The vision is compelling: a million people living on Mars by 2050. Humanity, a multi-planet species! The eternal growth of our population envisioned by the longtermists would be guaranteed. Once the ball of consciousness starts rolling across the universe, it might never stop. And we can be the generation to make it happen.

Elon Musk says we also need a backup planet, in case we carelessly break this one. He’s clearly a longtermist—he believes that we have a duty to future people equivalent to our duty to those alive now.

In sharp contrast to Musk’s purely theoretical longermist justifications for settling Mars, we have empirical data that predicts an imminent Earthly catastrophe due to global warming. We’re already in the catastrophe, so that argument is settled. Last year was the hottest in recorded history, and the previous one had been too—and we know exactly why. The trendlines are clear, and unlike the stock market, they’re not subject to the whims of human folly. They are inexorable. Global warming is locked in.

The cracks in the proposed mitigation policies are starting to show. Nearly every country has blown right through its Paris Climate Agreement goals. Renewables are going to make fossil fuels obsolete but do nothing to remove the CO2 that’s already in the air. And it’s infinitely less energy-intensive to pour CO2 into the atmosphere in a concentrated stream than it is to pull it out after it’s fully diluted, as Carbon Air Capture proponents propose to do. CAC will take more energy than humanity has ever used, and how will we prevent that from heating the planet simply from the thermal inefficiencies of generating it? I was the first to point out that the only technology we have that scales to the required levels is nuclear bombs, if we used them to do enhanced weathering on a planetary scale. I also pointed out that this would cause problems of its own.

We used to have choices, but we’re now out of time. Fixing the problem through emissions reductions has failed. Renewables have not arrived fast enough. Carbon capture and removal is, as I just said, a non-starter. There is still a solution, but as far as I know, only one, to us preventing a mass extinction caused by anthropogenic global warming—and that is to temporarily reduce the amount of energy Earth receives from the sun. Solar Radiation Modification (SRM) is now an absolute necessity.

One way to do SRM is by pumping vast quantities of sulfur into the upper atmosphere, where it will reflect sunlight and lower global temperatures. We have to keep doing it, year after year. It has to be funded and done uninterrupted for generations. We have to commit to a new kind of global pollution to offset the one we already did. And to support it, entire new industries will have to be created, which will, for all those generations, need to mine and transport huge quantities of material—all of which uses energy, needs roads and other infrastructure, and causes its own environmental problems.

There is an alternative, but everybody dismisses it. It reads like science fiction—a supreme irony, because compared with settling Mars, it’s easy.

Doing What’s Never Been Done

To really address the crisis that atmospheric CO2 has put us in, we need to do new things—take actions that we have never taken before. This truly is a ‘moonshot’ effort, but that’s not how the international community views it. They are using the logic of ‘granted what we have now, what can we do?’ to look for solutions, when a realistic appraisal of our needs should tell anyone with common sense that we cannot fix this with any systems that we have now.

But we do have everything we need to build the systems that will fix the problem of anthropogenic global warming. For instance, in the early 2020s industries have a lot of experience manufacturing thin-film electronics in large batches (think flat-screen TVs and solar panels). Solar sails had been tested and are well understood; ditto for very large orbital satellite constellations. If SpaceX’s Heavy Lift launcher can reliably reduce orbital launch costs by a factor of 1000 from the Space Shuttle’s, all the pieces will be in place to build a new kind of satellite constellation: the orbital sunshade.

The idea of literally shading the planet has been widely mocked, but this is a case of something seeming to be impossible only because it hasn’t been done before. Each sunshield will be a thin-film solar sail manufactured using technologies long ago perfected for the mass production of electronic consumer goods. Launched by the millions into the stable L1 gravitational point between the Earth and the sun, this mega-constellation can dilute exact frequencies of infrared radiation and cool the Earth directly. Deployed over ten years, the sunshade will directly counterbalance the greenhouse effect caused by all the extra CO2 in the atmosphere. It will not eliminate that CO2; it’s a symptomatic treatment, and ending fossil fuel emissions is still necessary. Most importantly, while the sunshade solves the problem of direct insolation, ocean thermal inertia and acidification are still major problems. What the sunshade does is provide breathing room for artificial and natural systems to remove all that extra CO2. Carbon air capture continues to get cheaper and more effective, but without the sunshade it won’t make a difference in time.

Maybe. But maybe the best way to judge that would be to compare it with a project that is (apparently) being seriously undertaken right now by the world’s richest man. How does the cost of launching a sunshade—and thereby halting global warming, mass migrations, water wars, and a global extinction event—compare with the cost of colonizing Mars?

Which Costs More, Saving the Earth, or Colonizing Mars?

Elon Musk wants to build a city of million people on Mars by 2050. His proposed spacecraft for transporting these new colonists is called Starship; at present, SpaceX estimates that they might be able to crowd 100 people into one of these, for a cattle-car-to-the-stars odyssey to the beige planet.

Starship isn’t able to make this journey as soon as it gets to orbit. It needs to be refueled first. Since the mass a Starship can carry to orbit is minuscule compared with the amount of fuel it needs to get there, each tanker that rendezvous with the Mars-bound ship can only carry a fraction of the needed fuel. It’s estimated that it will take up to eight refueling flights to top up the Starship before it can leave Earth orbit. Counting the ship itself, that’s nine launches for one colony ship.

