCat (Ferrand & Safi-Harb 2012).HAWC l [ ◦ ] b [ ◦ ] 4FGL ( ◦ ) Class ATNF ( ◦ ) SNRCat ( ◦ ) SNR Type3HWC J0621+382 175.44 10.97 4FGL J0620.3+3804 (0.22) bcu J0622+3749 (0.42) ... ...3HWC J0630+186 193.98 4.02 ... ... J0630+19 (0.94) ...3HWC J0631+107 201.08 0.43 4FGL J0631.5+1036 (0.15) PSR J0631+1036 (0.14) ... ...3HWC J0633+191 193.92 4.85 ... ... ... ... ...3HWC J1739+099 33.89 20.34 4FGL J1740.5+1005 (0.22) PSR J1740+1000 (0.13) G034.0+20.3 (0.13) filled-centre3HWC J1743+149 39.13 21.68 ... ... ... ...3HWC J1844 −
001 31.95 1.50 4FGL J1848.2 − − − − − − − classifications obtained from the fourth Fermi -LAT source catalog, 4FGL. Two new sources, ( T S = 38 .
9) and ( T S = 31 . is within 0.95 ◦ of the pulsar PSRJ0630+19 . is associated with PSR J1918+1541 with a separation distance of 0.26 ◦ . Theage and spin-down luminosity of these objects are not available.3.4.1. Unassociated New TeV Sources
We observe four sources that do not have an apparent counterpart in any of the catalogs that we scanned forpotential associations: ( T S = 37 . ( T S = 27 . ( T S = 29 . ( T S = 25 . is in a dense region of known pulsars including Geminga. is near , which itself is a 1 ◦ extended source. is inthe Cygnus-X region with a number of star-forming clusters nearby, most notably the Fermi -LAT cocoon (Hona et al.2020). Without a detailed morphological study of their respective regions, we cannot definitively exclude the abovenew sources as appendages of existing sources. Such a study, however, is beyond the scope of this paper. (TS = 25.9) is the only new unassociated source that is not in spatial proximity of a region ofknown TeV sources. It is also notably distant from the Galactic plane with a Galactic latitude of b = 21 . ◦ .3.4.2. Pulsars and TeV Halo Candidates in the 3HWC
A significant fraction of 3HWC sources are associated with pulsars in the ATNF catalog (c.f. Table 3). Figure 10shows all the 3HWC sources potentially associated with pulsars in the galaxy for which the distance information isavailable.5 [ ] N s o u r c e s All 3HWCNew TeV
Figure 8.
Distribution of HAWC sources (excluding the known blazars) with respect to galactic latitude for -5 ◦ < b < ◦ .The darker shaded histogram shows the new TeV sources in 3HWC that were not present in 2HWC. [ ] N s o u r c e s S e n s i t i v i t y F l u x [ T e V c m s ] All 3HWCNew TeVSensitivity
Figure 9.
Distribution of HAWC sources as a function of galactic longitude. The darker shaded histogram shows the newTeV sources in 3HWC that were not present in 2HWC. The blue solid line shows the sensitivity at b = 0 ◦ . Due to its location,HAWC is most sensitive towards the Galactic anticenter region, l ≈ − ◦ and, to the inner Galaxy at l ≈ +50 ◦ . Most knownTeV gamma-ray sources are located in the inner Galaxy. After the discovery of extended gamma-ray emission around the Geminga and Monogem pulsars by HAWC (Abey-sekara et al. 2017a), and the discovery of several extended TeV PWNe by H.E.S.S. (Abdalla et al. 2018), it has beensuggested that extended pulsar “Halos” are a common feature of older pulsars (Linden et al. 2017; Linden & Buckman2018; Giacinti et al. 2020; Sudoh et al. 2019; Fleischhack et al. 2019). The observed gamma-ray emission is thought6
The HAWC Collaboration
Table 4.
