Near Gigapixel Imaging of Exoplanets from the Gravity Lens Regions

Exoplanets are planets in other solar systems. They are very faint but they are even harder to see because of the much brighter star they are orbiting. If we send a telescope to the solar gravitational lens (SGL) point on the opposite side of our sun then light from objects like exoplanets will be focused to provide 100 billion times more magnification. The Sun becomes a telescope that is 1.4 million kilometers wide for the SGL regions.

We could resolve exoplanets around Proxima B to 450-meter resolution using a one-meter telescope SGL mission. If there was an earth-sized planet around Proxima B, we could resolve to 800 megapixels. We would only be able to resolve 10 square kilometers at a time. The space telescope would have to roam around the einstein-ring image of the target object to assemble the full image. The image would need to be converted from an einstein ring back into the image of the exoplanet. A giant 1.3 kilometer focus line diameter space telescope would be able to resolve an entire einstein-ring image of an earth-sized exoplanet at 100 light years from the right SGL location.

There are ten known stars within ten light years. There are over 150 stars and white and brown dwarfs known to be located within 20 light-years.

One light year is 63,241 AU (astronomical units, distance from Earth to the sun.) The gravitational lens points are 200 to 420 times closer than the nearest star and would be 4000 to 8000 times closer than starts that are 100 light years away.

Voyager 1 is escaping the solar system at a speed of about 3.6 AU per year. Voyager 2 is escaping the solar system at a speed of about 3.3 AU per year. If we went 10 times faster than Voyager then we could have missions reach the SGL areas in about 20 to 40 years. Ideally, we would want to decelerate and keep space telescopes moving around the lensing locations for each target solar system. You would only be able to resolve 10 square kilometers at a time. You would need to precisely move around the image area to take millions of picture to form the full picture of the planet. You could move along the focus line to form the full picture. You would not have to completely stop the telescope. The SGL reduces the image of the exo-Earth by a factor of ~6,000. An orbital radius of 1 AU becomes ~2.25´104 km and an orbital velocity of 30 km/s translates into 5 m/s. Solar gravity accelerates the Earth at 6 mm/s2. The imager spacecraft needs to accelerate at ~1 µm/s2 to move in a curved line mimicking the motion of the exoplanet. Even if the probe mass is ~1,000 kg, the corresponding force is 1 mN, which may be achieved with electric propulsion. Electric propulsion should be able to move the space telescope to sample the pixels in the Einstein ring needed for imaging. JPL we could potentially reach velocities of roughly 60 to 70 kilometers per second with planned solar sails. This would be 5 times faster than Voyager. A rocket could perform a slingshot around the sun to send a mission to the SGL in 27 to 35 years. An new megawatt laser and lithium-ion drive combination could reach 45 AU per year in speed. This could send missions to the SGL with 15-25 year trip times and the beamed power could be used to slow down and move around. There are a few thousand stars within 100 light years. Light from the parent star is typically focused many tens of thousands of kilometers away from the focal line that corresponds to the instantaneous position of the exoplanet. Light contamination due to the parent star is not a problem when imaging with the SGL. A regular telescope has an optical lens with a single focal point. A gravitational lens has no single focal point but a focal line. A mission to the SGL carrying a modest telescope and coronagraph opens up a possibility for direct megapixel imaging and high-resolution spectroscopy of a habitable Earth-like exoplanet at a distance of up to 100 light years. The entire image of such a planet is compressed by the SGL into a region with a diameter of ~1.3 km in the vicinity of the focal line. The telescope, acting as a single pixel detector while traversing this region, can build an image of the exoplanet with kilometer-scale resolution of its surface, enough to see its surface features and signs of habitability.

The gravitational lens regions are from 550 AU to 1200 AU. We would need to send one or more mission for each solar system we would want to study.

Propulsion Technology Possible by 2030



There is a NASA NIAC phase 2 study to use lasers to beam 10 megawatts of power to power new ion drives. This will enable a system to go ten times faster than any previous space mission. This will go 40 AU per year. It would take less than year to get to Pluto.



