Reading the articles in order is highly recommended, as the first two articles provide important context for the ideas presented here. However, be warned that some inconsistencies may exist between them, as my thinking has evolved over time.

The second one is especially important since it explains how large military satellite constellations could be built to provide jam-resistant navigation and to accurately track warships, vehicles or even stealth aircraft by leveraging commercial satellite technology already in mass-production.

Large satellite constellations, especially when paired with the global strike potential of reusable rockets that I discussed in the first article, could have a similar effect on both naval and air warfare that drones have had on ground warfare in Ukraine.

If enemy satellite constellations aren’t quickly disabled after the breakout of a conflict, surface warships, airbases, military installations and crucial civilian targets could be accurately targeted by salvos of thousands of satellite-guided ballistic (conventional) warheads.

Therefore ensuring that enemy satellite constellations are quickly rendered inoperative in a conflict is critical. I’ll call this concept orbital denial. (I prefer to use orbital over space here since ballistic missiles travel through space as well).

However quickly disabling constellations of thousands of satellites designed to be as resilient as possible is no easy feat. In this article I’ll explore which Anti-Satellite (ASAT) weapons may prove most effective while discussing the ways satellites might be designed and maneuvered to maximize survivability.

This will offer a glimpse into what future orbital warfare might look like.

Current Surface to Space Interceptors

The most common anti-satellite weapons in active service are commonly called direct-ascent ASAT interceptors: these missiles are designed to be launched from either a warship or a ground launcher to then quickly ascend through the atmosphere to strike a satellite within its operational range and altitude. This term however can be a bit misleading since there’s an overlap with Anti-Ballistic Missiles (ABM) designed to operate outside the atmosphere: for example the SM3, one of such ABM interceptors, destroyed a low-altitude satellite in a 2008 test. I’ll use instead an extension of the nomenclature used for surface-to-air and surface-to surface-missiles and call these interceptors surface-to-space missiles.

Surface-to-space interceptors currently in service are largely inadequate for future orbital denial. A missile that costs 10 million $ and is available only in limited numbers isn’t a weapon well-suited to destroy thousands of cheap satellites.

The debris created during interceptions is another huge issue: these missiles impact the satellites at multiple kilometers per second with the entirety of the last rocket stage, breaking it up into thousands of pieces of various sizes that spread out to occupy many nearby orbits.

Kessler syndrome and AI satellites

Debris is extremely dangerous because it can collide with other satellites at very high speed, causing significant damage and generating even more debris. This process could trigger a chain reaction that would result in the destrucion of thousands of satellites, rendering many orbits unusable for years. This phenomenon is commonly known as Kessler syndrome.

This has been a worry for years but the planned huge trillion dollar AI satellite constellations would massively increase the likelihood and the economic damage of such an event.

I have long argued that space is an inherently military domain, civilian infrastructure like data centers that could be built on Earth instead should never be placed in orbit, the risks just aren’t worth the slightly lower power costs.

Valuable civilian constellations in orbit might force a military to fight with with its hands tied during a war, limiting attacks on enemy satellites out of fear of runaway debris and retaliation, while the enemy would be free to use more aggressive tactics.

AI satellites also open up blackmail opportunities for rogue nations: these satellites need to occupy a very specific orbital plane to have access to continuous solar power and that means that a single North Korean or Iranian rocket could disperse hundreds of millions of small untrackable particles into trajectories intersecting the orbits of these AI satellites, potentially causing hundreds of billions of dollars in damage.

These kind of tactics could be used to quickly disable military satellites constellations as well, especially if the large lift capacity of military reusable rockets is exploited.

Ultra Low Earth Orbit

However there are ways to defend satellite constellations from this threat, the most promising relies on designing satellites to be able to operate at extremely low altitudes, lower than 200 km (120 miles).

This region is called Ultra Low Earth Orbit (ULEO) and is has rarely been used due to the high atmospheric drag at low altitudes, which forces satellites to expend significantly more fuel and power to maintain a stable orbit.

When space launch was very expensive using such orbits was impractical but with reusable rockets it’s now possible to design heavier satellites with powerful propulsion system and large fuel tanks that could potentially be refueled in flight. Minimizing the area perpendicular to the air flow can also help reduce drag and fuel consumption, for example building satellites with larger solar panels that are always kept parallel to the surface of the Earth.