HAWC Sources with the corresponding associated TeV halo candidate pulsars. The age of the pulsar in kyr and thespin-down luminosity, ˙ E , in erg s − are also given. The Separation column indicates the angular distance between the HAWC sourceand the ATNF pulsar (Manchester et al. 2005). The TeVCat column lists the previously detected TeV counterpart of each source.HAWC l [ ◦ ] b [ ◦ ] Pulsar Age [kyr] ˙ E [erg s − ] Distance [kpc] Separation [ ◦ ] TeVCat3HWC J0540+228 184.58 -4.13 B0540+23 253.0 4.09e+34 1.56 0.83 HAWC J0543+2333HWC J0543+231 184.67 -3.52 B0540+23 253.0 4.09e+34 1.56 0.36 HAWC J0543+2333HWC J0631+169 195.63 3.45 J0633+1746 342.0 3.25e+34 0.19 0.95 Geminga3HWC J0634+180 195.00 4.62 J0633+1746 342.0 3.25e+34 0.19 0.38 Geminga Pulsar3HWC J0659+147 200.60 8.40 B0656+14 111.0 3.8e+34 0.29 0.51 2HWC J0700+1433HWC J0702+147 200.91 9.01 B0656+14 111.0 3.8e+34 0.29 0.77 2HWC J0700+1433HWC J1739+099 33.89 20.34 J1740+1000 114.0 2.32e+35 1.23 0.13 ...3HWC J1831-095 22.13 0.02 J1831 − to be due to inverse Compton up-scattering of cosmic microwave background photons by relativistic electrons andpositrons diffusing freely in the vicinity around the pulsar. TeV halos are thought to form around older pulsars (atleast several tens of thousands of years old) that have either left their SNR shell or whose SNR shell has already dissi-pated. They are thus distinct from (classical) PWNe, where the electron-positron plasma is confined by the ambientmedium.We produce an updated list of pulsars that are likely candidates to have a TeV Halo detectable with HAWC,following similar criteria as Linden et al. (2017). We select pulsars from the ATNF with ages between 100 kyr and400 kyr, declinations between -25 ◦ and +64 ◦ , and an estimated spindown flux of at least 1% of that of the Gemingapulsar. We find sixteen such pulsars, out of which eight are coincident with at least one 3HWC source (within 1 ◦ ).Table 4 lists the 3HWC sources that are coincident with these TeV Halo candidate pulsars. Some pulsars have morethan one 3HWC source nearby. This is not unexpected as our source search sometimes finds multiple point sourcesassociated with the same extended emission region. Two of these pulsars, PSR J0631+1036 and
PSR J1740+1000 ,have not previously been detected at TeV energies. LIMITATIONS AND SYSTEMATIC UNCERTAINTIES4.1.
Background Fluctuations and Spurious Detections
It is possible for mere fluctuations in the background and/or the Galactic diffuse emission to pass the selectioncriteria and produce a spurious source. In order to estimate the frequency of false positive sources, we create twentysimulated significance maps using the background counts from the original source search. For each map, we obtainthe simulated number of signal events in each pixel by Poisson-fluctuating the number of background events in thecorresponding pixel. We then run each of these randomized background maps through the full analysis pipeline,including point and extended searches. In the 20 total randomized background maps, we find 15 local maxima with a
T S >
25. Therefore, the estimated number of false positive sources is 15 /
20 = 0 .
75. The fluctuations typically occurjust above the threshold value of
T S = 25.4.2.
Limitations of the Source Search
As in the 2HWC catalog, we conduct blind source searches for four different fixed morphological assumptions (pointsources, and 0.5 ◦ , 1.0 ◦ , 2.0 ◦ extended sources). We then combine these results, with preference given to sources foundin the point source search and the smaller radius searches to avoid double counting of sources.This approach can lead to sources being misidentified or missed. First, some extended sources may be significantenough to be detected in the point source analysis. Poisson fluctuations of the signal could potentially lead to several7 Figure 10.
