Satellite imagery from Wednesday reviewed yesterday, May twenty-ninth, shows China constructing new launch complexes within ten kilometers of at least three known nuclear silo fields in Gansu and Xinjiang provinces. That's the headline. Daniel's asking us how China got here — how it built one of the world's largest nuclear arsenals in under two decades.
The timing is not subtle. New START expired February fifth of this year. There is no replacement. China is not bound by any bilateral limits, and what the satellite imagery is showing us is that they're not slowing down — they're accelerating. The Planet Labs images from this week caught launch control centers, garages for transporter-erector-launchers, what looks like new access roads. This is not a static silo field. This is an active, expanding, dual-track nuclear posture being built out in real time.
"Dual-track" meaning some of these missiles live in silos, some ride around on trucks.
Some are silo-based, some are road-mobile on TELs. And building launch complexes near the silos suggests they're not just parking missiles in holes. They're building the infrastructure to sustain operations, reload, reconstitute after a strike. That's a shift from "we have a few missiles to deter an attack" to "we intend to fight through one.
The imagery is the visible tip of a much larger industrial and strategic story. Let's unpack how China got here.
The numbers tell the story better than anything. In twenty ten, the Stockholm International Peace Research Institute estimated China had about two hundred fifty nuclear warheads total. By early twenty twenty-six, SIPRI's yearbook puts the estimate at roughly twelve hundred. That's nearly a fivefold increase in sixteen years. And the projections have them reaching fifteen hundred by twenty thirty. That puts China in the same weight class as the United States and Russia — not yet in total numbers, the US and Russia each have over five thousand, but in operational, deliverable warheads, China is rapidly closing the gap.
The pace of silo construction specifically — what are we actually seeing on the ground?
Over three hundred fifty new silos since twenty twenty-one. Three main fields. Huitongxiang in Xinjiang: one hundred twenty silos built in eighteen months starting in twenty twenty-one. Yumen in Gansu: about one hundred ten. Hami, also in Xinjiang: over one hundred twenty. And now the new imagery shows launch complexes being added near at least three of these fields. This is not a slow, bureaucratic program. This is industrial mobilization.
The core question is how. How does a country go from two hundred fifty warheads to twelve hundred plus in sixteen years, with the physical infrastructure to support many more? What's the machine behind this?
I think there are three pillars. First, industrial mobilization — China built dedicated production lines for solid-fuel ICBMs between twenty fifteen and twenty twenty, and they can now produce forty to fifty missiles per year. Second, dual-use technology leverage — their commercial space sector, companies like LandSpace and Galactic Energy, provides a civilian talent pipeline and a component supply chain that feeds directly into military programs. Third, a strategic doctrine shift — in twenty nineteen, China's Defense White Paper moved explicitly from "minimum deterrence" to "credible deterrence." That doctrinal change justified the budget, the silos, the production lines.
Let's start with the industrial piece. Forty to fifty ICBMs a year — what's the benchmark for that? What does it compare to?
During the peak of the Cold War, the United States was producing something like thirty to forty Minuteman missiles per year in the early nineteen sixties. China's current rate exceeds that. The key is that China's state-owned enterprises — CASIC, the China Aerospace Science and Industry Corporation, and CASC, the China Aerospace Science and Technology Corporation — they built dedicated solid-fuel motor production lines specifically for ICBMs between roughly twenty fifteen and twenty twenty. These are not dual-purpose factories making commercial rocket motors that can also do military work. These are purpose-built military production lines.
Solid-fuel is important because...
Solid-fuel missiles don't need to be fueled before launch. A liquid-fueled ICBM, like China's older DF-5, can take hours to fuel — it's vulnerable during that window. Solid-fuel missiles, like the DF-41, are basically ready to go. You can silo them, you can put them on a truck, and they launch in minutes. The production of solid-fuel motors is the industrial bottleneck for any nuclear modernization program. China solved it.
They built the factories to build the motors, and then the missiles, and then the silos to put the missiles in. What about the silos themselves? Three hundred fifty plus in five years sounds fast.
It's extraordinarily fast. The Huitongxiang field is the case study here. One hundred twenty silos in eighteen months. The way they did it was modular construction. Prefabricated concrete segments manufactured at a dedicated plant about two hundred kilometers away, trucked to the site, assembled on location. Traditional silo construction — you dig a hole, you pour concrete, you let it cure, you install the launch equipment — that can take eighteen months per silo. China reduced it to roughly six months per silo by prefabricating.
It's the nuclear equivalent of IKEA furniture. Flat-pack missile silos.
I hate how accurate that is. And the plant that manufactured those segments — it's not a general construction supplier. Satellite imagery analysts tracked it because it appeared specifically to support the silo program, with rail spurs connecting it directly to the construction sites. The logistics chain is visible from space.
"Visible from space" is doing a lot of work here. We're talking about commercial satellite imagery — Planet Labs, Maxar — that anyone can access. The physical infrastructure is an open secret. The warhead counts are not.
That's the transparency paradox, and we'll get to it. But let me finish the industrial picture. The DF-41 ICBM is the centerpiece of this buildup. First tested in twenty twelve, deployed by twenty seventeen. Range of twelve thousand to fifteen thousand kilometers — that puts the entire continental United States in range. MIRV capability, meaning it can carry up to ten independently targetable warheads on a single missile. Current estimates are sixty to eighty launchers, and that number is growing. Each launcher can carry multiple warheads, so the warhead count multiplies quickly.
The commercial space sector — how does that feed into this?
This is the dual-use piece that often gets missed. Over the last decade, China has developed a robust commercial launch industry. Companies like LandSpace, Galactic Energy, iSpace — they're building and launching orbital rockets. LandSpace successfully launched its Zhuque-2 methane-fueled rocket in twenty twenty-three, first methane rocket to reach orbit. These companies employ thousands of engineers who trained on solid and liquid propulsion, guidance systems, stage separation, re-entry — all skills directly transferable to ballistic missile programs.
The civilian talent pipeline is also a military talent pipeline.
It's not just talent. The component supply chain — guidance systems, avionics, materials science for heat shields and nose cones — a lot of that is dual-use. A company that builds a guidance computer for a commercial satellite launcher is, in many cases, using components and expertise that can be repurposed for a missile guidance system. China's military-civil fusion strategy explicitly encourages this. The commercial companies get state support and in return, their technology and talent are available for defense applications.
"Military-civil fusion" — that's the official policy term?
And it's not hidden. It's been a stated national strategy since roughly twenty fifteen. The idea is to erase the boundary between civilian and military technological development. In the US, there's a fairly clear line between SpaceX launching Starlink satellites and Northrop Grumman building ICBMs. In China, that line is deliberately blurred.
Which gives them a larger industrial base to draw from. It's not just the dedicated military factories — it's the entire commercial space ecosystem.
And this connects to something I've been tracking. The academic exchange that used to exist between US and Chinese researchers in aerospace and materials science — it's largely dried up in the last decade. We're more blind to their internal developments than we were. So when satellite imagery shows new launch complexes appearing near silo fields, it's often the first concrete signal that something has changed.
The industrial capacity is there. But capacity alone doesn't explain why China chose this specific mix of silos and mobile launchers. That's where doctrine comes in.
The twenty nineteen Defense White Paper is the key document. For decades, China's nuclear doctrine was "minimum deterrence" — keep a small number of warheads, enough to inflict unacceptable damage on an attacker, but not enough to fight a nuclear war. The classic formulation was "no first use" and a force just large enough to survive a first strike and retaliate. The twenty nineteen white paper shifted the language to "credible deterrence." That's not a semantic change. It means a force large enough, diverse enough, and survivable enough that an adversary cannot doubt its ability to retaliate even after absorbing a first strike.
"credible" means you need enough missiles, in enough different basing modes, that no adversary can be confident they've taken all of them out.
That's why the new imagery matters so much. It's showing launch complexes — launch control centers, TEL garages — being built near the silo fields. This suggests a mixed basing strategy. Some missiles in silos, which are hardened and can launch quickly but whose locations are known. Some missiles on road-mobile TELs, which are harder to locate and target but take longer to bring to readiness. Having both creates a targeting problem for any adversary.
The silos force an adversary to expend warheads on known locations. The mobile launchers ensure a surviving force regardless.
The spacing tells us something interesting. US Minuteman III silos are spaced eight to ten kilometers apart. China's new silos are spaced three to five kilometers apart. Tighter spacing means they're less concerned about a single warhead destroying multiple silos — which suggests they're calculating that modern warhead accuracy makes wider spacing less relevant, or they're prioritizing the ability to defend the silo field as a complex, or both. The point is, it's a different survivability calculation than the one the US made in the nineteen sixties.
They're not just copying the American playbook. They're adapting it to their own strategic assumptions.
Their own geography. Gansu and Xinjiang are vast, sparsely populated regions. They can spread silos across hundreds of kilometers without the political complications you'd have in, say, the American Midwest. The Huitongxiang field alone covers an area roughly the size of Connecticut.
Which is where you're from, originally.
And I can tell you, if someone proposed building one hundred twenty nuclear missile silos across Connecticut, there might be a town hall meeting or two.
Might be a strongly worded Nextdoor post.
But in Xinjiang, the political calculus is entirely different. The state owns the land, the population density is low, and there's no public consultation process.
Let's talk about the knock-on effect. China building twelve hundred plus warheads and hundreds of silos — that doesn't happen in a vacuum. What does it do to India?
India is the most directly affected. China and India share a contested border, they fought a war in nineteen sixty-two, and they had a lethal border clash as recently as twenty twenty. India's nuclear arsenal has historically been sized against Pakistan, with China as a secondary concern. That's changing. India's Agni-VI program — an ICBM with a range of over ten thousand kilometers — was accelerated in twenty twenty-four and twenty twenty-five specifically in response to China's buildup. The Agni-VI puts Beijing and Shanghai in range from Indian launch sites.
China builds up, India responds, Pakistan responds to India, and you get a cascade.
The cascade is already happening. Pakistan is expanding its plutonium production capacity. India is diversifying its delivery systems, including submarine-launched ballistic missiles. The Indo-Pacific nuclear dynamic is becoming a three-way — really four-way, if you count North Korea — arms competition. And none of it is governed by any treaty framework.
The US angle — missile defense in the Pacific?
China's buildup directly complicates US missile defense planning. The US has Ground-Based Interceptors in Alaska and California, designed primarily against a limited North Korean or Iranian attack. The system was never sized for a Chinese arsenal of this scale. If China can launch dozens of ICBMs with multiple warheads each, the math of missile defense breaks down. You'd need hundreds of interceptors, and each interceptor costs something like seventy million dollars. It's not economically sustainable.
The cost-imposition strategy. China spends on offensive missiles, the US has to spend multiples on defense to keep up — if it even can.
This is where the MIRV capability of the DF-41 really matters. A single DF-41 can carry up to ten warheads, plus decoys. To defend against one missile, you might need four or five interceptors. Against sixty DF-41s with full MIRV loads, you'd need hundreds of interceptors. The US currently has forty-four Ground-Based Interceptors deployed. The math doesn't work.
The doctrine shift, the industrial mobilization, the dual-use pipeline — all of it adds up to a force that changes the strategic equation with both India and the United States. What's the misconception you most want to knock down here?
The biggest one is that China's arsenal is still a "minimum deterrent." I see this in commentary all the time — people who haven't updated their mental model since twenty ten. Twelve hundred warheads and three hundred fifty silos is not minimum anything. It's a major nuclear power building toward parity. The twenty nineteen white paper explicitly abandoned minimum deterrence as a concept. The physical construction confirms it.
The second misconception I'd flag is that silos are obsolete. The argument goes: fixed locations, known coordinates, vulnerable to a first strike — why build them? But China isn't building silos as its only basing mode. It's building them as part of a mixed strategy. Silos force an adversary to allocate warheads to known locations. That's warheads not available for other targets. And a silo-based missile can launch faster than a road-mobile one — it's already upright, already connected to launch control, already targeted. In a crisis, that matters.
The third misconception is that this buildup is purely a response to US actions — missile defense, the Pacific pivot, whatever. US moves are a factor, but the primary driver is China's own strategic ambition. China wants to be treated as a global power, and global powers have large nuclear arsenals. It's a status marker as much as a military requirement. The twenty nineteen white paper is not framed as "we must respond to American aggression." It's framed as "China's national rejuvenation requires a military commensurate with our standing.
"National rejuvenation" — that's the phrase that does a lot of work in Chinese strategic documents.
It's the umbrella justification for everything from aircraft carriers to nuclear silos.
Let's dig into the new launch sites specifically — the May twenty twenty-six imagery. What exactly are we seeing?
The Defense News piece from May twenty-ninth, citing the Planet Labs imagery, describes launch control centers and TEL garages under construction within ten kilometers of the silo fields in Gansu and Xinjiang. A launch control center is the facility that actually fires the missile — it's the nerve center connected to the silos by hardened communication lines. A TEL garage is where you store and maintain the road-mobile launchers. The fact that both are appearing near the silo fields suggests these are not standalone silo complexes — they're integrated nuclear bases with multiple basing modes co-located.
Which means an adversary can't just target the silos. They'd have to target the launch control centers, the TEL garages, the access roads — and even then, the mobile launchers might have dispersed.
The NBC News piece from the same day quoted an analyst from the James Martin Center for Nonproliferation Studies saying this represents "a maturation of China's nuclear posture from a force-in-being to an operational warfighting capability." That's a significant line — "from a force-in-being to an operational warfighting capability." It means they're building the infrastructure to actually command and control a nuclear war, not just to have missiles that exist as a theoretical deterrent.
"Force-in-being" is a naval term originally — a fleet that exists and deters by existing, but isn't necessarily ready to fight tomorrow. What they're describing is a shift to a force that can fight.
That's what the launch complexes signal. You don't build launch control centers near silos unless you intend to use them in a coordinated way. You don't build TEL garages unless you plan to disperse mobile launchers and sustain them in the field. This is operational infrastructure, not just symbolic infrastructure.
Let me ask about the transparency paradox. China publishes no official warhead numbers. They don't participate in arms control treaties that would require declarations. But the physical infrastructure — the silos, the launch complexes, the production plants — is visible to commercial satellites. So we have this strange situation where the warhead count is secret but the delivery system count is an open secret.
It's a deliberate ambiguity. By not publishing numbers, China retains flexibility. They can build at whatever pace they want without being accused of violating a treaty obligation, because they have no treaty obligations. But the visible infrastructure sends a signal. It says "we are a major nuclear power" without having to say it officially. Other countries' intelligence agencies can see the silos. Their defense planners have to account for them. The message is delivered without a press release.
It's the geopolitical equivalent of wearing a bulky jacket and letting people guess what's underneath.
A bulky jacket with three hundred fifty visible bulges.
What should someone who wants to track this themselves actually watch for?
The leading indicators are not warhead counts — those are always estimates. The leading indicators are construction logistics. New access roads. If you see a dedicated concrete plant appear in a remote part of Xinjiang with a rail line connecting it to a known silo field, something is happening. If you see new roads being cut into previously undeveloped areas near existing missile bases, something is happening. The analysts who spotted the Huitongxiang field early were tracking construction patterns, not missile deployments.
There are open-source intelligence accounts that do this work.
The James Martin Center for Nonproliferation Studies publishes regular satellite imagery analysis. There are researchers on X — the platform formerly called Twitter — like at nukestrat and at armingdonkeys who track this stuff meticulously. The Federation of American Scientists puts out estimates. It's a remarkably transparent field for something that involves the most secretive weapons on earth.
"Armingdonkeys" — that's uncomfortably close to home.
I have no affiliation. Although if they're looking for a mascot, I'm available.
Let's talk about the next five years. Where does this trajectory lead?
If current construction rates continue — and there's no reason to think they won't — China will have between fifteen hundred and eighteen hundred warheads by twenty thirty-five. That makes it a peer competitor to the US and Russia in deliverable warhead numbers. The US and Russia will still have larger total stockpiles because of non-deployed warheads in storage, but in terms of warheads on launchers ready to go, China is approaching parity.
Diversity of delivery systems — where does China still lag?
That's the qualifier. The US has the triad: land-based ICBMs, submarine-launched ballistic missiles, and air-delivered bombs and cruise missiles. Russia has the same. China has land-based ICBMs — now very capable ones — and a nascent submarine-launched capability with the JL-3 missile on the Type 094 and Type 096 submarines. But China's bomber force is not in the same league as the US B-2 and B-21 or the Russian Tu-160. The air leg of their triad is the weakest.
The submarine leg?
Improving but still the least mature. The JL-3 missile, tested in twenty nineteen, has a range of roughly ten thousand kilometers, which means a submarine in the South China Sea could reach the US West Coast. But China's submarines are noisier than US or Russian boats, and they haven't demonstrated the kind of continuous at-sea deterrence patrols that the US and UK have maintained for decades. That said, they're investing heavily, and the Type 096 submarine, expected around twenty twenty-eight, is designed to be much quieter.
The triad is under construction, but the land-based leg is doing the heavy lifting right now.
Which is why the silos and mobile launchers matter so much. The land-based force is China's deterrent backbone, and they're building it to a scale that can absorb a first strike and still retaliate massively. That's the "assured retaliation" part of "credible deterrence.
Let's address the question of whether China will eventually join arms control talks. The US and Russia have been trying to bring China into the process for years. China has consistently said no — their stockpile is much smaller, they're not in the same league, arms control is for the big boys. That argument gets harder to sustain at twelve hundred warheads and counting.
It does, and there's an active debate among arms control specialists about at what point China's own strategic logic flips. At some warhead number, China might find it in its interest to cap the US and Russia through a treaty framework, because an unconstrained three-way arms race is expensive and destabilizing for everyone. The counterargument is that China views arms control as a tool the US uses to preserve its advantages, and they won't join until they've achieved unambiguous parity.
The historical parallel would be the Soviet Union, which didn't seriously engage in arms control until it had achieved rough parity with the US in the late nineteen sixties and early nineteen seventies. SALT One was signed in nineteen seventy-two, after the Soviets had spent a decade catching up.
That's probably the model China is following. Build first, negotiate later — from a position of strength. The question is whether "later" means twenty thirty or twenty forty.
One thing I want to pull out — you mentioned that academic survey collaborations inside China have largely dried up. What does that mean for our understanding of their program?
It means we're more reliant on satellite imagery and technical intelligence than we were a decade ago. In the two thousands and early twenty tens, there was a fair amount of academic exchange — Chinese researchers at international conferences, joint projects, that kind of thing. You could get a sense of where the technical frontier was. Under Xi Jinping, that's largely stopped. The research environment has become more closed, more securitized. So when we see new launch complexes in satellite imagery, it's often the first indication — we're not getting early warning from the academic grapevine.
Which makes the open-source intelligence work all the more important. It's not just hobbyists on the internet — it's a genuine verification mechanism.
Commercial satellite imagery is now the primary verification tool for arms control, full stop. The New START treaty had its own verification regime — inspections, data exchanges, notifications. That's gone. In its absence, Planet Labs and Maxar and the analysts who pore over their images are doing the work that treaty inspectors used to do.
The silo fields in Gansu and Xinjiang — these are not hidden facilities. They're not buried under mountains. They're out in the open, visible to anyone with a satellite subscription.
Which is itself a signal. If China wanted to hide this program, they could — they have extensive underground facilities, tunnel networks, the whole underground Great Wall concept. The fact that they're building silos in the open, with launch complexes visible from space, suggests they want the capability to be seen. It's deterrence by demonstration.
"We are building this, and we want you to know we are building this, but we're not going to tell you how many warheads are in it.
The ambiguity is the point. An adversary has to plan for the maximum possible capability, not the minimum.
All of this adds up to a new strategic reality. Here's what I think is worth watching in the coming months.
First, track the construction logistics. If you see new cement plants or rail spurs appearing near known missile fields, that's the leading indicator of another expansion cycle. Second, watch for the Type 096 submarine — its first deployment will signal that China's sea-based deterrent is reaching maturity. Third, pay attention to India's response. The Agni-VI program and India's own submarine developments are the mirror image of China's buildup, and the two are locked together.
The open question — the one that'll shape the next decade — is whether China eventually joins an arms control framework or continues building until it reaches unambiguous parity with the United States. The next five years will determine whether we enter a three-way nuclear arms race or some new stability regime emerges. I don't think anyone knows the answer, including Beijing.
The one thing I'm confident about is that the era of transparent, treaty-verified nuclear arsenals is over. New START is gone. No replacement is in sight. The primary verification tool going forward is going to be commercial satellite imagery and the analysts who interpret it. If you want to understand the nuclear balance, you need to learn to read satellite photos — or follow the people who do.
To recap the three pillars we walked through: industrial mobilization — dedicated production lines building forty to fifty ICBMs per year, modular silo construction that cut build time from eighteen months to six. Dual-use leverage — the commercial space sector feeding talent and components into military programs through military-civil fusion. And the doctrine shift — from minimum deterrence to credible deterrence, justifying the largest nuclear expansion since the Cold War.
The satellite imagery from this week is just the latest data point in a program that's been running for a decade and will run for at least another decade. China is not sprinting. It's running a marathon, and it's well into the middle miles.
Now: Hilbert's daily fun fact.
Hilbert: In nineteen sixteen, German zoologist Ernst Marcus described a species of tardigrade found in a sample of moss from the roof of a church in Bremen that measured one point seven millimeters in length — making it the largest tardigrade ever recorded, nearly triple the size of the average water bear.
One point seven millimeters. Truly, a giant among microbes.
A tardigrade you can almost see without a microscope. I'm not sure how I feel about that.
This has been My Weird Prompts. Our producer is Hilbert Flumingtop. If you enjoyed this episode, leave us a review wherever you listen — it genuinely helps other people find the show. I'm Corn.
I'm Herman Poppleberry. We'll be back next week.
