Solar Gravitational Lens Telescope
The largest telescope humanity will ever use is already built. Park a small spacecraft 550 astronomical units from the Sun, hold it on a line, and the Sun's gravity will focus light from a distant exoplanet onto its sensor with an amplification of 10¹¹ and an angular resolution of 10⁻¹⁰ arcseconds. That is sharp enough to image a 25 km feature — a coastline, an ice cap — on a world 98 light-years away. The lens is free. Getting to its focal line is the problem.
How it works
General relativity says mass bends light. A photon grazing the Sun's limb is deflected by 1.75 arcseconds, the prediction Eddington measured at the 1919 eclipse. The same bending makes the Sun a converging lens with a focal line that begins at roughly 542 AU (rays just skimming the photosphere) and extends to infinity. Slava Turyshev's JPL team, funded by two NASA Innovative Advanced Concepts grants (Phase II in 2018, Phase III in 2020), has worked out the optics in detail since 2017.
Unlike a glass lens, the focal length is fixed by the Sun's mass (1.989 × 10³⁰ kg) and is largely wavelength-independent. The amplification at 1 μm is 10¹¹. A 1-meter telescope sitting on the focal line is optically equivalent to a kilometer-scale aperture in Earth orbit, for the one point of sky behind the Sun.
The catch: the spacecraft, the Sun's center, and the target exoplanet must be collinear. The exoplanet orbits its star. The Sun has its own proper motion. The spacecraft is itself drifting at whatever speed got it there. So the imager has to scan a roughly 1.3 km × 1.3 km image patch in the focal plane to assemble a megapixel of an Earth-sized world — a deconvolution problem, not a single-exposure photograph. A 2024 study proposed tethered spacecraft pairs to cover multiple image-plane positions in parallel.
The numbers that matter
- Focal line: ~542–548 AU minimum, extending to infinity along the radial direction.
- Amplification: ~10¹¹ at optical wavelengths.
- Angular resolution: ~10⁻¹⁰ arcsec, roughly 10⁶× sharper than tech-james-webb-space-telescope.
- Target benchmark: a 25 km surface element on an exo-Earth at 30 parsecs (98 ly), in ~6 months of integration.
- Distance scale: 550 AU is 14× Pluto's mean orbit. Voyager 1, the most distant human artifact, sits at ~163 AU after 48 years of coasting.
The corona is the lens flare
The signal you want — focused photons from a planet — arrives smeared along an Einstein ring around a Sun whose corona is emitting roughly 10⁶ times more light than the planet at every wavelength of interest. The coronagraph aboard the spacecraft has to suppress direct solar light by a factor of 10⁸ to 10¹⁰ while leaving the ring intact. This is the load-bearing engineering problem. It is hard in a way no current space coronagraph (Roman, HabEx, LUVOIR concepts) has demonstrated, because those instruments suppress a distant star, not a Sun filling a meaningful fraction of the field of view.
Clouds are the second problem. A 6-month integration averages over changing weather. A 2025 modeling paper (arXiv:2504.18630) argues that temporal-variability analysis can separate surface and atmospheric features for partly-clouded worlds, but cloud-dominated planets need substantially longer integration to recover any surface signal.
Getting to 550 AU
At Voyager 1's heliocentric speed of 17 km/s, 550 AU takes about 153 years. No mission proposal aimed at SGL imaging this century can use chemical propulsion alone.
| Approach | Plausible cruise speed | 550 AU transit |
|---|---|---|
| Voyager-class chemical + gravity assists | ~3.6 AU/yr | ~150 years |
| Nuclear-electric propulsion | 3–4 AU/yr | ~140 years |
| Oberth maneuver, 3–5 solar radii perihelion | ~6–8 AU/yr | 70–90 years |
| Laser-driven solar sail (Turyshev concept) | ~20 AU/yr | ~25–30 years |
The current JPL baseline is a swarm of small laser-pushed solar sails — same technology family as mission breakthrough starshot — that assemble a meter-class telescope at the focal region. The required ground-based laser array (hundreds of megawatts to gigawatt class, phased) does not exist. Neither do the sails at the required reflectivity and thermal tolerance. The physics is fine. The infrastructure is a decade-plus project that nobody has funded.