Face-on view of the galaxy showing positions of HAWC sources associated with (i.e., spatially coincident within1 ◦ of) pulsars for which distances are estimated. Spatial coincidence does not necessarily imply that the observed gamma-rayemission is (fully) powered by the pulsar in question. The color scale corresponds to the measured flux normalization from Table2. The annotated Milky Way background is taken from NASA/JPL-Caltech/R.Hurt (2008). hotspots being detected around the center of such an extended source. As HAWC collects more data, this issue isincreasing in prevalence, as evidenced by the five point sources detected inside the Geminga halo. Second, it is alsopossible that multiple smaller sources located near each other are detected as one source in the extended source searchif the individual sources are not strong enough to cross the detection threshold. This might be happening near theGalactic center where , found in the 1 ◦ extended-source search, overlaps with several known TeVsources. Third, weaker sources may be missed if they are located near a stronger source, as they may not produce awell-defined peak in the significance map. In-depth studies (such as Abeysekara et al. (2017a, 2018)) are needed toproperly resolve source-dense regions. Such studies include multi-source fits, fitting the extensions/shapes, locations,and spectra of several sources the same time. Additionally, measurements by other gamma-ray observatories as wellas measurements at other wavelengths might help disentangle the morphology of complex regions. Further studies ofselected regions of the Galactic plane are in preparation.4.3. Systematic Uncertainty on the Source Locations
Earlier publications (Abeysekara et al. 2017b) quoted an 0.1 ◦ systematic pointing uncertainty, which was estimatedusing simulations and verified through the observation of the Crab Nebula, Mrk421, and Mrk501. New studies ofHAWC’s pointing calibration as a function of source position suggest that the uncertainty could be larger than previ-ously thought for sources that transit near the edge of HAWC’s field of view. HAWC’s absolute pointing uncertaintyincreases to 0.15 ◦ for sources at -10 ◦ or +50 ◦ declination and could be as high as 0.3 ◦ at declinations of -20 ◦ or8 The HAWC Collaboration − −
10 0 10 20 30 40 50 60HAWC declination [ ◦ ] − . − . . . . . . . δ HA W C − δ I A CT [ ◦ ] VERITASMAGICH.E.S.S.
Figure 11.
Measured declination of HAWC sources relative to their TeV counterpart measurements from IACT experiments,MAGIC, H.E.S.S. and VERITAS. HAWC measurements agree with the source locations measured by IACTs within uncertaintiesfor most of its declination range. See text for discussion on source declinations below -10 ◦ . +60 ◦ . (There are no well-isolated point sources detected by HAWC that could be used to unambiguously verify theinstrument’s pointing at these declinations.)In Figure 11, we compare the measured declinations of 3HWC sources to the locations of their likely TeV counterpartsas measured by other experiments. For this comparison, we consider relatively well-localised 3HWC sources that havea TeV association within 1 ◦ detected by a different experiment. We do not include sources in regions of extendedemission or multiple components such as . It can be seen that for most of the declination rangespanned by HAWC’s sensitivity, the 3HWC positions agree with the literature values within statistical and systematicuncertainties. Below source declinations of about -10 ◦ , HAWC measures systematically higher values than the IACTs,between an offset of 0.1 ◦ and 0.4 ◦ .The trend observed in Figure 11 could indicate a bias in HAWC’s pointing at low declinations. However, all ofHAWC’s southern sources lie on the Galactic plane, in a region rich in sources and diffuse emission. HAWC’s angularresolution is poor for low-declination sources compared to sources transiting overhead. Accordingly, Galactic diffuseemission or emission from nearby unresolved sources might affect the peak position detected by HAWC, especiallyfor low-declination sources. IACTs tend to have better angular resolution and are thus affected less by large-scaleemission or neighboring sources. The shift could also be an indication of an energy-dependent morphology of somesources Hona et al. (2020). Future in-depth studies of some of these sources as well as an improved understanding ofHAWC’s pointing are needed to resolve this apparent discrepancy.4.4. Systematic Uncertainty of the Spectral Fits
