Space Debris and Space Sustainability

At a glance

Space debris is any human-made object in orbit that no longer serves a purpose: dead satellites, spent rocket stages, mission-related parts and, above all, fragments from explosions and collisions. Only the larger pieces can be tracked individually. Smaller ones are known from statistical models, so the table mixes catalogue counts with model estimates.

The population of objects in Earth orbit: key figures
Quantity	Value	Type / date	Source
Objects larger than 10 cm	68,450 (including about 11,300 active payloads)	MASTER-8 model estimate, February 2026	ESA [1] [2]
Debris from 1 cm to 10 cm	about 1.5 million	MASTER-8 model estimate, February 2026	ESA [1] [2]
Debris from 1 mm to 1 cm	about 230 million	MASTER-8 model estimate, February 2026	ESA [1] [2]
Objects regularly tracked and catalogued	about 47,270	Catalogue count, 23 October 2026	ESA [1]
Satellites still functioning	about 15,900	ESA count, 23 October 2026	ESA [1]
Active satellites (independent tally)	17,099, of which 11,148 are Starlink	As of 26 October 2026	McDowell [3]
Starlink satellites in orbit	11,190 (13,069 launched)	As of 28 October 2026	McDowell [4]
Total mass in Earth orbit	more than 17,400 tonnes	ESA estimate, 23 October 2026	ESA [1]
Fragmentation events since 1957	more than 660	Break-ups, explosions, collisions, anomalies; 23 October 2026	ESA [1]

Checked 29 October 2026. ESA values come from the Space Debris Office’s “Space debris by the numbers” (DISCOSweb, updated 23 October 2026) and the ESA Space Environment Report 2026 (September 2026, data to end-2025). The size-class figures are model outputs, not counts. “Active” is defined differently by each tally, so ESA’s 15,900 and McDowell’s 17,099 are not directly comparable.

Sources: [1] [2] [3] [4]

Where debris comes from

About 7,370 successful launches since 1957 have placed some 28,270 satellites in orbit, of which about 19,440 remain in space. Most catalogued objects, however, are not working hardware but rocket bodies, mission-related items and fragments. Fragmentations are the largest single source: ESA counts more than 660, and over the past two decades about 9.8 non-deliberate fragmentations have occurred per year on average, typically from residual propellant or batteries in old stages and satellites.

Speed is what makes small objects dangerous. In low orbit debris moves at about 7–8 km/s, and the average collision speed is about 10 km/s. The US Space Surveillance Network generally tracks objects larger than about 10 cm. Ground radars sample debris down to a few millimetres, and returned spacecraft surfaces record even smaller impacts.

Figure 1. Objects in orbit by size class. The count axis is logarithmic. Dots are model estimates; the amber line is the count of catalogued objects of all sizes. Fragment icons are not to scale. Data: sources [1] and [2] on this page.

Traffic is rising fast. In 2025 there were more than 300 launches and over 4,000 new payloads, and ESA counted 1,200 intact objects re-entering, more than three a day.

Sources: [1] [2] [5]

Interactive lab: impact energy of a small fragment

Start with the 1 cm aluminium sphere preset: weighing only 1.4 g, it carries as much energy at a closing speed of 10 km/s as a slow-moving car. Then shrink the diameter to 1 mm and grow it to 10 cm; notice that the energy rises with the cube of the diameter, and see why the 1–10 cm range marked by the amber band is considered the most troublesome.

Orbital regimes and lifetimes

International guidelines define two protected regions. Low Earth orbit (LEO) extends from the surface to 2,000 km altitude. The geostationary region (GEO) is a band 200 km above and below 35,786 km, within 15° of the equator. LEO hosts most satellites, including the megaconstellations. Navigation systems occupy medium Earth orbit in between.

In LEO the only natural cleaner is atmospheric drag, which weakens sharply with height. According to NASA, debris below about 600 km normally re-enters within several years, at 800 km decay often takes centuries, and above 1,000 km objects normally remain for 1,000 years or more. Drag also rises when solar activity heats the upper atmosphere. GEO has no meaningful drag, so retired satellites are moved to a graveyard orbit that should stay clear of the protected band for at least 100 years.

Sources: [5] [6]

Major fragmentation events

Fengyun-1C (11 January 2007). China destroyed its retired weather satellite with a ballistic interceptor. NASA called it the worst single contamination of LEO in 50 years: debris spread from about 200 km to over 4,000 km, and 1,613 fragments were tracked by 31 March 2007.

