Low-frequency observations of giant pulses from ordinary pulsars
LLow-frequency observations of giant pulses fromordinary pulsars
A. N. Kazantsev ∗ , M. Yu. Basalaeva , P. N. Lebedev Physical Institute of the Russian Academy of Sciences,Pushchino Radio Astronomy Observatory, Pushchino 142290, Russia Astrophysical school Traektoria,107078, Moscow, Russia
Abstract
We present our results of investigation of the rate of emission of thegiant radio pulses (GRPs) from several second period pulsars, observedwith Large Phased Array radio telescope of Pushchino Radio AstronomyObservatory at 111 MHz. It was found that for all pulsars detected ratewas not constant and changed with time significantly. Abrupt jumps inthe rate of GRPs generation were detected for PSR B0950+08 and PSRB1112+50. We found that GRPs of all pulsars have demonstrated verydifferent clustering properties. Finally, we have carried out the phaseanalysis of the time of arrival for GRPs and detected their phase withreference to the averaged pulse time of arrival.
The vast majority of radio pulsars show small pulse-to-pulse intensity varia-tions within the limits of 10 times of the intensity of averaged pulse. However,some pulsars demonstrate unpredictable short-duration outbursts of pulsed ra-dio emission. These pulses were named Giant Radio Pulses (GRPs).At the present day, only several about 16 pulsars are known to be GRPemitters [1–14], i.e. emit individual pulses that satisfy GRP criteria. Thesepulsars can be divided into two groups: pulsars with strong magnetic fields attheir light cylinders (over B LC > Gauss) and millisecond periods (includingB0531+21, Crab) and ordinary pulsars with B LC from several to several hundredGauss. Well-known representatives of the former group are Crab pulsar [1]and millisecond pulsar B1937+21 [2]. There are a lot of theoretical models,concerning GRPs from bright, fast-rotating pulsars of the first group [15–18]but much less is known about pulses from the ordinary pulsars with GRPs. ∗ E-mail:[email protected] (ANK) a r X i v : . [ a s t r o - ph . H E ] M a y able 1: Parameters of Pulsars.Pulsar name RAJ DECJ P dP/dt DM PEpochepoch 2000(1950) (J2000) (J2000) s s/s cm − pc MJDJ0304+1932 (B0301+19) 03h04m33.115s ◦ (cid:48) . (cid:48)(cid:48) ◦ (cid:48) . (cid:48)(cid:48) ◦ (cid:48) . (cid:48)(cid:48) ◦ (cid:48) . (cid:48)(cid:48) ◦ (cid:48) . (cid:48)(cid:48) ◦ (cid:48) . (cid:48)(cid:48) Our observations were made with The Large Phased Array transit radio tele-scope of Pushchino Radio Astronomy Observatory (LPA LPI) of Astro SpaceCenter, P.N.Lebedev Physical Institute.This is a low-frequency radio telescopewith central frequency equal to 111 MHz, and effective bandwidth equal to 2.3MHz. During our observation sessions, the telescope effective area was 20 000 ± m in the zenith direction. The Single linear polarization was used.For the majority of observations we used the sampling interval 1.2288 ms. Thepulsars can be observed only during their culmination when they cross the beamof the LPA LPI. As a result, the duration of observations depends on the decli-nation of a pulsar. For example, the duration of observations of PSR B0809+74is equal to 11 minutes, and PSR B1237+25 - 3.5 minutes.Our data processing pipeline can be described as follows. The digital re-ceiver get phased analog signal from the radio telescope. After that,using thesynchronizer that generates trigger pulses, the signal is split into a sequence ofsegments with the duration equal to a period of a given pulsar. The synchro-2izer has ±
100 ns accuracy when converting GPS time to the universal timescale and ±
10 ns accuracy when setting the start time of the receiver. Theseaccuracies exceeds any reasonable requirements for accuracy of observations atthe central frequency of LPA LPI.The signal is digitized at a frequency of 5 MHz for each pulse. The digitizedpulse is accumulated in the receiverâĂŹs buffer. After digitization, all readingsfor given pulsar are loaded into the fast Fourier transform hardware processorfrom the buffer. The processor returns digitized signal as a raw-file observation.The file consists of a header with common information(name of pulsar, starttime of observation, time sample, etc.) and array of time-series spectra. Dataof time-series spectra are recorded as 32-bit floating-point numbers.This alghorithm does not