Hacker News Reader

Comment by shantara | Hacker News Reader

Any explanation how it affects the asteroid detection? It is not a manual sky search it was decades ago, but a completely automated process with a computer controlled telescope, orbit calculation software, check against a database of known objects and a submission of the new findings.

View on HN

"So far, ZTF science operations have not yet been severely affected by satellite streaks, despite the increase in their number observed during the analyzed period"

So it requires processing for mitigation, but isn't as dire as the title suggests.

It's probably seen more as a herald of things to come. SpaceX will not be the last private entity throwing things in orbit. There have already been concerns about the amount of space debris that currently exists and coordinating it to avoid costly and time consuming repairs/replacements (or in the case of ISS, life threatening) accidents.

It's probably seen more as an opportunity to acknowledge a problem early to try and prevent issue before the problem grows too large and to consider mitigation strategies now.

From the paper linked in the other comment:

>approximately 4 × 10−4 of all pixels would be lost over the course of a year.

I don't believe this is a significant enough impact

If you have to remove the streaks, you're losing data behind those streaks. It's possible that in that frame, you miss the data that would be needed.

Of course an expanding space economy is just what we need to be able to survive an asteroid on a collision course.

Those streaks are what, 0.0000001% of the sky? Probably less?

It’s like if there were just a few thousand cars and boats on the entirety of planet earth, evenly spaced out across the land and sea. And then someone pipes up to complain about the heavy traffic, or the constant sound of horn honking.

The problem isn't so much missing pixels, its that bright satellites raise the noise floor of the image significantly, the more satellites in shot, the worse it is.

Its the same reason telescopes can't just "filter out" stuff like city lights and human radio sources.

The satellite streaks affect a limited number of pixels. The optics are not perfect, so a bright object can smear light over all the pixels of the sensor (noise floor).

Do you have any facts to show that the streaks raise the noise floor significantly (say more than 10%)?

I would guess from your lack of hard information that you are just suggesting it could potentially be a problem.

Edit: for a quick and dirty calculation, I estimated the pixels on a horizontal line in the image in the article, ignoring the galaxy, and the streak is not significant compared with the light from the stars. I presume because the satellites move fast across a slow time-lapse image capture, your hypothetical problem is not actually significant.

Your metaphor is poor. Imagine 2000 vehicles travelling at 25000 km/h across your location ocassionally from seemingly random directions.

Orbital period is ~90 minutes, and the surface area of the planet is ~510 million km2, so they would only occasionally pass you closely. The Doppler effect would prevent you hearing the horn, but the sonic booms would be rather noticeable (the speed of sound is 1080 km/h).

Your metaphor is poor. Imagine 2000 vehicles travelling at 25000 km/h across your location ocassionally from seemingly random directions.

But are they actually random? I thought satellite orbits (even LEO constellations like Starlink) were well know and tracked?

To a human observer, the orbits may be 'seemingly random', but a computer should know if a Starlink satellite crossed its field of view.

Incidentally, even to a human observer they are very far from random (though they aren't entirely regular either).

Go to https://droid.cafe/starlink (disclaimer: my site), click the gear icon in the top right, and drag the speed slidebar all the way to the right to see...

Seems a little silly. Streaks are small and a tiny part of an image. So long as the satellites don’t block a constant part of the sky, if they block < 1% of the frame you can probably just ignore the problem entirely. Assume that by chance anything missed on photo n will be captured on n + 1.

I’m assuming it’s totally fine to delay asteroid detection by one search with regards to whatever we can do about asteroids.

Also I thought that Sentinel was doing most of the near Earth asteroid work.

https://en.wikipedia.org/wiki/Sentinel_Space_Telescope

EDIT: Oops, I was actually thinking of WISE but found the wrong thing Googling. https://en.wikipedia.org/wiki/Wide-field_Infrared_Survey_Exp... but apparently that isn't nearly as important as land based surveying.

"This article is about the cancelled Sentinel Space Telescope for detecting asteroids."

Don't think it's doing a lot of monitoring honestly.

Reply on news.ycombinator.com