How precisely can the splash-down point of re-entering capsules (from Mercury onwards) be known in advance? Wasn't there a risk that capsules would hit a ship or an island?

Likewise, Russian and Chinese capsules that re-enter on dry land seem to have a risk of landing on a house, tree, cliffside, river, or other inopportune spot. On land, there are a lot more potential obstacles.

The same could be asked about rocket components falling to the sea or about military test missiles. For these, there might be a lot a lot less precision. How well is the landing spot known, and that are the risks of hitting something?

The converse question: Does the predictability of the landing always allow recovery boats to be close enough to reach the capsule in a few minutes? Is there a chance that that won't be possible?

  • 1
    $\begingroup$ I’m voting to close this question because it's based on false premises $\endgroup$ Commented Aug 3, 2020 at 12:10
  • 7
    $\begingroup$ It's a fair question about space exploration @CarlWitthoft . There is a risk that re-entry could go off track, it's happened before, think Mercury 4. Sure, it's not likely but it's still possible to hit something. Returning Soyuz spacecraft have occasionally found themselves in a pickle. $\endgroup$
    – GdD
    Commented Aug 3, 2020 at 12:22
  • 10
    $\begingroup$ Please don't confuse space debris and controlled spacecraft. Entry of space debris, indeed, is unpredictable because they are slowed down by atmospheric friction only. A tiny change in atmospheric conditions can result in 1000s km difference for entry point. But controlled spacecraft are different. They make entry burn to slow down for entry at planned point. Atmosphere still add some uncertainty but far less, about tens of kilometers maximum. $\endgroup$
    – Heopps
    Commented Aug 3, 2020 at 12:22
  • 2
    $\begingroup$ @OrganicMarble I actually don't know how much uncertainty. I will appreciate any information on that. $\endgroup$
    – Joshua Fox
    Commented Aug 3, 2020 at 12:45
  • 15
    $\begingroup$ @CarlWitthoft "based on false premises" is not a close reason. Also, the question has been edited. $\endgroup$
    – uhoh
    Commented Aug 3, 2020 at 13:13

4 Answers 4


Partial Answer:

From Apollo by the Numbers we can see that even in the 1960s and 1970s the splashdown point was quite predictable.

Page 305 shows the maximum miss distance to the target point was 3 nautical miles. The maximum distance to the recovery ship was 13 nautical miles.

Note that the Apollo capsule was actively flown during entry by rolling the lift vector.

I have no reason to suspect that in the 2020s it would be worse, assuming active control of the capsule.

An emergency deorbit to an unplanned site is quite another case and could end up landing anywhere along the path of the capsule's orbit. As @MSalters points out, that's limited by the orbital inclination.

  • $\begingroup$ Thank you. What about landings on dry land? Unless it is some very flat steppe-land, there seems a much bigger risk of hitting something, even with such a maximum miss distance. $\endgroup$
    – Joshua Fox
    Commented Aug 3, 2020 at 13:12
  • $\begingroup$ @JoshuaFox I don't know about those. Thanks for the comment, edited answer to show that it's a partial answer. $\endgroup$ Commented Aug 3, 2020 at 13:15
  • 5
    $\begingroup$ @JoshuaFox, there's a reason why Russia's landing targets are in Kazakhstan, while China's are in Inner Mongolia. $\endgroup$
    – Mark
    Commented Aug 3, 2020 at 22:37
  • 1
    $\begingroup$ An emergency deorbit will land somewhere "near" the equator, where "near" is limited by the inclination of the initial orbit. For the ISS, that would be 51 degrees or less. "Anywhere" might be too strong; a returning Soyuz won't hit the North Pole. Still an awful lot of ocean to cover. $\endgroup$
    – MSalters
    Commented Aug 4, 2020 at 11:00
  • $\begingroup$ @MSalters that's a good point, thanks for pointing out that imprecision. Editing. $\endgroup$ Commented Aug 4, 2020 at 13:03

Partial answer:

According to russianspaceweb.com, there are a set of large, pre-selected landing sites for the Soyuz capsules.

Soyuz can land with an accuracy of only 28 kilometers, (with a probability of 0.9997), in the automated aerodynamic descent mode, AUS, relative to the center of the projected landing area.

The main reason for such a low precision is the suseptibility of the parachute landing to winds. Moreover, in case of a ballistic return, the capsule can end up as far as 600 kilometers short of the primary landing site for the aerodynamic mode.

As a result, all Soyuz landings have to be planned over a flat and open areas without any structures, rivers or even trees. A total of 13 areas currently meet all the requirements for the Soyuz landing. Ironically, all of these sites are in Kazakhstan and none of them are in Russia.

The website also lists seven of the potential landing sites; but I would take these coordinates with a grain of salt, since the first one is awfully close to the Volga river and the second one is smack in the middle of the village of Peschanoe.

See also How close to a building has a Soyuz landed?

  • $\begingroup$ I imagine when these landing sites were chosen, Kazakhstan was still part of the USSR, so it's not really all that ironic. Kazakhstan also has the advantage of being further south that most of Russia, so it made sense to have their space program centered there, as it's more convenient to launch (and land) closer to the equator. $\endgroup$ Commented Aug 5, 2020 at 16:41

The first of your several questions is about controlled reentry of devices designed for reentry. You then ask about devices not designed for reentry, which I address (partially).

Undesigned reentry can be rather imprecise. Taco Bell made a sizeable bet on this imprecision. The Progress M1-5 deorbit of Mir had relatively large location uncertainty. (This was a controlled reentry of a device that was not designed for reentry. Consequently, its performance during reentry was, at best, only approximately understood.)

Mir reentry map

(source: https://en.wikipedia.org/wiki/File:Mir_reentry_map.svg )

The "X" marks the estimated point of main debris impact. Red indicates expected debris impacts. Green is the track of the final orbit and orange is an estimate of the region above which the device entered the atmosphere (the entry interface). "The debris spread about ±1,500 kilometres (930 mi) along track and ±100 kilometres (62 mi) laterally ..." (from Wikipedia: Deorbit of Mir -- the source, having more details and figures). Notice this is an area of $600\,000$ square kilometers.

  • $\begingroup$ NZ "suddenly acquired" some spherical gas tanks a few hundred mm in diameter longish ago. They had cryllic writing thereon. Russia denied prior ownership. || And here is (probably) Kosmos 2430 breaking up over NZ in 2019 (annoying 40 second pre-ad) $\endgroup$ Commented Aug 5, 2020 at 0:18

The probability of hitting a ship in the ocean is very small. Ships are tiny compared to an ocean.

enter image description here

I assumed a landing zone (blue circle) of 10 km diameter. The black circle is a security zone of 20 km diameter.

There are nine ships hidden in the forbidden landing zone. The ships are drawn as a filled circle of 50 m diameter. Do you see those tiny (8 green and 1 red) spots? The red ship is located in the center of the circles.

We assume the capsule will land with a probability of over 99 % within the landing zone. What is the probability of hitting the red ship in the center?

For a rough estimate we assume the probability is constant within the landing zone. The probability of hitting the red ship is then the ratio of the circle areas, that is the ratio of the squares of the diameters. $$(50/10000)^2 = 0.000,025$$ So the probability of hitting the red ship is only 25 ppm (part per million). Or 24.75 ppm if we use the 99 %. For the 9 ships it is only 225 ppm.

So it is save when there are no ships within the blue landing zone and only recovery ships within the black security zone. To avoid hitting the ground the landing zone should be at least 50 km away from any coast or island.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.