Video Observations of Meteors: History, Current Status and Future Prospects

Video meteor observations have been performed by amateur astronomers for more than ten years. They enjoy a rapidly increasing interest in the meteor community and will evolve into a powerful tool for amateur observers in the near future. Video meteor observation is the key to a fundamental increase of our knowledge about meteoroid populations and their interaction with the Earth's atmosphere.
In this paper we want to summarize the history of video meteor observation and describe the current state of affairs. We discuss problems and limitations and propose future projects. The paper is intended to serve as basis for the foundation of appropriate organizational structures within the International Meteor Organziation.

History and Current Status

Professional astronomers started to use image intensified systems in connection with film equipment already in the sixties and seventies of our century [Hawkes 1993]. Among amateurs, Japanese (1986) and Dutch (1987) observers have been the first using low-light-level video systems [de Lignie & Jobse 1989, Fujiwara 1993].
At the beginning, there were only XT personal computer and rudimentary frame grabber cards available. Thus, most of the video tape analysis had to be done manually. The main advantage compared to visual observations was the increase in positional accuracy by orders of magnitude. In addition, video systems where much more efficient than photocameras, since they could record meteors down to magnitude +6 and fainter.
In the beginning of the nineties, new amateur groups started indepently to use that technology in several European countries like Austria, Germany and the United Kingdom. Thanks to the increasing power of computer hardware, more and more problems could be solved computer aided. First attemps for the automatic video tape inspection were reported in 1993, and the analysis software for video meteors became much more efficient. However, the major breakthrough did not happen before 1995/96, when image intensifiers became cheap enough to be efforded by a larger group of amateurs [Molau & Nitschke 1996].
Scientific results of video observations from several observer groups were published in different journals. So far, video systems have been used for the determination of meteoroid stream orbits, shower radiants, calibration of visual observations, cluster analysis, recording of spectra and many more research projects. A major event was the 1995 outburst of the alpha- Monocerotids, which has been completely recorded by two video teams [Rendtel et al. 1996, Jenniskens et al. 1997]. It was for the first time, that such an outburst was observed with an appropriate method.
Currently, there are about fourty video systems operated by amateur astronomers around the world. At least fifteen systems are in operation in Japan. We know of approximately ten video systems in Germany and nearly five in the Netherlands. Video meteor observations are carried out in the United Kingdom, the United States and Austria. Beside that, professionals in Canada, the Czech Republic and Tadjikistan do use this technology.
There are at least three groups who deal with computer-aided meteor detection on video tapes. One of those systems has already proved to work in practice [Molau and Nitschke 1996].
Several software packages exist for the digital measurement of recorded video meteors [de Lignie & Jobse 1989, Hawkes 1993, Molau 1995]. In the near future it is intended to provide a standard software package containing solutions for all tasks of video observers.

Why do video observations?

Video systems combine the advantages of visual and photographic meteor observation and can even compete with telescopic observers.
Current systems allow positional accuracies down to one arc minute. Thus, they are more accurate than visual meteor plottings. When used with wide angle objective lenses they can have a field of view of more than 100 degrees in diameter, which is almost comparable to all-sky photography. The limiting magnitude depends strongly on the field of view of the camera, the focal ratio of the lens and the intensifier's gain. However, modern video systems record on average more meteors in the same time than visual observers. They obtain by some orders of magnitude more meteor recordings than photographic systems.
Video systems achieve a high time resolution (25 or 30 images per second depending on the video standard used) and record the evolution of a meteor directly. All events can easily be timed down to an accuracy of less than one second. The angular speed of meteors can be determined accurately without mechanic shutters due to the short exposure times for each video frame. The brightness of video meteors varies from bright fireballs down to the level of telescopic meteors. Thus, such systems can provide uniform data over a much larger spectrum of meteoroid sizes than any other method. In addition, video systems record the light curve of a meteor, leading to important results about the properties of meteoroids and their interaction with the Earth's atmosphere.
A major disadvantage are the costs, that are still relatively high compared to photographic or visual equipment. In addition, video systems depend on the availability of electrical power.

Limitations

A real limitation for video systems is the amount of time needed for data processing compared to the number of events that can be recorded. Currently, the video tapes are inspected visually and meteor positions are then measured with the help of a computer. A working solution for automated meteor search is only a question of time. However, fully autonomous analysis systems seem impossible from the current state of affairs. With the help of specialized computer software the measurement of meteors can be accelerated, but in practice it still requires five to ten minutes for each meteor to be analysed. Thus, it is impossible to analyse all meteors in detail, that can be recorded by video systems. Depending on the actual aim of investigation, one either has to restrict the amount of information to be derived, or the meteors to be analysed have to be selected.
Video systems are not as portable as photographic equipment. Even though newer cameras are more robust than earlier systems, they still are highly integrated electronic devices with some limits. Most systems are not meant to operate at temperatures far below the freezing point or when dew turns up. This, together with their power dependency, makes them only partly useful under expedition environments.

