in this paper http://arxiv.org/pdf/1412.1626.pdf it says that Lemma 3.1/(3.1) follows from Theorem 1.3 in http://arxiv.org/pdf/math/0402192.pdf without extra details. Can somebody please explain that?

I can see, that these two estimates are every similar as $q=2$, $r=\infty$ , $n=3$ are admissable in Theorem 1.3. Furthermore the $\gamma=1$ in Theorem 1.3 fits perfectly with the $\lambda$ in Lemma 3.1.

The only real difference seems to be that Theorem 1.3 concerns the wave equation and Lemma 3.1 the Klein-Gordon-Equation, I know, that the wave equation is just the Klein-Gordon-Equation with $m=0$ but I still don't see why this local Klein-Gordon-Estimate follows directly from the corresponding local Wave-Estimate.

Thanks a lot for your help!


1 Answer 1


I don't think they mean that you can literally just plug in Theorem 1.3 of Sterbenz-Rodnianski to get Lemma 3.1. I think they mean that the proof is basically the same, with suitable adjustments.

A rough sketch:

The key to Theorem 1.3 of Sterbenz-Rodnianski is the derivation of the $L^\infty$ estimate (14). If you follow the same argument for Klein-Gordon, you will have that in the equation between (13) and (14) you should have, instead of $$ \exp( 2\pi i (t \pm r + k/4)\rho) $$ (which is derived from the wave propagator $e^{it\sqrt{-\triangle}}$) the expression $$ \exp(2\pi i ( c(\rho) t \pm r + k/4)\rho) $$ (which is derived from the Klein-Gordon propagator $e^{it\sqrt{-\triangle + m}}$) where the wave-speed, coming from the dispersion relation of Klein-Gordon, is $$ c(\rho) = \sqrt{1 + \frac{m}{\rho^2}}.$$

Now, since we have inserted a $\rho$ cut-off, we can restrict the $L^\infty$ estimate for $\psi^\pm_k$ to the region where $$ \{(t,r): \exists \rho \in [\lambda/2,2\lambda], [\rho c'(\rho) + c(\rho)] t \pm r + k/4 = 0 \} $$ and its complement. In the complement we can integrate by parts as in the wave case to get basically equation (14) with a wave-speed $C(\lambda)$ inserted in the estimate. Within the region we don't integrate by parts, and bound the $\exp$ term by $1$. This gives us a space-time $L^\infty$ estimate for $\psi^\pm_k$. Plug this in the equation (13) and you can get $L^2$ integrability in time.

  • $\begingroup$ sorry but I don't understand how you get the expression exp(2πi(c(ρ)t±r+k/4)ρ). Maybe I have to understand why (8) [that is the idenity for the wave propagator with the Bessel function] is true first. This is taken for granted in all the papers I found. Do you have a proof/reference for that? $\endgroup$
    – Tom
    Mar 12, 2016 at 11:23
  • $\begingroup$ That is a standard expression for the fourier transform of radial functions. You can find this by a google search, or in many standard Fourier analysis texts (Stein's Fourier Analysis on Euclidean Spaces comes to mind.) $\endgroup$ Mar 13, 2016 at 0:31

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.