Iridium 33 – Cosmos 2251 (10 February 2009). The first accidental collision between two intact satellites occurred at about 790 km over Siberia, at more than 11 km/s. The working Iridium (about 560 kg) and the defunct Cosmos (about 900 kg) produced 823 catalogued large fragments by the end of March 2009.

Kosmos 1408 (15 November 2021). A Russian direct-ascent missile destroyed the satellite in an orbit of about 490 × 465 km. By 7 March 2022, 1,604 fragments were catalogued, and NASA estimated that the ISS collision risk settled at roughly twice its pre-test level. The low altitude meant many fragments re-entered quickly; ESA attributes the 2022 spike in re-entering debris to this test.

Intelsat 33e (19 October 2024). This Boeing-built GEO satellite lost power and broke up; trackers reported dozens of fragments, and the cause remains under investigation. Overall, 2024 added over 3,000 catalogued fragments, mostly from propulsion-related break-ups, while 2025 saw comparatively few.

Sources: [2] [7] [8] [9] [10]

The Kessler syndrome, stated carefully

In 1978 Donald Kessler and Burton Cour-Palais of NASA argued in the Journal of Geophysical Research that as satellite numbers grow, collisions become likely, and each collision creates fragments that make further collisions more likely. Over time this could form a debris belt around Earth. The label “Kessler syndrome” was coined later at NORAD.

It is often misread as a sudden chain reaction that seals off space within hours. In 2010 Kessler and NASA colleagues stressed that cascading is slow: at present densities it takes decades to change the small-debris environment significantly. The real concern is that some bands are already above a “critical density”, where collisions would keep increasing the population even with no new launches. Their modelling pointed to instability between roughly 600 and 1,700 km.

Figure 2. Left: the logic of the Kessler process (schematic). Right: orbital lifetimes that lengthen with altitude in low Earth orbit, the band the models found unstable, and the altitudes of two fragmentation events (linear scale). Data: sources [5], [8], [9] and [12] on this page.

ESA’s 2026 report agrees that debris would keep growing even without launches. Its “health index”, which compares projected strain with a sustainable benchmark, jumped from about 4 to 50 in one year. In the same edition the projection horizon was cut from 200 to 100 years and the benchmark moved from the 25-year to the 5-year disposal target, so the value is best read with the method notes.

Sources: [2] [11] [12]

Megaconstellations, traffic and the night sky

Starlink satellites now make up about 65% of active spacecraft: 11,148 of 17,099 in McDowell’s tally on 26 October 2026. With thousands of satellites in narrow shells, collision avoidance has become routine.

The US Space Force’s 18th Space Defense Squadron screens all active satellites against the catalogue several times a day and issues Conjunction Data Messages (CDMs) through Space-Track.org, free for registered users. Operators decide whether to manoeuvre based on collision probability; the ISS, for example, normally moves when the risk exceeds 1 in 10,000. Space traffic management aims to turn this into coordinated rules: shared data standards, better orbit determination and agreed procedures when two manoeuvrable spacecraft meet.

Astronomy is affected too. A study of Hubble archive images found satellite trails in 2.7% of exposures taken in 2002–2021, with the fraction rising over time; ground-based surveys and radio astronomy are hit harder. The IAU’s Centre for the Protection of the Dark and Quiet Sky (CPS) coordinates mitigation work.

Sources: [3] [4] [5] [13]

Re-entry risk and mitigation rules

Most of a re-entering satellite burns up, but dense parts can reach the ground, usually in the ocean or sparsely populated areas. The IADC guidelines and the US ODMSP (2019) cap the expected casualty risk of an uncontrolled re-entry at 1 in 10,000. Above that, a controlled re-entry over an unpopulated ocean area is expected.

The Inter-Agency Space Debris Coordination Committee (IADC), a forum of 13 space agencies, revised its guidelines in January 2025 (Rev. 4). They call for limiting operational debris, passivating stored energy, avoiding intentional destruction, managing collision risk, and disposing of LEO objects within 25 years at the latest with at least 90% reliability, adding that large constellations may need much shorter lifetimes. The UN General Assembly endorsed the voluntary COPUOS Space Debris Mitigation Guidelines in 2007.