include absolute calibration of intensity. Thus,units of intensity are analog-to-digital converter (ADC) units.The next stage of data processing is offline de-dispersion procedure. Thede-dispersion process is done at a fixed DM for a given pulsar (see Tab. 1). Thede-dispersed pulses for a given observation are saved for subsequent analysis.For each session of observation, the dynamic average pulse of pulsar is builtby summing individual pulses from the observation. The individual pulse isconsidered to be a giant radio pulse if its peak flux density exceeds the peakflux density of dynamic average pulse by 30 times or more. For each detectedGRP the next parametrs were calculated: • Time stamp of first sample (MJD), • Time of arrival (MJD), • Number of pulse in the observation (arbitrary units), • Amplitude of backgroung(ADC units), • Amplitude of GRP (ADC units), • STD noise (ADC units), • Width of pulse at 50% of peak (ms), • Width of pulse at 10% of peak (ms).The rate of GRPs generation was calculated in 30 days long bins and normal-ized to the total durations of observations in corresponding bins. Additionaly,we have performed search for clusters of GRPs, trying to detect consecutivepulses with peak flux density satisfying our criteria.The phase analysis was carried out in several stages. The phase distribu-tions of giant and non-giant pulses by the longitude of the averaged pulse wereanalyzed at the first stage. The phase was detected simply as a maximum pointof individual pulse. During the second stage, the averaged giant radio pulseand the averaged pulse were calculated by summing all detected pulses from3able 2: Duration of Observations of Each Pulsar.Pulsar name Years Sessions of observations Total observations time, hB0301+19 2012-2016 44 2.42B0809+74 2012-2017 587 16.39B0950+08 2012-2017 606 32.48B1112+50 2011-2018 1214 105.85B1133+16 2013-2018 781 42.98B1237+25 2012-2018 606 37.92Table 3: Rate of GPs generation.Pulsar name Maximum, min − Average, min − B0301+19 0.59 0.09B0809+74 0.03 0.01B0950+08 3.10 0.51B1112+50 0.70 0.24B1133+16 0.30 0.07B1237+25 0.38 0.07pulsars. For the calculation of a location, a half-width and amplitude of the av-erage pulses, they were fitted by Gaussian functions. For fitting one componentof the average pulse profile the next function was used: f ( t ) = ae − ( t − b )22 σ (1)where: a is the height of the curve’s peak, b is the position of the center ofthe peak and σ is the standard deviation.The double- or three-component average pulse of the pulsar was fitted bythe sum of two or three Gaussian functions respectively. Summarized information about statistical properties of rate of GRPs generationis presented in Tab. 3. Also we separately present information about clusteredGRP in Tab. 4. Hereinafter , we provide detailed information about each pulsar..
PSR B0301+19 demonstrates an unstable rate of GRPs generation (see Fig. 1).The average rate is approximately 5 giant pulses per hour. Around epochMJD 57000 a leap in the rate of GRPs generation can be seen. The maxi-mal rate reached a value of 35 impulses per hour. For a significant number ofsessions, no individual pulses which can be classified as GRPs were detected.4able 4: Clusters of GRPs. Numbers of detected pairs of GRPs are in thesecond column, number of instances of clusters with >2 GRPs are in the thirdcolumn, and in the last column we show the maximal size of the clusters foreach pulsar.Name 2 GRPs >2 GRPs Max size of clusterB0301+19 1 0 2B0809+74 2 0 2B0950+08 29 4 3B1112+50 95 7 4B1133+16 1 0 2B1237+25 3 0 2Only one cluster of GRPs from PSR B0301+19 was detected: two consecu-tive GRPs were registered at the MJD 57106. Fig. 2 shows this cluster.Phase distribution of individual pulses from PSR B0301+19 is shown in theFig. 3. The giant pulses of the pulsar are located along the longitudes of themain components of the average pulse of the pulsar. The GRPs mostly occurduring phase interval of the second component rather than the first (n relationof 3 to 1). However, the GRPs from the first component are more concentrated,located in a narrow interval 25.19 ms ± ± ± As we have already mentioned, PSR B0809+74 is usually not considered to bea GRPs emitter. Nevertheless, it sometimes generates individual pulses thatexceed the amplitude of the average profile by 30 or more times. The rate of ab-normally strong pulse generation is the lowest of studied pulsars (see Fig. 1). Aswith PSR B0301+19, there are no strong individual pulses in most observationsessions. On average, PSR B0809+74 generates approximately 1 abnormallystrong pulse in 4 hours. The maximum detected rate of GRPs generation is 2pulses per hour.We find cluster of 2 consecutive strong pulses which were registered at theMJD 56685 (see Fig. 5);Strong pulses are located in the middle of the longitude of the average pulse(see left panel of Fig. 6). Despite large number of detected strong pulses, theaverage pulse is highly jagged (see the right panel of Fig. 6).5 .3 PSR B0950+08
