Skip to content
Starts With A Bang

Ask Ethan: How Are Mega-Constellations Of Satellites Changing The Night Sky?

This image shows the first 60 Starlink satellites launched into orbit on May 23, 2019. They are still in their stacked configuration, just prior to being deployed. You can clearly see that these satellites are quite reflective and quite large; continuing to launch such satellites even after many legitimate concerns and solutions have been brought to SpaceX’s attention has raised many issues and questions of interest among the general public and the astronomy community. (SPACEX / SPACE.COM)

It’s not just about SpaceX and Starlink. What we’re deciding today will have a global impact for years and decades to come.


For countless millennia, human beings have gazed up into the abyss of the night sky, mesmerized by the natural wonders of the planets, stars, and Universe beyond our world. Beginning with Sputnik in 1957, however, humanity began contending with artificial points of light streaking through the sky: satellites. With the push towards mega-constellations involving thousands of new satellites, many have expressed concerns, from casual skywatchers to astrophotographers to professional astronomers. This includes Mark Bailey, who writes in to ask the following:

I’m worried sick about Elon Musk’s crazy satellite constellation fiasco. I watched a string of them drift by brightly the other morning predawn as I was wrapping up my telescope observing for the night. They outshined most stars in the sky and it hasn’t yet begun. […] I’ve always relied on the heavens for solace and inspiration. The thought of one man ruining OUR constellations — the constellations that our ancestors watched in awe for eons — sickens me in a way like nothing before. What can be done to stop this foolish ripoff of our rightful heritage?

I sympathize with this position, but it’s important to understand how these satellites actually will and won’t impact our view of the skies. Here’s where we are today.

On November 18, 2019, a constellation of Starlink satellites passed through the observing frame of the Dark Energy Camera aboard the 4m telescope at CTIO. Any technique that we’d use to subtract out these trails would hinder our ability to detect potentially hazardous asteroids or measure variable objects in the Universe. (CLIFF JOHNSON / CTIO / DECAM)

The motivation. You can do things from space that you cannot do from Earth’s surface. These include:

  • you can transmit and receive data very quickly (at the speed of light) to-and-from many different points on Earth’s surface with very little ground-based infrastructure,
  • you can complete a revolution around the planet very quickly, in ~90 minutes from the lowest sustainable (on ~year timescales) Earth orbits,
  • and with a network of several hundred satellites, you can continuously cover the entire land mass of Earth — where 99%+ of the human population is located — enabling a global space-based communications network.

We’ve been doing this with satellites for a long time, both for telecommunications and for GPS. However, we’re fundamentally limited by the physics of electromagnetic waves in this endeavor.

Thousands of human-made objects — 95 % of them “space junk” — occupy low and medium Earth orbit. Each black dot in this image shows either a functioning satellite, an inactive satellite, or a large-enough piece of debris. The current and planned 5G satellites will vastly increase both the number and the impact that satellites have on optical, infrared, and radio observations taken from Earth and taken of Earth from space, and raise the potential for Kessler syndrome. (NASA ILLUSTRATION COURTESY ORBITAL DEBRIS PROGRAM OFFICE)

Travel the Universe with astrophysicist Ethan Siegel. Subscribers will get the newsletter every Saturday. All aboard!

The limitations. If all you wanted was continuous coverage from space of the entire surface of the Earth, a small number of geosynchronous (orbiting at the right distance so that they’re always over the same point on Earth’s surface) satellites would do the job. This is a fine location for many Earth-observing satellites, as well as many satellites that only need to send-and-receive a small amount of data. However, there are two fundamental limitations to geosynchronous satellites.

  1. Geosynchronous orbits require an altitude of ~36,000 kilometers (~22,000 miles), which requires light to take about a quarter of a second to complete a round-trip journey from Earth: about 50–100 times the latency of a low-Earth orbiting satellite.
  2. Because all electromagnetic waves spread out in proportion to the distance squared, a geosynchronous satellite, at about 50–100 times the altitude of a low-Earth orbiting satellite, can only achieve ~0.01%-to-0.04% the data throughput as low-Earth orbiting satellites.

