Characteristic for long GRBs with high energy component presence, which not required cosmological corrections

Several thousands of gamma-ray bursts were observed by various experiments. During several GRBs very high-energy photons were detected both in space and ground-based experiments (up to some tens of GeV and up to some TeV, respectively). Usually 2 classes of bursts are considered: short and long GRBs separated by t90~2s. Because of several hundreds of GRBs located at high redshift, its sources’ origins nature concluding as cosmological. Therefore correction to cosmological dilation of GRBs t90 should be considered during any analysis of bursts duration. Firstly very high-energy component was observed during GRB 970417a: 18 photons with energy ~650 GeV were registered by Milagrito within t90 interval of this burst. Now several tens of GRBs reveal activity in energy bands up to some tens of GeV and up to some TeV accordingly data of space and ground-based experiments correspondingly. Unfortunately redshift is unknown approximately for half of GRBs with high energy component presence. Here we introduce new parameter Rt is ratio of maximum energy photon arrival time to burst duration and it not required cosmological correction. At least 2 groups of long GRBs could be separated using parameter Rt: for 25% events highest energy gammas detected within t 90 interval, but for other 75% of bursts it registered more than 10 sec. later than one. Moreover, preliminary results of analysis allow concluding 2 subtypes of second group GRBs. For one μ-quantum with maximum energy arrived within t90. For other such photon was registered later than t90. Therefore, the results of preliminary analyses allow conclude long GRBs population inhomogeneity.


Introduction
GRBs observed since the end of 60 th of previous century and now several thousands of events were listed in more than 20 catalogues as results of more than 40 satellites and ground experiments. During several GRBs very high-energy photons were detected both in space and ground-based experiments (up to some tens of GeV and up to some TeV, respectively). High energy (HE) GRB emission was firstly registered by the satellite experiments onboard the Compton Gamma Ray Observatory (CGRO) in time interval 1991-2000 [1]. Four experiments onboard CGRO: BATSE, OSSE, COMPTEL and EGRET [2] provided the widest energy range since of 10 keV up to more than 20 GeV. Spectra of most part of such bursts were well described by two components Band model [3]. But in some GRB spectra the new spectral components not corresponded to Band model was found up to 200 MeV, for example during GRB 941017 [19]. So, it was possible to introduce 2 spectral breaks in prompt emission of GRBs E1 -between 2 components of Band model and E2 -corresponds to HE component -see figure 1. Firstly confirmation of high energy afterglow was found during analysis of 2 data of GRB940217: γ-emission with energy more than 50 MeV was registered till 1.5 hours after start of burst, highest-observed energy of gamma was 18 GeV [5]. CGRO registered 15 GRB with E>120 MeV, but mostly no prompt emission E>200 MeV [6].
Firstly evidence of subTeV emission was observed during GRB 970417a: several photons with energies ~650 GeV were detected by Milagrito during this burst duration t90 [20]. This fact allow to conclude a possibility of prompt emission with very high energy during GRBs.
Than GRBs high energy emission was observed in Russian experiment AVS-F onboard satellite COROPNAS-