Lower fuel consumption could also be achieved by having only a fraction of the satellites in lower orbits, moving all satellites to the safety of low altitudes when a conflict is believed to be imminenent.

But how does low altitude protect satellites? The drag affects debris far more than satellites, since they have higher surface-to-weight ratios and no propulsion. This means that small debris can spend little time in ULEO before losing enough speed and altitude that they no longer threaten the constellation.

Another advantage is that the atmosphere quickly gets denser at lower altitudes. This means that when they are low enough to threaten satellites their velocity needs to be pretty much horizontal to the surface of the Earth or they would quickly deorbit. (A vertical velocity component would mean that their trajectory includes points of even lower altitude and much higher drag).

Since the area perpendicular to the air flow is small to reduce drag that means that debris hits would be relatively rare and concentrated on the sides of the satellites. Therefore little mass would be needed to effectively armor the satellites against them. Whipple shields made up of a number of thin spaced armor layers would be especially effective here: if a small hypervelocity projectile impacts the outer layer it disintegrates, harmlessly dumping its energy on the deeper layers.

Armoring the satellites this way would ensure that larger fragments would need to be used to disable satellites, but these would be much heavier and could only be released in far smaller numbers, massively lowering hit probabilities. Therefore the combination of ULEO and saetllite armor would make the use of tactics involving the release of large amounts of debris into the path of military satellites much less effective.

Improved sensor performance and electronic warfare

Low altitude does not just improve survivability but would also have a positive effect on the operation of satellite sensors and antennas: for example, radar performance is inversely proportional to the fourth power of distance and lowering altitude from 300 km to 200 km (190 to 125 miles) would improve it by a factor of 5.

Occupying significantly lower orbits than enemy satellites also opens up up the opportunity to blind Earth observation sensors on enemy satellites while identifying which satellites are actively transmitting radio signals.

This is such an overwhelming advantage that I think most military satellites will be placed near or in ULEO.

Cheap anti-satellite weapons

Armored satellites in ULEO remain vulnerable to more accurate fragmentaion weapons dispersing small hypervelocity projectiles in the path of a targeted satellite.

One reason is cost: a weapon dispersing fragments over an area of a few tens of square meters in the path of a targeted satellite needs to be far less accurate and complex than legacy surface-to-space missiles. Smaller impacts also reduce the amount of debris produced, helping minimize risk to other satellites while still causing crippling damage to solar panels, antennas and the propulsion system.

Armor is less effective against this kind of weapons since fragments can be bigger and their trajectory more vertical, striking areas that can be difficult to defend without making the satellite too heavy or compromising the performance of antennas and solar panels.

One reason ASAT weapons have rearely been built with fragmentation warheads is that it can be hard to verify that a satellite has been disabled. However next-gen satellite constellations could monitor the radio emissions and the spin of enemy satellites: each hypervelocity impact transfers significant momentum to the satellite, causing it to spin. A spinning satellite suffers a lot more drag and would struggle to use its engines, antennas or laser links and would need to slow its rotation before resuming operations.

Therefore follow up strikes could be focused on satellites actively trying to reduce their spin.

Cheap and Stealthy Orbit to Orbit Missiles

There are many of ways that these fragmentation weapons could be built, one I find especially promising is a small and cheap space-to-space missile carried in large numbers by satellites. It would maneuver with ammonia gas thrusters: ammonia is an especially attractive fuel here because it can be stored as a liquid, meaning smaller and lower pressure tanks can be used compared to more common nitrogen high pressure systems. Low pressure tanks have the advantege that if the missile is struck by a fragment before launch, damage to the host satellite is going to be limited.

Another advantage of ammonia is that before it can be used for propulsion it must first evaporate, cooling the missile and making it less visible to infrared sensors. A stealthyshape coupled with the small size would make radar detection very difficult as well.

To reduce costs the missile wouldn’t carry any external sensors, completely relying an inertial guidance system and tracking and navigational data provided by satellites to reach its target.

I emphasized stealth here because these kind of weapon would be best suited to surprise attacks, quickly achieving orbital superiority.

Intercepting ICBMs, SLBMs and other rockets from VLEO

Similar missiles could be used to intercept rockets travelling in the the high atmosphere at altitudes over 60 km (30 miles). Adding thermal protection, more powerful propulsion and a warhead designed to more effectively damage the walls of solid fueled rockets wouldn’t make them much more expensive and would enable the interception of a significant number of ICBMs, SLBMs and other rockets in the early stages of a launch.