In what follows I’ll only count the number of flights needed to carry the colonists, not the cargo hops that will have to bring heavy machinery, robots, drills and refining equipment to the new colony. So consider the numbers I’m giving here for the Martian escapade to be maybe half what they’d really be.

If each Starship can carry 100 people, then we need to launch 10,000 of them, and 80,000 refueling launches must be done to get them to Mars. That’s a total of 90,000 launches between now and 2050 (again, not counting cargo-only flights).

Mars Colonization Effort (Assuming SpaceX Starship)

• Goal: 1 million people to Mars.

• Each Starship carries 100 passengers.

• Each crewed Starship requires ~8 refueling launches.

• Total Starships needed: 1,000,000 / 100 = 10,000

• Total Refueling launches: 10,000 × 8 = 80,000

• Total Launches: 90,000 (10,000 crewed + 80,000 refueling)

Now consider the launch capacity required to put an orbital sunshade in place that will reduce global temperatures by 1.5C for several decades. The most innovative thinkers in this space right now seem to be Olivia Borgue and Andreas M. Hein. In a 2022 study, they propose using diffractive solar sails to reduce the radiation pressure on the sunshade. This design lets us move the shade closer to the Earth, which means it doesn’t have to be as big as systems deployed at L1. Those actually have to be slightly on the other side of the L1 point to counterbalance solar pressure, which means they have to be much bigger.

There are other designs, which you can review at Wikipedia. Borgue and Hein’s proposals are the state of the art, and their estimated number of launches over a ten year period to deploy a diffractive sunshade range from 400 to 3600 (for a total of between 4000 and 36,000 launches) depending on the design. Other designs such as the “space bubbles” proposal have similar ranges; the estimated mass for a sunshade ranges between 100,000 and 620,000 tonnes, and the number of launches from 3,990 to 8,990. We’ll use those numbers rather than Borgue and Hein’s to be conservative.

Sunshade Deployment (Conventional & Diffractive Designs)

Case 1: Traditional Occulter Sunshade

• Estimated mass: ~100,000 to 620,000 metric tons depending on design.

• Launches needed: Estimates range between 3,990 and 8,990 launches (based on Starship capacity at ~100 tons to L2).

Case 2: Diffractive Sunshade (zero-radiation-pressure scenario)

• Advantages: Can be placed closer to Earth, reducing the required area and mass.

• Expected mass reduction: Unclear without precise numbers, but preliminary studies suggest ~10%-30% smaller compared to traditional designs.

• Estimated launch count could therefore fall in the range of ~3,000 to 6,000 launches conservatively, possibly lower as technology matures.

I made a rough chart comparing the requirements to save our world from global warming compared with what would be needed to put a million people on Mars. Here it is:

The top chart compares the number of launches needed to colonize Mars with the number needed to save the Earth. The diffractive sunshade option is practically a rounding error on the Mars colonization number.

The bottom chart compares required time depending on the number of launches per year (100 versus 1000). You can see that once the production line is rolling, it only takes a few years to deploy a diffractive sunshade, which in turn means that global temperatures start returning to preindustrial levels as quickly or slowly as we want. (The sunshade is a tunable system and theoretically can be switched on or off, meaning we can react quickly to any adverse climatic effects it produces. It’s a giant terraforming machine, after all, it’s going to powerfully impact life on Earth; the highest likelihood is that its unintended consequences will be less than letting runaway global warming continue.)

The numbers are so wildly disproportionate that if we committed to colonizing Mars, we could easily fix global warming along the way with a few extra launches per year. Once the launch capability is there, it can serve both ends simultaneously.

Of course, there’s another interpretation we can make—namely that Elon Musk’s Martian colonization plan is bullshit. The number of launches he’s talking about is crazy and how is the whole thing to be funded? On the other hand, by comparison the sunshade requires a large but not inconceivable number of launches, and even if it costs a trillion dollars it will save multiple times that much, while also restoring priceless ecosystems on Earth.

A Question of Priorities

I’ve seen a silly headline that worries that the prospective new head of NASA, Jared Isaacman, might piss off Elon Musk by having suggested that we should go back to the moon before attempting Mars. Looking at the above chart, it’s obvious how ridiculous the idea of such a conflict is. If Musk’s system can be rolled out at the scale needed to settle Mars, Earth can be saved with a sunshade and we can settle the moon as well, with trivial extra effort. We can also orbit solar power satellites if that’s your thing (I’m not a fan of those).

All of this is dependent on the superheavy launcher working reliably. It does fly, and it lands; the current issues with the Starship second stage are actually largely irrelevant to the question of whether it can be used to orbit a sunshade; Starship is not needed for that. We can use disposable second stages that are just packed with self-deploying solar sails. We launch those to orbit and the sails find their own way to L1. So the issues come down to the launch cost, turnaround time, and how many flights you can get out of each superheavy launcher. If each one can fly as many times as the Falcon 9, then we have a winner, and everything described here is possible.

Of course, those 90,000 flights have only landed humans on Mars, not the systems needed for them to actually build the colony. I picture the fresh-faced would-be Martians stepping out of their Starship onto the landing field and each being handed a shovel.

The bottom line here is: if you believe that putting a million people on Mars is possible, then you have to believe that building a solar sunshade that can restore Earth’s climate to preindustrial levels is even more possible.

And if that’s the case, the question becomes: which is more important?

—K