Thousands of working satellites orbit the Earth — and alongside them, many times more of their dead remains. Retired satellites, spent rocket stages, fragments from explosions and collisions, and small parts released during launches form an environment in low Earth orbit through which every new mission must learn to move. The problem scientists warned about as early as the 1970s has accelerated in recent years: thousands of new satellites head into orbit every year, and the number of tracked objects keeps growing.
How Much Debris Orbits the Earth
According to the ESA 2024 report on the state of the orbital environment, surveillance networks at the time tracked approximately 35,000 objects. Of these, about 9,100 were active satellites; the rest — roughly 26,000 objects — was debris. Catalogued pieces larger than ten centimetres are, however, only the tip of the iceberg. Models by ESA and NASA estimate that there are more than one million fragments larger than one centimetre in orbit, and tens to hundreds of millions of millimetre-scale particles that neither radars nor telescopes can detect. Detailed statistics are collected by the ESA Space Debris Office, and scientific findings are published by NASA’s Orbital Debris Program Office (ODPO). The 2024 ESA report states that the amount of debris continues to grow rapidly and that natural decay through drag from residual atmosphere cannot offset that growth.
The 1978 Warning
The theoretical basis for today’s concerns was laid in 1978 by NASA scientists Donald Kessler and Burton Cour-Palais. In the paper “Collision frequency of artificial satellites: The creation of a debris belt”, published on 1 June 1978 in the Journal of Geophysical Research, they calculated that as the number of satellites grows, so does the probability of mutual collisions — and each collision produces fragments that further increase the probability of subsequent collisions. Above a certain critical density, the process could become self-sustaining: debris would breed further debris faster than the atmosphere could pull it down. The authors wrote at the time that, under certain conditions, such a belt could begin to form before the end of the 20th century and become a serious problem in the following century, while timely technical and operational measures could substantially delay its formation. The text of the paper remains available in the NASA technical reports archive. The term “Kessler syndrome” later came to denote a cascade in which some orbital bands would become practically unusable for new satellites.
Collision Over Siberia: Iridium 33 and Kosmos 2251
Theory became practice on 10 February 2009. On that day, over northern Siberia at an altitude of around 790 kilometres, the functioning American telecommunications satellite Iridium 33, weighing 560 kilograms, collided with the decommissioned Russian satellite Kosmos 2251, weighing approximately 900 kilograms. It was the first accidental high-speed collision of two intact satellites. A detailed analysis of the collision was summarised in a 2010 NASA analysis: the relative velocity at impact was 11.647 kilometres per second, and by June 2010, 1,267 fragments of Kosmos 2251 and 521 fragments of Iridium 33 had been catalogued, with cataloguing of smaller pieces continuing. The scale of the event is also described in NASA’s 2009 orbital debris quarterly, according to which 783 larger fragments had been identified by the end of March 2009 and ground observations confirmed a far larger quantity of smaller wreckage. A NASA presentation for the 2009 IAC congress states that the space surveillance network eventually captured more than 1,800 new fragments — about one-tenth of the tracked population at the time — and estimates around 200,000 collision fragments between one and ten centimetres in size. Some of the debris from both satellites will remain in orbit for many decades.
Megaconstellations: Orbit Is Filling Up
Whereas the world once launched dozens of satellites per year, today it launches thousands — chiefly because of commercial communications constellations. The largest of them is being built by SpaceX under the name Starlink. According to SpaceX’s own 2025 progress report (operator figures), the company flew more than 120 Falcon 9 missions dedicated to launching V2 Mini Starlink satellites in 2025, has completed more than 300 such launches in total, and operates more than 9,000 active satellites in orbit. Independent records of launches and constellation status are kept by astrophysicist Jonathan McDowell in his Starlink network overviews. Each Starlink satellite has propulsion for avoidance manoeuvres and is meant to burn up in the atmosphere at the end of its life, but the sheer number of objects increases the frequency of close approaches and forced manoeuvres. Starlink is joined by competing projects — for example Amazon’s Kuiper constellation or the Chinese GuoWang networks — so traffic density in low Earth orbit will keep growing in the coming years.
Risk to Crewed Flights
In low Earth orbit, bodies move at around 7 to 8 kilometres per second. At such speeds, even a centimetre-sized fragment can pierce a satellite’s shielding, and a millimetre-sized particle can damage a solar panel or a pressure wall. After the 2009 collision, NASA identified untrackable fragments as the main risk to operational satellites. The warning also applies to crewed stations. As early as April 2009, NASA’s CloudSat satellite had to perform an avoidance manoeuvre because of a fragment from Kosmos 2251, and in September 2009 the International Space Station prepared for a manoeuvre for the same reason — it was ultimately called off because the updated risk calculation did not exceed the threshold for performing it. Both events are documented in a 2009 NASA presentation. Since then, the station has manoeuvred regularly, and its modules bear traces of microscopic impacts. As the number of objects grows, so does the number of close-approach warnings, which means more frequent manoeuvres, higher fuel consumption, and a greater burden on control teams — and, in the extreme case, the need to evacuate the crew into the return vehicle.
The European Answer: Zero Debris by 2030
The European Space Agency responded to the worsening trend with an initiative meant to stop the creation of new debris. Under the Zero Debris Charter, ESA has committed that its own missions will be “debris neutral” by 2030 — that is, they will leave no wreckage in orbit after the end of operations — and it promotes the same philosophy among European industry and in international standards. In practice, this means satellites capable of controlled demise in the atmosphere, upper rocket stages reliably removed from orbit, and stricter limits on the release of operational small items such as covers and separation elements. The charter builds on ESA’s annual space environment reports, which document that without actively removing large dead bodies, the long-term cascade cannot be stopped. The 2030 goal is ambitious, and meeting it also depends on commercial operators, which today launch most new satellites.
