The Space Reviewin association with SpaceNews

space debris
Some satellites appear in one catalog but not in another, while the identity of others is unknown. (credit: ESA)

Tracking unknown satellites

Bookmark and Share

Tracking satellites used to be something that only large organizations could do, but today enthusiastic amateurs track many satellites. But to track many of the more interesting satellites it helps to have an observatory with professional support. This is the story of how some people are doing just that, and the contribution to safety in space that they are making.

But first, when we think of the satellites that are orbiting above us, we distinguish them by one of several ways. One is by their names. Another unique way to identify them is by their “COSPAR ID,” which is an alphanumeric designator, and then each satellite has a unique number. There isn’t room to give the history of those here, but there are a number of sites with more details.

One fascinating project is to compare the various satellite catalogs that exist, which quickly reveals that all of these catalogs have quirks. There are several lists of satellites that are in space, designed for similar goals, but they are very different.

These satellites are listed in a number of “satellite catalogs” and one example is the one maintained by the US Space Force. They are the (default official) keeper of the official list of satellites. This is often called the Space Track catalog and it lists most of the larger satellites that have been launched. The Space Track catalog lists the satellite number, the COSPAR identifier, a “country of origin” for most objects, and some of the orbital parameters.

One fascinating project is to compare the various satellite catalogs that exist, which quickly reveals that all of these catalogs have quirks. There are several lists of satellites that are in space, designed for similar goals, but they are very different. The most interesting sites are those that collect unique observations of satellites so they have data that they collect and process. There are also sites that repeat the satellite catalog but don’t actually accept tracking of satellites and generate unique information; they are not a subject of this article.

Satellite catalogs

But what major satellite catalogs exist? There is Space Track, which is the public face of the default official world’s satellite catalog. As I said, Space Track assigns the official satellite numbers and COSPAR identifiers and these are used by most satellite catalogs. They do an adequate job of accounting for most larger satellites and they have the goal of listing every satellite that has been identified. Then there is the Mini-MegaTORTORA site, which is a project of the Kazan Federal University. The “ISON satellite catalog” is provided by the JSC “Vimpel Interstate Corporation” and the Keldysh Institute of Applied Mathematics in Russia. ISON stands for International Scientific Optical Network and there are several good sites that give more information about that organization. They assign their own satellite numbers and do not use the ones assigned by Space Track. The Ukraine Network of Optical Stations (UMOS) catalog also uses numbers as shown in Space Track.

Most of the efforts to create and maintain this catalog have been made by the Research Institute of the Mykolaiv Astronomical Observatory under the direction of Dr. Oleksander Shulga.

There are certainly other satellite lists and catalogs, including several commercial ones now. Numerica Corporation has one and so does ExoAnalytic Solutions, but the ExoAnalytic catalog appears to not have any public access (that I have found.) There is LeoLabs and also ComSpOC, the original Commercial Space Operations Center. These are certainly very comprehensive lists and catalogs but these are companies that have to charge for access to their (expensive to obtain) information and so any publically available information mostly duplicates Space Track. Europe has begun a satellite catalog (the EU SST), but they say that it might one day have some part of it that is public.

Various catalogs leave objects out, or do not give orbital parameters for, some satellites. For instance, the Mini-MegaTORTORA catalog does not give parameters for any satellites that are from, according to their country of origin, the Commonwealth of Independent States (CIS) as they say on their web page. One interesting fact is that there is not an identifier for Russia, only for the CIS.

The ISON catalog mostly lists objects that have an orbital period of more than 200 minutes (I have not checked each orbit in it, so some may be in orbits with shorter periods.) The ISON list apparently gives parameters for objects regardless of which country claims them, though I have not checked that.

So, who can find these unknown satellites? If you have a professional observatory, you are likely to have both a well-qualified team and excellent equipment.

There are a number of objects that are listed on the Space Track catalog but their orbital parameters are not available, the rationale for which ones are not there is not applied consistently. I have written about this situation before. The satellites whose orbital parameters are not available from Space Track are also certainly not available from the commercial operators, such as Numerica or LeoLabs, probably since that might offend the US hovernment. If you go to the Numerica satellite listing and try to display recent orbital parameters for satellite 23609, COSPAR 1995-034A, you will find that it is not available even though those parameters are available from other sources.