This is not a novel idea: it bears some similarities to the Brilliant Pebbles concept discussed in the 1980s. At the time high launch costs and less sofisticated electronics made orbital interceptors unpractical but that’s no longer true.

Ultra low Earth orbit is very useful in this application as well, reducing interception times and interceptor propulsion requirements. Lower satellites can more effectively guide the missiles as well, as previously discussed.

Conclusions

Stealthy space-to-space missiles make surprise attacks extremely attractive. With little to no warning entire constellations could be disabled, quickly establishing orbital superiority. Orbital superiority could then potentially be exploited to completely deny the use of space by the enemy, including the launch of intercontinental ballistic missiles.

Evaluating the consequences on nuclear deterrence is a very complex topic that I will discuss in the next article, newer tactics and weapons will likley need to be developed like powerful ground-based electronic warfare systems and/or sprint missiles carrying nuclear warheads designed to punch holes in enemy satellite constellations.

But as I said, that’s the topic of my next article so consider subscribing if you don’t want to miss it.

Here I’ll explain how these warheads could hit both fixed and mobile targets with the support of next-gen satellite constellations.

Weapon guidance sensors and reentry

Weapons can use a variety of sensors to determine their relative position to the target and accurately steer towards it. The harsh conditions encountered by long-range ballistic weapons when they reenter in the atmosphere however complicate the use of many sensors like optical and radar, driving up costs and complexity.

For example, long-range Anti-Ship Ballistic Missiles (AShBMs) like the Chinese DF-26 need to slow down to speeds below Mach 5 before they can use their onboard radar seekers. In fact, at higher speeds, a plasma sheath surrounds any object moving through the atmosphere, interfering with the transmission and reception of the radio waves used by radars.

Since an AShBM warhead can only see its target after it has slowed down significantly and has only about 15-20 seconds left to correct its trajectory, it may miss the target entirely even if it’s designed (at high expense) to be highly maneuverable.

In general guidance systems reliant on external sensors are a bad fit for relatively inexpensive warheads to be produced in the thousands: small and light sensors robust enough to survive reentry aren’t cheap.

Inertial navigation

This is also one of the reasons why the most common guidance system used on ballistic missiles relies instead on an Inertial Navigational System (INS) paired with a Global Navigation Satellite System (GNSS). Let’s explain these terms.

An inertial navigational system is a device that can accurately measure its acceleration, using this data to then calculate an estimate for its speed and position. Since successive integration steps are used even small accelerometer errors compound with time resulting in warheads missing the targets by up to a couple of kilometers.

Current navigational satellite constellations

A way to accurately determine the position of a warhead in flight is therefore needed to periodically reset the errors accumulated by the INS.

This can be achieved through the reception and processing of multiple radio signals each encoding a precise timestamp and data on the position of the transmitter: this is working principle behind GNSS constellations like GPS.

Current GNSS satellite constellations however have a number of weaknesses: since they use expensive atomic clocks each constellation only has a few dozen satellites, this not only makes them a very attractive target for anti-satellite weapons but forces them to use Medium Earth Orbit (MEO) to provide adequate worldwide coverage.

Since the MEO orbits used by GNSS satellites are 20000 km from the Earth and each satellite must broadcast signals to a large surface, GNSS signals are very weak and susceptible to jamming, a vulnerability exacerbated by the fact that these satellites use signals with frequencies in the L band, a very crowded section of the electromagnetic spectrum, and so only have around limited frequency bands available for their transmissions (the electromagnetic spectrum is split in a number of frequency bands each reserved for a specific use to avoid mutual interference). This means that a jammer can focus a lot of power in a limited section of the spectrum, drowning the weak GNSS signals in noise.

A way to reduce the impact of jammers is using composite antennas called CRPA (Controlled Reception Pattern Antenna). These antennas are capable of reducing the reception of signals coming from certain directions by processing the signals received by multiple smaller antennas, however if there are jammers transmitting from many different directions the CRPA needs to be larger: in Ukraine some Russian drones fly with CRPAs of more than 30 elements to have a chance to intermittently receive GNSS signals.

However these large CRPA antennas can get pretty big and are unsuitable for warheads with limited available space: elements of a GPS CRPA needs to be spaced by about 10cm, the length of the half wavelength of the radio signals they are designed to receive.