Modern results of observations of high energy gamma-emission during GRBs
Next step of HE gamma-emission observed observation was start with launch of satellite experiments Fermi/LAT [8,9] and AGILE/MCAL [10]. Also it followed by Cherenkov detectors MAGIC [11] and H.E.S.S. [12]. Now ~170 GRBs were registered by Fermi/LAT [13] and ~70 bursts observed by AGILE/MCAL [14]. But third spectral breaks should be introduced in very low energy band based on data of these experiments -see figure 3. And it is very interest fact that spectral indexes are similar in very low and in very high energy bands. So, very interesting question occur about extension of HE component to low energy region down to tens of keV.
For most part of GRBs sources' origins nature is cosmological -see redshift corresponding columns in catalogues, for example, [13,14,15]. Therefore correction to cosmological dilation of GRBs duration should be consider because of real cosmological sources time properties should be investigated only taking into account its redshift. Typically short and long GRBs classes considered using the results of burst duration distribution analysis. GRBs duration characterized by t90 and t50 accordingly to BATSE data analysis [16]. These parameters are the intervals for 90% and 50% of burst statistics accumulation (i.e. durations where the integrated burst's counts raise from 5% to 95% and for 25% to 75% respectively). Short GRBs are defined to have durations t90 of <2s accordingly to BATSE data analysis [16], but after correction to cosmological dilation several long events could interpreted as short in this classification. For example, t90 of long GRB160625B is 2.8 sec, z~1.4 and after correction it became t90z ~1.2 sec which correspond to burst of short class. Also the same values for GRB110731A are t90~7.3 sec, z~2.83, t90z ~1.2 sec as relate to short class burst again. LAT GRBs distribution on burst duration t90 and cosmologically corrected t90_z is presented in figure 3. Of course, several correlations exist between burst duration and t90_z, but it is impossible to     LAT GRBs distribution on burst duration and maximum registered energy is shown at figure 4. It is seen that after correction to cosmological dilation most part of GRBs were shifted in time interval with t90 from 2 to 30 sec. Unfortunately only several tens of LAT GRBs has information both about redshifts and t90. The bursts distribution on high energy episode duration and t90 is presented at figure 5. It is possible to conclude 2 long GRBs subgroup existence separated by limit where maximum energy photon arrival time is equal to event t90.
For next analysis we introduce new value Rt is ratio of maximum energy photon arrival time to burst duration and this value not required cosmological correction. The distribution of LAT GRBs on Rt and t90 is shown at figure 6. The investigation results conclude 2 long GRBs subgroup existence separated by limit where maximum energy photon arrival time is equal to event t90. During events of the first subtype high energy emission duration interval is smaller than t90. Second subtype characterized longer period of high energy emission than t90.   [17]. Unfortunately, only 3 bursts were registered in subTeV region during Fermi operation: GRB 190114C, GRB 180720B and GRB 160821B [18,19,20]. Now GRB high energy gamma-emission was observed both during short and long bursts, but photons in the band E>0.1 TeV usually are observed only during long GRBs (now only one short GRB appear such emission). Long GRBs with subTeV emission has characteristics the same than 2b type and 1 type accordingly to preliminary results of analysis. As 2b type events we can consider GRB 190114C during which MAGIC start registration of about 50 s after the trigger and detected photons with E > 300 GeV for the first 20 min from this burst with a significance higher than 20σ [18]. GRB190114C is near long burst (z = 0.4245 and t90~120 s in low energy band [21,22]).
Also GRB180720B reveals similar properties: H.E.S.S. began observation of this burst at about 10 h after the burst trigger and detected 100−440 GeV photons at such late time interval [19]. GRB 180720B is near long burst (z = 0.6535 and t90 ~150 s in low energy band [23,24]). We can consider GRB 190829A as 1 type event: prompt emission during 190829A was detected by H.E.S.S. to in subTeV band. It is very near long burst z = 0.0785 +/-0.005 [25].

Conclusion
Several thousands of gamma-ray bursts were observed by various experiments. During several GRBs very high-energy photons were detected both in space and ground-based experiments (up to some tens of GeV and up to some TeV, respectively). For example, GRB 190114C was detected by Fermi and MAGIC in very wide band up to subTeV energies. Now GRB high energy -emission was observed both during short and long bursts, but photons in the band E>0.1TeV usually are observed only during long GRBs (now only one short GRB appear such emission).
GRBs mostly located at cosmological distances and cosmological correction should be used in duration investigation. But here we introduce new value Rt as ratio of maximum energy photon arrival time to burst duration in low energy band and it not required cosmological correction. At least 2 groups of long GRBs could be separated using parameter Rt: for 25% events highest energy gammas detected within 90 interval, but for other 75% of bursts it registered more than 10 seconds later than one. Moreover, preliminary results of analysis allow concluding three types of GRBs with high energy emission registration without dependence of burst duration value. During events of first subtype high energy emission duration interval smaller than 90. Second subtype characterized longer period of high energy emission than 90. But second subtype bursts divided to 2 subgroups: a) -quantum with maximum energy arrived within 90, b) such photon was registered later than 90. Long GRBs with subTeV emission has characteristics the same than 2b type (for example, GRB190114C and GRB180720B) or 1 type (for instance, 190829A) on preliminary results of analysis.
Therefore, results of preliminary analyses allow conclude long GRBs population inhomogeneity.