Redshifts, what frequencies are they looking at?

[dimensionless]

I looked at a spectrogram of our sun several months ago. It contained a whole plethora of frequencies. It appeared that there were many thousands of frequencies in the visible spectrum alone.

When scientists investigate redshifts, do they look at a specific frequency, or do they just examine the distribution of spectral energy?

 

[Garth]

I looked at a spectrogram of our sun several months ago. It contained a whole plethora of frequencies. It appeared that there were many thousands of frequencies in the visible spectrum alone.

When scientists investigate redshifts, do they look at a specific frequency, or do they just examine the distribution of spectral energy?

If you can obtain a spectrum then you compare wavelengths of specific transitions. With some very distant and faint objects a spectrum as such cannot be obtained and the Planck spectrum itself has to surfice, if you think you know its efffective temperature.

One example of such an object is the Hubble ultra deep field object UDF033238.7-274839.8 aka HUDF-JD2 , a 6 x 10¹¹M_solar galaxy at an inferred z = 6.5.

Dan Stark, a graduate student at Caltech, attempted to pinpoint the galaxy's distance by obtaining detailed spectra using both the 10-meter Keck II and the 8m Gemini telescopes, but he was unable to do so. "We have now reached the point where we are studying sources which lie beyond the spectroscopic capabilities of our current ground-based facilities," he says. "It may take the next generation of telescopes … to confirm the galaxy's distance."

Instead, the team estimated the distance to HUDF-JD2 indirectly by looking at its energy at different wavelengths.
"It's not quite as convincing as a spectrum, and we estimate a 75-percent chance that we have the right distance," Ellis tells Astronomy. "There's no real prospect of getting a more accurate distance with current facilities."

Garth

 

[chroot]

They look at spectral lines. You can think of each element having a sort of "fingerprint" encoded in the spacings of its various spectral lines. This "fingerprint" can positively identify a specific element's emissions. When the spectral lines of the light from some known element from a distant astronomical object all appear at lower frequencies than samples of the same element in a laboratory, the redshift can be calculated.

- Warren

 

[ratfink]

We look at them all - and they are consistent: twice the wavelength gives twice the redshift from the Hydrogen line to visible

 

[SpaceTiger]

When scientists investigate redshifts, do they look at a specific frequency, or do they just examine the distribution of spectral energy?

Both answers given so far are correct in some circumstances. Most redshifts that are of high precision are obtained by examining the positions of spectral lines. We know exactly where the lines should be located in the rest frame, so we can calculate the redshift to within (or often better than) the resolution of the spectrograph.

Other times, when there is no spectroscopic data available or there aren't enough lines to identify the transitions they correspond to, we use the broadband spectral properties. This means, basically, that we compare the relative amounts of light of different colors (say the total blue light vs. the total red light) to what we would expect from a particular type of source. High-redshift sources (like quasars or galaxies) don't have Planck spectra, but still have relatively predictable spectral properties.

 

[matt.o]

We look at them all - and they are consistent: twice the wavelength gives twice the redshift from the Hydrogen line to visible

This is not quite correct. Redshift is defined by the following;

$$z\equiv \frac{\Delta\lambda}{\lambda}=\frac{\lambda_{obs}-\lambda_{lab}}{\lambda_{lab}}$$

So if the $$H\alpha$$ is observed at $$\lambda_{obs}=13126$$ angstroms (ie twice the laboratory wavelength), then the redshift of the observed object is z=1.

 

[Chronos]

Good catch matt.o. Math is good! To be fair, however, ST is talking about spectroscopy wrt faint galaxies. It is very difficult to sort out the 'lines' with all the intervening gas and dust.

 

[SpaceTiger]

Good catch matt.o. Math is good! To be fair, however, ST is talking about spectroscopy wrt faint galaxies. It is very difficult to sort out the 'lines' with all the intervening gas and dust.

Matt.o is correct, but it doesn't contradict anything I said (as best I can tell, anyway). I was just describing a method of approximating a redshift without access to the spectral lines, the definition he gave is still correct at any point in the spectrum.

 

[matt.o]

Hi Chronos. I wasn't contradicting anything ST was saying, just what ratfink said was slightly incorrect.

I am well aware of the difficulties in obtaining redshifts with weak to no lines! I have tryed to get redshifts out of crappy, cloud affected and under-exposed data before! Damn cloudy weather...

 

[ratfink]

This is not quite correct. Redshift is defined by the following;

$$z\equiv \frac{\Delta\lambda}{\lambda}=\frac{\lambda_{obs}-\lambda_{lab}}{\lambda_{lab}}$$

So if the $$H\alpha$$ is observed at $$\lambda_{obs}=13126$$ angstroms (ie twice the laboratory wavelength), then the redshift of the observed object is z=1.

In what way am i incorrect?

twice the wavelength gives twice the redshift from the Hydrogen line to visible

 

[ratfink]

In what way am i incorrect?

aH! i SEE WHAT YOU MEAN.

sorry, twice the wavelength gives twice the 'shift' or redshift. It all boils down to z is called 'redshift' and delta lambda is also called 'redshift'. I am using redshift 'delta lamba' as being proportional to wavelength here. 'z' is also known as redshift.
If a photon is redshifted it is redshifted. No probs.

 

[matt.o]

No, that is the way redshift is properly defined. I have never seen redshift defined as just $$\Delta\lambda$$.

 

[Chronos]

Apologies for the confusion. My comment was wrt what ratfink said, not ST.