Anisotropies in the cosmic radiation observed with ARGO-YBJ
aa r X i v : . [ a s t r o - ph . H E ] D ec Anisotropies in the cosmic radiation observed with ARGO-YBJ
R. Iuppa
Department of Physics, University of Tor Vergata and INFN,Sezione di Roma Tor Vergata, via della Ricerca Scientifica 1, 00133 Roma , Italy on behalf of the ARGO-YBJ collaboration
Important informations on the origin and the propagation mechanisms of cosmic rays may beprovided by the measurement of the anisotropies of their arrival direction. In this paper the obser-vation of anisotropy regions at different angular scales is reported. In particular, the observation ofa possible anisotropy on scales between ∼ ◦ and ∼ ◦ may be a key-detection for speculations onthe presence of unknown features of the magnetic fields the charged cosmic rays propagate through,as well as to potential contributions of nearby sources to the total flux of cosmic rays. Evidence ofnew weaker few-degree excesses throughout the sky region 195 ◦ ≤ R.A. ≤ ◦ is also reported. Introduction
As the most part of cosmic rays (CRs) are chargednuclei, their arrival direction is deflected and madeisotropic by the action of galactic magnetic field(GMF) that they propagate through before reachingthe Earth atmosphere. In such a field, the gyroradiusof CRs is given by r a.u. = 100 R TV , where r a.u. is inastronomic units and R TV is in TeraVolt. It must betaken as a reference value, because the GMF is thesuperposition of regular field lines and chaotic contri-butions, the strength of both them being still underdebate. Data available today gives for the local totalintensity the value B = 2 ÷ µ G.Actually, deviations from the isotropy are expectedto occurr as a consequence of the particular realiza-tion of the random distribution of cosmic ray sourcesand magnetic field lines in the galaxy [1]. In thisframework, the amplitude of the anisotropy is propor-tional to the rigidity, that is why it is expected to beseen at high energies (above few hundreds TeV). How-ever, different experiments [2–7] observed an energy-dependent ”large scale” anisotropy in the siderealtime frame, well below that threshold. The ampli-tude is about 10 − - 10 − , suggesting the existenceof two distint broad regions, one showing an excess ofCRs (called ”tail-in”), distributed around 40 ◦ to 90 ◦ in Right Ascension (R.A.). The other a deficit (the”loss cone”), distributed around 150 ◦ to 240 ◦ in R.A..The origin of these anisotropies is still unknown. Someauthors claim that the observations may be due to acombined effect of the regular and turbolent GMF [8],or to local uni- and bi-dimensional inflows [9]. Otherstudies suggest that it can be explained within thediffusion approximation taking into account the roleof the few most nearby and recent sources [1, 10].Easy to understand, more beamed the anisotropiesand lower their energy, more difficult to fit the stan-dard model of CRs and GMF to experimental results.That is why the evidence of the existence of a mediumangular scale anisotropy contained in the tail-in re-gion by the Tibet AS γ [11] and Milagro [12] collabo- rations in rcent years was rather surprising. Similarsmall scale anisotropies has been recently claimed tobe observed by the Icecube experiment in the South-ern hemisphere [7]. So far, no theory of CRs in theGalaxy exists which is able to explain few degreesanisotropies in the rigidity region 1-10 TV leaving thestandard model of CRs and that of the local GMFunchanged at the same time.From the experimental viewpoint, observinganisotropy effects at the level of 10 − with an airshower array is a difficult job, because of the intrin-sic difficulties that this kind of apparatus has to copewith in estimating the exposure.Finally, the observation of a possible small angularscale anisotropy region contained inside a larger onerely on the capability for suppressing the anisotropicstructures at larger scales without, at the same time,introducing effects of the analysis on smaller scales.In this paper the observation of CR anisotropy atdifferent angular scales with ARGO-YBJ is reportedas a function of the primary energy. I. THE ARGO-YBJ EXPERIMENT
