Video Observations and Results in 1995/96

1. Introduction

The last twelve months have been very fruitful for the meteor observers at Archenhold- Observatory Berlin. On the one hand, we were able to follow the activity of most major meteor showers both visually and with video techniques. We obtained accurate radiant plots for different streams and witnessed the incredible outburst of the alpha-Monocerotids on the morning of November 22, 1995. On the other hand, we could provide a basis for extensive video observations in the future in Germany.
This paper deals with the results of our observations as well as some aspects of the new video meteor systems. The camera series itself is described in detail in other papers ([1],[2]).

2. Meteor Shower Observations

The observations and results of the 1995 Quadrantids and Lyrids have been discussed at the last IMC [3]. In autumn and winter we focused our work on the Orionids, Leonids and alpha- Monocerotids.

2.1 Orionids

In October we could observe the Orionids in two nights, our video system MOVIE (see [4] for details) was operated on October 23, 1995. The activity level at the maximum was promising: NITMI and MOLSI observed all together 88 meteors in 3.03 hours effective observing time, whereas MOVIE recorded 76 meteors in 2.31 hours. We changed the camera's field of view in the middle of the observing session. Thus, we recorded meteors in all directions around the radiant (Figure 1), which is an important condition to obtain an accurate radiant plot (Figure 2).

It turned out, that the Orionids produced the narrowest meteor shower radiant we obtained so far. It is obvious, that all shower meteors radiated from a single, sharply defined point in the sky. Table 1 compares our position of the Orionid radiant with the results of other investigators.

2.2 Leonids

A comparison between the radiant position we obtained and other investigations is given in Table 1.
In parallel to MOVIE we operated VK1, the prototype camera of the new video camera series, for the double station project. However, the video tapes of that camera are not yet inspected.

2.3 alpha-Monocerotids

This minor meteor shower was almost unknown before 1995, but it produced undoubtedly one of the most spectacular meteor event that we have ever witnessed.
From a WGN paper [5] by Peter Jenniskens we learnt, that the alpha-Monocerotids showed a brief, but intense outburst in the years 1925, 1935 and 1985, and that there were good chances for another outburst on the morning of November 22, 1995. So we tried to watch the questionable event together with many other meteor observers in Europe.
The weather circumstances were favourable, since we had clear skies promising good limiting magnitudes. On the other hand, winter begun early in 1995, so we had to cope with snow drifts, strong winds and temperatures far below the freezing point. After the odyssey to reach the observing site south of Chemnitz, we were soon rewarded with a most exciting meteor display.
As during the Leonids, we operated two video systems, MOVIE and VK1. Both cameras pointed towards Taurus and started to record the sky before 1:00 UT. When I begun to observe visually at 1:20 UT I immediately realized, that the outburst was just in progress.
It is impossible to describe the emotions when you witness such an unique event after a tough fight with the elements, but it was one of the happiest meteor observations I ever enjoyed. The display with an activity level comparable to the memorable 1993 Perseid maximum lasted only for some 30 minutes, after that the zenith rate jumped down to almost zero. Figure 5 shows the raw number of alpha-Monocerotids that were recorded by MOVIE, which reflects the activity profile very good. Whereas the author counted 52 alpha-Monocerotids between 1:22 and 1:47 UT, MOVIE recorded 36 shower meteors during the outburst, and the prototype camera VK1 provided another sample of 25 shower meteors.

The immediate inspection of the video tapes a few hours after the event indicated, that we had recorded the entire outburst with both video cameras. Thus, together with the DMS observers we are the first group ever, that obtained video observations of an meteor outburst. In fact, due to the good predictions, the 1995 outburst of the alpha-Monocerotids is by far the best ever documented event of that type.
One of the first analysis results was the accurate radiant position of the alpha-Monocerotids, which proved older results of occasional outburst witnesses to be in error by several degrees. Although the plot (Figure 7) has the same longish shape as for the Leonids, because we could not change the camera's field of view during the brief event, we could pin down the radiant with an precision of about two degrees. The position was impressively confirmed by the much more accurate results of double station photography obtained by the DMS group in Spain ([6],[7]).

