...
As part of this experiment, we hope to train numerous astronomical telescopes in Australia and New Zealand on the expected path of the ATV5 during its early interactions with the upper atmosphere. Although resolved imagery is a nice feature to have, our primary objective of these images from astronomical instruments will be to get a very accurate record of the ballistic path of the spacecraft as a function of time, and in some cases, good spectral data from the spacecraft’s wake. (We have some specialty tracking assets that will attempt the resolved image…this is not your goal!)
The following web video illustrates just how tight a problem this will be, illustrating the ISS transiting the ½ degree disk of the sun 400km distant. For the low-flying ATV5, the transit of an equivalent ½ degree field of view (with reference stars in the background, instead of the sun) will be twice as fast, due to the much closer range. It will also be MUCH fainter, self-illuminated in the Infrared. Note that the ATV5 itself is >>IN<< this recent video, currently docked to the aft end of the ISS.
http://www.youtube.com/watch?v=AtEVzRx9ktw
The photographer behind this remarkable image sequence is Thierry Legault in France who is an expert in this technique (cc’d above). We have many challenges to move towards low-light, faster-moving images against reference star background. Thus there is lots of work left to do. I know that your students will appreciate the challenge.
Depending upon resolution of the telescope imaging systems, I hope to be able to characterize the elevation of the trajectory to within a meter or so. The azimuth however is VERY dependent upon the precision and accuracy of the timing subsystem associated with the imaging system. (an error of 1 millisecond results in a position error of 8 meters. However, the azimuth far more sensitively reveals the subtle physics of the deceleration of the spacecraft, so great azimuth vs time accuracy is a good goal, once we capture the basic issue of pointing the telescope to capture the transit.)
Because the spacecraft will be transiting the sky at an exceptionally fast rate compared to the normal tracking capability of most astronomical telescopes and will be VERY faint, we are expecting that the most effective and accurate technique to record the ballistic path will be to take a streak image of the spacecraft across the field of view. It is for the perfection of this technique that I request the assistance of the Berkshire astronomy students. However, I am eager for them to additionally explore the issues of repeating the experiment in low-light infrared transits, and if they choose, real time tracking once the primary test is well-characterized.
We have secured the assistance of Chris Peat--the developer of the Heavens-Above website at
www.heavens-above.com also cc’d above--to facilitate the posting of the final trajectory estimate of the ATV5 on the day of the experiment just minutes after the final engine burn, and 30 minutes before observations begin. This globally-accessible website already includes the key information that will allow a telescope to pre-position its field of view along an expected spacecraft trajectory, although such a task is only easy in theory. Between now and February, we will be looking to perfect the step-by-step procedure for operators of the Australian and New Zealand observatories to get from the heavens-above data to a precisely pointed and timed image. If there is some simple adjustment to the data posting technique at Heavens-above that would assure greater success in capturing the image, I encourage you to work with Chris to explore whether this can be accomplished.
In principle, this should be straightforward, but as we have learned in countless complicated procedures in the past, it is important to practice every part of the process, to record lessons learned, and to perfect the procedure with best practices. Fortunately, Heavens-Above provides identical data formats on thousands of orbiting objects of varying brightness and size, which provide plentiful test cases for the basic technique on any clear night, just after sunset. (Also just before dawn, and for very faint viewing, whenever the moon is up). A candidate list of targets is available.
I propose that the Berkshire students practice imaging such low-orbiting spacecraft that are illuminated by the sun (or even the moon in the extreme low-light scenario). I encourage you to experiment with all available recording media at the image plane, from normal telescope CCD to CMOS Digital cameras and ultimately to chilled IR detectors. In particular, we are interested in the lowest light level (spacecraft and reference star visual magnitudes) that can be recoded using the proposed technique. We will want to understand the limitations and results of the following techniques:
1. Take a short-duration streak image (with telescope fixed) with known start and stop times scheduled while the spacecraft is in the expected field of view. This will result in a long faint streak image of the spacecraft, and very short streak images of the reference stars in the background. Record to greatest possible precision for each exposure the telescope pointing AND the GMT (to a fraction of a second, if possible). Note that it is important to the astrometry of the experiment that one or more known reference stars are visible, and that the resolution of the image be maximized such that we can best determine the displacement of the spacecraft trail from its predicted path. (There should be very little deviation from the expected path in your shots, but in the actual experiment, we expect to see deflections of several meters. In general, I believe that 30 arc minutes will be our limiting narrow field of view for non-tracking assets.
a. Once the technique is comfortable, attempt to do this on teacher-selected spacecraft whose trajectories are revealed to the students on progressively shorter countdowns. Record any issues and bottlenecks in the process, and any induced human errors as the time is compressed before first exposure.
b. Instead of tracking a sunlit object, attempt to capture an infra-red image of a LARGE spacecraft just after sunset (say ISS), when it will still be very warm from its sunlit pass. Then see how small an object can be so resolved.
