Author Topic: Discussion: Green Peas: A Class of Compact Extremely Star-Forming Galaxies  (Read 24910 times)

Alice

  • Global Moderator
  • Hero Member
  • ****
  • Posts: 31782
    • View Profile
A different question, where did "green" come into the name? They're just "peas"!

Mark OConnell

  • Sr. Member
  • ****
  • Posts: 418
    • View Profile
So basically such a huge difference means the galaxy is highly defined or has precise edges?

It means that the emission line is very powerful compared to the underlying continuum of starlight.

Ok then. So alot of one element.

Rick Nowell

  • Hero Member
  • *****
  • Posts: 1333
  • Bristol, U.K.
    • View Profile
And what about the most distant OIII's, out to z=0.9? Are they just more distant Green Peas or something else?

I venture that the 5 nights Steven Bamford and Seb Foucaud have secured at the European Southern Observatory
in November in Chile might well go a long way to helping answer that question.

One of the comments by the European Space Agency noted that this was somewhat speculative, but as some very
interesting results might be obtained, they gave the time. FIVE nights!
http://www.galaxyzooblog.org/2009/07/03/more-peas/
http://www.eso.org/public/

What might be interesting would be to compare the dataset Seb Foucaud has with the list of Peas of varying colours and
redshifts that Starry Night put together here:
http://www.galaxyzooforum.org/index.php?topic=3638.msg115486#msg115486

I guess that the Peas must have 'evolutionary ancestors', as it seems unlikely that they would appear out of nothing. In
Carrie's paper, it is speculated that "These knots may be different star-forming regions, suggesting a morphology typical
of merger events" (page 6). Some better images of lots more Peas would certainly help a more certain ancestral heritage
be established. Why these small low-density galaxies are extremely star-forming when compared to other mergers, is a
question better answered by someone else other than me!

starry nite

  • Hero Member
  • *****
  • Posts: 1783
  • Pea-Wrangler
    • View Profile
I posted a comment on zooblog asking if they'll be looking at some of our "red peas" as well, but I haven't seen a response. Of course, I know they're busy people, but I hope they look at the really distant ones with clean charts we've found.
Good news everyone!

NGC3314

  • Global Moderator
  • Hero Member
  • ****
  • Posts: 1590
    • View Profile
So basically such a huge difference means the galaxy is highly defined or has precise edges?

It means that the emission line is very powerful compared to the underlying continuum of starlight.

Ok then. So alot of one element.

Sometimes. More precisely, strong emission from one atomic species (yes, that sounds like a tautology). The strength of an emission line depends not only on how many atoms there are of that species, but on how easily they are excited to radiate in the transition you're observing. That in turn can be changed by the temperature, density, and radiation field. For [O III] (emission frm O++) in particular, there is a cool paradox that to a certain point, the stronger the [O III] emisison, the less O++ there is. How can this be? The oxygen provides an important part of the processes that cool the gas, and these emission lines get stronger at higher tempetures. Less cooling, balance point reached at a higher temperature, and that effect compansates for having fewer oxygen atoms until you look at things down below 10% of the Sun's abundances. (This is related to the calculations that Cecilia Payne-Gaposchkin did almost a century ago, showing that the Sun and other stars are mostly hydrogen no matter which lines dominate their visible spectra).
« Last Edit: August 03, 2009, 06:49:34 pm by NGC3314 »

Rick Nowell

  • Hero Member
  • *****
  • Posts: 1333
  • Bristol, U.K.
    • View Profile
To quote Carrie's paper:  "We note that for many of our objects near z=approx 0.3, sky lines fall on top of the [OIII] line
and the Hbeta line, and we removed all of these objects from our sample."

What are 'sky lines', and why does this mean that over half the original 250 should be dropped
from the final analysis and list?

EigenState

  • OotD posters
  • Hero Member
  • *****
  • Posts: 1334
    • View Profile
Greetings,

To quote Carrie's paper:  "We note that for many of our objects near z=approx 0.3, sky lines fall on top of the [OIII] line
and the Hbeta line, and we removed all of these objects from our sample."

What are 'sky lines', and why does this mean that over half the original 250 should be dropped
from the final analysis and list?

I would interpret "sky lines" to be contributions from the Earth's atmosphere--light pollution if you prefer.