Later investigations suggested, that the comet C/1943 W1 is very unlikely the parent of that meteor shower as assumed so far [8].
Other analysis are still undertaken. Several observers reported a double peak, which is also supported by our visual and video observations. Peter Jenniskens concluded from the data, that no such double peak occurred on a global scale [7], whereas Luis Bellot-Rubio found indications for it in visual, video and radar data [9]. In fact, we might have observed a sorting effect of smaller and larger particles in the meteoroid stream, since there is a small shift in the activity profiles of MOVIE and VK1. MOVIE (field of view 60 degrees, lm ¯ 6.5 mag for stars) recorded on average brighter meteors than VK1 (field of view 20 degrees, lm ¯ 8.0 mag for stars), which are caused by larger meteoroids. Therefore, a shift in the activity profiles hints to sorting of meteoroids of different sizes.

2.4 Meteor Observations in 1996

We did not obtain any further observations in 1995, and all of the 1996 video observations are not yet analysed. MOVIE broke down before the maximum of the Lyrids, which could be visually observed under good conditions near Berlin.
Together with several members of the German AKM meteor group I took part in a major double station campaign during the Perseids in 1996. Since the new video camera series was distributed among the meteor observers in spring ([1],[2]), it was the most effective video observation campaign we had so far in Germany.
Again, the weather was quite poor around the Perseid maximum. A crash expedition to the Baltic Sea failed to give us the chance to witness the new Perseid filament another time (only a small group travelling behind the Polish border experienced clear skies), but especially the double station work in the days before and after the maximum seems to have been very fruitful. In two camps west of Berlin, four video systems were operated in parallel: in Ketzür MOVIE and two cameras from the new series, and another camera from the series in Golm. So far, only the video tapes from August 10/11 were inspected. Whereas the author visually observed 110 meteors in 4.19 hours effective observing time that night, I found almost 180 meteors recorded in 5.43 hours by AVIS (Advanced VIdeo System).
AVIS is a system that slightly differs from the other cameras of the new series. It consists of a 0,75/65 mm lens, a 25-mm second generation image intensifier, and a Phillips video camera. The field of view is about 20 degrees in diameter; the camera records stars down to magnitude 8.5. The average meteor recorded by AVIS is much fainter than visual or meteors recorded by MOVIE. In addition, the percentage of shower meteors is smaller, since the Perseids lack faint meteors like many other visual streams. However, the number of meteors recorded in parallel from Golm and Ketzür is amazing.

2.5 Summary

We carried out several visual and video meteor observations in the last few months. So far, we have been able to investigate the radiants of six different meteor showers in detail with MOVIE, the results are given in Table 1. With the availability of new video systems it is expected, that many valuable double station observations will be obtained in the near future.

