scientific article; zbMATH DE number 3905197

From MaRDI portal
Publication:3682984





If \(f: {\mathbb{R}}^ n\to {\mathbb{C}}\) is continuous and supported in \(| x| \leq R\), then its Radon transform \(\hat f(\omega,t)\), thought of as a function of \(\omega\) in the (n-1)-sphere \(S^{n-1}\) and \(t\in {\mathbb{R}}^+\), is supported in \(S^{n-1}\times [0,R]\). The author investigates the question of what constraints on supp\((\hat f)\) are entailed by constraints on supp(f). There exists a non-zero entire function f which is integrable over every straight line L and has \(\int_{L}f=0\) for each such L [cf. \textit{L. Zalcman}, Bull. Lond. Math. Soc. 14, 241-245 (1982; Zbl 0464.28006); cf. also \textit{R. S. Strichartz}, Am. Math. Mon. 89, 377-384 and 420-423 (1982; Zbl 0499.44003)]. Hence no direct converse of the first result is valid. But if a growth condition is imposed on \(\hat f\) outside its support, then positive results can be obtained, even if f is not supposed a priori to have compact support. The author shows (theorem 1) that if (1) f is continuous and rapidly decreasing, if (2) \(\hat f(\omega,t)\) is \(O(e^{-\epsilon t})\) for \(\epsilon =\epsilon_{\omega}>0\) and each \(\omega \in S^{n-1}\) and if (3) \(\hat f=0\) on \(\Omega\times [R,\infty)\) for some nonempty open \(\Omega \subset S^{n-1}\) and some \(R<\infty\), then f has compact support. He also gives an example to show that some growth condition like (2) is generally essential. On the other hand (theorem 2), if f has compact support, then hypothesis (2), and of course also (3), can be dropped and the conclusion reached that f vanishes identically on every hyperplane not in the support of \(\hat f.\) If (2) is strengthened to \(\sup p(\hat f)\subset S^{n-1}\times [0,R]\) for some \(R<\infty\), that is, if \(\Omega =S^{n-1}\) in (3), then we recover the support theorem of S. Helgason. In the proof of Theorem 1 a lemma of mixed derivatives is used to show that the Fourier transform of f extends to an entire function of exponential type and then the Paley-Wiener theorem is engaged. Theorem 2 is based on a proof which the author gave in the preceding volume of the same journal [Indagationes 46, 95-111 (1984)] for the theorem of \textit{M. Plancherel} and \textit{G. Pólya} [Comment. Math. Helv. 9, 224-248 (1937; Zbl 0016.36004)].











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3682984)