PSR B0950+08 showed a very stable GRPs generation rate (see Fig. 1). Thereare onle several ’empty bins’ – bins without registered GRPs. The rate ofgeneration varies between from 0.1 min − and 0.9 min − with the averagevalue equal to 0.5 min − or, i.e. 30 GRPs per hour. There is a jump in ratearound epoch MJD 56500 where it increased approximately sixfold, up to 180GRP per hour.We detected 33 clusters of GRPs from PSR B0950+08. Four of these clustersinclude 3 consecutive GRPs. The example of one of these groups is presentedon Fig. 7. It can be seen that these GRPs are part of even large group groupof strong pulses, although not all of them have reached the threshold level (30average profiles).The average pulse of pulsar B0950+08 is fairly complex (see the right panelof Fig. 8). It includes precursor area (0 - 25 points) and main pulse (25 - 40points). Most frequently, GRPs are generated on the second component of themain pulse of pulsar. As a result, the shapes of averaged GRP and averagepulse are quite different. Still, distribution of GRPs is not too narow. Of all researched pulsars, PSR B1112+50 demonstrates the most stable rate ofGRPs generation (see Fig. 1). In particular, this is especially noticeable in therange from epoch MJD 56500 to the last analyzed observation session. Thisincrease in the rate of GRPs generation began in approximately MJD 57350and lasted around 90 days. This is the longest increase in the rate of GRPsgeneration registered in our work. At a maximum rate of GRPs generationfor this pulsar reached 42 pulses per hour. That should be compared with theaverage value around 14 pulses per hour.Also we detected the largest number of clusters from this pulsar – 102. In7 cases number of GRPs in them exceeded 2 and the largest cluster contained4 events. In Fig. 9 we show one of the largest cluster registered in MJD 57489.This group was generated only 21 pulsar period after another cluster of GRPswhich included 2 pulses. In another case, presented in Fig. 10, clusters of 4GRPs and 2 GRPs are separated by faint but still clearly visible individualpulse.The simple gauss-like shape of an average pulse is typical for the pulsarB1112+50. However, there is an asymmetry of the averaged pulse shape (seeright panel of Fig. 11). GRPs from B1112+50 are distributed randomly alongthe average pulse longitude which is wider that the distribution of non-giantpulses. The averaged GRP is 1.4 times narrower than the averaged pulse whichincluded all detected pulses from the pulsar. There is a significant offset betweenthe peak of the averaged GRP and the peak of the average pulse.6 .5 PSR B1133+16
The rate of GRPs generation form B1133+16 is visibly unstable (see Fig. 1).The maximum of the rate generation was 18 GRPs per hour. On average, thepulsar generates around 4 GRPs per 1 hour. But as it for the pulsar B0301+19,there were no detections in numerous sessions.During large number of observational sessions, the pulsar B1133+16 gener-ated only single cluster of GRPs on MJD 57658 (see Fig. 13 ).The pulsar B1133+16 has double-component average pulse (see right panelof Fig. 14). The distribution of non-giant pulses shown that weak pulses can berecorded between the main components of the profile. GRPs were detected onthe main components longitudes only. The fitting results have shown that thedistance between the main components for the averaged GRP and the averagepulse is significally different (see Fig. 15). For the average pulse, the distanceis equal to 30.34 ms ± ± The pulsar B1237+25 showed a quite unstable GRPs generation rate (see Fig. 1).Giant radio pulses are detected in 70% of the observational sessions. On averagePSR B1237+25 generates approximately 4 impulses per 1 