Components and Classification of Video Systems

In general, all video meteor systems consist of a fast lens, an image intensifier and a video camera.
It could be shown, that image intensifiers are absolutely neccessary for recording faint meteors. Considering the number of photons reaching the photocathode, a charge coupled device (CCD) alone may in theory be sensitive enough to detect faint objects. However, when operated at video frame rates of 25 or 30 Hz, the readout noise by far overwhelms the number of electrons generated by the meteor's light.
The least requirement is a first generation image intensifier with multiple amplification stages. The gain should be >1.000, and the diameter of the photocathode needs to be larger than 15 mm. First generation intensifiers with three sequential amplification stages can reach a higher gain than other intensifier generations, but suffer significantly from strong image distortion, a variable sensitivity within the field of view and strong noise. This is why future automatic meteor detection systems will probably not work for those cameras.
Second generation image intensifiers (micro channel plates - MCPs) do usually contain a single amplification stage. Therefore they do not reach as a high gain as first generation devices. However, they are prefered for meteor observation because of their high quality image with less noise, distortion and sensitivity variations. New MCP devices fulfiling military specifications are still very expensive (above US$ 2000), but nowadays second hand MCPs are also available at reasonable prices (below US$ 500) from several dealers around the world.
Third generation image intensifiers are not especially useful for meteor observations, since they reach their maximum sensitivity in the infrared.

Different video cameras are used to record the intensifier's phosphorous screen: A camcorder has the advantage, that it is usually able to automatically record the time. Other systems involve cheaper video cameras. They either mark the time with audio signals or insert it electronically into the video signal. In the analysis process, the video tapes need to be digitized by frame grabbers. Currently those are available at prices between US$ 200 and US$ 4000. The use of conventional CCD cameras as used for astronomical imaging has been discussed. However, major disadvantages like the loss of the high time resolution have prevented observers so far from using this technology.
In the future, the data stream may be stored digitally. With currently expensive hardware it is possible, to digitize the enormous data amount of a video signal (>9 MByte/second or >33 GByte/hour) in real time and save it on a computer's hard disc. This requires good data compression, which is possible for video meteor observations with almost no changes from one video frame to the next. The technology will become cheaper with further technological progress and is an alternative to the use of VCRs and the loss of information caused by that. Also digital camcorders may help to improve the quality of data storage and transfer in the future.
Today's high end system do not involve any signal conversion between digital image aquisition and storage. They also contain sensors with much higer resolution.

The lens is most important for the recording properties of a video system. Generally it should be as fast as possible (low f ratio) to get best limiting magnitudes. According to the focal length we can distinguish between three types of video cameras:

• wide angle video systems: They apply wide angle photo lenses and have a field of view of more than 40 degree in diameter. The limiting magnitude of such systems is usually between +5 and +7 mag for stars. A special type are all sky video meteor cameras, which use convex mirrors and provide total field coverage.



• standard video systems: Using standard or moderate tele lenses, those video cameras record a field of view between 10 and 40 degrees in diameter. Their limiting magnitude for stars is usually between +7 and +9 mag.
  Most currently operated video systems belong to this class.



• tele video systems: On application of fast and long focus tele lenses a field of view smaller than 10 degree in diameter is achieved. The limiting magnitude gets better than +9 mag for stars.

All image intensified video meteor detection systems are technically limited in one of three ways (see Hawkes and Jones [1986] for a more detailed treatment of this topic):

• the "quantum limit": If too few photons arrive during a video frame's intregration time, a useful meteor image cannot be obtained. For a typical system with a 50 mm diameter lens this would correspond to approximately +12 mag. This is the ultimate detection limit using optical techniques, unless the collecting area of the lens, the integration time or the quantum efficiency of the image intensifier photocathode is increased.



• the "background illumination": Because of the limited resolution of image intensified video systems, a variety of faint stars and other skyglow form an unresolved background illumination. There is a simple relationship between the angle subtended by a pixel and the background brightness [Hawkes and Jones 1986]. Most current video systems are background limited.



• the noise of image intensifiers and video detectors: Each electronic device involved into the image aquisition process introduces some noise, which will limit the ultimate sensitivity of a typical MCP to +10 or +11 mag.

Video Observation Projects

From the described properties of video systems we would like to derive the following three key projects:

• minor meteor shower radiant determination: Visual meteor observations reach their limits as soon as the activity of a meteor shower becomes very small. Then, the low plotting accuracy cannot be overcome by a larger statistical sample of recorded meteors. Photographic systems, however, only record very few minor shower meteors. Video system provide the ultimative solution for that problem: They are able to provide an accurate radiant position from only a few recorded meteors [Molau and Arlt, 1997]. Thus, they can be used for the investigation of radiants from known minor meteor showers as well as for the search of new showers.
  Radiant positions can be obtained from single station video observation. Both wide angle and standard video systems have certain advantages for this task. Due to the higher spatial resolution, standard video systems provide more accurate meteor positions. Reliable radiants can be derived from smaller numbers of recorded events. Some meteor showers, however, lack faint meteors. They give impressive displays for visual observers and equivalent wide angle video systems, but are virtually not detectable in the fainter magnitude range due to their low population indices. For such showers, the ratio of sporadics to shower meteors increases rapidly with the camera's limiting magnitude favouring video systems with a large field of view.



• determination of meteoroid orbits: So far, most meteoroid orbits have been computed from double station meteor photographs. However, with the availability of modern video systems, the orbital database can be largely extended. Double station video observations with standard video systems are necessary for that goal. They combine a suitable accuracy with sufficient large fields of view and limiting magnitudes. Typically the baseline for dual station setup is of the order of 20 to 100 km. The cameras need to be aligned so that the centers of the field of view aim at a height of 100 km. Accurate time information at both stations and precise alignment of the cameras are vital for the investigation. On average, 50 to 75% of the meteors will be in common between the two stations.
  Again, especially minor meteor shower will be the target of such observations, since their photographic records are naturally very rare. The determination of minor meteoroid stream orbits is probably the biggest challenge for video observers. Depending on the shower's population index even a double-station video setup may not be efficient enough to record enough meteors in parallel. Thus, coordinated observations from many video teams may be required. On the other hand, orbit determination is the only task that requires double station work. All other investigations can be carried out by single station observers!
  A number of double station video observation studies of selected meteor showers has been published so far [Sarma and Jones 1985, Ueda and Fujiwara 1996, de Lignie and Jobse 1995, for example]. However, each of these studies relied on observations in only a few nights. Especially the period from mid January to early July is poorly covered in that respect.



• derivation of flux densities: Video systems have the major advantage of being objective regarding the shower association contary to visual observers. Much less factors do influence the rate of meteors detected by video systems, so it is possible to directly determine flux rates for major and minor meteor showers. In addition, a much larger variety of meteoroid sizes can be investigated, which extends current visual studies to other particle populations. For that, all types of video systems can be applied. It is not necessary to analyse each meteor in detail, so the flux analysis can be done almost automatically.

Beside those key projects, we suggest a number of other observations to be carried out with video meteor cameras:

• observations of very faint meteors: The possibility to record meteors down to magnitude +9 or even fainter with tele video systems opens the door to a class of particles that has not been observed so far in the telescopic range. The behaviour of those very faint meteors needs to be investigated. Is there a minimum brightness for meteors?



• observations of exceptional high activity: Video systems provide the unique possibility of objective observations under meteor storm conditions. Therefore, the study of meteor outbursts should be a major goal for video observer. All types of systems should be involved to enable studies of mass sorting and selection effects as well as meteor cluster analysis.



• fireball patrol: Wide angle video systems regularly record a large percentage of bright meteors, among them fireballs. Special all-sky video systems can work almost autonomous and obtain valuable information about the rate of very bright meteors. They can fill the gap to routinely working camera networks like the European Fireball Network. For this task, an especially robust video system is required. On the other hand, cheaper equipment of lower power will be sufficient. Even systems without image intensifiers may be useful in this respect. All fireballs capable of producing a meteorite are bright enough to be recorded by an ordinary camcorder. Since the amount of data is enormous compared to the number of events actually recorded, such a system definitely requires an automatic meteor detection system.



• observations of wake, persistent trains, light curves and other meteor characteristics: Video systems give us the unique chance to study individual meteors in detail (see Robertson and Hawkes [1992], Fleming at al. [1993], for example). Once the characteristics of a video system are known well enough, reliable measurements of meteor properties and their statistical analysis for different meteor showers can be carried out. They may give insight into physical properties of the meteoroids and their parent bodies.



• recording of meteor spectra: So far, only a handful of good quality meteor spectra have been obtained by professional astronomers [Millman et al. 1971, Borovicka & Bocek 1995]. With little design modifications, video systems can be used to record spectra of much fainter, and thus, much more meteors regularly.



• calibration of visual and radio observations, teaching of new observers: With the help of wide angle video systems, which have characteristics similar to visual observers, it is possible to calibrate visual observations [Molau 1995]. In addition, observers can be trained with the help of video recordings for standard situations as well as exceptionable circumstances. If video systems are used in parallel to radio equipment, they can contribute to the important question of the calibration of radio data.

Conclusions and Outlook

We have presented a list of projects that can be work on with video systems. This list is certainly not inclusive of all possible projects. We would like to invite other observers to join a discussion about the future of video meteor observation and the focus of our work. We want to call for participation in the projects mentioned above, which will certainly improve our knowledge about small particles in the solar system and their interaction with Earth fundamentally.
We feel, that the importantance of this observing technique, which will be the key for new investigations in the future, should be reflected by an own commission within IMO. The main aim of an video commission should be the coordination of activities and the encouragement of further observers to apply this still rarely used observation method. The key to success for many of the proposed projects lies in the coordination between video observers and fruitful cooperation with other techniques like photographic, visual and telescopic observation. An commission should provide general information on the how and why of video observations, technical hints and construction plans for video cameras, suggestions for observation targets and support for the data analysis. Forums like WGN and the WWW homepage of IMO could be employed for that function. In addition, the maintenance of a video database and the provision of free access to the stored meteor data should be realized.
Last but not least, a video commission could serve as a contact address for everybody who has specific questions or problems. With joint efforts of the currently mostly uncoordinated working video groups, we may approach our scientific tasks more efficient than ever.

Literature References

Borovicka, J. and Bocek, J. (1995): "Television spectra of meteors", Earth, Moon, and Planets 71, p.237

de Lignie, M. and Jobse, K. (1989): "Accurate radiant determination from TV meteors", Proceedings of the IMC 1989, p.41

de Lignie, M. and Jobse, K. (1996): "Double station video observations of the 1995 Quadrantids", WGN 24, p.20

Fleming, D. E. B., Hawkes, R. L. and Jones, J. (1993): "Light curves of faint television meteors", In: Stohl J. and Williams, I.P., (eds.), Meteoroids and their parent bodies, Smolenice, Slovakia, p.261

Fujiwara, Y. (1993): "Television observations of meteors in Japan", In: Stohl J. and Williams, I.P., (eds.), Meteoroids and their parent bodies, Smolenice, Slovakia, p.265

Hawkes, R. L. and Jones, J. (1986): "Electro-optical meteor observation techniques and results", Quart. J. of the R.A.S. 26, p.569

Hawkes, R. L. (1993): "Television meteors", In: Stohl J. and Williams, I.P., (eds.), Meteoroids and their parent bodies, Smolenice, Slovakia, p.227

Hawkes, R. L., Mason K. I., Fleming D. E. B. and Stultz, C. T. (1993): "Analysis Procedures for Two Station Television Meteors", Proceedings of the IMC 1992, p.28

Jenniskens, P., Betlem, H., de Lignie, M. and Langbroek, M. (1997): "The detection of a dust trail in the orbit of an Earth threatening long period comet", Astrophy. J., scheduled for February 1997

Millman, P. M., Cook, A. F., Hemenway, C. L. (1971): "Spectroscopy of Perseid meteors with an image orthicon", Can. J. Phys. 49, p.1366

Molau, S. (1995): "MOVIE - Analysis of Video Meteors", Proceedings of the IMC 1994, p.51

Molau, S. and Nitschke, M. (1996): "Computer based meteor search - a new dimension in video meteor observation", WGN 24, p.119

Molau, S. and Arlt, R.(1997): "Meteor shower radiant positions and structures as determined from single station video observations", Planetary and Space Science, scheduled for 1997

Rendtel, J., Brown, P. and Molau, S. (1996), "The 1995 outburst and possible origin of the alpha-Monocerotid meteoroid shower", Month. Not. of the R.A.S. 279, p.L31

Robertson, M. C. and Hawkes, R. L. (1992): "Wake in faint meteors", In: Harris A. and Bowell T. (eds.), Asteroids, Comets, Meteors, p.517

Sarma, T. and Jones, J. (1985): "Double-station observations of 454 TV meteors", Bull. Astron. Inst. Czech. 36, p.103, 585

Ueda, M and Fujiwara, Y. (1995): "Television meteor radiant mapping", Earth, Moon & Planets 68, p.585