---------------------------------------------------------------------------------------------------------------
|   Meteor    | Year of | Shower  |                  Radiant Position                                         |
|   Shower    | Observ. | Meteors |      our result       |               other Investigations                |
---------------------------------------------------------------------------------------------------------------
| Quadrantids |  1995   |    39   | alpha = 229.4° ñ 1.5° | [10] alpha = 229.8° ± 1.1°  delta = +49.4° ± 1.1° |
|             |         |         | delta = +49.7° ± 1.5° | [11] alpha = 231.0° ± 2.1°  delta = +49.1° ± 0.8° |
|             |         |         |                       | [12] alpha = 229°   ± 1°    delta = +49°   ± 1°   |
|             |         |         |                       | [13] alpha = 230.6° ± 2.2°  delta = +49.1° ± 1.0° |
---------------------------------------------------------------------------------------------------------------
|   Lyrids    |  1995   |    31   | alpha = 271.6° ± 1.5° | [10] alpha = 271.9° ± 0.3°  delta = +33.3° ± 0.3° |
|             |         |         | delta = +32.9° ± 1°   | [13] alpha = 271.9° ± 0.9°  delta = +33.4° ± 0.4° |
|             |         |         |                       | [14] alpha = 271.4° ± 0.5°  delta = +33.4° ± 0.5° |
---------------------------------------------------------------------------------------------------------------
|  Perseids   | 1993/94 |   228   | alpha =  46.0° ± 2°   | [10] alpha =  46.8° ± 1.2°  delta = +57.7° ± 1.2° | 
|             |         |         | delta = +57.7° ± 2°   | [13] alpha =  45.6° ± 5.4°  delta = +57.5° ± 1.6° |
|             |         |         |                       | [15] alpha =  46.6° ± 2.3°  delta = +57.8° ± 1.1° |
---------------------------------------------------------------------------------------------------------------
|  Orionids   |  1995   |    27   | alpha =  93.6° ± 1°   | [10] alpha =  94.7° ± 0.8°  delta = +15.9° ± 0.8° | 
|             |         |         | delta = +14.9° ± 1°   | [13] alpha =  94.2° ± 3.6°  delta = +15.8° ± 0.9° |
---------------------------------------------------------------------------------------------------------------
|   Leonids   |  1995   |    35   | alpha = 154.5° ± 2°   | [10] alpha = 152.7° ± 0.3°  delta = +22.5° ± 0.3° |
|             |         |         | delta = +21.4° ± 1°   | [13] alpha = 153.2° ± 1.0°  delta = +22.0° ± 0.8° |
---------------------------------------------------------------------------------------------------------------
| alpha-Mono- |  1995   |    28   | alpha = 117.1° ± 3°   | [16] alpha = 109°   ± 5°    delta = -6°    ± 5°   |
|  cerotids   |         |         | delta = + 1.0° ± 2°   |                                                   |
---------------------------------------------------------------------------------------------------------------

3. Further Investigations

Beside the observation of meteors and their radiant analysis, we have carried out other investigations and experiments in connection with meteor data obtained earlier and the new video camera series.

3.1 Differences between Visual and Video Meteor Brightnesses

At the 1994 IMC I reported for the first time a systematic difference between meteor brightness estimates of visual observers and the values obtained from video meteors [17]. Later I investigated another set of observations, which showed a similar deviation of visual estimates [18]. In addition it was shown, that the difference is independent from both meteor velocity and brightness. This was another hint, that visual observers systematically underestimate the brightness of meteors.
In spring I have analysed the recording characteristics of MOVIE to find out, in how far properties of the camera and analysis process may influence the finding. The following aspects were taken into consideration:

• spectral sensitivity of MOVIE
• variable sensitivity in the field of view
• variable response to moving objects (afterglow, saturation)

In order to determine the spectral response of MOVIE, I have investigated the dependence of brightnesses of individual stars from their spectral class. For that, I measured a set of stars in the camera's field of view, and calibrated the brightness scale with a least squares fit to the values from the PPM star catalog. The resulting differences are shown in Figure 8.

It is obvious, that brightness measurements of individual objects show a significant scatter, mainly due to the changing sensitivity within the field of view of a first generation image intensifier. There is no systematic dependence from the spectral class, even though it seems, that red objects are measured slightly too faint. During the analysis process, reference stars are arbitrarily chosen independent from their spectral class, whereas meteors reach their maximum brightness in the blue light. However, that effect cannot explain the large differences between visual estimates and video measurements.

In a second experiment I investigated the radial sensitivity of the image intensifier for different objects brightnesses. I recorded a constant light source at different positions in the field of view and measured its brightness in the video record. The result is given in Figure 9.

The graph shows a complex pattern of dependencies. Whereas the sensitivity of the first generation intensifier is generally higher in the middle than at the edges, it has another minimum at the center. Faint objects are more effected by that effect than bright ones.
This is the main reason, why individual brightness measurements show a large scatter. The dependence is too complex to be corrected by the analysis software. However, also this effect does not result in systematic deviations of meteor brightness measurements, since both the measured meteors and reference stars are to be found everywhere in the field of view.