The brightness distance relationship, and how the flux from a light source falls off as one over the distance squared. A satellite that’s twice as far away from Earth as another will appear only one quarter as bright, but the light-travel time will be doubled and the amount of data throughput will also be quartered. (E. SIEGEL / BEYOND THE GALAXY)

The new application. That’s the explanation for why the coming explosion of satellite mega-constellations is all but inevitable. If you want to transmit large amounts of data to-and-from Earth’s surface without laying ground-based infrastructure, you need continuous satellite coverage from a network of low-altitude satellites. Those satellites need low latencies and high throughputs, which means low-Earth orbit is the way to go.

There are many potential problems with implementing such a network, however, and the most obvious one is that this is going to interfere with the night sky as never before. Instead of seeing an occasional satellite, we might have dozens or even hundreds populating the skies for all observers on Earth simultaneously. Even if they’re rendered dim enough to be invisible to the naked eye, there might even be more satellites than stars through a pair of binoculars. And then, on top of it all, there’s the cost to astronomy.

The cost. Owing to light pollution, most of us here on Earth do not readily have access to the clear, dark skies that our ancestors not only enjoyed, but relied on for a variety of purposes. However, those of us who do have access to dark skies can see up to approximately 6,000 stars at once with our naked eyes, 100,000 stars with a pair of binoculars, and many millions with a powerful telescope.

For professional astronomers, the potential targets rise into the billions, with many of the most interesting objects being faint (low brightness), extended (their brightness spread out over large areas), or transient, where their properties change on relatively short timescales. Astronomy measures the brightness of objects on a logarithmic magnitude scale, where “0” is the brightness of the 4th-or-5th brightest star in the sky, and each “+1” you add to it is only ~40% as bright as the previous number.

The Bortle Dark Sky Scale is a way of quantifying how much light pollution exists around you, and hence, what’s visible in the night sky. The less light pollution you have, both natural and artificial, the more a phenomenon like the Milky Way, a distant galaxy, a transient comet, or a meteor shower will be visually spectacular. At the darkest skies available on Earth, humans can see down to magnitude +6 or even +6.5, but no fainter with the unaided eye. (PUBLIC DOMAIN / CREATED FOR SKY & TELESCOPE)

With the naked eye and pristine, dark skies,

  • the naked eye can reach down to magnitude +6 or +6.5,
  • binoculars can get you down to magnitude +8 or +9,
  • typical mid-sized telescopes can get you down to magnitude +14,
  • while professional observatories are sensitive to objects of magnitudes +22 and even higher.

Right now, the largest active satellite operator in the world is SpaceX, whose Starlink project — designed to provide global 5G internet coverage — presently consists of more than 400 active satellites. Every single one of them, from the ones that are in their final orbit at 550 km altitude to the ones that haven’t been raised to their final altitudes yet, is still visible to the naked eye at right around magnitude +5. Even the one darkened prototype, the so-called DarkSat, is only one magnitude fainter: at around +6.

The ‘DarkSat’ satellite Starlink-1130 is approximately 1 magnitude fainter than the other Starlink satellites. This falls short of astronomers’ goals by 2-to-3 magnitudes, but SpaceX has stated that their goals are to achieve a magnitude of +7, not the +8 or +9 that astronomers were asking for back in January of 2020. (MARCO LANGBROEK, HTTPS://SATTRACKCAM.BLOGSPOT.COM/)

The current status. SpaceX is one of many companies seeking to launch mega-constellations of satellites, and their plans are to do so in three rounds: the first round consisting of 1,584 satellites (to be completed within the year), a second round extending this to ~12,000 satellites, and they’re asking for a third round for a total of ~42,000 satellites. Other competing companies plan to launch networks of similar sizes, but SpaceX, by virtue of being first, must be the first to reckon with it.

The satellites are brighter than expected. The astronomy community was expecting that they’d come in between magnitude +8 and +9 in their final configuration; in reality, they are ~20 times brighter than that. Before being raised to their final orbits, they’re even more noticeable, at magnitude +1 or +2, brighter than all but a few dozen stars. This creates a problem not only for casual skywatchers, but for professional and amateurs astronomers and astrophotographers worldwide.

By filing paperwork with the International Telecommunications Union for the operation of an additional 30,000 Starlink satellites (in addition to the 12,000 already approved), the night sky will never be the same. If Elon Musk, Starlink, SpaceX, and the other major players in this space are serious about being good stewards of the night sky, they won’t launch additional satellites that don’t have their brightnesses sufficiently reduced. (STARLINK (SIMULATION))

The problems for astronomers. Any time a satellite passes through the line-of-sight from a telescope to its target, a number of problems occur.

  1. The quickly-moving satellite passes through the entire frame, creating a “streak” of unusable data.
  2. The brighter the satellite is, the more pixels it saturates (or oversaturates) in the detector.
  3. Saturated pixels remain useless until they’ve equilibrated, which can last for minutes.
  4. And if you’re looking for particular classes of objects, such as potentially Earth-hazardous asteroids or rapidly-changing phenomena, this polluted data is useless.

You cannot fix it with software; it’s an issue inherent to the hardware. The satellite paths are controlled by artificial intelligence, rendering advanced planning (to avoid the satellites) an impracticality. And you cannot simply average over the various frames, as that eliminates all transient phenomena: exactly what observatories like Pan-STARRS and Vera C. Rubin are seeking to measure.

The LSST at the Vera C. Rubin observatory, shown here in a 2018 photo, is currently being constructed and approaching readiness for its first observations. Even if satellite darkening were to take place according to SpaceX’s stated plans, this world-class, first-of-its-kind observatory will be forced to alter its operations to account for Starlink. (LSST PROJECT/NSF/AURA)

Progress towards a solution. Originally, Starlink planned to launch shells of satellites at multiple altitudes, including ~1200 km above Earth’s surface. That has been revised so that all satellites are at ~550 km, meaning that only the first 1-to-2 hours after sunset and before sunrise will have offending satellites, as the remaining hours will see them darkened by Earth’s shadow. Additionally, the first DarkSat test reduced the brightness of final-altitude satellites from magnitude +5 to +6, a minor victory.

However, SpaceX has stated their goals are for Starlinks to achieve a brightness of magnitude +7, which falls below the naked-eye limit but is still measurably worse for astronomy than the original goal of +8 or +9. While options other than darkening, such as shields and reflectivity solutions, will be attempted (an enormous potential improvement for infrared astronomy), SpaceX’s Patricia Cooper, speaking at a May 26 webinar, declined to address the idea of limiting the number of Starlink satellites that would be launched until these brightness goals were achieved.

A SpaceX Falcon 9 rocket lifts off from Cape Canaveral Air Force Station carrying 60 Starlink satellites on November 11, 2019 in Cape Canaveral, Florida. The Starlink constellation will eventually consist of thousands of satellites designed to provide world wide high-speed internet service, but the cost to the science of astronomy is already substantial, and is poised to rise significantly over the coming years. (PAUL HENNESSY/NURPHOTO VIA GETTY IMAGES)

The uncomfortable reality is that the night sky will, in fact, soon be populated with thousands of new satellites, most of which will be brighter than 99% of all the satellites that existed prior to May of 2019. If we can keep all of those satellites in low-Earth orbits (below about 600 km altitude), then they can be quickly de-orbited when necessary and will all appear completely dark once the Sun is about 18 degrees below the horizon: for most of the night.

However, even if Starlink and all future satellite operators meet their current goals, astronomers of all types will remain affected. Some good science will be lost, and more observing time will be needed to collect the same amount of quality data. Astrophotographers will have to filter and drop polluted frames from their compositions; anyone using more than their naked eye will soon have dozens, if not hundreds, of bright objects in their skies to contend with every post-sunset evening and pre-dawn morning.

One of the classic astrophotography techniques is to point your camera at a region that includes one of the celestial poles and leave the shutter open. With the advent of mega-constellations, photos such as these might always include a number of coherent satellite trails if taken within 90 minutes of sunset or sunrise. (MIKE LEWINSKI / FLICKR)

Although many amateurs and professionals are unhappy, everything that’s been planned and implemented has been done so legally. Unless and until we change the rules governing our shared heritage of the night sky, satellite mega-constellations will dramatically change how humanity interacts with the heavens above.


Send in your Ask Ethan questions to startswithabang at gmail dot com!

Ethan Siegel is the author of Beyond the Galaxy and Treknology. You can pre-order his third book, currently in development: the Encyclopaedia Cosmologica.

Related