Future navigational satellite constellations

These issues can be solved by building a GNSS constellation using thousands of modified starlink V3 satellites.

This is quite practical since Starlink satellites are being cheaply mass-produced by SpaceX, and are already equipped with laser links that could be used to accurately measure the relative distance between satellites and constantly synchronize them so that no expensive atomic clocks would be required to provide GNSS capabilities. Since Starlinks are communication satellites they are also equipped with large antennas operating in the Ku and Ka band, in a region of the electromagnetic spectrum made up of waves of about 10-20 GHz frequency or 1-2 cm wavelength, these antennas could be used to broadcast the GNSS signals.

Since starlink V3 satellites are designed to orbit at 300km and they can focus their transmissions on a far smaller area compared to GPS satellites, the resulting signal could be more than a million time stronger, making jamming extremely impractical, especially since starlink has a lot of electromagnetic spectrum available in the Ku and Ka band, forcing a jammer to spread its power over a wide band.

The small wavelenght of the signals transmitted by starlink satellites is another element that improves jam-resistance since it enables the design of small CRPAs with many elements that can fit on the rear face of warheads.

Another advantage of this new architecture becomes apparent during reentry: the signal is strong enough that the plasma sheath would only block it only for a short amount of time. 10-20 GHz signals help here as well since higher frequency signals can more easily go through the reentry plasma, especially at the rear of the warhead, where it’s less dense.

Datalink and moving targets

Starlink V3 satellites were originally designed for communications and there’s no reason to not retain some of this capability. This could be used to redirect a warhead in flight to a new target of opportunity or to update the projected position of a moving target at time of impact.

This guidance method has the advantage of enabling cheap and simple warheads to threaten moving targets like warships but requires the target to be constantly tracked by external sensors with high accuracy, especially in the terminal phase.

This kind of global all-weather reconnaissance can be achieved with a large LEO radar satellite constellation.

Global reconnaissance and tracking constellation

Starlink V3 satellites could be modified into radar satellites as well. This might seem strange but radar and communication satellites do share a lot of features: both have large antennas capable of transmitting and receiving signals, the main difference is that a radar is designed to receive the transmitted signal after it’s reflected by a surface while in a communication satellite the transmitted and received signals are largely unrelated.

Starlink V3 satellites have a number of features that make them very attractive radar platforms: the most useful is the low operating altitude of 300km (other radar satellites operate at more than 600km), since the strength of the reflected signal is inversely proportional to the fourth power of the distance this alone improves the performance of the radar by 16 times.

The large Ka/Ku band antennas are also capable of focusing a lot of radio power in a small area, increasing radar return but do have a weakness: radio waves in those frequencies are more susceptible to atmospheric absorption, especially in bad weather.

This is the reason why ground based radars do not use these frequencies but it’s less of an issue for a satellite based system where the radio waves spend less time in the thicker regions of the atmosphere.

Finally the laser links between Starlink satellites have a lot of potential here as well, enabling multiple satellites to synchronize their emissions and work together, significantly improving the performance of the constellation.

Defeating radar stealth

This would be especially useful to defeat stealth: radar stealth mostly works by deflecting incoming radio waves away from the direction of the emitter. This is extremely effective unless distributed and coordinated radar systems are employed: radio waves deflected away from one radar could be detected by another.

If many radar satellites emitting and receiving from many different directions work together, a 6th gen fighter wouldn’t be much more difficult to detect than an unstealthy one. This is especially true since stealth planes aren’t optimized to defeat radars in the Ka/Ku bands since it’s unsuitable for ground-based radars.

Jamming resistance

This radar constellation, while extremely capable, is still vulnerable to jamming. However the high number of satellites and the wide bandwidth available to them means that jamming them all wouldn’t be trivial. This is especially true since each jammer could only obscure an area of about one square km, meaning a large number would be needed to defend an area.

Another important fact is that while a jammer would be able to hide an area around it from observation, its emissions would be very visible to the constellation and could be tracked accurately enough for area-effect weapons to be dispatched to a jammer’s location.

Conclusions

Constant worldwide radar reconnassaince and tracking and jam-resistant navigation and datalink provided by next-gen satelllite constellations will revolutionize the ways wars are fought, especially if combined with the strike potential of reusable rocketry. Controlling Low-Earth Orbit will be crucial in future conflicts, arguably more than air superiority is today.