They are building and proving out the various components of this system. The sail and the ion drives are coming together. The hard part is the phased array lasers.



They are boosting the testing voltage up to 6000 volts so the lithium ion drives can be directly driven. Direct drive eliminates the need for a lot of heavy electronics which would kill the performance.



The phased array lasers will increase the power density over the solar density by 100 times.



Going from a laser wavelength of 1063 nanometers down to 300 nanometers would reduce the needed power and system size.





SOURCES- NASA, Arxiv, Slava G. Turyshev (Physicist, JPL), Louis Friedman (Co-Founder and Executive Director Emeritus, The Planetary Society)

Written By Brian Wang. www.nextbigfuture.com

16 thoughts on “Near Gigapixel Imaging of Exoplanets from the Gravity Lens Regions”

  1. Not if you want to take advantage of the 2 million km aperture diameter the Sun produces. Otherwise you have to keep individual telescopes’ positions accurate to a fraction of a wavelength over that distance, to use them as an interferometer.

  2. Now that we know the “scattered disk” region (50-2000 AU) has a significant population, we may not need to launch them all from Earth. Rather, we send a factory out there and crank out telescopes locally.

    The name of this region refers to objects scattered by the giant planets’ gravity to distant orbits. This differs from the classical Kuiper Belt, which ranges from 30-50 AU. Those are leftovers from the original Solar Nebula and are more or less where they first formed (ignoring giant planet migration).

    Beyond 2000 AU, we call it the Oort Cloud, and it is no longer disk-like. The Sun’s gravity is so weak at this distance that stars in the cluster the Sun formed in, and later passing stars, gas clouds, and galactic tides have scrambled the orbits into a sphere. The Scattered Disk objects still retain more of their original direction orbiting the Sun.

  3. Wouldn’t it be cheaper and easier to just construct a large telescope in space by flying in formation several large mirrors.

  4. “isn’t that one of the basic assumptions?” … staying in one place for a long time.

    Yes, it is, of course. 

    The hard problem is figuring out with essentially 1 receiving pixel, whether the target planet is veering off course, orrotating in view. Something of a conundrom. 

    Rather readily solved by another (you?) poster’s thought of having an array of 1 meter pixel-integrators at the image plane. You know, thousands, or ideally millions of them. Then, if it turns out that the thing-a-mabob being imaged is rotating, the moment-by-moment pickup of each receiver will correlate with rather high coefficients of linear agreement with its neighbors. Even if tangential. 

    Which is the reason why the SCIENCE mission would have hundreds (at least) of 1 m² circular pixel-receivers. 

    I, however, remain guardedly sanguine about the prospect of engineering a star-shade sufficiently far away (thousands to hundreds-of-thousands of kilometers), with a sufficiently narrow annulus (open ring) and another fall-away imagining shield at LEAST as large as Sol’s projection … to allow sufficient imaging resolution to actually divine something scientifically significant about the imaged exoplanet. 

    Might really be simpler (and scientifically more interesting) to finely image exo-stars. 

    You know, see their whole faces, their sunspots, their waning and waxing output. Could be quite the LANews topsheet. 

    Just saying,
    GoatGuy ✓

  5. Acknowledging my ignorance, isn’t that (remaining in one very tight spot) one of the basic assumptions of doing solar gravity lens astronomy?

  6. Starshade, starshade, starshade, starshade. 

    Has to be figured in, and has to be critically well positioned (to block Sol, and select the in-focus annulus) continually.  

    Just saying,
    GoatGuy ✓

  7. LOL, if you can keep the PV panels at the focal point! Interstellar gravitational scalar drift ought to be a beast.

  8. Actually, there is a subtly different problem that is something of a “great idea killer” unless adequately addressed. 

    Namely, the physics of the Einstein light deflection depends on the integration of gravitational field orthogonal pull over the light path. Since the pull is a function of k/R² (distance from centerpoint of the gravitational ‘lens’ object), the the only meaningful ‘pencil’ beam in focus would be that from an extremely narrow ring at radius R, width ε. In order that this be the only beam selected, you in turn need not just a star-shade to block Sol, but you need a star shade large enough to select the critical pass cylinder of light, and another one outside that to block less deflected light. 

