News for hackers. Stuff that matters.
Re:Compasses to point true north for first time in 360 years... by BurningFrog on Friday August 30, 2019 @11:22AM (#20839472)
No, it's almost always roughly aligned with the axis of Earth's rotation. I believe that's because it's related to the rotation of the internal molten iron core.

Though every few million years, the north and south magnetic poles swap, for unclear reasons.

Re:Compasses to point true north for first time in 360 years... by double0jimb0 on Friday August 30, 2019 @11:35AM (#20839603)
We are also about 350,000 years overdue for a magnetic pole swap. [1]

Also, it cycles at 300,000-400,000 years not millions.

[1] https://futurism.com/earths-magnetic-poles-overdue-switch

Re:Compasses to point true north for first time in 360 years... by 24gttghh on Friday August 30, 2019 @01:37PM (#20840714)
Looking at a map someone else linked to, the pole has shifted drastically in the last 50 years. Could that be an indication that it is in the process of shifting? It's been moving North in a straight line for the past 100 years, and now it's actually heading South towards Siberia.

https://maps.ngdc.noaa.gov/viewers/historical_declination/

Re:Compasses to point true north for first time in 360 years... by BurningFrog on Friday August 30, 2019 @05:11PM (#20842750)
> Could that be an indication that it is in the process of shifting?

It could be. Since we've never seen a shift, we don't really know how it looks.

And a shift takes a few thousand years, so we can probably be centuries into it before it's clear what's happening.

Re:Compasses to point true north for first time in 360 years... by jobigoud on Friday August 30, 2019 @05:56PM (#20843094)
Ah, too bad, it would have been great drama if it flipped instantly! Or even better, oscillated a few times over the course of a few years before settling in.
Re:Compasses to point true north for first time in 360 years... by BurningFrog on Saturday August 31, 2019 @12:22AM (#20844595)
I mean, for all we know it might oscillate like that.

I doubt the geological record is fine grained enough to tell fast oscillations apart from a slow drift.

Re:Compasses to point true north for first time in 360 years... by BurningFrog on Friday August 30, 2019 @11:50AM (#20839721)
Thanks for the timing correction!

I hoped "every few million years" was vague enough that I didn't have to look it up, but sometimes you gamble and lose :)

Re:Compasses to point true north for first time in 360 years... by inflatableDodo on Friday August 30, 2019 @12:10PM (#20839917)
Also, the swap apparently takes ages and during that time you have loads of magnetic poles floating about and the ionosphere dips by 80% or so. We currently have a little north pole floating around the south Atlantic. https://en.wikipedia.org/wiki/South_Atlantic_Anomaly
Re:Compasses to point true north for first time in 360 years... by bmm6o on Saturday August 31, 2019 @01:30PM (#20847626)
Is it really a "little north pole"? I don't know anything about it beyond what's in the WP article, but that doesn't sound like an accurate description.
Re:Compasses to point true north for first time in 360 years... by inflatableDodo on Saturday August 31, 2019 @04:11PM (#20848539)
I reckon it isn't that inaccurate, especially given the reported behaviour of compasses.

"From these models and extrapolating down into the Earth, it is known that regions of reversed flux at the core-mantle boundary have grown over time. In these regions the compass points in the opposite direction, in or out of the core, compared to that of surrounding areas. It is the growth in area of such a reversed flux patch under the south Atlantic that is primarily responsible for the decay in the main dipolar field. This reverse patch is also responsible for the minimum in field strength called the South Atlantic Anomaly, now centred over south America. In this region energetic particles can approach Earth more closely, causing increased radiation risk to low Earth orbit satellites."

http://www.geomag.bgs.ac.uk/education/reversals.html

Re:Compasses to point true north for first time in 360 years... by tialaramex on Friday August 30, 2019 @11:55AM (#20839790)
Does anybody have a link about that first part?

"Almost always" is a term of art in probability theory, where it means p = 1 despite the existence of possible outcomes that defy it. For example if you keep tossing a (fair) coin it will /almost always/ eventually come up "Heads". But I guess here you mean something else and I'm not sure what.

Re:Compasses to point true north for first time in 360 years... by inflatableDodo on Friday August 30, 2019 @12:06PM (#20839879)
It isn't just a theoretical supposition, I believe we measure alignment of magnetic fields in igneous rocks and the 'almost always' comes from that, the current field of the earth being fixed in the rock when it solidifies.
Re:Compasses to point true north for first time in 360 years... by Dylan16807 on Friday August 30, 2019 @04:06PM (#20842152)
> "Almost always" is a term of art in probability theory, where it means p = 1 despite the existence of possible outcomes that defy it.

That's a somewhat misleading term! I much prefer "probability one".

> But I guess here you mean something else and I'm not sure what.

Read it literally. The everyday definition of "almost". In other words, "the vast majority of the time".

Re:Compasses to point true north for first time in 360 years... by Someone on Saturday August 31, 2019 @12:23PM (#20847255)
I agree this should be read as “the vast majority of the time”, but if one wants to be precise, it is unavoidable to deviate from everyday definitions (is 99% the vast majority of the time, or does that need to be 99,9%, 99,99%,…?)

“probability one” is misleading, too, as laymen would assume the reverse cannot happen.

The interpretation of “almost always” as ”p = 1 despite the existence of possible outcomes that defy it” likely is based on http://mathworld.wolfram.com/AlmostEverywhere.html (which has “almost surely” as the mathematical expression, but I would immediately read “almost always” with the same meaning)

Re:Compasses to point true north for first time in 360 years... by Dylan16807 on Sunday September 01, 2019 @12:28AM (#20850477)
> “probability one” is misleading, too, as laymen would assume the reverse cannot happen.

In the real world it cannot happen. It can only happen in an unbelievably unlikely mathematical abstraction. So I would call that layman understanding closer to the truth.

Re:Compasses to point true north for first time in 360 years... by learnstats2 on Saturday August 31, 2019 @06:29AM (#20845542)
I would assume that "almost always" is a common English phrase that has been adopted by probability.
Re:Compasses to point true north for first time in 360 years... by Dylan16807 on Saturday August 31, 2019 @08:35AM (#20845932)
Sure, but the more accurate English phrase for that meaning is more like a simple "always".

"almost always" is more like a 95% threshold, sometimes higher but with an implication that failure is realistic.

Re:Compasses to point true north for first time in 360 years... by learnstats2 on Saturday August 31, 2019 @06:24AM (#20845534)
Perhaps. However - "the magnetic axis of Uranus, for example, is inclined by as much as 60°"