Orbital Cleaners
Preventing new debris does not, however, solve the wreckage already in orbit. That is why missions are being developed to deliberately remove large pieces of debris. The European ClearSpace-1 project, prepared in cooperation with ESA, was originally to demonstrate the capture and deorbiting of the retired Vega VESPA adapter — a body of around one hundred kilograms that has been orbiting since 2013. After the adapter was apparently struck by a fragment in August 2023 and new debris appeared nearby, the mission target was changed: the PROBA-1 satellite from 2001 is now to be removed (ESA ClearSpace-1). The company summarises information about its servicing missions on the ClearSpace website. Further along is the Japanese-British company Astroscale: its ADRAS-J probe, selected by the Japanese agency JAXA for the first phase of a commercial debris-removal demonstration programme, launched in February 2024 and became the first body to safely approach a large unprepared piece of debris and document it. According to the company’s December 2024 announcement (operator figures), the probe approached to within just 15 metres of a Japanese rocket upper stage and gathered data for the follow-on ADRAS-J2 mission, which is to actually capture the fragment and pull it down. An overview of the mission is on the ADRAS-J page. Both missions show that the technology for clean-up exists — the open question remains funding and who will pay for the clean-up.
What Comes Next
Experts agree that a single approach to fifteen metres will not save the orbit. Models by both NASA and ESA show that in the most congested bands between 700 and 1,000 kilometres, several large dead bodies will need to be removed every year to stabilise the environment — otherwise, accidental collisions will keep producing debris even if the creation of new debris is halted. Added to this is the need for better data: more precise tracking of smaller fragments, sharing of position data among operators, and automation of avoidance manoeuvres, because human teams can no longer handle the growing number of warnings manually. In 1978, Kessler and Cour-Palais hoped that timely launch and operating rules would substantially delay the formation of a debris belt. Almost fifty years later, it turns out that we have used that delay only partially — and that the coming decade will decide whether low Earth orbit remains an open transport route or becomes a minefield.
Sources
- Kessler, Cour-Palais: Collision frequency of artificial satellites (1978) – https://agupubs.onlinelibrary.wiley.com/doi/10.1029/JA083iA06p02637
- NASA NTRS: Collision frequency of artificial satellites – https://ntrs.nasa.gov/citations/19780057167
- ESA: Space Environment Report 2024 – https://www.esa.int/Space_Safety/Space_Debris/ESA_Space_Environment_Report_2024
- ESA: Space Environment Report (current PDF) – https://www.sdo.esoc.esa.int/environment_report/Space_Environment_Report_latest.pdf
- ESA: environment statistics (DISCOS) – https://sdup.esoc.esa.int/discosweb/statistics/
- NASA ODPO: Orbital Debris Quarterly News – https://orbitaldebris.jsc.nasa.gov/quarterly-news/pdfs/ODQNv28i2.pdf
- NASA: Analysis of the Iridium 33 – Cosmos 2251 collision (2010) – https://ntrs.nasa.gov/api/citations/20100008433/downloads/20100008433.pdf
- NASA: Iridium 33 / Cosmos 2251 – IAC 2009 – https://ntrs.nasa.gov/api/citations/20100002023/downloads/20100002023.pdf
- NASA: Orbital Debris Quarterly News 13-2 (2009) – https://ntrs.nasa.gov/api/citations/20090017680/downloads/20090017680.pdf
- SpaceX: Starlink Progress Report 2025 (operator figures) – https://starlink.com/public-files/starlinkProgressReport_2025.pdf
- Astroscale: ADRAS-J mission (operator figures) – https://www.astroscale.com/missions/adras-j
- Astroscale: 15-metre approach, December 2024 (operator figures) – https://www.astroscale.com/news/astroscales-adras-j-achieves-historic-15-meter-approach-to-space-debris
- ClearSpace (operator figures) – https://clearspace.today/
- ESA: Zero Debris Charter – https://www.esa.int/Space_Safety/Clean_Space/The_Zero_Debris_Charter
Content Transparency and AI Assistance
How this article was created:
This article was written by Hermes Agent — an open-source self-improving AI agent (Nous Research) — in collaboration with our editors. The editors set the topic, research direction, and primary sources; Hermes prepared the research, structure, and draft. Every key claim passed iterative fact-checking against two independent sources plus blind review audits.
Editorial review and fact-checking:
- ✓ The text was editorially reviewed
- ✓ Fact-checking: All key claims and data were verified against two independent sources
- ✓ Fact corrections and enhancement: Our editors corrected inaccuracies and added subject matter expertise
- ✓ Multi-round audit: The article passed repeated blind audits including a final shared review round
AI model limitations (important disclaimer):
Language models can generate plausible-sounding but inaccurate or misleading information (known as “hallucinations”). We therefore strongly recommend:
- Verifying critical facts in the primary sources (official documentation, peer-reviewed research, subject matter authorities)
- Not relying on AI-generated content as your sole information source for decision-making
- Applying critical thinking when reading
Spotted a mistake?
Let us know in the comments under the article — we check every report.
Technical details:
- Agent: Hermes Agent (Nous Research, open-source)
- Model: Muse Spark 1.3 Contributor (Meta)
- Method: Self-improving editorial skill — research, drafting, multi-round blind audit, human review
- Run: Self-hosted Hermes Agent instance run by our editors, connected to the model provider’s API (not a local model)
Be First to Comment