    Quite a significant demand. It doesn’t help much to have many at-focal-point sensors (the much talked about 1 meter collectors), if they don’t each have their own Sol-blocking open-ring aperture setup in place.

    For instance, if we choose the ring-of-exoplanet light at a distance of 110% of Sol’s radius, with an annular star shade at a distance of 25,000 km from the telescope, then the Einstein focal distance (for Sol) is 758 AU, (the 558 AU goes up as a function of R³)

    But an annular star shade is needed about 343 m in diameter, with a 1 cm annulus (ε), giving 2.0e-10 radian ecosystem spacial res, which is pretty much useless. 

    Just saying,
    GoatGuy ✓

  9. Good explanation.

    We tend to believe a small satellite would suffice. This reinforces my belief that any serious sky survey using SGLs would require an incredible amount of parallel work, to be able to build and launch that many telescopes so faraway. Because this will need thousands of transportable units, to scan only a handful of the possible destinations within 100 ly…

    Probably this will only happen in force, after we get a real industrial capability in space, and enough GDP from Earths’ and space activities, to create such a magnificent and probably fully automated project.

  10. That’s one of the funny things of gravity lenses: the stellar brightness at certain key spots could allow some previously unforeseen power and usage models, at least that far from the Sun.

  11. It makes sense to do shorter range test flights before we try interstellar. There are two kinds of useful missions to the 800-1000 AU region. First is visiting scattered disk objects, of which there are 13 known that orbit this far, and probably hundreds more yet to discover. Second is gravitational lens telescope. If we are to eventually travel to these other stars, we may as well get a detailed survey before we go

  12. Brian, your numbers are off, but that’s not unusual because you don’t really understand what you are copy/pasting.

    Photons that just graze the Sun’s surface come to a focus at 544 AU, but you will also be seeing the edge of the Sun itself, which makes it hard to see. Instead you want to back off to about 800-1000 AU, where photons that miss the Sun by half a radius, and are more weakly bent, reach a focus. Half a radius allows you to block out the Sun itself, and the Sun’s corona, and therefore see the photons you are looking for.

    Next, a one meter telescope won’t see much of any use. Simple optics tells us the image size in relation to the actual object is reduced by the ratio of the object distance to focal distance. Proxima is 268,400 AU from the Sun. If our telescope is 800 AU, the image will be 335 times smaller. Our 1 meter telescope thus images a 335 meter patch of a planet. Since the Earth moves at 30 km/s, if your telescope is stationary, your view will change 100 times a second.

    Instead, what you want is a line of telescopes, whose sideways movement is nearly matched to the planet. If the line was a kilometer long, you would image a 335 km swath of the planet. You could then adjust your position and scan more swaths until you map the whole planet. By the time we launch such a mission, we should have an ephemeris of the planet orbit, and know where to position ourselves to find it.

    The star’s image will be amplified to 36 times full moon brightness

  13. “I just want to know 3 things: How many jobs this will produce in my district/state? How much more money can be budgeted so a lot of graft and ‘waste’ can be created? How much of answers to numbers 1 & 2 are going to help me with re-election, PAC money and ‘fact finding’ trips to Hedonism Island?” — Your typical ‘science oriented’ politician in Congress on the relevant NASA committees/subcommittees.

    Of course, he will never say any of that out loud before reporters or kids visiting his office from his district/state. But that doesn’t stop NBF fantacists thinking this is not how congress and NASA budget-making really works, of course.

  14. If Jim Woodward et al are right, we could be going to other stars before the century finishes.

    It’s a long shot, though. They have a lot to prove before we can bet on that horse.

Comments are closed.