2

I would like to check if I understood well the proof of "$\frac{d^3k}{(2\pi)^3}\frac{1}{2 k_0}$ is Lorentz invariant".

My question is different from the others linked to this topic because I want to check the proof in details with all the steps.

I have an arbitrary function $f(k^0,\vec{k})$.

I can proove that : (*)

$$ \int \frac{d^3k}{(2\pi)^3}\frac{1}{2 \omega(\vec{k})} f(\omega(\vec{k}),\vec{k})=(2 \pi) \int \frac{d^4k}{(2 \pi)^4} \delta(k^2-m^2) \Theta(k^0) f(k^0,\vec{k}) $$


Indeed :

$$ \delta(k^2-m^2) \Theta(k^0)=\frac{\delta(k^0-\omega(\vec{k}))}{2\omega(\vec{k})} $$

And we thus have :

$$ (2 \pi) \int \frac{d^4k}{(2 \pi)^4} \delta(k^2-m^2) \Theta(k^0) f(k^0,\vec{k})=(2 \pi) \int \frac{d^4k}{(2 \pi)^4} \frac{\delta(k^0-\omega(\vec{k}))}{2\omega(\vec{k})} f(k^0,\vec{k}) \\=\int \frac{d^3k}{(2 \pi)^3} \frac{1}{2\omega(\vec{k})} f(\omega(\vec{k}),\vec{k})$$


We also can proove (I will not do this here) that $$ \frac{d^4k}{(2 \pi)^4} \delta(k^2-m^2) \Theta(k^0) $$ is Lorentz invariant.

Thus, if we do a Lorentz transform : $k=\Lambda k'$, we have, with $g(k'^0,\vec{k'})=f(\Lambda(k'^0, \vec{k'}))$, we have according to (*) :

$$ (2 \pi) \int \frac{d^4k'}{(2 \pi)^4} \delta(k'^2-m^2) \Theta(k'^0)g(k'^0,\vec{k'}) = \int \frac{d^3k'}{(2\pi)^3}\frac{1}{2 \omega(\vec{k'})} g(\omega(\vec{k'}), \vec{k'})$$

Thus, if we have $\vec{k}=\vec{\Lambda k'}$, we have for any function $f$:

$$ \int \frac{d^3k}{(2\pi)^3}\frac{1}{2 \omega(\vec{k})} f(\omega(\vec{k}),\vec{k})=\int \frac{d^3k'}{(2\pi)^3}\frac{1}{2 \omega(\vec{k'})} f(\Lambda(\omega(\vec{k'}), \vec{k'}))$$

I just used them l.h.s of (*) with replacing the r.h.s with the line calculated above.

And it is finally equal to : $$ = \int \frac{d^3k'}{(2\pi)^3}\frac{1}{2 \omega(\vec{k'})} f(\omega(\vec{\Lambda k'}), \vec{\Lambda k'})))$$

Because $f(\Lambda(\omega(\vec{k'}), \vec{k'}))=f(\omega(\vec{\Lambda k'}), \vec{\Lambda k'})$

Which proove that $\frac{d^3k}{(2\pi)^3}\frac{1}{2 k_0}$ is Lorentz invariant.

StarBucK
  • 1,350
  • 3
  • It is related but I really want to check if I understand the integrality of the proof. – StarBucK Mar 26 '17 at 12:24
  • To directly check it just note that you have length contraction for the spatial boost by a factor of $\gamma$ but have time dilation for time by the same factor so the $\gamma$ factors cancel out – Amara Mar 26 '17 at 13:18
  • In this case, the factor $\frac{d^4k}{(2 \pi)^4} \delta(k^2-m^2) \Theta(k^0)$ signifies the fact that the integral is for a single massive particle, which inhabits the hyperplane of the $4-k$ plane where $(k^{0})^{2}-(\vec{k})^{2}=0$ .

    The term $d^4k$ in itself is Lorentz covariant, and so is the term $\delta(k^2-m^2) \Theta(k^0)$.

    – Aritro Pathak Mar 26 '17 at 13:32
  • It seems you've got the gist of the proof right and all that needs sorted out is the details in intermediate steps. Also, not relevant to the question as a whole, but it may be useful for you to know that it is "prove," not "proove" (not trying to be pretentious, just saying that this will help scientific communication in the future). – Bob Knighton Mar 26 '17 at 13:34
  • I'm voting to close this question as off-topic because check-my-work questions are off-topic. – ACuriousMind Mar 26 '17 at 15:35

0 Answers0