4
$\begingroup$

This is a naive question. I have a temperature map of radio emission at 1420 MHz, in Kelvin. I want to extract the flux density (in Janskys) from an extended region of this map. Is the flux density ($S_{\nu}$) from this region equal to $$ S_{\nu} = \frac{2\nu^2 k_B T_{\text{mean}}}{c^2} \times A$$ where $\nu$ is the frequency of my map, $k_B$ is the Boltzmann constant, $T_{\text{mean}}$ is the mean value of the temperature inside my region, $c$ is the speed of light and $A$ is the area of my region in steradians?

$\endgroup$
4
$\begingroup$

This expression is valid for low frequencies, including the case of the 1420 MHz hydrogen line. It arises from treating the source as a black body with temperature $T(\theta,\phi)$$^{\dagger}$, and assuming $h\nu\ll k_BT$: $$I_{\nu}=B_{\nu}=\frac{2h\nu^3}{c^2}\frac{1}{e^{h\nu/k_BT}-1}\approx\frac{2h\nu^3}{c^2}\frac{1}{h\nu/k_BT}=\frac{2k_BT\nu^2}{c^2}$$ where we have made the approximation $$e^{h\nu/k_BT}\approx1+\frac{h\nu}{k_BT}$$ and I drop the explicit $(\theta,\phi)$ dependence. The flux density is then, by definition (and assuming the source is small, so $\cos\theta\approx1$), $$S_{\nu}=\int I_{\nu}\cos\theta\;d\Omega\approx\int I_{\nu}\;d\Omega=\frac{2k_B\nu^2}{c^2}\int T(\theta,\phi)\;d\Omega\tag{1}$$ If we calculate the mean of the function $T(\theta,\phi)$ over the source, we get $$T_{\text{mean}}=\frac{\int T(\theta,\phi)\;d\Omega}{\int d\Omega}$$ and some algebra and noting that $\int d\Omega=\Omega$ should convince you that we can make the desired substitution. We can essentially always take the mean of a well-behaved function when integrating over a patch of sky.


$^{\dagger}$ The brightness temperature $T_b$ is typically defined as the temperature of a black body which would yield the same specific intensity.

$\endgroup$
2
  • $\begingroup$ Thank you for your detailed explanation. I am aware of the connection between density flux and brightness temperature and I understand the Rayleigh-Jeans law. I was wondering if I could replace $ \int I_{\nu} \cos\theta d\Omega$ with $I_{\nu} \times A$, where $A$ is the surface of my ds9 region and $T_{mean}$ is the mean temperature obtained from all the pixels inside my region. $\endgroup$
    – Tutan
    May 27 '20 at 8:07
  • $\begingroup$ @Tutan Ah, my bad - then the answer is certainly yes; we're treating $T$ as a function $T(\theta,\phi)$ and we can certainly take the mean of that function over the source - answer edited. $\endgroup$
    – HDE 226868
    May 27 '20 at 14:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.