The All Sky Camera Fireball Detector


David A. Kenyon

Professor of Astronomy, Sierra College

5000 Rocklin Rd.

Rocklin, CA. 95677


Wayne T. Watson

Watson Laboratory

And James Clerk Maxwell Observatory

11467 Lothlorien Lane

Nevada City, CA 95959


A fireball (or bolide) is another name for a very bright meteor due to atmospheric entry of space debris, both natural and man-made. This paper will discuss the technology, methods and results from an “All-Sky” video camera and custom electronics to record fireball events that occur over Northern California. Images and  links to videos will also be presented.


The data was captured with Sandia Fireball Network cameras. These systems are operated by the Sierra College Astronomy Dept. and a private observatory in Nevada City, CA. These cameras are automated system which operates from just before dusk to just after dawn.

        The systems are video cameras which have a 180o field of view. There is a custom signal processor and memory system, called the Sentinel system, which buffers a few seconds of continuous sky video. The processor compares each video frame to the previous, and when it detects an event (fireball), a few seconds of video (before and after the event) are sent to a host UNIX system.  Only data relevant to motion is downloaded  to the computer.


The system provides a method of recording and study for meteor shower activity. Second, all events captured by both cameras provide the potential to compute the pre-earth-encounter orbit, but also to estimate the impact corridor of any meteorites the fireball might have produced. 

        Our experience thus far shows that they occur about once a day and a very bright one every few weeks. The cameras began regular operation around mid-April 2004. To date, no fragments have been recovered from detected events.



1.      Introduction

A fireball (or bolide) is another name for a very bright meteor due to atmospheric entry of space debris, both natural (meteoroid) and man-made.


Sandia National Labs in Albuquerque support a loosely connected network of meteor observers in the U.S. and Canada using  video equipment provided to the observers. Sub-networks of individuals within the Sandia network work to coordinate visual, radio, and camera meteor data collected within the observer’s general geographic area.  The only equipment not provided is a PC and Linux operating system.  Our program represents  two of the 6 (British Columbia, Nevada City, Rocklin, Albuquerque, Las Cruces, and El Paseo) observatories equipped with the latest cameras from Sandia. Older sites have video cameras and tape recorders, but no Sentinel software or electronics. Some sites have equipment developed by the owner.

      There are 21 Canadian sites and 22 U.S. sites . The U.S. sites are west of the Rockies, mostly in Colorado, with others in Texas, New Mexico, Washington, and California. Canadian sites are spread from Newfoundland to British Columbia, i.e., coast to coast.





         Figure 1. The Fireball Network



The most active programs are coordinated by Chris Peterson, who manages several sites in Colorado;  Ed Majden, with his camera and a large network of  observers in British Columbia; and Jim Gamble, a collection of observers in El Paseo.


 Our experience thus far shows that fireballs occur about once a day and a very bright one every few weeks. Our program began regular operation around mid-April 2004. It is an automated system which operates from just before dusk to just after dawn.



2.      Program Goals

Primarily, the system provides a method of recording and study for meteor shower activity. This program supports the Fireball Network as well as providing is a focal point for meteor observations in the north western United States. It also provides a local educational resource as well.


 Second, all events captured by both cameras provide the potential to compute the pre-earth-encounter orbit, as well as to estimate the impact corridor of any meteorites the fireball might have produced.  This could result in the recovery of fragments from the falling object.


Since the beginning of the program over a year ago, we have received observer support and interest throughout California: with SAIC (for trajectory determination), UCLA (Astronomy Dept.), SETI/NASA (meteor studies), individual visual observers as well as several people who are planning to construct equipment.



3.      The Hardware

The video system used is a Sandia National Laboratory Sentinel Video System. The system consists of a video camera with a wide field of view. The camera and lens are mounted pointed to the sky in a weather protected acrylic dome. This unit is secured to a stable platform with a clear view of the sky (see figure 1.). The system yields an “All-Sky” imaging system which has a 180o field of view.






Figure 2. All-Sky camera and protective dome.



The camera is connected (weather protected) to the Sentinel Video System (custom signal processor and memory) which buffers six seconds of continuous video of the sky. The processor compares each video frame to the previous frame and when it detects an event (fireball), six seconds of video (centered on the event) is sent to a host PC based Linux system.


The signal processor is connected to the PC via the printer parallel port. This port provides command and control interface as well as data transfer functionality.


There are two systems at different sites involved with this program. The first camera is located on top of the Sewell Hall Science Building on the Sierra College campus. Located at 38 deg. 47’ 25.9” of latitude and 121 deg. 12’ 49.3” longitude with an elevation of 90 meters in Rocklin, CA (USA), about 20 miles northeast of Sacramento, CA.





   Figure 3. Sierra College site.



The second camera is located at a private observatory in Nevada City, CA. (USA) located at

39° 15' 7" N of latitude and 121° 2' 32" W longitude with an elevation of 823 meters in the Sierra foothills.



                 Figure 4. Nevada City site.



4.      The Software

The Sentinel detection software is written in C and operates under the RT (realtime) Linux OS, which permits Sentinel to field events immediately and adjust data collection and related.  Realtime OS’s are widely used in science and industry to optimize data collection under quickly varying conditions.


Clocks at both sites are independent, and provide time stamps for the data. The Rocklin site provides better accuracy and is used for analysis when both sites observe the same event.


       The software provided with the  Sentinel Video

System includes control software (sentuser2) which

is used to control the hardware. The software

includes commands to initialize the system, set

detection thresholds, masks to block problem areas

of the sky and program on and off times.


       Also provided with the system is a viewer

(sdisplay) to examine the images and video clips

downloaded from the hardware. There are also

support programs to generate xxx.jpg [jpg image

format] images and [Quicktime video

formatted files] of detected events.



5.      Data Collection

When an event is detected, an all-sky image is captured at 640 x 480 pixels of resolution and downloaded to the host. Superimposed on this background image is a 125 x 125 pixel video clip centered on the event (see figure 5). This image and video are viewable with the sdisplay program.




Figure 5. All-Sky image and video overlay

                 (Top:North, Right:West).

To accurately identify the path of the fireball event on the celestial sphere, an autoexpose feature is available on the Sentinel System. This feature periodically collects a long exposure image of the sky which yields the brighter stars for orientation of image and video clip (see figure 6).





        Figure 6. Autoexposure image.



The most useful data occurs when both of the Sentinel Systems in the program capture the same event. This provides the raw data to calculate the pre-atmospheric entry orbit and estimate the impact location. (see figure 7).





         Figure 7. A simultaneously detected event

        (Top: Rocklin, Bottom Nevada City).


The coverage area for the program’s cameras is depicted in figure 8. 



                Figure 8. Program coverage map.



6.      Preliminary Results

A summary of the 2004 events detected can be found in table 1.


Table 1: Counts for Both Sites



An Example


 On June 26th 2004, both sites recorded a very bright fireball (see figure 8). The images attracted quite a bit of attention in the meteor community. Initially the event was thought to be a Russian rocket reentry, but the event time and predicted re-entry were in disagreement. Robert Matson of SAIC decided to do an analysis of the images to extract sky positional elements.



      Figure 8. June ‘04  simultaneous detection

    (Top: Rocklin, Bottom: Nevada City).



From the two images and a calibration of the sky at each site, he was able to produce a 3-D track. It indicated a potential impact site southeast of Sacramento near Mosquito Ridge Road, a sparsely populated area in the Sierra Foothills sometimes used by sport motorcyclists. Robert decided there was enough potential to visit the area from July 17-18 to determine if a search was warranted. The terrain and estimated size of the fall indicated it did not merit putting together a search team.  No fragments were found.


The program is being recognized as a primary contact for fireball observers in western U.S. Observers from Washington to California have contacted us for support, and others have provided visual information on sightings captured by the program cameras. We expect to continue to develop a network of observers on this coast for fireball sightings.






Detected Events


The images below are some of the brighter fireballs detected by this program.




              Figure 9. Event at 09/09/04 02:00:12.




   Figure 10. Event at 09/21/0401:58:24.




   Figure 11. Event at 09/23/04 22:02:10.


            Figure 12. Event at 02/10/05 21:30:23.



Real time videos of detected events can be viewed from the Sierra College web site at :



7.      Table of Detected Events


                   Table 1. Sample Excel table.



8.      Conclusion

The Program has made very good progress in establishing the two sites for fireball and meteor monitoring. We have been able to exchange information between the two sites and demonstrate that we can capture simultaneous events to predict a potential impact corridor. Working with other observers in the local area and throughout the west, we are becoming a focal point for meteor sightings in our local area.


For the future, we will continue to expand the program capabilities by improving upon the software and tool set which support the systems. Specific areas of focus will be: image stacking, operational ease, improve  star overlays on images, and early estimates of tracks from dual observations of an event are some of the possibilities.



9.      Acknowledgements

We would like to acknowledge the support provided by Richard  Spalding and   Joseph Chevez with the Sentinel hardware and software.  Also, the support of Sierra College student Chris Giorgi for his Linux support of the systems.



10.  References

Video Sentinel Users Manual. J.C. Chaves, Sandia National Laboratories


Meteor and Aurora Detection Using Modern Video Technologies.  Martin Connors, et al.