But some decisions just do not make sense. Why are parameters for object 41941, a DSN upper stage from Japan, not given? It will be interesting to see if the EU SST catalog lists the ones that are not available from Space Track. The UMOS satellite catalog gives parameters for some of these satellites such as objects 23609 and object 29249, named USA 184.

Finding unknown satellites

Two satellite catalogs list satellites that do not have sequential satellite numbers (those that are sequential since the first satellites) and do not have COSPAR IDs. Space Track has a list of satellites that do not appear in the Satellite Catalog, those are the “analyst satellites” and you have to get a separate list of them. These are objects that have been tracked but have not been officially added to the catalog for various reasons.

UMOS has not (yet) tracked many unique satellites. They have tracked four (as of this writing) consistently enough to calculate good orbital parameters and they have tracked another, but they are not confident of its orbits yet. They have used two telescopes and currently track with the KT-50 telescope and generate orbital parameters. Ugeen Kozirev of the Mykolaiv Astronomical Observatory installed and runs software originated by Igor Kara of Odesa National University, named after I. Mechnikov. This accepts observations and calculates orbital parameters using methods developed by Edgar Everhart, PhD, (and others) of the USA.

The two best tracked satellites are shown in their satellite catalog at the time of this writing are numbered 90103 and 90083. These are apparently satellites that do not appear in Space Track. One good question that we don’t know the answer to is whether these possibly duplicates of satellites in the commercial catalogs, like those created by Numerica, LeoLabs, ExoAnalytic, and ComSpOC

The catalog at the UMOS web site was developed by Alexander Mazhaev of the Mykolaiv Astronomical Observatory. This catalog provides up-to-date orbital parameters in a very common format for a number of satellites, many of which are found in this format only in this catalog.

Types of orbits and why some may not be in catalogs

There are several ways to find satellites. Most satellites orbit not far above the Earth’s surface and they can easily be tracked by radar. Satellites that orbit below about 1,000 kilometers can be tracked with radar, assuming that they have the right inclinations. But many of them will be too small to routinely track and they may never be added to a catalog. There are many satellites that are geosynchronous or geostationary; those are almost only tracked by telescopes. These satellites are of high interest because many of them are commercial satellites and generate a lot of income. Then there are medium earth orbit satellites, like the GPS satellites, that can generally be tracked by radar.

One of the most fascinating families of orbits is those with high eccentricities, which means they can have a low perigee but high apogee. Some of these orbits are used as geotransfer orbits: the payload is launched into a low orbit but then an upper stage propels it into a high eccentricity orbit that takes it to geosynchronous orbit. These normally have a low inclination and often upper stages (or debris) are left in the geotransfer orbit. These are not normally too difficult to track because at some point their orbit near perigee will be visible to a radar station and perigee will be near the equatorial latitudes.

By checking potential orbits against all other objects in space, safety for the planned satellites will be improved.

Then there is a group of orbits which are known for their inclination of about 63 degrees. This inclination (the explanation is beyond the scope of this article) is useful because the payload can have an apogee which is set over the northern hemisphere and tends to remain there. The common version of this orbit provides a near-geosynchronous payload that dwells over the higher latitudes, and is useful for communications and surveillance. This set of orbits can be difficult to track: perigee is often over broad ocean areas in the southern hemisphere where there are few trackers and the objects are at the high altitude of apogee when they are visible. Only a sensitive telescope operated by a professional team is likely to track it. Many of these highly eccentric objects have gotten a lot less attention that those in the other orbits and apparently there are large unknown satellites in those orbits.

New satellites

So, who can find these unknown satellites? If you have a professional observatory, you are likely to have both a well-qualified team and excellent equipment. The UMOS network had been tracking satellites since 2012 and were happy to talk about tracking some challenging objects. Starting with some high eccentricity satellites, one technique to find unknown satellites would be to modify the orbital parameters by changing the right ascension of the ascending node (RAAN). One of my contentions is that flights tend to have similar rates of precession of that parameter, so modifying that can give an orbit to search. Another technique is to get orbital parameters that are old, and no longer usable, and update them to give a search orbit. Of course, updating an orbit without new observations can give a good, or a totally wrong, orbit.