Rules are tightening. On 29 September 2022 the US FCC voted 4–0 to require LEO satellites to deorbit no later than five years after mission end, with a two-year transition (Via Satellite). ESA’s Zero Debris approach aims to significantly limit debris production in Earth and lunar orbits by 2030, and the non-binding Zero Debris Charter, drafted in 2023, has been joined by more than 200 organisations and over 20 countries. ESA reports that over the last decade 15–60% of payloads reaching end of life attempted to meet the 25-year limit, and 70–90% of those succeeded.

Sources: [2] [6] [14]

Active debris removal

Kessler and colleagues argued in 2010 that disposal rules alone will not stabilise the most crowded bands, so some large objects must be removed. This is hard: the target has no attitude control or docking fixtures and may be tumbling.

Astroscale’s ADRAS-J, part of JAXA’s CRD2 programme, launched in February 2024 towards the H-IIA upper stage that launched GOSAT in 2009. It made fly-around observations at about 50 m in July 2024 and reached about 15 m on 30 November 2024 before an autonomous abort. Astroscale’s 520 kg ELSA-M, designed to capture a prepared end-of-life OneWeb satellite, was contracted in March 2026 to fly on Isar Aerospace’s Spectrum, with no published launch date. ESA’s ClearSpace-1 now targets the PROBA-1 satellite, with launch planned for 2029.

Sources: [12]

Data sources for researchers

ESA DISCOS (DISCOSweb): objects, launches, fragmentation events and physical properties; registration required.
ESA Space Environment Report: annual statistics and projections (2026 edition, PDF).
Space-Track.org: the US catalogue, orbital elements, decay predictions and CDMs.
CelesTrak (SATCAT): searchable catalogue with CSV downloads.
planet4589.org: Jonathan McDowell’s independent catalogues and constellation statistics.
NASA Orbital Debris Program Office: the Orbital Debris Quarterly News archive of fragmentation reports.

Counts differ between sources because of definitions (catalogued or tracked, active or manoeuvrable) and update times; always quote source and date together.

Sources: [1] [2] [3]

Known vs open questions

Established

The catalogued population and mass in orbit keep growing; fragmentations produce most debris objects.
Orbital lifetime rises steeply with altitude, from years below 600 km to a millennium or more above 1,000 km.
Some LEO bands are above critical density, where debris would grow even without launches.
Guidelines cap re-entry casualty risk at 1 in 10,000 and require passivation and timely disposal.

Open

How accurate are the model estimates of the sub-centimetre population?
Will 5-year disposal suffice with tens of thousands of constellation satellites, or is routine removal unavoidable?
What are the atmospheric effects of so many re-entries?
Who coordinates traffic globally, and how is liability assigned for unattributable debris?
Can active removal become affordable at the scale needed?

Sources and further reading

ESA Space Debris Office — Space debris by the numbers (updated 23 October 2026) — European Space Agency
ESA — ESA Space Environment Report 2026 (September 2026) — European Space Agency
McDowell, J. — Active satellites statistics — Jonathan’s Space Pages (planet4589.org)
McDowell, J. — Starlink statistics — Jonathan’s Space Pages (planet4589.org)
NASA Orbital Debris Program Office — Frequently asked questions — NASA
IADC — Space Debris Mitigation Guidelines, IADC-02-01 Rev. 4 (2025; A/AC.105/C.1/2025/CRP.9) — UNOOSA
NASA ODPO — Orbital Debris Quarterly News 11(2), April 2007 (Fengyun-1C ASAT test) — NASA
NASA ODPO — Orbital Debris Quarterly News 13(2), April 2009 (Iridium 33 – Cosmos 2251 collision) — NASA
NASA ODPO — Orbital Debris Quarterly News 26(1), March 2022 (Cosmos 1408 ASAT test) — NASA
Johns Hopkins Space Security — Intelsat 33e breakup (28 October 2024) — Johns Hopkins University
Kessler, D. J. & Cour-Palais, B. G. — Collision frequency of artificial satellites: The creation of a debris belt (J. Geophys. Res. 83(A6), 1978) — NASA Technical Reports Server
Kessler, D. J., Johnson, N. L., Liou, J.-C. & Matney, M. — The Kessler Syndrome: Implications to Future Space Operations (AAS 10-016, 2010) — University of Western Ontario (author archive)
Kruk, S. et al. — The impact of satellite trails on Hubble Space Telescope observations (Nature Astronomy 7, 262–268, 2023) — Nature
ESA — The Zero Debris Charter — European Space Agency