This is the second in a series of posts I’m making on technology and engineering challenges.
Satellites are the bedrock of our technological civilization, and a vast array of technologies, such as GPS, map imaging, and telecommunications rely on the growing fleet of satellites in orbit.
There are many different satellites out there, the ones I’ll be covering are commercial imaging satellites and spy satellites in low earth orbit, I’ll talk about higher orbiting and geosynchronous satellites in a future post.
Satellites are usually arrayed in constellations, a network of satellites that can cover a significant portion of the Earth’s surface collectively, most Low earth orbit telecommunications satellites, such as Iridium, use this technique.
Imaging satellites, like their name suggests, primarily take images of Earth using a variety of technologies, these images are then transmitted to a receiving station ground side. Now you might say it can’t be that hard! why can’t you strap a gopro in an air tight box with an antenna and some solar panels and you got a satellite. Voila you now have a nano satellite, these, in fact, already exist, the most popular form being CubeSats from companies like Planet Labs, Nanosatisfi, and Skybox. Building a satellite is not rocket science, getting it to where it should be is.
These extremely small satellites have several major advantages like the obvious ones of cost, small size, and increased network resilience and harder to quantify ones such as rapid deployability, and very flexible launch vehicle requirements.
What I mean about flexible requirements is that a single nano sat doesn’t require a specific payload package, it could be stored in a small volume in the fairing (the cargo container on top of rockets) and deployed without the need for propulsion. Furthermore or containers of these satellites can be launched as one package and the satellites can then deploy in any number of ways from the container.
And now thanks to Moore’s Law each one of these small satellites can capture, process, and transmit more data than what all the Apollo spaceships, combined, could.
I believe that nano Sats can also be improved with several small gyroscopic flywheels that would act as momentum stabilizers in lieu of actual rocket motors, this would allow for even smaller, controllable, satellites. Somewhat analogous to the KERS system on F1 cars, but instead of transferring the kinetic energy to a drive axle, it would transfer the kinetic energy to increase the angular momentum of the satellite itself through a gimbal locking mechanism. Ideally 3 of these would be used for the 3 spatial dimensions.
The future of imaging systems, I believe, is arranging imaging satellites in constellations, allowing for continuous 24/7 coverage of the vast majority of the planet surface. Essentially an eye in sky that never blinks. This would be impossible, or at least require more money than exists in the world, if you were to try it with regular imaging satellites. However with nano satellites which could cost less than $10 000 each, and powerful cameras with a high megapixel count (like 4k cameras), could cover the entire earth from around 80 degrees south to 80 north. The resolution would be in the range of 5m per pixel, not terrific but more than adequate for general imaging. The reason for the latitude restriction economic in that the population, at the north and south pole is negligible so lowering the required coverage area by over 20% would save a lot of money, especially as high inclination polar orbits are more expensive to launch.
Some back of the envelope math:
Assuming the earth is perfectly spheroidal.
surface area of earth between 80 north and 80 south, a bounded zone, or quadrangle.
A = |2*pi*R^2(1-sin(lat2)) – 2*pi*R^2(1-sin(lat1))|
= 2*pi*R^2 |sin(lat1) – sin(lat2)|
= 392 192 104 km^2
At 5mx5m for one pixel, a 200×200 pixel camera could capture 1 square km.
Assuming an effective resolution of around 4000 x 4000 on the camera, each satellite could capture a 20km x 20km area, or 400 km^2.
Therefore we would need 980 481 cameras, or satellites to capture the entirety of the earth without any blind spots, in reality we would need some overlap as the orbits and position of the satellites aren’t mathematically perfect, as well as some extras to be deployed in event of micrometeorite impacts or debris disabling a camera. Let’s call it safe and say 1 million satellites are needed.
At $10 000 per satellite that’s $10 billion for the entire constellation. Conventional satellites would still be retained for on demand higher detailed images, but for most use cases, such as Google maps, a 5m per pixel resolution would be sufficient. And as imaging technology advances, higher megapixel sensors can be placed in the satellites allowing for higher detailed coverage with the same number of satellites. Plus expanding to the polar regions would simply be sending up more satellites.
Once the constellation is completed, stitching together a simultaneous image capture would produce an exact record of Earth at that moment, every cloud, plant, animal, person, car, airplane, and tree, on the planet, from the Himalayas to the smallest chunks of ice. Everything that exists on and above the surface of the earth would be in one picture.
Now how cool is that?