In a third experiment I have tested the relation between brightness measurement and the angular velocity of an object. This relation may influence the measurements systematically, since meteors are moving objects, whereas the reference stars are fixed. I turned the camera at a constant angular speed and recorded a constant light source. The result is shown in Figure 10.

It is clear, that the measured brightness depends from the angular velocity: Slow moving (or fixed) objects appear slightly brighter than fast ones. This is because of a certain saturation of the image intensifier, if a light source shines constantly at the same point.
Due to an equipment failure I could not repeat the measurement for very bright objects, where the afterglow could compensate the effect. However, the graph suggests that brightness measurements from moving meteors are systematically too faint, since they are calibrated with fixed stars.

I conclude from the experiments, that individual brightness measurements of MOVIE are less accurate than we thought before. However, there is only one effect that systematically influences the brightness measurements. Moving objects like meteors are measured systematically too faint, so the difference between visual and video observations becomes even more evident. I hope, that we can support this finding by more accurate measurements with the new meteor cameras and their second generation image intensifiers.

3.2 Meteor Light Curves

The set of meteor light curves presented at the last IMC ([3]) could now be completed with the graphs from other meteor showers observed in 1995 (Figure 11). Especially interesting is the unique behaviour of the Taurids, which was determined from meteor records during the Leonid and alpha-Monocerotid campaigns.

3.3 Time Recording with Video Cameras

On problem that has to be solved by every video observer is the provision of time stamps on the video tape. As long as a Camcorder is involved (MOVIE, for example), you can superimpose a clock to the image, which is adjusted before the observation. However, a normal video module as used in our new video camera series ([1],[2]), for example, does not have this feature, and different video observers came up with various solutions for the problem. Now we found a very neat way of time recording, which we suggest to other video observers.
We became aware, that lunar occultation observers face the same problem: They use image intensified video cameras to record events with high timing accuracy. Some years ago it was proposed in the IOTA/ES (International Occultation Timing Association / European Section) to develop an electronic time inserter. Such an device reads a video signal, superimposes a clock, and gives the modified video signal out. Thus, it can be used in connection with every video camera.
Dr. Cuno from the IOTA/ES designed the time inserter and distributed the first devices in 1995. We ordered a complete set for the video camera series and tested the inserter extensively during the Perseids . It turned out, that they ideally suit our needs.
The time inserter comes in a small cigarette size box. It has three cinch sockets (video in, video out, time signal in) and a fourth socket for the power supply (12V DC). The time is feed into the inserter in form of Germans official DCF-77 time signal. This time signal is broadcasted at 77.5 kHz from a station near Frankfurt/Main; it can be received almost everywhere in Central Europe. You can obtain the input signal for the time inserter from a simple DCF-77 controled clock.
The inserter decodes the time signal and superimposes a black bar with the actual date, day of week and time in the lower part of the image. In addition, a frame counter shows the number of the video frame in the actual second, which makes the analysis of the video tape especially easy.
There is a problem with interferences between TV monitors and DCF-77 receivers, because 77.5 kHz equals a multiple of the monitor’s line scan frequency. However, if you put the antenna some meters away from your video control monitor, you can receive the time signal without problems. Even if you cannot receive it, the time inserter is not useless. In that case it simply counts the time starting from zero, so that you have at least a relative time base. In addition, the inserter may support the American WWV time signal in the future.
Another minor problem is, that the inserter always shows the public German time (i.e. CET with or without daylight saving time). You cannot switch it to UT, since it would require a complete calendar calculation algorithm.
We suggested to Dr. Cuno to make the position of the clock adjustable in the image, and he modified the design for us. In version 3 you now have a set of jumper to move the clock and minimize the obstruction of the field of view.
There is no literature reference for the time inserter available. However, you can contact Dr. Cuno directly, if you are interested in such a device:

	Dr. Hans-Hellmuth Cuno
	Schrammlhof 2
	D-93164 Laaber
	Germany

The price for the time inserter is currently 250 DM.

