Why the flyby is a poor answer
One pass, and an unforgiving geometry.
A flyby at 0.2c crosses 1 AU in 2,500 seconds. The entire Proxima system, out past Proxima c at 5.2 AU, is traversed in about 3.6 hours.
The encounter geometry is unforgiving. At a closest approach of 10 million kilometers the angular rate at periapsis is 6 × 10⁻³ radians per second, so resolving a 12,000-kilometer planet at even a thousand pixels requires exposures under a fraction of a millisecond, from an aperture measured in centimeters, at photon rates that make the downlink the binding constraint.
The flyby is the cheap answer and it is a poor one. It is also the only one that fits inside 2100, which is why most of this section is about it.
Magnetic sail deceleration
Works best when you need it least.
A superconducting loop deflects interstellar protons and generates drag. Drag scales with density times velocity squared, which means it works best at the beginning of the deceleration and fades exactly as it is most needed.
The Zubrin and Andrews analyses suggest a magsail can shed a substantial fraction of an interstellar velocity over decades but cannot bring a craft from 0.2c to rest in a useful time. Its natural role is a terminal brake from 0.05c rather than a complete solution.
Two caveats. Magsail drag coefficients have never been measured, only modeled. And the local interstellar medium density, on which the entire deceleration profile depends, has been measured in situ at exactly one location, by two spacecraft, over a few decades — see the interstellar medium. A magsail braking against the solar wind is in principle testable at 1 AU today, and nobody has flown one.
Photon braking and the tension it creates
Deceleration selects for bright stars.
If a craft arrives slowly enough, the target star’s own radiation pressure can decelerate it. Heller and Hippke showed in 2017 that a sail of roughly 100,000 square meters carrying a 100-gram payload, arriving at about 0.046c, can be braked by Alpha Centauri A into a bound orbit, on a transit of roughly 95 years.
The mechanism scales linearly with target luminosity, and that produces the sharpest architectural tension in the interstellar case.
| Target | Luminosity (solar) | Relative brake | Habitability interest |
|---|---|---|---|
| Sirius A | 25.4 | 17× Alpha Cen A | None |
| Alpha Centauri A | 1.52 | reference | Candidate giant, possible moons |
| Tau Ceti | 0.52 | 0.34× | Contested candidates |
| Alpha Centauri B | 0.50 | 0.33× | None detected |
| Barnard’s Star | 0.0035 | 0.0023× | HZ ruled out above 0.57 M⊕ |
| Luyten’s Star | 0.0036 | 0.0024× | GJ 273 b in HZ |
| Proxima Centauri | 0.0017 | 0.0011× | Proxima b in HZ |
| Teegarden’s Star | 0.00073 | 0.00048× | b and c near HZ, quiet star |
The deceleration architecture selects for bright stars. The habitability argument selects for dim ones. Photon braking at Proxima is roughly 900 times weaker than at Alpha Centauri A, and 2,000 times weaker at Teegarden. Every M dwarf in the target set — which is to say every star with a known rocky planet in its habitable zone — is effectively unbrakeable by its own light. The stars we most want to stop at are the ones that cannot stop us.
Alpha Centauri is the beachhead
Proxima is the destination.
There is one clean way through that tension, and it is a node-and-corridor argument rather than a propulsion argument.
Alpha Centauri A can brake a craft. Proxima lies 0.21 light-years from it, about 13,000 AU, which is a routine in-system transfer for a vehicle that has already stopped.
A decelerated arrival at Alpha Centauri A, followed by a leisurely 13,000 AU transfer to Proxima using the arrival vehicle’s remaining capability, reaches the nearest rocky habitable-zone planet with an orbiter rather than a flyby. No other target in the neighborhood has a bright neighbor.
This is the same logic that puts the anchor at Callisto rather than Ganymede and the staging at Phobos rather than the Martian surface: you do not stage at the destination, you stage at the node that lets you reach it.
Staged sails
Physics-permitted, and an array the size of the Earth–Moon distance.
Robert Forward’s staged-sail concept splits the sail so that an outer annulus separates near the target, reflects the home beam back onto the inner sail, and decelerates it. The physics is sound. The engineering is not, and it can be dismissed quantitatively rather than by hand-waving.
The home array must place a spot smaller than the retro-mirror at a range of 4.25 light-years. With a 100-meter mirror and a 1-micron beam, the required array diameter is the wavelength times the range over the mirror diameter, which is 4 × 10⁸ meters. That is a phased optical array roughly the Earth–Moon distance across, held in phase, pointed with sub-microarcsecond accuracy, at a target whose position is known only as it was 4.25 years earlier.
Physics-permitted, and an object four orders of magnitude larger in aperture than anything else in the architecture. Not a 2100 topic.
The hybrid, and the recommendation
Obvious mainly because nothing else closes.
For the second mission family — tons rather than grams, arriving rather than passing — the recommended architecture is the obvious hybrid.
Beam a heavier craft, tens of kilograms rather than one gram, to 0.05 to 0.08c using the same inner-system array. The array does not change; only the shot does, at a proportionally longer burn or a proportionally larger sail.
Carry a fission-fragment or deuterium–helium-3 stage sized for terminal deceleration from that speed, which is a mass ratio of order 10 rather than order 1,000.
Use magsail drag to shed the first portion of the arrival velocity, reducing the propulsive burden by whatever the interstellar medium actually delivers.
Accept a 60 to 90 year mission. This does not fit a career. It fits an institution, barely.
The relay chain
The option that may be the real answer.
There is a fifth approach almost never discussed as an interstellar architecture, and it may be the most defensible thing in the whole case.
You do not have to send anything to another star for an interstellar program to pay for itself. A chain of nodes outbound from the Sun at 550 AU, 800 AU and 1,000 AU performs astronomy that nothing inside the heliosphere can perform, provides the receiving infrastructure that makes any later probe’s data return possible, tests every propulsion and autonomy technology the star mission needs at a fraction of the risk, and establishes the departure corridor.
The first thousand AU of an interstellar program is fully justified on its own scientific terms, independent of whether anything is ever sent further. That matters enormously, because it means the program can be funded incrementally against near-term returns rather than requiring a single 60-year commitment made up front. It is the difference between a program that can survive an election and one that cannot.
The instrument at the end of that chain is the solar gravitational lens, and it is simultaneously the targeting system, the primary science return and the communications infrastructure of any outward program.