Matching new satellites to other satellites

Using various techniques, so far UMOS has tracked five satellites which appear to not be in other satellite catalogs. One has an inclination of 62.3 degrees, one has an inclination of 62.7 degrees, one has an inclination of 9 degrees, and one has an inclination of 27.9 degrees. All have high eccentricity. One they have not gotten enough tracking on to calculate a reliable orbit.

It is convenient to give a satellite number of 99999 to a new object until it can be matched with an existing object. So far we have assigned a likely satellite number of object 90103 by using the “RAAN Test” that I have been developing. Using this test we compared 99999 to an object that has not been tracked in a long time and they appear to have the same rate of RAAN precession. With the new parameters, object 90103 has recently been tracked numerous times. A second object appears to match an older object with the number 90083 and it has also been very predictable with the parameters generated from tracking.

Looking at Figure 1, where the values of RAAN have been plotted for several satellites, it is apparent that 99999 and 90103 are co-linear and, even though 90103 has a large gap in tracking ,it is reasonable to conclude that they are the same object.

Figure 1
Figure 1: Matching Unknown Satellites

The unknown satellite is plotted against several objects that it could be identified with, both an older unknown (90103) and several example satellites. The three other satellites were chosen because their orbital parameters are not available from Space Track: to find them you have to go to one of the other sites that carry them (they are easily available.) UMOS, for instance, regularly tracks these satellites so that they will not confuse them with unknowns. The unknown was also compared to several other satellites, such as some Molniya satellites, but did not appear to match any of them.

We have also assigned likely numbers to two other objects: they are 90109 and 99212 and they may be added to the catalog if they are predictable. They have also tracked one other object but have not gotten enough data on them to generate a quality orbital description yet.

These objects have high values of eccentricity and are in a few inclinations that are commonly used. Objects 90103 and 90083 have inclinations of approximately 63 degrees and so are in orbits that are used as communications satellites for northern latitudes. Since there are no unknown communications satellites, these could be aperture covers or similar debris. The other objects are in lower inclinations and appear to be connected to launches that put objects in geosynchronous or geostationary orbits; these are probably debris left in those orbits. Again, we have done a quick comparison with other satellite catalogs and think that these objects are not in Space Track and do not have COSPAR IDs. UMOS does track a number of satellites whose parameters are hidden in the Space Track catalog, this is so UMOS can compare the unknown satellites to as many other, cataloged, satellites as possible.

Future plans

With this early success, we are going to continue to try to find satellites. This could be useful to satellite builders who wanted to ensure that they compared potential future orbits to all objects which might collide with their payloads, not just the ones in Space Track. By checking potential orbits against all other objects in space, safety for the planned satellites will be improved.

We are going to continue to search for unknown satellites and will look at some of these satellites to estimate when they might reenter the atmosphere. Object 99212 has a perigee of only 360 kilometers and an apogee of under 35,000 kilometers, so that is the one to watch most closely. These high eccentricity objects have the odd characteristic of having their perigees increase as well as decrease, so this object will likely not reenter in the next few years.

An earlier article described the potential use of RAAN Test to assign unknown objects to which launches they might have originated with. Possibly the rate of RAAN precession can also indicate which launch a piece originated with, here 90103 has a very similar rate to satellite 23609 but a very different rate from the other two satellites that have normal catalog numbers. But to make an assignment like that an object would have to be compared with all objects in similar orbits (with similar inclinations and eccentricity).


Charles: plots for this project were produced with a very wonderful application, Datagraph (I have no association with them except as a happy user) and Pamela Schultz, PhD generously helped me with some questions. Mike Marston wrote most of an application that is used to do a lot of processing of orbits. He is far better at writing C++ code than I am.

Mykola: The Mykolaiv Astronomical Observatory provided great support. Thanks to the Director Oleksander Shulga, DSc, Alexander Mazhaev, Ugeen Kozirev, PhD, and the Laboratory Director, Nadiia Maigurova, PhD.

Note: we are using a new commenting system, which may require you to create a new account.