Space
Space debris: the problem nobody owns
Tens of thousands of tracked objects, millions too small to track, and a governance framework built when orbit was empty.
Every launch leaves something behind. Spent upper stages, deployment hardware, fragments from explosions and collisions, and satellites that stopped working.
The population has been growing for six decades and has recently accelerated sharply.
The scale
Space surveillance networks track tens of thousands of objects larger than roughly ten centimetres.
Statistical models estimate objects in the one-to-ten centimetre range in the hundreds of thousands, and millimetre-scale fragments in the hundreds of millions.
Only the largest are catalogued and can be avoided. The middle band is the dangerous one: too small to track, large enough to destroy a spacecraft.
Why small fragments matter so much
Orbital velocity in low Earth orbit is around seven and a half kilometres per second. Two objects on crossing orbits can close at ten to fifteen kilometres per second.
Kinetic energy scales with the square of velocity, which means a one-centimetre aluminium fragment carries energy comparable to a substantial explosion at ground level.
Shielding against millimetre-scale particles is practical — the Whipple shield, a spaced double wall that vaporises the impactor, is standard on crewed vehicles.
Shielding against centimetre-scale objects is not practical at any realistic mass. Those must be avoided, and they cannot be seen.
The cascade concern
Each collision produces fragments, each of which is a new collision hazard.
Above a certain density, the fragment population becomes self-sustaining: collisions generate debris faster than atmospheric drag removes it, and the environment degrades without further launches.
The scenario is usually named after the researcher who modelled it in the 1970s. Whether it is imminent is contested; that the risk rises with density is not.
Two events dominate the current debris population: a deliberate anti-satellite test that destroyed a defunct satellite at high altitude, and an accidental collision between an active communications satellite and a defunct military one.
Together these produced thousands of trackable fragments, many in orbits with very long decay times.
Altitude determines everything
The single most important variable.
Below roughly four hundred kilometres, atmospheric drag removes debris within a few years. Objects at these altitudes are self-cleaning.
Between six hundred and a thousand kilometres, decay takes decades to centuries. This band is also the most heavily used, which is unfortunate.
Above that, and in geostationary orbit, objects effectively remain indefinitely. Geostationary satellites are therefore boosted to a graveyard orbit at end of life rather than deorbited.
Which means a constellation at very low altitude is far less concerning than one at eight hundred kilometres, whatever its size.
Mitigation practice
International guidelines have converged on a set of measures, adopted with varying force in national licensing.
Passivation — venting residual propellant and discharging batteries at end of mission, to prevent later explosions, which have historically been a major fragment source.
Deorbit within a defined period after mission end. The long-standing guideline was twenty-five years; some regulators have shortened this substantially, and enforcement varies.
Collision avoidance manoeuvres, coordinated using tracking data. Active satellites now perform these routinely, and the number has risen sharply with constellation deployment.
Design for demise — building spacecraft so they burn up completely, rather than dropping surviving components.
Why governance is weak
Space law rests on treaties drafted in the 1960s, when the environment was empty and the actors were two states.
The core principles — that a launching state retains jurisdiction over and liability for its objects — mean that no one can remove another state's debris without permission.
There is no binding international rule requiring deorbit, no enforcement mechanism, no traffic management authority, and no agreed liability framework for a collision between two commercial operators.
Licensing happens nationally, which creates the possibility of regulatory arbitrage.
Active removal
Technically demonstrated at small scale, using nets, harpoons, robotic capture and magnetic docking.
The obstacles are cost, and the fact that the objects most worth removing are large uncontrolled bodies tumbling unpredictably.
The deeper obstacle is that nobody has an incentive to pay. Removing a spent rocket body benefits every operator and no one in particular, which is the definition of a public good.
A small number of publicly funded demonstration missions exist. A sustained programme does not.
What would actually help
Shorter mandatory deorbit periods, consistently enforced across licensing jurisdictions.
Better tracking of the centimetre-scale population, which is a sensor investment rather than a research problem.
Shared, machine-readable conjunction data so operators can coordinate automatically rather than by email — which, remarkably, is still how some of it works.
And a norm against destructive anti-satellite testing, which several states have now declared unilaterally and which is the single cheapest improvement available.
