[nl]

It's not explicitly stated but it seems that this island was not charted because the area it was in had previously been full of icebergs.

> On the satellite images analysed, the island could hardly be distinguished from the numerous icebergs drifting around in the immediate vicinity due to its ice cover.

[madaxe_again]

Also, it’s Antarctica. Observation at the poles is patchy compared to elsewhere on Earth. Low satellite coverage, illumination issues, lack of incentives, etc. - and that’s just satellite and aerial ops, never mind boots on the ground.

I was in the Antarctic about a decade ago, and this was underscored for me when we went to visit an island which has had maybe 20 humans visit, total - only to find it wasn’t where any of the charts said it was - it was about 3 miles away.

Fortunately we could just see it, as we had fine weather - which, upon further reading, neither of the previous surveys could, which explains the error - they had gone by dead reckoning in the era before GPS.

[arcticfox]

I have no interest in finding islands, but it seems like it would be pretty easy to find icebergs that never move.

[rtpg]

reading this I was wondering about this.

My sort of childlike mental model of satelite imagery of the planet is that we've "covered" everything but does anyone know at what frequency we do get new satelite imagery for places like the antarctic (or, say, the dead middle of the atlantic ocean?)

I imagine that satelite imagery is a bit needs based but maybe every square meter of the earth is captured at least once a couple of months

(Not the same thing but am reminded of how despite the importance of the internet and undersea cables for fixing things, there are _very very few_ boats that can actually repair them. Maybe there aren't that many satellites pointing at some parts of the globe)

[wongarsu]

Around the poles is a bit of a blind spot in satellite coverage. The angle with which the satellite orbit is offset from an equatorial orbit is called the orbital inclination. Because the earth rotates under the satellite, a 0° orbit would give you perfect coverage of the equator and not much more, a 10° orbit would give you good coverage of a band around the equator, etc. The closer to 90°, the more coverage you get of the northern and southern latitudes.

Now there's a neat trick you can pull where you go into a special 98° orbit (so like a 82° orbit, but in the other direction). At that point the slight bulge of the earth twists your orbit around just so that for any given point on earth you always pass over it at the same time of day, giving you identical shadows. That's called a sun-synchronous orbit, and is obviously immensely helpful for optical observations. But those missing 8 degrees prevent you from observing extreme latitudes. Usually we don't care because not much is happening there anyways

Even satellites without optical instruments usually suffer from the same blindspot. For example if you look at the Starlink constellation almost all satellites only reach up to about the middle of Great Britain. Everything further North is only served by a much smaller number of high inclination satellites. And there don't seem to be any Starlink satellites going directly over the poles

[cyberax]

> I imagine that satelite imagery is a bit needs based but maybe every square meter of the earth is captured at least once a couple of months

Probably, but likely not as thoroughly as you'd think.

The problem with most high-resolution imaging satellites is that they are not designed to work over the ocean. They can't track the Earth perfectly, so they use a lot of image processing to "unsmear" the images. These algorithms rely on tracking recognizable features moving across the frames. Which obviously fails with the ocean.

So you often get hilarious results with images of offshore drilling platforms or ships.

That being said, there are satellites specifically designed for ocean observation, so they likely won't miss something as big as a new island.

[bigiain]

I watched a youtube vid recently (so use that to calibrate your bullshit detector here) that said there are a bunch of companies and even freely accessible satellites with Synthetic Aperture Radar covering the entire earth every 12 days.

( https://youtu.be/UKLuei1CnZY )

[eru]

Well, they often have pre-discoveries for astronomical stuff. Where they find whatever they just discovered already being on old photographs. Why would satellite pictures be any different?

[DoctorOetker]

Such a search could still be run to predict new islands before they are discovered in the same way?

[ithkuil]

Perfect place for the fortress of solitude!

[Aardwolf]

Not if it's on a map anymore