What an image would actually show
At 25 km resolution on a 30-pc Earth-analog:
- Continents, oceans, large ice sheets — yes.
- The "red edge" spectral jump of chlorophyll-like pigments at ~700 nm — yes, and simultaneously across the disc.
- Seasonal vegetation and ice changes over the integration window — yes.
- Cities as resolved structures — no. A New-York-sized urban region is ~50 km across, one or two pixels, only detectable as an anomalous spectral or thermal patch, not as a recognizable shape.
- Atmospheric O₂ + CH₄ in chemical disequilibrium — yes, via simultaneous spectroscopy.
So this is a biosignature imager, not a technosignature one. It would not photograph an alien city. It might force the concept-fermi-paradox to mutate from "is there life?" to "we are looking at green continents 98 light-years away — now what?"
What's contested
Three open questions sit unresolved.
First, whether the coronagraph can actually reach 10⁹ suppression of an extended, nearby, time-variable source. Lab demonstrations exist for distant point-source contrast at ~10¹⁰. The Sun is not a point source from 550 AU.
Second, whether the Einstein ring's outer structure carries enough recoverable information after passing through the Sun's corona itself, which scatters and refracts light in ways that change on minute-to-hour timescales. The 2020–2024 JPL papers argue it is recoverable. Critics (notably Ed Turner at Princeton, and several SPIE conference responses 2022–2024) argue the noise floor is underestimated.
Third, whether the political case survives the budget. A single-target SGL mission images one star system, slowly. A future generation could re-aim with new probes, but each pointing is a separate decades-long mission. Compared to a Habitable Worlds Observatory imaging dozens of nearby systems at lower resolution, the SGL is a deep telescope for a shallow sky.
Why this has to do with other realms
The SGL is a physics result — Einstein 1916 — that becomes an engineering problem only at the propulsion frontier of concept interstellar propulsion. It is also a concept-fermi-paradox instrument: if it finds a green world, the Drake equation's biological terms collapse to "yes" and the search shifts entirely to the L term (civilization lifetime). And it is a concept-information-theory dressed as optics: how much information can you extract from a 10⁹-to-1 background, given a finite integration time and a moving target? The answer dictates whether a 6-month exposure produces a continent or a hash.
An open question
If a swarm of laser-launched probes reached 550 AU in the 2070s and produced the first resolved image of an exo-Earth at 30 parsecs, and the image showed clear chlorophyll signatures and seasonal ice — would that change anything we do, or only what we believe?
Key sources
- Turyshev, Toth et al. — Direct Multipixel Imaging and Spectroscopy of an Exoplanet with a Solar Gravitational Lens Mission (NIAC Phase II final report, 2020). The canonical reference.
- Eshleman, V. R. — Gravitational lens of the sun: its potential for observations and communications over interstellar distances, Science 205 (1979). The 1979 paper that opened the file.
- Turyshev & Toth, Phys. Rev. D series (2017–2022) — to verify exact citations; these work out the diffraction-regime optics rigorously.
- arXiv:2504.18630 (2025) — cloud-cover deconvolution feasibility study.
- Eddington's 1919 eclipse measurement — the empirical foundation. Discussed in Daniel Kennefick, No Shadow of a Doubt (2019).
Further reading
- Einstein's Telescope by Evalyn Gates (2009) — gravitational lensing for cosmology readers, useful for intuition before the SGL papers.
- Slava Turyshev's NIAC presentations on the NASA NIAC archive — the source talks, with all the optical diagrams the papers compress.
- The Fabric of the Cosmos by Brian Greene — chapter on GR's tests, for readers who want the 1919 eclipse and the modern checks in one place.
- Centauri Dreams (Paul Gilster's blog) — running coverage of SGL mission concepts and laser-sail propulsion since the mid-2010s.
See Also
- mission breakthrough starshot — the laser-sail program whose hardware roadmap the SGL mission concept depends on.
- concept fermi paradox — the question an SGL image of a green continent would reframe.
- tech solar sail — the propulsion stack between Earth orbit and 550 AU.
- concept metamaterials — the sail-material problem shared with Starshot.
- mission voyager 1 — the only object that has covered even a third of the distance to the focal line, and the benchmark for how slow "chemical" really is.
- concept information theory (extracting a planet from a 10⁹-to-1 background is a Shannon problem before it is an optics problem)