Why leftover junk in orbit can make even more junk
In 1978, two NASA researchers, Donald Kessler and Burton Cour-Palais, published a paper about how often artificial satellites would collide with each other. Their reasoning was simple. As more satellites go up, collisions become more likely. A collision breaks things into fragments, and each fragment is one more thing that can hit something else. They suggested this could grow into a belt of debris around Earth [1].
The paper was a prediction made with a mathematical model, not a report of something that had already happened. It said such a belt could form under certain conditions, and that early launch rules and operating habits could delay it [1]. The idea is now often called the Kessler syndrome, though that name was not in the abstract.
That is why cleanup is a real question and not only a tidy-up chore. The worry is not that the sky is full today. It is that junk, left alone, could keep multiplying in the busiest orbits.
A real collision in 2009 showed how one crash creates many pieces
On February 10, 2009, a working US communications satellite, Iridium 33, and a retired Russian satellite, Cosmos 2251, collided about 790 kilometers (roughly 490 miles) above northern Siberia. NASA's Orbital Debris Program Office described it as the first accidental high-speed collision of two intact spacecraft. Their orbits crossed at nearly a right angle, so they met at more than 11 kilometers per second (over 24,000 miles per hour) [2].
By the end of March, the US Space Surveillance Network had cataloged 783 of the larger fragments, and special observations showed many smaller ones too [2]. One broken satellite had become hundreds of tracked objects.
This does not prove a runaway chain reaction has started. It does show the first step of the chain: a collision turns one object into many.
The first way to reduce junk is to stop adding to it
Cleaning up what is there is hard, so the main rule is not to add more. The European Space Agency (ESA) says its missions must dispose of satellites safely, by letting them fall back into the atmosphere or by moving them to a safe altitude, with a success probability above 90%. ESA also notes that even if all launches stopped today, larger pieces breaking up, at a typical rate of four or five a year, would keep the number of objects growing [3].
Two different tricks are used depending on where the satellite flies.
Low orbits: come down and burn up. A satellite in low Earth orbit can be steered to a lower altitude. The thin upper air drags on it, and it eventually re-enters the atmosphere. For ESA's new missions, the longest a satellite may stay in the protected low orbits after its mission ends was cut from 25 years to five, in rules that took effect in November 2023 [3]. An example: in 2011 ESA lowered its ERS-2 satellite's remaining orbital lifetime from more than 200 years to well below 15 years, and it re-entered on February 21, 2024 [3].
Very high orbits: move over. A geostationary satellite sits about 36,000 kilometers (22,000 miles) up, where it stays above one spot on the equator. Bringing it down would take a lot of fuel. Instead, ESA says its orbit should be raised by about 300 kilometers at the end of its mission, a safe distance from working satellites. That needs a speed change of only about 11 meters per second, roughly the fuel normally used for three months of holding position [4]. This higher parking zone is often called a graveyard orbit.
Junk that is already up there cannot take itself away
Those rules only help for new satellites that are still working. A dead rocket stage or an old satellite cannot fire an engine, so someone else has to go and get it. That is called active debris removal. ESA also mentions building satellites with interfaces that would make it easier to remove them from orbit if self-disposal fails [3].
The first problem is not grabbing. It is simply getting close safely. A dead object is usually tumbling and does not send out its position, and nobody can ask it to cooperate. Space engineers call this a non-cooperative target.
Japan's ADRAS-J practiced approaching a real dead rocket stage
ADRAS-J, built by the Japanese company Astroscale, was Phase I of a Japan Aerospace Exploration Agency (JAXA) program called the Commercial Removal of Debris Demonstration (CRD2). It launched in February 2024. Its target was the upper stage of an H-IIA rocket, about 11 meters (36 feet) long, 4 meters wide, and about 3 tons, left in orbit after launching Japan's Ibuki greenhouse-gas satellite in 2009 [5] [6].
In July 2024, ADRAS-J flew around the stage at about 50 meters and took pictures from different sides. JAXA said such images are meant to reveal the motion, damage, and wear of debris that has been in orbit a long time, information that is scarce worldwide [6]. Astroscale reports that the spacecraft later came within 15 meters, and that its autonomous collision-avoidance ability was checked. Astroscale also announced in March 2026 that ADRAS-J had begun deorbiting itself, which it describes as coming after 293 days of mission life [5].
ADRAS-J did not remove anything. It showed that a private spacecraft can approach and study a real, uncooperative piece of junk. A second mission, ADRAS-J2, is planned for fiscal year 2027 to approach, observe, and remove the same stage [5]. These are Astroscale's own plans, and a plan can change.
Why cleanup is still mostly an open question
So, can anyone clean up the junk? For now, the careful answer is: we can slow its growth, and we have started to learn how to reach single large objects. We cannot yet clear the sky. The sources here do not say how many removal missions would be needed or what they would cost, so this article does not guess.
ESA also says that space is a globally shared resource, and that its own rules cannot be enough alone [3]. Which large objects would you remove first, and why? Try searching for "active debris removal" on a space agency's website; and the sources below are good places to read more.