[CliffordK (OP)]

On TV, they often say that they need to "move a satellite into position". 

Does the military actually move satellites?

The only places they can maintain satellites in a single position would be over the equator, so a geostationary orbit near Moscow would be over Kenya, and would suffer from wicked shadows. 

I would imagine the military would design a grid somewhat like the GPS grid where satellites are placed in a low earth orbit, with an orbital period of about 2 hours.  Then once enough are put in the sky, there is always a few in the sky everywhere.

Moving a satellite out of its orbit would take an extraordinary amount of resources,  and holding it in one place over an arbitrary point on Earth would be essentially impossible.

The important part would be directing the surveillance equipment.

[syhprum]

I believe the military have several polar orbiting "spy satellites" of a similar design to Hubble and when they talk of moving them into place it is a matter of programming their surveillance instruments to cover the area of interest.
I doubt that any large change can be made to the orbital parameters.
 

[evan_au]

When there is an area of particular interest, a satellite can be launched into an elliptical orbit which spends more time above a particular point on the Earth - but during this part of the orbit, it is further away, and the images have less resolution.

Or you can launch it into a low, circular orbit, so it gets high resolution images, but it spends very little time over one position. In this orbit, you can only see a small region below the satellite for a very short period of time, and the satellite rarely passes over a particular point of interest - unless you burn a lot of fuel. (You need to burn some fuel anyway, otherwise atmospheric drag will quickly drop the satellite out of orbit.)

For the most part, it is a matter of waiting until a suitable satellite is in the right position (unlike the movies where it sounds like you can "park" a satellite over a particular target).

Polar satellites passing overhead near dawn and dusk are useful because the suns rays cast long shadows, which assists estimating of the heights of objects.

Drones have the advantage that they can loiter over an area of interest for hours, and also take "close up" photographs from much lower altitude than a satellite.

[SeanB]

Moving a satellite really is a waiting game, you wait until the existing orbiting unit is going to orbit such that it can get an image of the area, and depending on things like cloud cover and wavelengths in use you might have to wait for either daylight and no cloud cover or night time and no cloud. Clouds really do reduce resolution and make it difficult to get images, even light haze can do the same.

You can change the orbit a small amount, but this takes a lot of very expensive and irreplaceable fuel to do so, so moving orbit means you have just taken years off the satellite life, as when it runs out of fuel it becomes unable to maintain attitude and starts to tumble. Consider that they typically have 50-200kg of fuel, and will use about 5kg a year for attitude control so that they have a 10 year life in orbit, and changing orbit by 1degree will use most of it up. At a good few million dollars just to launch, and with the launch being the cheapest cost, you do not want to toss it away for a few photos.

Google images has data from satellites in polar orbit, and these are arranged so that they image a strip of the entire earth every 90 minutes such that you get a complete map every few days, depending on the resolution you need. The unclassified images can have a resolution of 1m or better, and if you aim the camera ( which does not require fuel) at a specific area as you pass either over it or within sight of it in orbit you can get a better image. INMARSAT takes a real time image of an entire side of the earth as it orbits, but there you have 10km resolution at best.

In the middle of the Cold War the USA did launch satellites into orbit and those were used a few times, as they actually had film cameras on board that took photos and then dropped the film in a capsule to deorbit where they would be caught by a plane towing a catch hook. About 12 capsules then you had a big flying junk that was deorbited deliberately. Very expensive, and you could not have one "just lying around" but had to plan a launch so images would take up to a week to get. Also very visible when you launched it, and easy to detect if you wanted to see where it was, the USSR certainly knew when they were launched, and roughly which direction, so they could look for it with RADAR and see where it was looking. They then did things like have inflatable planes to confuse the image, and camouflage the launch areas during the passes.