[yencabulator]

Consider a Hubble-size lens at a distance of 200-1000 km. If I did my quick math right, turning just 20 degrees moves the focus by up to 350 km on the ground. And typically orbits have a pattern of gradually shifting, so the next orbit will capture a different slice.

Reconnaissance planes were largely obsoleted by satellites when digital cameras were invented (before that, satellites physically dropped film to ground to be developed), and as Russian anti-air missiles got better. The primary use for optical reconnaissance planes these days is for up-to-the-minute coverage of rapidly developing situations, such as in war, where the next satellite flyover might not be until several hours later.

(RF & radar reconnaissance planes are very important to modern warfare.)

[chefkd]

Whoa so many mind blowing facts. Does the orbit change or does the lens rotate? If one knos the rotation degree can the total number of satellites it would take to cover the entire earth be calculated?

[yencabulator]

Here's an example of an non-"changing" orbit goes over different parts of the planet over time: https://www.scienceabc.com/eyeopeners/why-do-satellites-orbi...

It's highly unlikely that a satellite-based telescope would rotate/turn the lens. Turn it relative to what? It's in 0g, there's no part that's held in place by friction to the ground. Any rotating joint would make the two sides move in opposite directions. And if you think about that, let one side be much smaller, and allow rotating by multiple turns instead of just 0-360 or such, you might just invent the gyroscope.

They turn the whole satellite with gyroscopes (flywheels) and/or propellant.