This is why in the next article I’ll discuss how to deny the enemy the use of orbit while protecting your own assets, especially from the debris generated by the destruction of thousands of satellites.

This is the first in a series of articles that will explore the ways armed forces could exploit reusable rockets and large satellite constellations, building up a coherent doctrine for their employment. I’ll first focus on how reusable rockets can be used to deliver weapons at intercontinental distances while in the next article I plan to talk about future military satellite constellations and orbital warfare.

Rocket technology innovations in Starship

Cost savings

To minimize launch costs with the next-generation Starship rocket SpaceX has both reduced manufacturing costs and moved toward full reuse of both the first and the second stage.

The most-important cost savings come from the engines: Raptor 3 produces thrust at a very low price, burning liquid methane and liquid oxygen (so-called “methalox” mixture) very efficiently. (The cost to thrust ratio for Raptor 3 is apparently less than one fifth of other comparable engines)

Lower cost per thrust is very consequential since it fundamentally changes the engineering decisions done during design: it’s much less important to shave off mass from the rocket when you could reach the same performance target more easily by just adding an engine and making the whole rocket bigger.

This helps explain why the fuel tanks of Starship are made of steel with internal reinforcement, a cheaper (but heavier) solution compared to traditional machined aluminium-lithium alloy structures. Many other similar decisions have helped keep the production costs low.

Reusability

Much of Starship’s design revolves around full reusability, ensuring no new hardware needs to be built for each flight and both the first and second stage can be flown again with as little work as possible. The reuse process starts with recovering both stages: to save weight on landing legs, SpaceX has decided with Starship to add “catching” arms to the launch towers, designed to grab the rocket stage in the final stage of deceleration.

The most important innovation SpaceX has achieved here is designing the first stage, the booster, to be quickly reflown after each launch with no need for refurbishment, a process called “fast reuse”.

I won’t spend much time talking about the reuse of the second stage since it’s still being figured out and I suspect that the refurbishment process is going to take at least a couple of weeks, longer than what I think the decisive stage of a peer war will last.

So I’ll mainly focus here on a semi-reusable architecture that lands and reuses only the booster, built fully exploiting Starship technology. This can be done simply using an expendable version of the Starship second stage but limitations in an architecture designed for civilian use makes the design of a smaller rocket built for military use from Starship tech very attractive (I’ll come back to this later).

This does not mean that fully reusable architectures do not have place in warfare, reusing the second stage could be ideal in the asymmetric conflicts the US has been fighting to help keep costs low. It also would have the added benefits of preserving stockpiles of expendable second stages built up for peer conflicts.

Starship as a weapon delivery platform

Every space launch vehicle can put payloads on ballistic trajectories, since ballistic trajectories are less energetic than even Low-Earth Orbits. Each kg of space launch capacity can therefore be converted into 1.5 to 2 kg of payload to intercontinental ballistic trajectories, depending on the location of the target and the ballistic arc employed.

Historically this has been very impractical: launch costs were far too high and total launch capacity to orbit was limited and could not be quickly expanded during a war.

However these limitations don’t apply to Starship and other rocket using similar technology, since they have very low launch costs and very high launch cadence: in 2030 SpaceX plans to have 15 Starship launch pads in operation each one capable of launching once per day. Total payload to intercontinental ballistic trajectories using expendable second stages could reach 5000 tons per day at a 100$/kg cost.

The capacity to launch a lot of mass wouldn’t be enough to make Starship a useful military system: it should also have the ability to launch multiple payloads on many different targets, functioning like an ICBM with Multiple Indipendent Reentry Vehicles (MIRV), just on a much larger scale and with the booster going back to the launch pad instead of being expended.

An advantage of using Starship as a large liquid fuel ballistic missile compared to more common solid fuel ICBMs is that the second stage can accurately and quickly change its thrust, enabling the fast deployment of warheads into many different trajectories, even if the targets at hundreds of kilometers from each other. That means that every Starship launch could accurately shower China or Iran with one thousand 300kg warheads travelling at 5-6 km/s. (The US struck about targets during the first day of Epic Fury, 15 starships could have destroyed 10 times the targets)

This high speed (Mach 15-20) is both a problem and an opportunity. The payload needs to be protected by a reentry vehicle (RV) from the heat generated during the reentry process while maintaining some capacity for maneuver, this while not a trivial problem doesn’t require expensive and complex “hypersonic” technology. (In military circles hypersonic refers to technology designed to enable extremely high maneuverability at speeds higher than Mach 5 to complicate intercepts).

It’s far cheaper to use MaRVs (Maneuvering Reentry Vehicles), fairly simple conical reentry vehicles coated with phenolic resins designed to ablate to protect the warhead, with some limited maneuver capabilities to ensure high accuracy. (I’ll talk more about weapon guidance when I describe navigational satellite constellations in the next article)

This approach uses large numbers of cheap warheads (I estimate total cost for each weapon to be about half a milion dollars) to quickly overwhelm enemy air defense systems: here speed the high reentry speed is a huge advantage, compressing the time available for intercepts and making decoys and jamming more effective.

High speed is also incredibly useful to defeat hardened structures: for example the best weapon the US has currently available to destroy deep bunkers is the MOP (Massive Ordnance Penetrator) a 14 ton gravity bomb that requires the use of large bombers, Starship could launch faster and heavier penetrators capable of destroying deeper bunkers built to withstand current US weapons.

The Iran war underlines another advantage of Starship over current US weapons: US armed forces took weeks to move carrier strike groups and other assets to the region, making them vulnerable to Iranian retaliation. Starship could have struck while the protests were still ongoing, catching the Iranians completely unprepared with a salvo of thousands of weapons.

Weaknesses of Starship as a weapon system

Starship has a huge potential as an offensive system but it’s a glass cannon, incredibly vulnerable to attacks to the fixed infrastructure needed for its operation. Fueling infrastructure is by far the biggest vulnerability, since it’s often shared by multiple launch pads and it’s full of flammable methane that could be ignited even by small drones. The catch arms could be disabled by small but well placed explosions as well.

Finally the large size of Starship makes it especially vulnerable during the initial flight phase, either from orbital interceptors or even a fragmentation warhead launched by an enemy rocket and guided by overflying satellites.

Military reusable boosters

To deal with these issues a military reusable rocket needs to be designed and mass-produced.

To solve problem connected to fixed launch infrastructure this new rocket should use landing legs so it could take off and land from improvised launch pads made up of patches of specialized concrete.

These kind of improvised launch pads have been proven to work during the early stages of starship development, simple concrete only pads only failed under the combined thrust of 33 raptor engines during the first launch of the full starship rocket.

A hypothetical reusable rocket with a scaled down starship first stage using only 7 engines could be capable of launching from hundreds of different pads simultanously, after being refueled by mobile cryogenic tanker trucks.

This would significantly improve the offensive power of the system as well, making surprise attacks even deadlier.

To protect against attacks of similar weapon systems camouflage, decoys and jamming are very important. All fixed infrastructure like fuel depots should be hidden inside deep bunkers with secret locations and layouts, with fake and redundant entrances built.

Consequences on nuclear war

The kind of massive intercontinental ballistic attacks I described puts ICBMs and long-range bombers under serious threat from conventional attacks: for the first time in history it may possible to disable large part of the enemy nuclear arsenal without using nuclear weapons.

However under the current nuclear doctrine of launch on warning every large scale ballistic strike, nuclear or not, would result in immediate nuclear retaliation while the warheads were still in flight.

I have always thought that launch on warning is the dumbest legacy of the cold war: to protect land based nuclear weapons that are more and more vulnerable to a growing number of weapons millions of people are under constant threat from a hair trigger.

However other nuclear doctrines exist: until not so long ago China’s nuclear doctrine was to retaliate only after nukes were used on its territory. This a far safer option and mutually assured destruction is still ensured by nuclear ballistic missile submarines, by far the most survivable leg of the nuclear triad.

Conclusions

Reusable rockets have the potential to be the most effective way to launch long range strikes, reducing the need for expensive and vulnerable carriers and foreign bases.

The only real issue that their adoption may face (other than corrupt procurement officers) are the consequences on nuclear war. However as I explained at length the moment Starship becomes operational the US armed forces could immediately exploit it to launch strikes, there’s no going back now.

I also think that reusable rockets provide a nice simmetrical response to Chinese investments in ballistic missiles designed to destroy US bases: after all why shouldn’t the US be allowed to strike Chinese bases back with ballistic weapons?