February 12, 2023

Thoughts on the Chinese Balloon

Last week, the media was captivated by the story of a Chinese surveillance balloon floating across the US. Questions were asked about how we could let this happen, and many were extremely indignant about the affront to US sovereignty, so I thought I'd weigh in with a longer perspective.

First, the idea of sending a balloon over your enemy's country isn't new. In WWII, the British launched almost a hundred thousand balloons at Germany to short out the power grid and start fires, while the Japanese sent thousands of balloons into the jet stream, hoping they would cross the Pacific and set fires in North America. One of these bombs, found by a group on a church picnic, caused the only fatalities from enemy action on the US mainland in WWII. Later, the US used camera-carrying balloons to photograph the Soviet Union before the U-2 entered service.1 These balloons were responsible for many early UFO sightings, with a related project to detect nuclear testing with microphones on high-altitude balloons being responsible for the Roswell Incident. More recently, US fighter pilots have reported encountering "UFOs" that sound a lot like balloons over the last 5-10 years.

This is an obvious extension of all that, and from a strategic perspective it changes very little. We already knew that China was a serious threat to our strategic interests, and you should only update on this incident if you weren't already thinking that way. I'm not even particularly upset that China tried this. We and they both play the espionage game, and frankly I'm a lot more worried about their HUMINT efforts, which have the FBI opening a new case every 12 hours, than anything they can do with balloons. Almost as bad is their extremely successful campaign to persuade Americans to voluntarily install Chinese spyware on their phones in the form of TikTok.

But that still leaves the question of why the Chinese decided to go to all the trouble of building a massive balloon and launching it at the US, at which point we descend into the realms of speculation. In practical terms, balloons have the major disadvantage that they're at the mercy of the wind, although modern meteorology means that a balloon that can change its altitude (usually by pumping air in and out of a bladder inside the envelope) has some degree of maneuverability, depending on the winds at different altitudes. Their big advantage is that they can stay up essentially indefinitely, providing continuous surveillance over a given area. The US military has been looking at lighter-than-air craft for surveillance purposes, although these are intended to operate over areas they control, using solar power for theoretically infinite endurance.

Obviously, the balloon in this case didn't stay on station indefinitely, but it was able to keep whatever sensors it had over the US for a lot longer than Chinese satellites (which are very much a thing) could. Aircraft could do similar work, but China has no friends nearer to the US than it is, which means any surveillance mission would need a refueling plan that would make Black Buck look simple and if they tried to get over Montana, some fighters would insist that they leave. My best guess is that the balloon was intended to collect electronic intelligence over an extended period. Photographic recon seems unlikely given the capabilities of satellites and the uncertainties of cloud cover, while a persistent ELINT platform has the potential to capture signals that are timed to avoid satellite passes, or to record and analyze LPI signals that satellites pass too rapidly to get a good handle on.

That in turn brings us to the US response, which on the whole seems to have been pretty decent. China has been doing this for a while (at least four previous Chinese balloon passes over US territory are known, although some of them were only detected in retrospect) and we presumably had a pretty decent idea of what sort of payload it carried. If nothing else, I'm sure someone with a targeting pod flew around the thing and took pictures at pretty close range, and it's very plausible that our ELINT satellites were monitoring its transmissions. At that point, it's a question of the tradeoff between what information they will gather (and what we can get from watching it) and the risk of damage or harm to those on the ground from shooting it down. This assumes, of course, that we weren't jamming it to prevent anyone from signalling home, which the DoD claims it did. The US tends to be very careful about risks to bystanders, and using weapons is always a chaotic business. Even though Montana is mostly empty, it's not so empty that you can be confident that nobody would be hurt over a quite large area (the final debris field is reported as covering 7 square miles, and that assumes you do the shootdown with a missile and not a gun). Even if it did pass over an appropriate area, that would have to have been communicated in advance to the shooter, and the boundaries checked very carefully to make sure that it was still over the area and that shifting winds wouldn't suddenly dump the debris on a town.

There are also some practical aspects of a potential shootdown that would push towards an overwater shootdown. Even if we can broadly say what sort of intelligence it is gathering, we'd like to look at the actual hardware. Shooting it down over the sea makes a lot of sense in that case. First, the impact with the water is going to do less damage. Second, it's much easier to keep souvenir-hunters at bay when you're off the coast, and the number of people in the country with the ability to covertly retrieve anything from 80' down who aren't associated with the military in some way is plausibly zero. Third, we have spent decades learning how to find stuff on the ocean bottom, and the bottom in that region is pretty flat. It's very plausible that it's actually easier to search for the fragments in the ocean than it would be on land.

The bottom line of all this is that while the balloon was concerning, it wasn't nearly as big a deal as it has been made out to be. I'm sure that NORTHCOM is working to upgrade its sensors and come up with a better plan to deal with future balloon incursions, and rightly so. But anyone who has been paying attention already knows that China is a rival who would dearly like to unseat us at the top of the global order. If it takes a balloon to get that message to the general public, then it was clearly an own-goal for China. But it really shouldn't have taken that to get us to take the threat of China seriously. And if you're really worried about their spying, uninstall TikTok.

1 Interestingly, the Soviets used film retrieved from some of those balloons on the Luna 3 mission, the first to photograph the far side of the Moon.


  1. February 12, 2023Anonymous said...

    But that still leaves the question of why the Chinese decided to go to all the trouble of building a massive balloon and launching it at the US, at which point we descend into the realms of speculation.

    That even assumes it was a spy balloon, not just a weather balloon that went somewhere other than intended, the Soviets were shooting down a lot of them even long after the US stopped deliberately sending balloons over their territory.

