We present an analysis of the activity of the Perseid meteor stream in the periods July 5-16 (early Perseids) and August 25-31 (late Perseids) based on data collected by the observers of the Comet and Meteor Workshop (CMW) during 1996-2000. Perseids activity before July 13 was at the level ZHR <= 1.4 and increased to ZHR = 2.3 at by July 14, (solar long. = 112^o.). We cannot find a clear radiant position related to the Perseid shower before July 10. The activity of the late Perseids was at the constant level ZHR = 1.2 in the period from August 25 to 31. For the late Perseids we detect a probable radiant position at alpha = 50^o, delta = +56^o, i.e. about 10^o SW from the expected place.

Perseids activity is usually assumed to last from July 17 to August 24 [1]. Zenithal Hourly Rate (ZHR) profiles published for this period (for instance Figure 7 in [2]) indicate that, at its limits, activity is still noticeable, at the level of ZHR ~ 3. There are also some Perseid observations outside assumed activity period (for instance Figure 3 in [3] and Figure 3 in [4]). Thus we decide to search for such events in the data collected by observers of the Polish Comets and Meteors Workshop (CMW) . We selected nights from July 5 to 16 and from August 25 to 31. We analyze such periods in five consecutive years: 1996 to 2000.

In our analysis we used visual observations with plotted tracks collected during 1996-2000 in the periods July 5-16 and August 25-31, selected according to the following criteria:

effective observing time T_eff >= 30 min,
limiting magnitude LM >= 5.00,
radiant altitude h_R >= 20^o,
cloud correction factor F <= 1.50.

We classified meteors as Perseids according to following (well known) criteria:

The distance (l_br) from the beginning of the meteor track to the radiant position should be larger than 1.8 times the meteo track lenght (l_t).
The distance (x) from the back-projected line of the meteor track to the radiant position should be smaller than limiting value (x_max) which depend on the l_br. For x_max we assumed x_max = 7^o for l_br <= 15^o, x_max = 7^o + 0.1*(l_br-15^o) for 15^o < l_br <= 30^o , x_max = 8.5^o + 0.075*(l_br-30^o) for 30^o < l_br <= 70^o and x_max = 11.5^o for l_br > 70^o.
The meteor angular velocity omega in degrees per second, calculated from the equation omega = 0.573 * V * sin h_b * sin l_er, colud differ by at most one velocity class from the velocity of the velocity class estimated by the observer. Here V is the geocentric velocity of the meteoroid stream in km/s, h_bis the altitude of the beginning of the meteor track, and l_er is the distance between the end of the meteor track and the radiant position. The velocity classes are defined by 0^o/s < A <= 5^o/s < B <= 10^o/s < C <= 15^o/s < D <= 20^o/s < E.

We assumed that the Perseid radiant at 0^h00^m UT on August 13 was at alpha = 46^o and delta = 58^o, that its daily drift in ecliptic longitude was Delta_lambda = 1^o, and its geocentric stream velocity V=59 km/s.

This leaves some ambiguity, because according to the above criteria some meteors belong to two different showers. Many of these double assignations were for Alpha Aurygids and late Perseids. In such cases we selected the most probable shower according to a valocity criterion. When both showers were equally good, for instance when the calculated valocity in both cases differed by one velocity class from that estimated by the observer, the meteor was not counted as a Perseid.

Tables 1 and 2 give information about the effective observing time (T_eff), the number of observers (N_obs), the number of Perseids according to the above criteria (N_PER), and the number of sporadic meteors (N_spor), for individual years and for all years aggregated.

Table 1: Early Perseids (July 5-16)

Year	Teff [h]	Nobs	NPER	Nspor
1996	68.44	23	58	462
1997	64.99	23	45	366
1998	75.70	25	55	444
1999	219.44	22	149	1411
2000	30.66	8	8	105
Total	459.23	78	312	2344

Table 2: Late Perseids (August 25-31).

Year	Teff [h]	Nobs	NPER	Nspor
1996	10.83	2	2	38
1997	31.60	8	18	244
1998	17.53	3	14	81
1999	60.05	4	15	236
2000	66.77	9	61	372
Total	186.78	18	110	971

Activity profile

In calculating the ZHR for early and late Perseids we assumed thet r=2.6, similar to the value presented in Figure 1 of [6]. If the population idex is changed to r = 2 the calculated ZHR is 10-30% lower, depending on the limiting magnitude of the observations. We divided the observations into sutable solar longitude bins, and for every bin we estimated ZHR as: ZHR = (N_PER + 1)/T_EFF, where T_EFF=T_eff*sin h_R/(r^6.5-LM*F). One sigma error bas was estimated as .............

Radiant position

The equatorial coordinates of start and end points and the velocity classes of meteors from analyzed periods were obtained using the CooReader package [7] and input into the RADIANT software [5]. To obtain the radiant probability distribution (RPD) maps presented, we used the following parameters: geocentric stream valocity V = 59 km/s, meteor angular velocity 1^o/s < omega < 30^o/s, maximum distance between start of meteor track and radiant position l_max=50^o.

The most important year for early Perseid ZHR estimation was 1999 (see Table 1), but we present the average ZHRs from all years. The results are presented in Figure 1.

One can see that early Perseid activity is at the level of ZHR ~ 1 before sol. long. = 112^o, and by sol. long. = 112^o is incerasing to ZHR = 2.4 +\- 0.3.

References

Arlt R. (1992), WGN , 20 , 62.

Babadzhanov P.B., Kramer E.N., (1965), Smits. Contr. Astrophys. , 11 , 67.

Denning W.F., (1919), J. British Astron. Assoc. , 29 , 161.

Molau S., (2000), Proceedings of the International Meteor Conference, Frasso Sabino, Italy, 1999, 31-37.

Jenniskens P. (1994), Astron. Astrophys. , 287 , 990.

Olech A., Gajos M., Jurek M., (1999a), Astron. Astrophys. Suppl. , 135 , 291.

Olech A., Jurek M., Gajos M., (1999b), Proceedings of the International Meteor Conference, Stara Lesna, Slovakia, 1998, 58

Olech A., Juerk M., (2000), Proceedings of the International Meteor Conference, Frasso Sabino, Italy, 1999, 83.

Samujllo M., Olech A., (1999), Proceedings of the International Meteor Conference, Stara Lesna, Slovakia, 1998, 65