The Space Review

Figure 4
An illustration of a set of maneuevers that brought China’s SJ-12 satellite into at least close proximity to another Chinese satellite, SJ-06F, this summer.

Dancing in the dark: The orbital rendezvous of SJ-12 and SJ-06F

<< page 1: the facts of the matter

An analytical pinch of salt

It is vitally important to understand the shortcomings of the above analysis of the SJ-12/SJ-06F rendezvous. First and foremost is the inherent accuracy (or lack thereof) of the TLEs used as the primary source of data for analysis. The TLEs produced by the JSpOC are based on observations collected by the US military’s Space Surveillance Network (SSN). The process behind converting observations to TLEs has been covered here previously (see “The numbers game: What’s in Earth orbit and how do we know?”, The Space Review, July 13, 2009). The TLEs are issued without any information on their accuracy or the amount or quality of the underlying observations. Various groups have published studies that attempt to estimate TLE accuracy by comparing the positions of certain objects determined by TLEs with the positions for the same objects as determined by the satellite owner-operator or precision techniques, such as laser ranging. A paper presented at the Fourth European Conference on Space Debris in April 2005 estimated that for objects in orbits below 800 kilometers, TLEs describing their orbits can have errors of 100 to 600 meters.

It is somewhat of an art form to be able to calculate an accurate post-maneuver elset.

Based on these results, any analysis of the locations of objects or separation between two objects that is based on TLEs could be off by a few hundred meters or more. Thus, a calculation based on TLEs that predicts the satellites will come within 200 meters of each other could mean they actually collided or that they missed by hundreds of meters. An example of this is the collision between the Iridium 33 and COSMOS 2251 satellites on February 10, 2009. Calculations made prior to the collision using TLEs predicted that they would miss each other by about 600 meters.

Another factor to consider is the potential role that space weather could have played in either affecting the satellite’s orbit or in the data processing. The Earth’s atmosphere extends for considerable distance in space, and even though it is far too thin to breathe, the upper atmosphere exerts drag on satellites in orbits out to about 800 kilometers in altitude. Thus, atmospheric drag is an important perturbation in accurately calculating the orbits of satellites in LEO. The density of the atmosphere, and thus how much drag it causes, varies according to the energy output of the Sun, certain solar events, and their interaction with the Earth’s magnetic field.

Specific, short-term space weather events that impact atmospheric density can cause errors in the estimation of an orbit. If the computer algorithms assume these changes will last into the foreseeable future, it can result in the creation of inaccurate elsets. Once the event is over and the atmospheric density returns to its previous state, the algorithms will adjust accordingly and usually start producing accurate elsets again. According to the warnings and alerts published by the National Ocean and Atmospheric Administration (NOAA) Space Weather Prediction Center, there appear to have been no significant space weather events that could have caused the anomaly in the orbit of SJ-06F on August 19.

Aside from their inherent errors, there have also been times in the past when TLEs were published by the JSpOC that did not accurately reflect an object’s real location in space. Sometimes this is due to operator error, such as using too few observations to create the TLE, or including inaccurate observations in the calculation. This can also happen as a result of cross-tagging, where the orbits for two different objects are accidently swapped. In most cases, it is easy to see these errors when looking at the elset history of a space object. The bad elset sticks out from the normal perturbation pattern, and after it is corrected, the space object usually goes back to its normal pattern. If the tracking data for the space object shows a new pattern, then its orbit was changed by some force outside of the normal perturbations.

It appears as though the anomaly on August 19 does reflect an actual change in the orbit of SJ-06F, although only time will tell for certain.

Maneuvers, defined as deliberate changes in an orbit, also cause breaks in the normal perturbations of an orbit and present a special problem for orbit determination. The orbital determination algorithms used by the US military to create the TLEs assume ballistic flight, meaning the object is coasting and not under any sort of powered thrust. Attempting to create a TLE using observations that were collected while a satellite was under thrust will almost certainly result in a poor TLE. Thus, none of the TLEs published for SJ-12 contain the actual maneuvers it performed. They only contain the satellite’s orbital position before and after each maneuver. Therefore, the TLEs published right after each maneuver can be inaccurate.

It is somewhat of an art form to be able to calculate an accurate post-maneuver elset. The analyst needs to ensure that there are enough observations collected from sensors on the post-maneuver orbit, and that the calculation does not use any observations taken when the satellite was under thrust. Thus, the more time that elapses from any single maneuver, the more accurate the satellite’s calculated position is likely to be.

Although we cannot rule out a bad elset as the cause of the anomaly in the orbit of SJ-06F, it does not appear to be the cause from the currently available data. The anomaly only occurs in mean motion, while the other orbital parameters are largely unaffected. A bad elset is likely to have one or more incorrect orbital parameters. After the anomaly, SJ-06F appears to have settled on a new perturbation pattern, thus indicating that it is being tracked repeatedly by the SSN on a new orbit. This adds weight to the accuracy of the post-anomaly position.

