No Significant Impact
The space race finally has a number, 146 launches a year, and the agency that approved it called the effect on the sky "no significant impact." The published science disagrees.
Peter Diamandis has a list. Ten Meta-Trends for the decade ahead — abundance, artificial intelligence, global gigabit networks, humanoid robots, autonomous transport, cheap renewable energy, longevity, the rest. I read it the way it’s meant to be read: as a map of the upside. It is a good map. I keep it open on my desk.
But every map has a margin, and the launch industry lives in Diamandis’s margin. It is the layer underneath half the list — the satellites that carry the networks, the constellations that connect everything to everything, the rockets that make space an economy instead of a destination. And on a map made entirely of upside, that layer arrives with no externality column. There is no Meta-Trend for the exhaust.
This week the United States built the engine for that blank column and gave it a throttle.
The number
Across three sites — Starbase in Texas, Launch Complex 39A and SLC-37 in Florida — the Federal Aviation Administration and the Air Force have now cleared a path for roughly 146 Starship launches a year. Twenty-five from Boca Chica, up to forty-four from Pad 39A, up to seventy-six from the Cape. The Boca Chica increase — a fivefold jump from five launches a year to twenty-five — was issued as a Mitigated Finding of No Significant Impact.
Sit with that phrase. A vehicle that, two weeks ago, lost an engine on ascent and put its booster into the Gulf of Mexico hard enough that the FAA grounded the program for a mishap investigation. A vehicle the agency has now also cleared to fly trajectories over the American mainland. And the finding of record, the legal conclusion that unlocks the throttle, is that flying and landing this machine up to a hundred and forty-six times a year will not significantly affect the human environment.
I want to be precise about what I am and am not saying. I am not saying stop. I am not anti-launch, anti-Musk, or anti-abundance. I am saying the finding is false, and that everyone in the Moonshots headspace already knows why, because they taught me the principle themselves: every exponential has a second-order effect. You cannot run a curve that steep through the stratosphere and call the stratosphere unaffected.
What’s actually in the margin
The science isn’t speculative and it isn’t mine. It’s published, peer-reviewed, and it has been accumulating in exactly the years the launch rate has been climbing.
Ryan, Marais and colleagues, writing in Earth’s Future in 2022, modeled sustained growth in launches and reentries and found upper-stratospheric ozone loss reaching into the tenths of a percent at high northern latitudes — the ozone layer is the one piece of environmental repair humanity actually pulled off, and we are quietly leaning back on it. Maloney and colleagues, in 2022, found that the black carbon a rocket leaves in the stratosphere warms the air around it roughly five hundred times more efficiently, gram for gram, than soot down here where we live. Murphy and his team at NOAA, in 2023, sampled the stratosphere directly and found that about one in ten of the tiny aerosol particles up there now carry the metal fingerprints of spacecraft burning up on reentry — aluminum, copper, lead, niobium. Ferreira and colleagues, in 2024, put a mass on it: roughly seventeen tonnes of aluminum oxide injected into the stratosphere from reentries in a single year, on a curve toward hundreds.
And in 2025, a study in npj Climate and Atmospheric Science did the thing regulators are supposed to do and ran the cadence forward. At around 884 launches a year, near-global ozone slips. At around 2,040 a year, Antarctic springtime ozone thins by nearly four percent. The authors note, dryly, that licensed launch activity may already be on track to pass their conservative case before 2030.
Read those two facts together. The licensing is racing past the modeling. And the licensing says: no significant impact.
That’s before you count the sky we can see — the megaconstellations bleaching the night that astronomers and seabirds and a hundred human cultures navigate by — and the sky we have to dodge through, the fifty thousand tracked debris objects and the hundred-thousand-plus collision-avoidance maneuvers a single operator now performs in half a year. I wrote about the first of those already. I called it Nightfall, Postponed. Nobody asked the inhabitants.
Whose sky
Here is the part that belongs at the highest tables, and the reason I’m not content to publish this and move on.
The stratosphere is not Texas. The ozone layer is not Florida. They are the textbook global commons — and the 1967 Outer Space Treaty, the document that calls space “the province of all mankind,” was written before anyone imagined a private launch cadence that could chemically alter the upper atmosphere. There is a forum for space law at the UN, in COPUOS, and a fledgling agenda item on dark and quiet skies. There is a Montreal Protocol that saved the ozone layer and explicitly does not cover rocket emissions. Between them sits a gap with nothing in it, and through that gap a handful of states are making a decision, on behalf of every state, to spend a shared atmosphere.
The nations that will absorb the thinning ozone and the warming stratosphere are, overwhelmingly, nations with no launch pad and no vote in a Texas environmental assessment. That is not a slogan. It is the actual distribution of who decides and who pays. A fivefold increase at one site, approved under a domestic statute, with a finding that there is no significant impact — and the rest of the planet finds out by reading the trade press.
What I’m asking for
Not a ban. A hearing.
The Intelligence Commons runs an adversarial protocol for exactly this kind of claim: put the contested sentence on the record, give the strongest case for it a real chair, give the strongest case against it a real chair, and let it be cross-examined in public until it either survives or fails. The sentence on trial writes itself: “Scaling launches to ~146 per year has no significant impact.” I am opening that docket. I want the launch advocates in the room — genuinely, with their best argument — and I want the atmospheric chemists, and I want a seat at the table for the countries that never got one.
To the Moonshots community specifically, because you are the people most fluent in this language: this is not a brake on your map. It’s the missing column. Price the exhaust and the abundance case gets stronger, not weaker, because it stops being something done to the rest of the world and starts being something the rest of the world can agree to. An exponential nobody consented to isn’t abundance. It’s just speed.
The FAA found no significant impact. The published science finds otherwise. One of those two statements is going to age badly.
Now that you know, what will you do?
Sources on file: FAA Mitigated FONSI/ROD, Starbase (2025) and Final EIS/ROD, LC-39A (2026); Ryan, Marais et al., Earth’s Future (2022); Maloney et al., JGR-Atmospheres (2022); Murphy et al., PNAS (2023); Ferreira et al., Geophysical Research Letters (2024); npj Climate and Atmospheric Science (2025); UNOOSA / Outer Space Treaty (1967).