The ARGO-YBJ experiment, located at the Yang-BaJing Cosmic Ray Laboratory (Tibet, P.R. China,4300 m a.s.l., 606 g/cm ), is an air shower array ableto detect the cosmic radiation at an energy thresholdof a few hundred GeV. The full detector is in stabledata taking since November 2007 with a duty cyclegreater than 85%. The trigger rate at the thresholdis 3.6 kHz. The detector characteristics and perfor-mance are described in [13]. II. DATA ANALYSIS AND RESULTS
In order to study the anisotropy at different angu-lar scales the isotropic background of CRs has beenestimated with two methods: the equi-zenith anglemethod [14] and the direct integration method [15]. eConf C110509
The equi-zenith angle method, used to study thelarge scale anisotropy, is able to eliminate various spu-rious effects caused by instrumental and environmen-tal variations, such as changes in pressure and temper-ature that are hard to control and tend to introducesystematic errors in the measurement. The methoduses data coming from all the angular scales, so thatpotential small structures are not separated from theunderlying large scale modulation.The direct integration method, based on time-average, rely on the assumption that the local distri-bution of the incoming CRs is slowly varying and thetime-averaged signal may be used as a good estimationof the background content. Time-averaging methodsact effectively as a high-pass filter, not allowing toinspect features larger than the time over which thebackground is computed (i.e., 15 ◦ /hour × ∆ t in R.A.).The time interval used to compute the average spans∆ t = 3 hours and makes us confident the results arereliable for structures up to ≈ ◦ wide. A. Large Scale Anisotropy
The observation of the CR large scale anisotropy byARGO-YBJ is shown in the figure 1 at different pri-mary energy up to about 25 TeV. The data used in thisanalysis was collected by ARGO-YBJ from 2008 Jan-uary to 2009 December with a reconstructed zenithangle ≤ ◦ . The so-called ‘tail-in’ and ‘loss-cone’ re-gions, correlated to an enhancement and a deficit ofCRs, are clearly visible with a statistical significancegreater than 20 s.d.. The tail-in broad structure ap-pears to dissolve to smaller angular scale spots withincreasing energy. It should be stressed that the ener-gies reported in the figure 1 refer to the median energyof all nuclei triggering the experiment.To quantify the scale of the anisotropy we studiedthe 1-D R.A. projections integrating the sky maps in-side a declination band given by the field of view ofthe detector. Therefore, we fitted the R.A. profileswith the first two harmonics. The resulting ampli-tude of the first harmonic is plotted in the right plotof figure 2 where is compared to other measurementsas a function of the energy. The ARGO-YBJ resultsare in agreement with other experiments suggestinga decrease of the anisotropy first harmonic amplitudewith increasing energy. B. Intermediate Scale Anisotropy
The figure 3 shows the ARGO-YBJ sky map inequatorial coordinates. The analysis refers to eventscollected from November 2007 to May 2011 after thefollowing selections: (1) ≥
25 shower particles on thedetector; (2) zenith angle of the reconstructed showers
FIG. 1: Large scale CR anisotropy observed by ARGO-YBJ as a function of the energy. The color scale gives therelative CR intensity.FIG. 2: Amplitude of the first harmonic as a function ofthe energy, compared to other measurements. ≤ ◦ . The triggering showers that passed the selec-tion were about 2 · . The zenith cut selects thedeclination region δ ∼ -20 ◦ ÷ ◦ . According to thesimulation, the median energy of the isotropic cosmicray proton flux is E p ≈ ≈ eConf C110509
011 Fermi Symposium, Roma., May. 9-12 FIG. 3: Intermediate scale CR anisotropy observed by ARGO-YBJ. The color scale gives the statistical significance ofthe observation in standard deviations. multiplicity r e l a t i ve excess -0.100.10.20.30.40.50.60.70.80.9 -3 · p50 E region 1region 2 FIG. 4: Size spectrum of the regions 1 and 2. The verticalaxis represents the ratio between the events collected. Theupper scale shows the corresponding proton median energy(see text).