The analysis depended upon quantitative comparisons of certain emission line intensities, including both O[III] and Hβ.  Correcting those line intensities for contributions from the Earth's atmosphere introduces uncertainties in the desired line intensities which propagate throughout the entire analysis.  Such uncertainties are good reason to discard those objects because they constitute unreliable data points.

Best regards,
EigenState

zookeeperKevin

  • Galaxy Zoo Keeper
  • Admin
  • Hero Member
  • *****
  • Posts: 2020
    • View Profile
Yep. Sky lines are the `glow' of the atmosphere. If they are right on top of an emission or absorption line that you want to measure, then they can make it difficult or even impossible.


starry nite

  • Hero Member
  • *****
  • Posts: 1783
  • Pea-Wrangler
    • View Profile
Airglow lines are vertical purple dashes on the spectral charts.
Good news everyone!

NGC3314

  • Global Moderator
  • Hero Member
  • ****
  • Posts: 1590
    • View Profile
Here's an example. This is the two-dimensional spectrum of a familiar object, with wavelength running from bottom to top and distance along the spectrograph slit left to right:



NIght-sky emission lines (airglow and light pollution) are uniform along the slit. When we use this spectrograph mode, we can interpolate along the slit across the objects of interest to get a clean spectrum:



Notice how much better-defined the spectral features have become, especially near the bottom where the sky lines are relatively stronger.

The SDSS doesn't use slits, but hundreds of individually placed optical fibers, to collect data on large numbers of widely spread objets at once. This limits the accuracy of its sky subtraction, so that spectral lines falling right on the wavelength of the strongest airglow lines (5577, 5893, 6300 A) are suspect unless you inspect each data set in detail.



Rick Nowell

  • Hero Member
  • *****
  • Posts: 1333
  • Bristol, U.K.
    • View Profile
Thanks for the example. I recognise the concept of airglow, but sky lines is not a term I'm familiar with. Getting more
familiar with the paper now after constant reading!

Metallicity is an important concept to get right I feel. Astronomers regard everything after Hydrogen and Helium as
Metals: this can be hard to reconcile if one thinks of Oxygen as a metal, which for me is counter-intuitive. As the
Peas have a lot of ionised oxygen, does this not mean that they have a high metallicity? As I'm not familiar with
with the method involving log[O/H]+12 to estimate metallicity, perhaps that could be explained a tad.

Reddening is also somewhat mysterious: Is the reddening caused by high-dust content (the Peas don't have much
dust?), or by the redshift? Surely they are green because of the redshift of the [OIII] emission line being moved into
the r-band filter. Having said that, I remember Fermat's Brother writing that if one was next to a Pea, it would
appear red. (An explanation of the constants (?) E(B-V) would be appreciated).

So, if one was looking at the Peas from very near so that there was no redshift, what colour would they be? Obviously
not green- they are not intrinsically that colour. For a while when they first came to light, I thought (incorrectly) that
ionised oxygen was responsible for some of the greeness- electrical sparks can generate a green light as the Oxygen
gets ionised. Red to me would suggest 'age'- something the Peas lack. Confused...

As an aside, this from Carrie's list is the most star forming: 58 solar masses a year, which is a lot for any object
I would guess. http://cas.sdss.org/astro/en/tools/explore/obj.asp?id=587728906099687546

zookeeperChris

  • Admin
  • Hero Member
  • *****
  • Posts: 723
    • View Profile
Thanks for the example. I recognise the concept of airglow, but sky lines is not a term I'm familiar with. Getting more
familiar with the paper now after constant reading!

You and me both!

Metallicity is an important concept to get right I feel. Astronomers regard everything after Hydrogen and Helium as Metals: this can be hard to reconcile if one thinks of Oxygen as a metal, which for me is counter-intuitive. As the Peas have a lot of ionised oxygen, does this not mean that they have a high metallicity? As I'm not familiar with with the method involving log[O/H]+12 to estimate metallicity, perhaps that could be explained a tad.

If you think it bothers you to have Oxygen as a 'metal', think what it does to the poor chemists I work with! [O/H] is shorthand for 'Take the abundance of Oxygen and divide by the abundance of Hydrogen'. The idea is that a low metallicity system is one which has loads of hydrogen and not much of anything else and vice versa. The '+ 12' is just a conventional constant to make the numbers nicer to deal with.