3.4 Image Distortion of the New Video Cameras

I have used a record of a chess-board type test pattern (Figure 12) to determine the image distortion of VK1, the prototype of the new video camera series.

It turned out, that also second generation image intensifiers show a considerable image distortion . Contrary to MOVIE, the image is compressed near the edges. However, the distortion can be described reasonable with an exponential function of the same type ( Figure 13).

3.5 Automatic Meteor Search

As described in WGN ([2]), I finally succeeded to automate the meteor search on video tapes. With the availability of more and more video systems, the development of an automatic search system becomes extremely important: As long as we have to inspect the tapes manually, video observation of meteors will never become really effective.
The underlying principles for computer based meteor detection were already proposed by the author in 1993 ([4]). However, the meteor search was infeasible for video systems like MOVIE, which apply first generation image intensifiers with strong noise. The new MCP video systems are much less noisy, so that the results of computer based tape inspection where much more promising.
I updated the analysis program in spring 1996 and carried out several tests with the alpha- Monocerotids recording of the prototype camera VK1. In order to achieve an appropriate number of video frames per second, the search program needs to run four times in the current configuration, each time inspecting another part of the field of view (Figure 14).

The software was able to analyse every third non-interlaced video frame (8.3 frames/s) running at a 486 PC with 66 MHz clock rate, and every second frame (12.5 frames/s) when started at a Pentium machine with 90 MHz clock rate. The computer achieved a very good detection rate of almost 75%, the number of misidentification was neglectible. The remaining 25% of undetected meteors occurred either in the small uninspected corners of the field of view, or they were just too faint. In both cases they are not suited for further analysis anyway, since their positions are inaccurate due to the proximity to the border, or they would be lost in the noise.
It turned out, that the CPU speed is not anymore the main problem. The real bottleneck is the image transfer speed from the frame grabber card to the computer’s main memory, i.e. the bus system. It is to expect, that the number of necessary test runs will go down to two or even one, once a faster PCI bus frame grabber card replaces the currently used slow AT bus frame grabber.
I found, that the suggested inspection rate of about 10 frames per second is appropriate for meteor detection. Since even faint video meteors last on average at least 0.2 seconds, all of them are visible on at least one inspected video frame. It would be even critical to further increase the frame rate, since especially slow moving meteors near a shower radiant would be missed due to their almost stationary appearance.
Currently, the software has prototype status. It is intended to replace the hardware specific parts of the software and support other frame grabber cards. The program will then be freely available to other video observers.

4. Conclusions

1995/96 was one of the most successful years for the meteors observers at Archenhold- Observatory Berlin. We were able to carry out many interesting visual and video observations with MOVIE and new video metor cameras. An outstanding event was the outburst of the alpha- Monocerotids in November 1995, which could be completely recorded on video tapes.
The number of detailed with MOVIE investigated meteor showers grew to six. I determined the properties of the video system and found further evidence for systematic brightness estimate errors of visual observers.
A new series of relatively cheap and powerful video cameras was developed and distributed among German meteor observers, and significant progress in the development of automatic video systems could be achieved. It is expected, that video meteor observation will now be carried out regularly in Germany.

5. Acknowledgements

As in the last years, our work was supported by the public Archenhold-Observatory Berlin. I would like to thank Felix Bettonvil and Marc Neijts from the NVWS for another attempt to obtain double station video observations. We are in debt to Dr. Cuno from the IOTA/ES , who provided us with the time inserters for the camera series and even changed their design for the needs of meteor observers.
Last but not least special thanks to Mirko Nitschke, who was a great partner during the observations and the development of new techniques. We had a very effective time together in Chemnitz, and our success would have been impossible without his devotion.

6. Literature References