Comment by thunderbird120 | Hacker News Reader

Comment by thunderbird120 | Hacker News Reader

Parent Full thread

Satellites have essentially fixed and predictable orbits, are hard limited in how close they can get to targets, and soft limited by realistic payload limitations of satellite constellations.

You can have an extremely high end satellite in LEO but it may only pass over the desired target once per day. You can put it in GEO and have it see the whole area but it's going to be so far away it's going to be limited in what it can see. You can build more satellites so that one of them passes over areas of interest more frequently but you can't make a Starlink sized reconnaissance satellite constellation if you're putting super high end hardware in each one. Those sensors cost tens to hundreds of millions of dollars each.

Reconnaissance aircraft are a dramatically more economical way of finding mobile targets than massive LEO constellations of super high end sensors. Instead you just put the high end sensor in an aircraft and have it overfly the area you're trying to surveil. This is why many countries are dramatically increasing investment in them. Actually knowing where things are is critical to destroying them.

> but it may only pass over the desired target once per day

Plus the people you're surveilling might discover its timing, and just put everything in garages or under tarps at the necessary times.

If you want to somehow alter the timing, you have to burn irreplaceable fuel and reduce the lifetime of the satellite-investment.

> Reconnaissance aircraft are a dramatically more economical

I really don't see how they could be reality. Do you have any specific arguments or sources for this claim?

> Those sensors cost tens to hundreds of millions of dollars each.

What sensors are you talking about that cost hundreds of millions and requires multiple?

Very high resolution cameras.

I think the fundamental misunderstanding here is about use case. A SR-71 successor would not be snapping pictures of static known sites where nothing is changing. That can happen at a leisurely pace which means that satellites do that fine. Satellites and reconnaissance aircraft do different things. There are many cases where you need to go look at a specific area RIGHT NOW, not in 7 hours. You can't wait for your high end LEO satellite's orbit to pass over the correct spot of the earth. Reconnaissance aircraft can actually maneuver and fly over specific places making them dramatically more flexible when you need to find something that may be in a completely different place 1 hour from now.

>What sensors are you talking about that cost hundreds of millions and requires multiple?

Nearly all sensors worth putting on a satellite in a military context are going to cost at least $10 million and those would be pretty low quality and of limited usefulness. The NRO regularly launches individual satellites with sensors which cost >$1 billion dollars. Their yearly budget is >$6 billion despite having almost no personnel. High end sensors are very very expensive. You need multiple because, as previously stated, if you only have 1 you might be waiting an entire day for it to orbit over the correct place on the earth. You can build 1,000 of those and get near real time tracking but it's going to be absurdly expensive. Much cheaper to just fly a single aircraft over the correct part of the earth you're interested in. The sensor payloads that RQ-4s carry still cost >$100 million but you can just fly them exclusively over areas of interest meaning you don't need nearly so many of them. They're also a lot closer to the ground meaning that an equivalent quality sensor can see a lot more than it would be able to from LEO.

Reply on news.ycombinator.com