| Relativistic heavy-ion collision experiment,which studies the behavior of nuclear matter under extreme conditions of high temperature and high density,is an important way to explore new forms of matter.Early stage of collisions of non-centripetal high energy heavyions produce magnetic fields with an peak intensity in the interaction region estimated to be as high as several times or even 100 times mπ2(~1018 Gauss).The photons are produced through the gluon scattering and splitting processes when strong magnetic field is included.The photon is chosen as the probe for the study of relativistic heavy ion collisions.It is arguably the most common tool for studying relativistic heavy-ion collisions.Firstly,photons can be emitted at every stage of the relativistic heavy ion collision.Secondly,the photon is color neutral,its mean free path is large,and the colliding system is almost transparent relative to the photon.In addition,photons do not participate in the strong interaction,which can be relatively pure to reflect the dynamic information of a certain stage.The photon is a good electromagnetic signal,is a penetrating probe,so that the information reflected by the photon can be studied by us.The dynamic of system at very early stage can be describe as an overoccupied gluonic system,The gluon fields are initialized by the McLerran-Venugopalan(MV)model and evolved via the Classical Yang-Mills(CYMs)equations.We present an event-by-event study of photon are produced through the gluon scattering and splitting processes when strong magnetic field is included in the early stage of high energy nuclear collisions.We study the spectra and collective flows of the photons and show their dependence on transverse momentum qT.We use the boost invariant assumption to dress the highly anisotropic property of initial stage,and introduce the temporal profile to mimic the evolution of the short lived strong magnetic field.We study the photon spectra of Au-Au collision at RHIC energy,and the spectra of the photons are studied in different scenarios,and we find the yield of the photon is sensitive to the magnetic field intensity,magnetic field attenuation mode,magnetic field lifetime,the central classes as well as saturation momentum.The emission of photons favors stronger magnetic fields and more nucleons involved in the interaction.There is an IR singularity,and it is removed by a soft cutoff,and this results in the non-trivial behavior of the collective flows of the photons.The contribution on total v2 is large.We weight our results on top of Parton-Hadron-String Dynamics(PHSD)model in Au-Au 20%~40%centrality collisions at(?)=200GeV,and find that when including the photon produced by gluon,the result of v2 agrees the experiment measure even better,this can effectively explain the”direct photon puzzle",that is,the theoretical value of v2is lower than the experimental value.We weight our results of v3 on top of PHSD model,in qT(?)3GeV region our computation does not affect PHSD result and both of them are in agreement with the experimental results;v3 of the photon produced by gluon fusion and splitting is non-zeroand,which provide the right tendency in qT≥ 3GeV region,and this phenomenon has not been shown anywhere else.In the case of reasonable energy and magnetic field strength,Our calculation results can prove that the presence of magnetic fields in the early stages of high-energy heavy ion collisions provides a channel for photon production.The experimental results of photon spectrum and collective flow at RHIC energy can be better described. |