  2. February 12, 2023bean said...

    True. On the other hand, this was a lot bigger than a normal weather balloon, and the US has said that it has found surveillance equipment in the wreckage.

  3. February 12, 2023CmdrKien said...

    Weather balloon doesn't make sense. Actual rawinsonde ballons are much smaller, and generally just go up and down, because the point of weather balloons is to sense the temperature, humidity, and winds along your column of air. I suppose you could use satellite type remote sensing equipment, but those generally benefit from being able to take a swath of data from a large area, not repeatedly sensing the same relatively small area.

  4. February 12, 2023Lambert said...

    Does China also make their balloons from cereal packets?

  5. February 12, 2023ike said...

    frankly I’m a lot more worried about their HUMINT efforts[...]

    How much better on that front do you think the DoD-proper is doing over industry and the DoE(the joke used to be everyone who works for them is a spy for someone)?

    Maybe GE &al are getting prioritized becuase, not only do you get miltary inteligence, but also valuable trade secrets.

  6. February 12, 2023Ariel said...

    I don’t believe that searching for debris at sea is cheaper than on land - people that can look for stuff in terrain come pretty cheap, and ROVs are expensive.

    But it is certainly easier to close a debris field to unauthorized people, either cowboys or Chineese spies, when that debris field is at sea.

  7. February 13, 2023EngineOfCreation said...

    China insists that its "weather balloon" was blown off-course. How plausible is that anyway, assuming it was an actual weather balloon? Apparently the balloon in question had the ability to adjust its altitude, which should enable it to ride a more favourable wind current, perhaps at least one that doesn't happen to bring it over rival territory. My impression is that high altitude wind flows are broadly predictable at any give time of year.

  8. February 13, 2023Brian said...

    Collect electronic intelligence ? What are they hoping to get? Millions of cell phone calls? That's a lot of data to sift through in hopes that the balloon just happened to be over an area at the right time to catch a meaningful call. An embedded spy with a land based receiving station at a critical fixed location seems much more effective. Besides that - shouldn't we be using good encryption for all sensitive electronic communications - rendering interception useless?

  9. February 13, 2023bean said...


    I said "easier", not cheaper. The dollar cost of the search is definitely higher (although maybe not as much when you include security) but it's a lot more predictable what you'll end up with.


    It's not very plausible. This is way bigger than any weather balloon I've heard of, and while I'm sure that atmospheric scientists could use the data, if they really wanted it, this would have been done long ago. Also, why would they be running a secret weather balloon program?


    I'd point you towards the link marked "collect electronic intelligence", where I lay out the basics of SIGINT/ELINT. It was far more likely to be radar signals or encrypted comms traffic than cell phones, and those might be hard to gather from the ground, because we tend to do them in areas where there aren't a lot of plausible reasons for someone to be stringing new antennas, and anyone who does will get a visit from some government agency. As for encryption, knowing the structure of the signals can be quite valuable, even if you can't read them.

  10. February 13, 2023Anonymous said...


    My impression is that high altitude wind flows are broadly predictable at any give time of year.

    There's still enough unpredictability for balloons to get blown where they aren't intended.


    Besides that - shouldn't we be using good encryption for all sensitive electronic communications - rendering interception useless?

    Tell that to the idiots who think weakening encryption is a good idea.

    But even with encryption traffic analysis can still be useful, if there's a lot of traffic from an 'uninhabited' point in the middle of the desert there's a pretty good chance that there's something very interesting there.

  11. February 13, 2023Basil Marte said...

    "I’m a lot more worried about their HUMINT efforts, which have the FBI opening a new case every 12 hours"

    Are you willing to say more about this (what sort of article should an outsider read)? The quote only implies that the FBI is worried, is looking, and is finding ...possibly nothing, as far as I would know.

  12. February 13, 2023EngineOfCreation said...

    @Bean I'm not arguing they are weather balloons. I'm just asking, if they were, how hard or easy would it be to mess up their trajectory so badly on accident. I'm aware the meteorology community would be a better place to ask that, but just taking a shot here.

    Or I could frame it in military terms: Of the military balloons Japan sent towards the US in WW2, how many did arrive on or over US soil, how many were known to land somewhere else entirely, and what would these percentages be like with modern technology?

  13. February 14, 2023bean said...


    I don't have any inside information on that. I've just been paying attention to the news enough to know that China is doing a lot of espionage, and the FBI is worried. Obviously, every case doesn't pan out, but I think the counterintel people are generally doing a decent job. (Modulo the one taking bribes from Russian oligarchs, that is.)


    I don't really know, unfortunately. There are reports that a lot of these are loitering near sensitive targets like missile fields, which doesn't seem particularly possible to do by accident.

    Re the Japanese balloons, 300 of 9300, wiki doesn't mention anything beyond "a few in Mexico" and a lot higher. We have some ability to detect position, and much better understanding of flow in the upper atmosphere.

  14. February 14, 2023Andrew Sumwalt said...

    Well, I think if it were harmful, the ballon would probably would have been intercepted earlier.

  15. February 14, 2023AlanL said...


    if there’s a lot of traffic from an ‘uninhabited’ point in the middle of the desert there’s a pretty good chance that there’s something very interesting there.

    Surely you know that already from following the guards' strava accounts?

  16. February 15, 2023EngineOfCreation said...

    Latest development: It's still a spy balloon, but was maybe not intended for the US mainland.


  17. February 17, 2023Anonymous said...


    Surely you know that already from following the guards' strava accounts?

    Good point.

    It seems if you want a secret base to be secret you have to confiscate all electronic devices from everyone who goes there.

Comments from SlateStarCodex:

Leave a comment

All comments are reviewed before being displayed.

Name (required):

E-mail (required, will not be published):


You can use Markdown in comments!

Enter value: Captcha