As for the peas having a lot of ionized oxygen, it's probably fairer to say that the Peas are low metallicity, and so have relatively less oxygen, but what is there is much more ionized than we might expect. (Like walking into a party and noticing that only 5% of the people there have black hair, but all of them are wearing green t-shirts).

Reddening is also somewhat mysterious: Is the reddening caused by high-dust content (the Peas don't have much
dust?), or by the redshift? Surely they are green because of the redshift of the [OIII] emission line being moved into
the r-band filter. Having said that, I remember Fermat's Brother writing that if one was next to a Pea, it would
appear red. (An explanation of the constants (?) E(B-V) would be appreciated).

When astronomers talk about 'reddening' we don't mean redshift, but rather the effect of the dust. Light is absorbed and re-emitted by dust grains, but it gets reemitted at a longer wavelength than it was initially. To measure this we look at the relative change from the 'B' band to the 'V' band - E(B-V).

Karen is working on a paper which tries to measure the effect of reddening on Zoo spirals.

So, if one was looking at the Peas from very near so that there was no redshift, what colour would they be? Obviously
not green- they are not intrinsically that colour. For a while when they first came to light, I thought (incorrectly) that
ionised oxygen was responsible for some of the greeness- electrical sparks can generate a green light as the Oxygen
gets ionised. Red to me would suggest 'age'- something the Peas lack. Confused...

As an aside, this from Carrie's list is the most star forming: 58 solar masses a year, which is a lot for any object
I would guess. http://cas.sdss.org/astro/en/tools/explore/obj.asp?id=587728906099687546

I'll need to think about the colour question - Bill will probably have an instinctive answer. Just a quick reference - the Milky Way forms stars at a rate of 1-3 solar masses per year, so 58 is indeed huge!
« Last Edit: August 06, 2009, 01:58:55 pm by zookeeperChris »

Rick Nowell

  • Hero Member
  • *****
  • Posts: 1333
  • Bristol, U.K.
    • View Profile
Thanks for the explanations. What colour they are as if one was next to them is an intriguing and
necessary question. How the human eye would see them would be interesting- the many headlines
that say GZ has discovered 'green' galaxies is, in a way, misleading, as they are not actually that
colour, I would guess.

As a ballpark figure, if one takes the average mass of a Pea as around 9.5 etc, then the galaxy that
I gave as the highest star-former in Carrie's list is only slightly more than that. So, if the mass of that
Pea is a 100th of the Milky Way (an average figure as quoted in the paper), then if this Pea was the
same mass as the Milky Way, the Pea would be forming 5800 solar masses a year (not that it works
that way), but... Divide this by 3, the rate at which the Milky Way forms solar masses, then We can
see that this Pea forms stars at around 1900 times the rate of our Milky Way. I think this is right...
Our galaxy is pretty typical, although it is old at around 13 billion years.

How old is the average Pea I wonder...
http://cas.sdss.org/astro/en/tools/explore/obj.asp?id=587728906099687546

Alice

  • Global Moderator
  • Hero Member
  • ****
  • Posts: 31782
    • View Profile
According to Budgieye's spectra tutorial, wouldn't they be red if you saw them in optical light? (Which is so ironic, since red is supposed to mean no star formation :D)

NGC3314

  • Global Moderator
  • Hero Member
  • ****
  • Posts: 1590
    • View Profile
Some of them might actually look green if we were sitting next to one (well, at a respectful distance of a few tens of thousands of light-years, but close enough). The [O III] emission line at 5007 is pretty green, maybe a tint bluish; it's one of the easiest colors to see in nebulae (like planetary nebulae), because that wavelength is pretty close to where the human eye is most sensitive. The wild cards in perceived color are what other lines are strong, and surface brightness. If H-alpha is comparably strong, you might see something yellowish or orange due to its crimson color (although you'd need pretty strong H-alpha, come to think it through, since it's so red that or eyes aren't very sensitive to that wavelength). But the ones with [O III] ten times as strong at H-beta would probably look greenish to the eye. That is, if the light has high enough surface brightness for color vision, where the cone cells kick in; below a certain brightness per square degree, we generally only see in monochrome via the more sensitive rod cells. It's a bit ironic, after a the color discussion, that for redshifts around z=0.2, like many peas, the SDSS color mapping comes close to what we'd see with the redshift removed.