2. Take the same shots with the telescope actively tracking the sky, and the exposure terminated during the transit of the FOV. This will result in the reference stars being brighter single points, compared to faint short streaks as recorded in task 1. This technique MAY be necessary if reference stars of sufficient visual magnitude are hard to come by. It will come at the expense of critical timing since only one end of the spacecraft streak is expected to be visible in the FOV.
3. Attempt a second or even third exposure after the first attempt, by slewing the telescope at maximum rate to additional positions along the intended trajectory. Record and issues and limitations.
4. Bonus points: Suggest and attempt techniques to image a low-orbiting spacecraft in sharpest possible resolution with an astronomical telescope. (The ISS is a good very bright target that may allow a short exposure. A short exposure should help in this effort, but telescope motion control is the real desired objective.) I suspect Thierry Legault would be a good contact for this and the other tasks, but the real challenge is that you do not have a spacecraft tracking servo system, so creativity is necessary!
I will caution that one of the hardest problems in some attempts has been the accurate coordination of the telescope and camera timers to true GMT. Since the spacecraft is moving at nearly 8000 meters per second, even a millisecond of error can have large effects. In Australia, the spacecraft at closest approach to some telescopes will be only 200 km away. Moving at 8km per second, it will this cover 8/200 1/25 of a radian per second, or over 2 degrees of motion. A ½ degree field of view must therefore record the entire event within ¼ second, at closest range.
I attach a catalog of the orbiting spacecraft with inclination greater than 30 degrees (such that they are visible to NZ and Australia) with perigees below the apogee of the ISS, so that they will be moving reasonably quickly in the FOV. . This will give you a range of targets to select in Heavens-Above for sighting opportunities just about any night. Use the NORAD_CAT_ID column to tell Heavens-Above which satellite you wish to track. (enter a range of NORAD Cat numbers in the menu under the “Satellite Database” tab on the home page.
I have sorted this first by the radar cross section of the orbiting satellite: this is a fair indication of its approximate size and therefore brightness.
I suggest that that you start with any blue row: these are objects lower than ISS in near-circular orbits. Of these, the ones near the bottom of the list are any spacecraft with LARGE cross section (The ISS is a huge target, but not always visible when you have clear skies). Then work your way down to medium and then small ones to get the low-light, fast moving stuff.
The small objects are generally going to be VERY faint. I have highlighted in yellow the known cubsesats, which will have the best characterized areas and surface materials (most of the other stuff is debris), but these are TINY (10 cm x10 cm x 10cm at the smallest). The “Flock” constellation are 3x longer than the others. You can get images of the actual spacecraft by web-searching the names. Note the large number of items with a launch date of Nov 20, 1998. That’s the day we launched the first element of the ISS (I was THERE, in Baikonur!) and anything sent away from the ISS now carries that launch date, for historical reasons, no matter what spacecraft brought that object to ISS.
Objects with a near-vertical pass over you will be moving much faster in your field of view than objects closer to the horizon, so challenge yourself with High-elevation passes.
ONLY if you want this extra challenge: The RED rows are pretty difficult, and I wouldn’t recommend trying these until everything else is perfected. These are more elliptic orbits with low perigees, and can have somewhat faster transits near perigee than the low circular orbits. However, this is only true for the near-vertical passes with perigee right over you. Any slant range takes away the advantage of the faster orbit velocity in making a very fast pass. (THAT perigee alignment and near-vertical pass in appropriate lighting takes some very rare luck…) Such objects are also a real challenge because often the predictions for these objects are not as accurate as the predictions for the lower orbiting stuff. If we get good with the blue and yellow small cross-section targets, I think we are good enough!
Jack Bacon Ph.D., P.E.