Figure 10
Figure 10: Prior anomalies in the mean motion of SJ-06F. (larger version)

The recent orbital history of SJ-06F shown in Figure 10 reveals two other anomalies. One occurred around April 6, 2010, and the other around May 2, 2010. However, in both of those cases, SJ-06F stabilized and resumed its normal periodic variations, leading to the conclusion that the anomalies were a result of the processing of the tracking data and did not reflect changes in the satellite’s actual orbit. It appears as though the anomaly on August 19 does reflect an actual change in the orbit of SJ-06F, although only time will tell for certain. Analysis of its orbital position over the coming weeks and months will provide further evidence as to whether its orbit was changed or whether it was simply an anomaly in the data.

Technically valid reasons for rendezvous operations

Although China has not stated publicly why the rendezvous between SJ-12 and SJ-06F took place, the technical details of the event establish two broad reasons why it might have. The first is to develop and practice rendezvous techniques. China is currently in the middle of an ambitious human spaceflight program, including building and operating a space station in low Earth orbit (LEO). Achieving this goal requires developing the skills for on-orbit rendezvous and docking. In the past, China has performed these techniques with modules launched simultaneously, which then slowly separated and then recombined (see “China’s BX-1 microsatellite: a litmus test for space weaponization”, The Space Review, October 20, 2008). Assembling a space station in LEO and regularly transporting people and goods to it requires expertise in maneuvering one space object to rendezvous with another from different orbits in a controlled and safe manner.

Rendezvous operations fall into one of two categories: cooperative and uncooperative. Cooperative rendezvous are done with some amount of feedback data from the target object. This feedback can take the form of radar reflectors, transmission of telemetry, or other methods. Uncooperative rendezvous is done without any assistance from the target, although the satellite being maneuvered may have onboard radar or optical sensors to help guide it. There is a large difference in the degree of difficulty between cooperative and uncooperative rendezvous, with uncooperative being much more challenging. All rendezvous operations are notoriously tricky, given the distances from human controllers and speeds involved, and it is safer to practice with unmanned satellites than spacecraft with humans on board.

A second possible reason to conduct such a rendezvous is for satellite formation flying: operating multiple satellites in close proximity to each other. Typically this is done so that the satellites overfly the same location on the ground in quick succession or even have the same ground location in their field of view at the same time. Overflight in quick succession allows the different sensor payloads on each satellite to collect data on the same ground location in a coordinated manner, and simultaneous collection allows for applications such as interferometry. However, satellites flying in formation are typically spaced by hundreds of kilometers or more apart. Theoretical concepts have been developed for flying satellites in much closer formations, such as in a fractionated spacecraft concept (see “Breaking up may be good to do”, The Space Review, November 2, 2009), but these have not been demonstrated operationally.

A lack of transparency in spacecraft activities can lead to misperceptions and mistrust among space actors thus leading to instability and potential conflict.

A third possible reason for doing such a rendezvous would be to perform a close-up inspection of another satellite, which has in fact been done by other countries in the past, most notably by the American XSS-11 and MiTEx satellites (see “Mysterious microsatellites in GEO: is MiTEx a possible anti-satellite capability demonstration?”, The Space Review, July 31, 2006). The XSS-11 was designed to be able to rendezvous with another satellite and observe it at close range with a variety of sensors, including high resolution LIDAR mapping. Shortly after its launch in July 2005, the XSS-11 conducted a series of maneuvers to rendezvous with the Minotaur upper stage that placed it in orbit.

The two MiTEx satellites were placed in geostationary orbit (GEO) on June 16, 2006, and reportedly contain optical sensors, which could be used to inspect other satellites. Much of what they have been doing is not publicly known, since their position in orbit is not released publicly by the US military as part of the satellite catalog. However, it was reported in January 2009 that at least one of the MiTEx satellites was maneuvered to inspect the United States’ inactive DSP-23 missile launch detection satellite after its failure in fall 2008 (see “The ongoing saga of DSP Flight 23”, The Space Review, January 19, 2009).


The orbital rendezvous of the Chinese SJ-12 and SJ-06F satellites is just the most recent example of the ever-increasing number of space actors that are performing complex operations in space. Space is a global commons, and all states share the use of space for the many benefits it provides. Irresponsible behavior or accidents by any space actor can create situations where satellites are damaged, and in some cases generate additional space debris. This can lead to degradation in the long-term sustainability of Earth orbit, and potentially higher costs to use space. Thus, all states should ensure that the utmost care is taken when planning activities in space and that they are conducted in as safe a manner as possible.

Orbital rendezvous are complex operations, but are also seen as necessary by many space actors for a number of reasons, including inspecting satellites to determine causes of failure, on-orbit servicing to repair or refuel damaged satellites, and removing debris from crowded orbital regimes. A lack of transparency in spacecraft activities can lead to misperceptions and mistrust among space actors thus leading to instability and potential conflict. Operating outside of best practices, which are usually the result of experience and hard-won lessons, can lead to costly mistakes and accidents. When space actors undertake these activities it is important that they do so in a transparent and responsible manner and in accordance with established best practices and norms of behavior.



Enter your email address below to be notified when new articles are published: