The h Bar - 2020-09-16

« first day (3604 days earlier) ← previous day next day → last day (1317 days later) »

1:12 AM

1

Q: What is electron precipitation? Why would STEVE care?

Trying to decide if the unanswered question What was Steve? Is it called something else now? is in need of attention I checked Wikipedia's Steve (atmospheric phenomenon); Occurrence and cause which says: Occurrence and cause STEVE may be spotted closer to the equator than the aurora, and as of M...

atmosphere physics planetary-science steve

2 hours later…

3:33 AM

Scientific American has never endorsed a presidential candidate in our 175 year history—until now. The 2020 election is literally a matter of life and death. We urge you to vote for health, science and Joe Biden for President.

1 hour later…

4:41 AM

@uhoh this?

5:21 AM

Suppose I start a bounty and there is no answer eligible for automatic award at the end of the bounty and I am not satisfied with any answer to award the bounty manually. Will I be refunded the bounty back?

5:35 AM

Apple’s New Series 6 Watch To Perform COVID Test.

> “Apple has kept this new feature of rapid COIVD Test in its new series 6 watch considering the poor people who can’t afford COVID test” told the Apple Marketing manager to The Fauxy.

6:14 AM

@Azmuth fake news. The Apple watch can detect blood flow, and Apple are looking into whether this can be used to detect covid. But so far this is just an idea and may come to nothing.

The idea that people are too poor to afford a covid test but can afford an Apple watch seems unlikely to me.

Superfast Jellyfish

6:50 AM

@Azmuth hahahaha haha

7:04 AM

@JohnRennie I just now realized that this was a fake news and that was a satire website!

:P

My bad!!

@SuperfastJellyfish hheheheheh XD LMAO

It is true that Apple are looking into whether their watch can help with a covid diagnosis

Oh okay... I don't know about it.

Spoiler: it probably can't.

Okay... I knew it.

Otherwise it'd have been created already..

Though Apple might be on to something with the healthcare aspect of wearing an Apple watch.

7:08 AM

It is doing... (Oxygen levels of Blood and things like that)

For example atrial fibrillation is a very common in old people, and if untreated it can cause strokes.

@TonyStark No, you've paid with the bounty for the exposure, you don't get the rep back.

If the watch can detect when atrial fibrillation starts, and the wearer gets treatment for it then the watch really could save lives.

yep

don't need an Apple watch for that, there's already dedicated health monitors for old people who want that

7:11 AM

yap

The problem with a health monitor is that you have to make the decision to wear one.

If everyone wore Apple watches because they wanted to know the time then you get the AF detection as a free extra without any effort on your part.

As it happens my mother has atrial fibrillation and she only found out when it caused her to faint. Before that there had been no symptoms and it could have caused a life threatening stroke.

(Though she is unlikely ever to wear an Apple watch)

@JohnRennie fair enough - but then again how many of the people at risk here make the decision to wear a smartwatch in the first place?

Should I order iPad for the next stock?

@ACuriousMind I would guess it's a tiny percentage because smartwatches are a niche product and the older generation tend to be technophobic anyway.

the functionality as such doesn't seem to hurt anyone, but then again I'd personally be wary of giving a corporation like Apple a live feed to my bio data

7:17 AM

But that hasn't stopped Apple publicising the potential health benefits, and for once they may actually have a point.

For the record I have never known Apple to act for the good of anyone except Apple.

2

yep

agree

@JohnRennie that's it - the 'marketing is "it could save your parents lives!" but the cynic in me suspects it's just more data to aggregate. I can already see the algorithms being trained that correlate your heart rate to what ad you're currently watching to figure out which ones you really hate/like...

I shall make sure I take the watch off before navigating to PornHub.

Musk also started NuraLink

or something like that

a heavy machine that prevents depression.

@JohnRennie might at least throw off their data analysis if you keep it on :P

7:27 AM

twitter.com/bharathd/status/1305724384850583552?s=20

anyway, I'm not saying this would be a particularly powerful way to use such data but history so far shows that most data that is gathered will be used at some point and that "We promise we won't do anything with it" is usually a lie unless there would be painful penalties for doing so.

^That's what I want to do in life

I work in the data field so I can tell you that if they didn't need the data, they wouldn't bother gathering it

It is extremely annoying to do data collecting

@Azmuth in my wilder years I rode a motorcycle from Cambridge to Rome and back. That was a lot of fun if sometimes a bit scary.

I'll start from Kashmir

@JohnRennie Wow! I just want to do that from India to London

7:31 AM

That's a long motorcycle ride! :-)

Yep. my girlfriend's gonna company me (maybe XD)

Everyone see her: twitter.com/KanganaTeam/status/1306051324732088321?s=20

I'm already in love!

Her hair is best <3 :)

@Slereah now imagine living the dark ages where people had to enter data manually into computers!

now at least your "customers" mostly do that for you already :P

If you are not paying for the product, you are the product!

@Azmuth Welcome to stackexchange.com

@FadedGiant Hahaha... Hilarious! LMAO

That cracked me like hell!

BEST!

7:41 AM

I don't like that platitude because it suggests you're always being duped when something is freee. But e.g. Wikipedia is a non-profit that you can use without ads and without giving them any data you don't want to give them

wasn't expecting that... honestly...

It is possible to build free services dedicated to common good. It's just not what private corporations tend to do.

@FadedGiant That's a good example though. I have benefited enormously from being a member of this site and I'm happy for them to use what little personal data I've given them in return.

@JohnRennie same here, no difference XD

@JohnRennie SE doesn't want your personal data, they want your contributions to the site to attract more eyeballs. Not saying that is nefarious, but in this case "personal data" isn't the endgame, it's getting volunteers to provide content for them for free

7:44 AM

@ACuriousMind Data entry in the dark ages was a bit more primitive : en.wikipedia.org/wiki/Domesday_Book

(well, actually maybe they want your personal data but they don't try very hard at collecting it :P)

@ACuriousMind it's still fine by me. When I cease to get any benefit I will stop using the site.

user image

Ancient database

where do I put the sql

'OR' '='

'Select * from Ancient DB Where A*= '='

7:47 AM

@ACuriousMind just hire some monks

I post this in the certain knowledge that no-one under 50 will have the faintest idea what it's about, but it just made me literally laugh out loud:

@Slereah But will they optimize my query properly? My census needs maximal throughput!

user image

pls patch the monks they've invented beer and got drunk

@JohnRennie now you've gone full boomer - posting memes entirely inscrutable to the younger generation :P

I am officially an old fart.

7:51 AM

lol

can we not post "jokes" glorifying Putin (or any other authoritarian, really), please?

Okay

Please remove those... I'm afraid, he may see me...

I took them to be satire rather than glorification, but I agree that we should not be commenting on politics here.

I posted them as satire

shit! That timeout for delete button....

Is Putin authoritarian? A google search says that there are elections in Russia too...

@JohnRennie a lot of genuine glorification hides being "It's just a joke". Poe's law is in full effect - if you want to do satire, you need to be better than just looking exactly like the thing you're trying to make fun of

7:56 AM

@ACuriousMind Please remove that image

I can't delete that now

5 messages deleted

I didn't think it was so terrible as to need outright deletion, but if you wish...

Relax guys

It wasn't that bad

Superfast Jellyfish

8:20 AM

@JohnRennie typical boomer. Can’t accept the current terminology even while agreeing :P

:-)

Weren't "boomers" born just after WW 2?

Early 1950s

He is a Gen X

All about that MTV

@skullpatrol 1945 to 1964. I was born in 1961 so I'm just included.

I see, I was taking the average :-)

8:25 AM

Y'all watch those Rodies and MTV stuff?

Did they get there name from the baby boom that occurred after the soldiers returned from WW 2? @JohnRennie

@skullpatrol yes

So why the 19 year gap?

generations are long

Is someone 18 years younger or older apart of your generation?

8:33 AM

I assume it's because at 18, you can have your own children

Y'all kids, I've dated a girl 12 years elder than me.... '__'

Good for you?

Was that a question?

It was likely an attempt to hint that this information was completely irrelevant to the conversation and no one is interested in it.

We can't use tone here so it doesn't really matter @Azmuth :-)

8:41 AM

okay

I got the signals, so, I'm silent now

@Slereah bit weird to still use that as the reason when the average parent age in most developed nations is far higher than 18

@ACuriousMind I mean, same order of magnitude

1945 to 1964 were different times

I'm not sure that we don't actually do the generations more after relevant social changes - certainly the 60s with all the counter culture were sort of a water shed for generational experience

There aren't really any standards, anyway

8:49 AM

yup, generations X,Y, and Z do overlap

@Slereah but I need my neat little boxes to put people in

can't have them milling around

Everyone is put into a neat little box eventually

6

We are in generation COVID

@Slereah don't take life too seriously. You'll never get out of it alive anyway.

5

Just a question

What Netflix series y'all watch?

I want to peek into your suggestions....

I'll be from 9 PM, so....

9:36 AM

user image

That's sad :'-(

1 hour later…

10:41 AM

@JohnRennie that could be. I don't understand why it was important to discuss the relationship between STEVE and electron precipitation. A good answer will highlight the meaning of there being, or not being a connection.

3 hours later…

1:56 PM

Does Lorentz invariant mean it’s invariant under passive and active transformations?

I guess yea!

I told ya all, My chats in CSE are being tracked!

Now, I was taking about Putin and USSR here before a while

and this popped up in my Twitter feed: twitter.com/palkisu/status/1306229749929316352?s=20

2:08 PM

Did you watch "The social dilemma" yet?

@JakeRose Active and passive transformations are essentially equivalent so I would imagine so yes

Otherwise a bigger deal would be made about whether we're talking about active or passive transformations if lorentz invariance meant something different in either case

Got it thanks

2:35 PM

@uhoh I'm reluctant to post an answer because this isn't my area and I know only what that nice Mr. Google has told me. However it looks so me as if STEVE is the same physics as aurorae i.e. particles ionising atoms in the upper atmosphere and those atoms emitting light as they recombine with electrons.

Roughly speaking an active transformation is picking a new field $\phi'$, but a passive transformation is performing a diffeomorphism on your underlying spacetime $f$

The relationship between the two is that $\phi \circ f$ is equal to $\phi'$

The difference seems to be that we get aurorae only where the magnetic field lines dip into the atmosphere, i.e. near the poles, while STEVE occurs nearer the equator. The reason is that the ionisation is being done by electrons trapped around field lines and ejected from that trapped state by EM waves.

who's Steve

The suggestion in what I've read is that when the frequency of the EM wave matches the gyromagnetic frequency of the electron around a field line there is a resonant absorption and the electron can be ejected into the upper atmosphere with considerable energy. It then ionises atoms causing the glow.

The phenomenon called electron precipitation is this resonant absorption of energy by electrons rotating around field lines.

@uhoh But this all seems rather speculative to me. It is unclear from anything I've read that there is firm evidence for any of this.

@Slereah STEVE

Hi Steve

2:41 PM

A VETLA

Mohamed Obeidallah

Hi everyone,

Can a nuclear explosion generate powerful gamma rays in the gigaelectronvolt range? If it can, is it a fission or a fusion?

A quick google search suggests even fusion reactions are on the 200MeV energy scale so, maybe?

@MohamedObeidallah yes

In fact gamma ray bursters were discovered by a US satellite designed to detect the gamma rays from Russian nuclear tests.

scary times

Does "the diffeomorphism group" depend on the manifold in question?

I expect it does

@MohamedObeidallah No.

@Charlie 200 MeV for fusion? What kind of fusion is that?

2:52 PM

idk I just googled it lol

Mohamed Obeidallah

@FadedGiant can you give a detailed explanation on why it can't?

@Charlie about 200 MeV is the total energy released by fission, not fusion

@JohnRennie I think there are at least several distinct phenomena that fit that general description, it needs a physics-based answer but with attention to those distinctions

@FadedGiant Yeah after re-googling I see it's talking about fission, it matched the word from different parts of the article

fusion produces less energy right?

at least per nuclei

@MohamedObeidallah If you look at fission, maybe 204 MeV is the total energy release. So gigaelectronvolt range is right out.

Most of this is carried away as kinetic energy of the fission products, about 167 MeV.

The produced neutrons get about 5 MeV.

Prompt gamma radiation is about 6 MeV (not GeV).

Delayed gamma radiation of the fission products about another 6 MeV

plus about 8 MeV for the beta radiation of the fission products

and finally about 12 MeV for the neutrinos

Mohamed Obeidallah

2:59 PM

@FadedGiant ok, so if I'm following you a fission bomb generate 12 MeV gamma rays (6 Mev prompt gamma plus delayed gamma radiation). Correct?

a bit more if you have fast fission

@Charlie yes, about 17.6 MeV for a D-T reaction

Mohamed Obeidallah

then according to this paper https://ieeexplore.ieee.org/document/767415

"Hermes III is a 16 Terawatt, 22-MV, 730-kA, 40-ns pulsed power
accelerator, which resembled a short subway train in size and shape, that drives an electron beam diode/converter to
generate an intense pulse of bremsstrahlung radiation. In
Hermes III, eighty individual I.I-MV, 220-kA pulses are
generated and added in groups of four to develop twenty
I.I-MV, 730-kA pulses which are then fed through
inductively isolated cavities and added in series by a

3:29 PM

@FadedGiant can perhaps answer that.

22 MV bremsstrahlung? Yes, that's a lot.

But more lethal?
The effective dose per fluence is 15 pSv cm^2 for 6 MeV photons and 38.2 pSv cm^2 for 20 MeV photons.

But that doesn't really matter if someone is shooting with 16 terawatt.

Anyway, that still has nothing to do with the above-mentioned "gamma rays in the gigaelectronvolt range".

3:52 PM

@ACuriousMind I kept the text brief so that the reader would look at the example code. Is that not working for you?

As for floating point issues: they are a thing. Not much to be done about it.

You just can't really expect two circles made in different ways to be equal, but then again that would be true even if I did something like this:

c1 = Circle(radius=1)
c2 = Circle(radius=2/2)

@DanielSank After looking at the example I mainly wonder whether making the constrained class immutable is necessary or whether it would somehow uphold the constraint when I assign to one of the two values

Superfast Jellyfish

4:51 PM

@BioPhysicist it’s been on my watchlist for a week now. How is it?

5:08 PM

@ACuriousMind Unless thoroughly convinced otherwise I would tend to avoid that kind of consistency guarantee with mutable data. It's possible but... ewwwwww.

@SuperfastJellyfish I enjoyed it. I have had the view for a while that social media has greatly helped in all of the polarization of views we see today; the documentary definitely strengthened that view

If, for all abelian lie groups, their algebra has trivial commutator ($[J_i,J_j]=0$), and also if the structure constants derived from the Lie algebra of a group uniquely define it, does this mean all abelian Lie groups have identical Lie algebras? And so all abelian Lie groups are locally identical

@Charlie no, they can have different dimensions

ah

if that's the only remaining degree of freedom then all n-dimensional abelian lie groups are locally (isomorphic?) similar

The generators of translations e.g. in the plane or spacetime (for example as part of the Poincare group) are an abelian subalgebra

5:20 PM

@Charlie I don't know what you mean by "locally similar", but the only Abelian Lie groups are $\mathrm{U}(1)^n$ and $\mathbb{R}^n$.

oh, really? well that's boring :c

or mixtures of that, i.e. $\mathrm{U}(1)^m \times \mathbb{R}^n$

ok then

also I definitely mean isomorphic, since they are literally the same algebra i guess

Superfast Jellyfish

5:45 PM

@BioPhysicist great! I’ll watch it this weekend then.

Plenty you can do with $U(1)$!

2 hours later…

8:05 PM

If we have a gauge symmetry like $U(1)$ in the QED lagrangian, does this mean that the target space of each of the fields involved has to admit a representation of $U(1)$?

So the Dirac spinor space and Minkowski space admit a representation of $U(1)$ as well as a a representation of the Lorentz group?

1 hour later…

9:11 PM

Yes, even normal QM wave functions being invariant up to phase factors can be framed in these terms, 'sections of a complex $U(1)$ line bundle' or something

@Charlie Yes, but just as Abelian groups are boring, their representations are also boring!

the only irreducible reps are one-dimensional, so "admitting" a $\mathrm{U}(1)$ representation is not a restriction

In general, if you have an internal symmetry group $G$ and a representation $V_G$ of it and you want to talk about a vector (or spinor or whatever tensor) field transforming under that group in that representation, the field target space just becomes $V_G \otimes V_\Lambda$, where $V_\Lambda$ is the Lorentz representation space for the field

This sounds very complicated but in reality it just means that the field has two indices - one Lorentz index and one index for the internal symmetry, i.e. you'll see things like $A^{\mu a}$ for a non-Abelian gauge field where $\mu$ is the index in $V_\Lambda$ and $a$ is the index in $V_G$.

@ACuriousMind I'm not sure why it follow that this isn't a restriction

@Charlie every complex vector space admits non-trivial $\mathrm{U}(1)$ representations because the irreps are one-dimensional, so every vector space can be written as the sum of one-dimensional irreps in arbitrary ways. But this is irrelevant anyway because as I just said, you don't look for a representation of $G$ on the space $V_\Lambda$ but you just tensor it with a representation space of $G$

Yang-Mills e.g. Peskin Ch. 15 is a good example of this

that this is the unique way in which internal symmetries and spacetime symmetries combine is the content of the Coleman-Mandula theorem

9:19 PM

oh ok I see what you mean

(In mathematical terms, CM says the full symmetry group is the direct product of Lorentz and internal symmetry group, and you get representations of a direct product by tensoring representations of the factors)

Given what you've said about $V_G\otimes V_\Lambda$ why can't we tensor arbitrary representation spaces onto our lorentz representation space?

what do you mean we can't?

you can take any $V_G$ you want

So $V_G$ is arbitrary in principle but we choose it to be the representation space of some internal symmetry

I guess I'm surprised to see that this is a necessary step

well, this is kinda a circular definition: We call the group the "internal symmetry group" whose representation we tensor here ;P

9:25 PM

hmm

the content of "internal symmetry group" is that it's a symmetry that acts on the fields but is not the spacetime Lorentz symmetry

Ok in that case I don't see why gauge symmetries aren't artificial, if we're just tensoring on representation spaces to the target space of our field

hmm

but for such a symmetry to act on the field they need to be in a representation of it

Are we saying that in qcd for instance the target space of the field isn't a representation of SU(3), we tensor on a representation of SU(3)?

if so, that step seems artificial

all of physics is artificial

there are no Lagrangians growing on trees

9:28 PM

:c

why can't we choose the gauge symmetry of qcd to be something crazy like SU(9) if all we're doing is tacking on a representation space to the field

you can

you just get a different theory that doesn't describe the world we observe

the standard model is not unique in any sense

doesn't the symmetry have to arise somewhat naturally from the lagrangian though?

how do you write down the Lagrangian with that symmetry before you have fixed the representations (and hence indices) the fields transform with?

It feels like that's what's been done when building up to QED in the notes I've seen, we haven't gone out of our way to ensure the Lagrangian has U(1) symmetry, it just does

well, $\mathrm{U}(1)$ is special

its irreps are one-dimensional, so the tensoring does not increase the dimension of the target space

and so we do not get the additional index

9:31 PM

I thought the point was we find all of these nice representations of the Lorentz/Poincare group, build the fields out of them, construct the lagrangian and then find the lagrangian has some special symmetry like U(1) or SU(3) and go from there

For U(1) you can just say that now everything transforms with phases $\mathrm{e}^{\mathrm{i}\alpha(x)}$ that not think about the representational stuff

but once you go to non-Abelian internal symmetries you must do what I said above

@Charlie nope

aww

you can write down the generic Yang-Mills Lagrangian with matter for any compact Lie group as its gauge symmetry you like, and it always looks the same, just your choice of groups/representations is different

hmm

it's odd to me that we write down the lagrangian and then decide what symmetry it has

it seems like it would unavoidably work the other way round

well, not really - it's a bit circular

I can say that I have a field strength tensor in the adjoint of some group $G$ and work everything out without ever naming $G$

so I can write down the Lagrangian and even compute a bunch of stuff and then say "Now $G$ is $\mathrm{SU}(3)$, what happens?"

9:35 PM

People clearly write down a Lagrangian possessing the symmetries we expect

That's how it was found originally

^ that seems to exclude the symmetries that the lagrangian doesn't have

but you can also start from "I want a Lagrangian with $\mathrm{SU}(3)\times \mathrm{SU}(2)\times \mathrm{U}(1)$ symmetry where the list of particles is charged like [...], go find one"

The motivating idea is that the proton and the neutron's mass are almost the same, so people thought maybe the difference in mass is due to an electromagnetic field which if turned off gives them equal mass so the proton and neutron should be interchangable i.e. the theory should be invariant if you rotate them into one another in a way that preserves norms of complex wave functions i.e. the wave functions transform under $SU(2)$

So a choice of lagrangian doesn't uniquely identify a theory we also need to say what symmetries it has, even if we've already chosen a lagrangian

Making SU(2) a local symmetry took nearly 30 years to see as an idea

9:38 PM

and historically we had something like bolbteppa is hinting that that people were looking at the particle zoo and trying to find some ordering principle that made sense of it all, and other people were doing precursors to generic Yang-Mills and it all ended up converging towards QCD but not in the neat way we present Yang-Mills or QCD today

Really the symmetries are what determines everything, it's all we have, even in classical mechanics one can only fix the Lagrangian by invoking Galilean symmetry or Einsteinian relativity

@Charlie that depends on what you mean by "Lagrangian"

:P

To me, part of "the Lagrangian" is specifying what representations the fields transform in

I.e. the generic YM-Lagrangian $\mathrm{tr}(F\wedge F)$ is a different thing depending on whether $F$ transforms in the adjoint of $\mathrm{SU}(2)$ or that of $\mathrm{SU}(3)$

to me this seems like we're picking a geometric shape and then deciding what symmetries we want it to have

How

which seems like having your cake and eating it

if the choice of lagrangian and choice of symmetry of the lagrangian are independent choices

9:41 PM

I don't know what you mean by the geometric shape but all model building is, at the end, educated guesswork

The choice of Lagrangian is not independent of the choice of symmetry obviously

we don't derive the Lagrangian or Hamiltonian from divine law, we guess it and see if its predictions correctly match the observations. If they do, hurray, we've found "the Lagrangian"

oh, that is the impression i was getting

with the whole "tensor on the representation of the symmetry" stuff

@Charlie not all Lagrangians work for all symmetries

i'll leave it for now if this isn't going to create immediate problems :P

9:43 PM

what I said holds for the standard YM Lagrangian

Saying fields transform under both Lorentz transformations and something like $SU(2)$ means in some sense tensoring those groups, but you can basically ignore that

there are plenty of other forms of Lagrangians that are not so generic w.r.t. the choice of $G$

ah

if we have to choose a symmetry that the lagrangian actually has then that makes sense

I said that we can choose arbitrary $V_G$ because in my world I first pick the fields and then pick the Lagrangian. If in your world you first pick the rough shape of the Lagrangian and then look for fields you can stuff in there that's fine, too

that's just two different approaches to model building

Yeah you basically should pick the fields first and if they transform under some group a Lagrangian invariant under that group will produce dynamics covariant under that group

9:45 PM

I would pick the field first too that makes sense

Even in Newtonian mechanics, you are picking coordinates that transform under a symmetry group (Galilean) then building a Lagrangian invariant under it

it depends what you want to model - if you have a list of particles and charges and so on, you have more information about the representations on the field than about the terms in the Lagrangian. If you have more things like "I know two of these and one of these should be able to interact" then that's data about what terms you need in the Lagrangian

It just seems unfamiliar when you have to specify some extra group ('internal symmetry group')

Is the QCD Lagrangian U(1) invariant too?

$U(1)$ is a subgroup of $U(3)$

9:49 PM

oh right

ok

there's many different $\mathrm{U}(1)$s QCD is symmetric under :P

I'm afraid to ask, it would probably add to the confusion :p

there's infinitely many subgroups of $\mathrm{U}(3)$ isomorphic to $\mathrm{U}(1)$!

10:02 PM

screams internally

« first day (3604 days earlier) ← previous day next day → last day (1317 days later) »

Transcript for

Sep '2016

The h Bar

General chat for Physics SE (physics.stackexchange.com). For M...

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile