9 Fragmentation

Throughout the entire video-recorded trail, the main body has been the brightest and the leading. The preceding sections dealt with the main body. However, in addition to the main body, numerous smaller fragments and a wake are visible on the video record. We were able to measure positions of 13 individual fragments and positions of the points of the wake termination, all relatively to the main body. The instantaneous direction of motion of the main body and its distance from the video-camera derived in the preceding sections were used to transform the measured angular distances into metric distances. The results are given in Figs. 5 and 6.

The total metric distance L of a fragment (or the terminal point of the wake) from the main body can be expressed by means of the following 3 components: the one along the trajectory, l, is the distance L projected perpendicularly onto the main body trajectory (counted positively behind the main body); the two components lateral to the trajectory can be expressed as d, the distance L projected perpendicularly onto the line which is perpendicular to both the trajectory and the instantaneous vision line (the + sign means that the point is above the trajectory, and the - sign means below the trajectory), and the other one, p, the distance L projected perpendicularly onto the vision line. Because the information on the individual angular distances of fragments from the main body is available from only the one video station, we do not know the lateral component p. On the other hand, dynamical reasons made l much larger than both d and p (the observed d does not exceed 0.3 km, while l are as large as 6 km. There is no reason to assume that the two lateral distances p and d would be too much different. If p and d do not exceed 10% of l (which is the case for almost all measured values), the difference L-l does not exceed 3%, a negligible difference in scope of all the other uncertainties. Thus the distance l behind the main body is very close to the total metric distance L.

  

Figure 5: Distance l of individual fragments behind the main body is plotted against time t=n/25, where n is the frame number. Fragments are numbered. The symbol "w" is used for the wake and means the terminal point of the wake behind the main body. Points represent direct measurements on individual frames; lines fitting these points represent a smooth change of l for the individual fragments. Fragment 3 was observed only on one frame. The wake consists of two parts, before and after the main fragmentation, which took place between 3.5 and 3.7 s

The fragments separated at different time instants (Fig. 5). Fragments 1 and 2 separated somewhere close to $t \doteq 1$ s: this early fragmentation was not included into the theoretical model described in the preceding section, because it was not visible on the light curve (except that the whole small increase of brightness before t=2 s might be related to fragment 1). Thus the initial mass might be somewhat (by $\approx\!10$%) greater than derived in the section on theoretical modeling. Also the first part of the wake, which seems to be mostly gaseous, could contain small fragments from this early fragmentation; this part of wake reached the maximum distance behind the main body of more than 5 km at t=1.7 s, and then quickly shortened to values below 1 km. Fragments 3, 4, 5, 6, 7, 9, 10, and 13 separated from the main body during at-least-two closely-separated events between t=3.5 and 4 s. This corresponds to the main double maximum on the light curve at t=3.6 and 3.7 s. Fragments 9, 10, 13 may be secondary fragments originating from fragment 5. Fragment 8 separated from the main body at $t \doteq 4.6$ s. Fragment 11 is a secondary one separated from fragment 8. Small fragment 12 separated at $t \doteq 5.3$ s; its significance for the theoretical model of mass-loss of the main body is negligible.

  

Figure 6: Lateral distance d of individual fragments above (+) and below (-) the trajectory is plotted against time. They represent the minimum lateral distance possible. Points representing measurements on individual frames are given only for the first part of the wake. They show a typical spread of our measurements. Only the average smooth lines are given for all fragments and the second part of the wake "w" (plotting points of individual measurements would make the figure uncomprehensible). The spread of the points for individual fragments is of the same order as the total change of d over the corresponding time interval; thus only average values of d are available for most of the fragments

Thus fragments 3 to 13 (except 12) originated from 3 major break-up events with the largest mass-loss rate (dm/dt) as is evident from Fig. 4. Later break-ups after t=5.5 s (at least two major events more) did not show individual fragment trails: it is demonstrated only by a very slow brightness decay after the main maximum and by a bump on the light curve between t=6 and 7 s. Theoretical mass-loss rate in Fig. 4 between 5 and 8 s is $3\times$ less than between 3.5 and 5 s. Fragments released after t=5.5 s were already so small that they contributed only to the light of the main body collectively (evaporating dust cloud), and were invisible (below the sensitivity limit) when they traveled to distances, where their images could be separated from the main body. The release time (from break-up to visibility) was about 0.5 s for all break-ups. This value proved also to be the best for the theoretical modeling of the light curve and was also in agreement with first visibilities of individual fragments (Fig. 5).

Lateral distances d shown in Fig. 6 are small values if compared to l. This means that spread of the d-values for any fragment measured at individual frames is relatively large and only average values or linear trends could have been determined. Most of the fragments were close to the trajectory or by most of 0.3 km below it. The initial wake termination started more than 0.2 km above the trajectory and moved then just onto the trajectory. Small fragment 2 is an exception: this was the only one with rather quick lateral motion; the average value in Fig. 6 originates from three consecutive values: 3.6 s, 0.6 km above the trajectory; 3.64 s at the trajectory; 3.68 s, 0.6 km below the trajectory.

Acknowledgements

Our sincere thanks are due to Dr. V. Tamazian from Astronomical Observatory "R.M. Aller" for his helpful comments, to Mr. J. Ramirez from Supercomputers Center of Galicia for his special work in video tape digitization. We are very much obliged to the authors of the video record of the bolide, Mr. A. Salazar, Mr. D. Neira, and Mr. J.A. Quiroga for their kind collaboration. Our special thanks are appointed to the Investigation Vice-Rectorate of the University of Santiago de Compostela for their financial support. We are pleased to announce that this work has also been supported by Contract AJ-4706 of the Sandia National Laboratories, and by the Grant Agency of the Czech Republic (205/97/0700).