hour. The maximumvalue of the GRPs rate was 22 pulses per hour. We detected 3 clusters of GRPs(see example in Fig. 16).The average pulse of B1237+25 is very complex and includes five compo-nents. Only three of the components (1, 3, and 5) can be reliably identified inone observational session. The sum of numerous individual pulses allowed us todetect all five components of the average pulse. As seen in Fig. 17, the giantradio pulses are emitted on the longitudes of the 1, 3, and 5 components. Wedetected only one GRP on the 4-th component of the average pulse. For fittingsuch a complex average pulse, we use the sum of 5 Gaussian functions (see leftpanel of Fig. 18). According to the results of the fitting, the positions of 1, 3,and 5 components for average pulse are equal to 20.84 ± ± ± ± ± ± The main result of our study is that the rate of GRP generation in all pulsars isnot constant (see Fig. 1). One can interpret this finding in several ways. Firstly,it may reflect the instability of mechanism of GRPs generation. However, theemission rate of typical individual pulses is also subject to strong variability. Inthis way, the detected rate of GRPs may reflect the common nature of pulseemission, both giant and normal. 7 small number of the detected GRPs clusters from PSR B0301+19 is likelyrelated to a short time span of observations. The GRP rate will be defined moreprecisely with increased observational time.The rates of GRPs generation of the pulsars B1237+25 and B1133+16 arevery similar –they have almost the same value of the average rate and themaximum rates are quite close – 18 and 22 GRPs per hour respectively.Another strikimg feature of PSR B1112+50 was the longest train of GRPs.At period 1.6564 s (see Tab. 1) duration of emission of this group of giant pulseswas approximately 10 seconds. It could mean that the process giving rise toGRPs in the magnetosphere of this pulsar has a lifespan of considerable duration.This is the only pulsar in our study which demonstrated such extrardinarybehavior.For the majority of our pulsars, the GRPs are located on the longitudes of themain component of the average pulse. The averaged GRPs is not significantlyshifted in phase with respect to the average pulse, meaning that the GRPs areemitted in the same region as non-giant pulses.Only for pulsar B1133+16 the distance between the main components foraveraged GRP is significally narrower than that for the average pulse. Thismay indicate that the GRPs are emitted from slightly different regions thannon-giant pulses. More research should be done for this pulsar.
It was found that pulsars B0301+19, B050+08, B1112+50, B1133+16, andB1237+25 demonstrate unstable rate of GRPs generation during all time ofobservations.Giant radio pulses from these pulsars are distributed along the longitudesof the main components of an average pulse. Abnormally strong pulses fromB0809+74 are strictly located in the middle of the average pulse of the pulsar.This fact shows, that such anomalous pulses are formed rather in the core areaof the emission region than in the drifting sub-pulse, which formes essentiallypart of the pulsar’s profile.For pulsars B0301+19, B050+08, B1112+50, and B1237+25 averaged GRPand average pulse have no significant difference by phase. However, the pulsarB1133+16 demonstrated a narrower average pulse for GRPs than for averageordinary pulses.We have detected a huge pack of giant radio pulses emitted by B1112+50.The pulsar is the only one that emitted more than 3 consecutive GRPs. More-over, pulsars B0950+08 and B1112+50 demonstrated jumps in the rate of GRPsgeneration – in a rather short time the rate increased in 6 and 3 times respec-tively.In summary, we have shown that the pulsars with very similar parametersdemonstrate significant difference in the rate of GRPs generation and GRPclustering. 8 cknowledgements
The authors are grateful to program committee PRAO ASC LPI for the timeavailable on LPA LPI. AK acknowledges the support by the Foundation for theAdvancement of Theoretical Physics and Mathematics «BASIS» grant 18-1-2-51-1. We acknowledge the Traektoria fundation for for valuable support of thiswork. We thank our colleagues V. A. Potapov (Pushchino Radio AstronomyObservatory) and M. S. Pshirkov (Sternberg Astronomical Institute) for usefuldiscussions and contributions during the preparation of this paper.