In Table 2, we report the best-fit fluxes and spectral indices of the 3HWC sources. These fits assume a power-lawspectrum; we do not test other spectral models including a curvature or cutoff term. The reported spectral indexshould be interpreted as an average or effective spectral index across HAWC’s energy range. For the two known extra-9galactic sources, we also do not account for absorption by the extra-galactic background light. Additional studies areongoing for sources that are detected with sufficient statistics to allow more sophisticated spectral models to be fit.In Table 2, we report systematic uncertainties related to the modeling of the HAWC detector response individuallyfor each source. More details about the sources of uncertainty considered here can be found in Abeysekara et al.(2019). In order to compute the uncertainties, we repeat spectral fits with certain properties of the detector modelshifted up or down. We assign an additional uncertainty of 10% to the flux normalization to account for effects, such asvariations in the atmosphere, that are not considered otherwise. We add the resulting positive shifts to the spectral fitparameters in quadrature to obtain the total upward systematic uncertainty, and add the negative shifts in quadraturefor the total negative downward systematic uncertainty.There are other systematic issues affecting the spectral fits. Some fluxes may be overestimated due to “leakage”from nearby (detected or unresolved) sources or from the Galactic diffuse emission. These may also affect the spectralindices. In cases where the apparent extent of a source is larger than HAWC’s angular resolution (0.1 ◦ at highenergies), but the source is strong enough to be significantly detected already in the point-source search, we onlyreport the spectrum assuming a point-source hypothesis. This leads to the flux normalization being underestimatedand the spectral index to be biased towards softer spectra (as the angular resolution improves for high energies andthus more of the high-energy emission is “lost”). Upcoming publications will provide better spectral fits for extendedsources. CONCLUSIONThe HAWC observatory has been conducting the most sensitive, unbiased survey of the Northern sky at TeV energiesfor over five years. We have presented the third catalog of steady gamma-ray emitters detected by HAWC using 1523days of data. The catalog consists of 65 sources, including two blazars. The most abundant source class among thepotential counterpart of HAWC sources in the Galactic plane is pulsars (56).The 3HWC catalog provides many targets for multi-wavelength and multi-messenger follow-up studies that arecrucial to several open problems in high-energy astrophysics. Detailed morphological and spectral studies of severalsources are being conducted and will be the subject of future publications. A dedicated survey to constrain theemission from various extra-galactic objects of interest is under preparation. Future gamma-ray observatories suchas CTA (Acharya et al. 2018) and SWGO (Albert et al. 2019) will be able to extend both the sensitivity and energyrange of this survey. ACKNOWLEDGMENTSWe acknowledge the support from: the US National Science Foundation (NSF); the US Department of Energy Officeof High-Energy Physics; the Laboratory Directed Research and Development (LDRD) program of Los Alamos NationalLaboratory; Consejo Nacional de Ciencia y Tecnolog´ıa (CONACyT), M´exico, grants 271051, 232656, 260378, 179588,254964, 258865, 243290, 132197, A1-S-46288, A1-S-22784, c´atedras 873, 1563, 341, 323, Red HAWC, M´exico; DGAPA-UNAM grants IG101320, IN111315, IN111716-3, IN111419, IA102019, IN112218; VIEP-BUAP; PIFI 2012, 2013, PRO-FOCIE 2014, 2015; the University of Wisconsin Alumni Research Foundation; the Institute of Geophysics, PlanetaryPhysics, and Signatures at Los Alamos National Laboratory; Polish Science Centre grant, DEC-2017/27/B/ST9/02272;Coordinaci´on de la Investigaci´on Cient´ıfica de la Universidad Michoacana; Royal Society - Newton Advanced Fellow-ship 180385; Generalitat Valenciana, grant CIDEGENT/2018/034. Thanks to Scott Delay, Luciano D´ıaz and EduardoMurrieta for technical support. REFERENCES