The most evident features are observed by ARGO-YBJ around the positions α ∼ ◦ , δ ∼ ◦ and α ∼ ◦ , δ ∼ -5 ◦ , positionally coincident with the re-gions detected by Milagro [12]. These regions, named“region 1” and “region 2”, are observed with a sta-tistical significance of about 14 s.d.. The deficit re-gions parallel to the excesses are due to a known ef-fect of the analysis, that uses also the excess eventsto evaluate the background, artificially increasing thebackground. On the left side of the sky map, severalpossible new extended features are visible, though lessintense than those aforementioned.The area 195 ◦ ≤ R.A. ≤ ◦ seems to be full of few-degree excesses not compatible with random fluc-tuations (the statistical significance is more than 6 s.d.post-trial). The observation of these structures is re-ported here for the first time and together with that ofregions 1 and 2 it may open the way to an interestingstudy of the TeV CR sky.To figure out the energy spectrum of the excesses,data have been divided into five independent showermultiplicity sets. The number of events collectedwithin each region are computed for the event mapas well as for the background one. The ratio of thesequantities is computed for each multiplicity interval.The result is shown in the figure 4. Region 1 seemsto have spectrum harder than isotropic CRs and acutoff around 600 shower particles (proton medianenergy E p = 8 TeV). On the other hand, the excesshosted in region 2 is less intense and seems to havea spectrum more similar to that of isotropic cosmicrays. The steepening from 100 shower particles on(E p = 2 TeV) is likely related to efficiency effects.Further studies are on the way.The figure 5 reports the amplitude of the region1and 2 anisotropies as a function of the UT. Time unitis the Modified Julian Date. It can be appreciatedthat no time evolution of the anisotropies are there,nor evidence of correlation of the emission. III. CONCLUSIONS
This paper reports the observation of CR anisotropyat different angular scales with ARGO-YBJ, as a eConf C110509
MJD54400 54600 54800 55000 55200 55400 55600 55800 r e l a t i ve excess -3 X 10
FIG. 5: Region 1 (upper plot) and 2 (lower plot) relativeintensity as a function of UT. function of the primary energy. The large scale CRanisotropy has been clearly observed up to about 25TeV. Evidence of existence of different few-degree ex-cesses in the Northern sky (the strongest ones posi-tionally coincident with the regions detected by Mila-gro in 2008) is reported. The time distribution of the phenomenon has been showed too. A discussion of thesystematic effects which may come from imprecisionsin estimating the reference level is outlined.In fact, a wrong estimation affect the significanceand the relative intensity sky maps, even creating ar-tifacts (i.e. fake excesses or deficit regions). Drifts indetector operating conditions or atmospheric effectson the air shower development are quite hard to bemodeled, then difficult to be accounted for to suffi-cient accuracy. As anisotropies of the order 10 − arelooked for, operating conditions must be known downto this level, all across the field of view and during allthe acquisition time.Background methods applied in this analysisdemonstrated to be effective for analysis of diffusefrom even weaker regions (e.g. see [16]).Given the importance of the topic, a joint analy-sis of concurrent data recorded by different experi-ments in both hemispheres, as well as a correlationwith other observables like the interstellar energeticneutral atoms distribution [17, 18], should be a highpriority to clarify the observations. [1] Blasi P. and Amato E. 2011 arXiv:1105.4521 and arXiv:1105.4529 .[2] Nagashima K. et al. 1998 J. of Geoph. Res.
Phys. Rev. D Science
ApJ
ApJ
L130.[7] Abbasi R. et al. 2011
ApJ
ApJ
L90.[9] Amenomori M. et al. 2010
Astrophys. Space Sci.Trans.
49. [10] Erlykin A.D. and Wolfendale A.W. 2006
Astrop.Phys. AIP Conf. Proc.
Phys. Rev. Lett.
Phys. Rev. D ApJ
ApJ
ApJ
Science