Inversion formulas and range characterizations for the attenuated geodesic ray transform (Q1696022)

For a simple Riemannian surface \((M,g)\) and for \(a \in C^\infty(M,\mathbb{C})\), the authors study in this paper the \textit{attenuated geodesic ray transform} \({I_a}:C^\infty(SM,\mathbb{C}) \to C^\infty({\partial_ +}SM,\mathbb{C})\) defined as \[ {I_a}f(x,v) = \int_0^{\tau (x,v)} f({\varphi_t}(x,v))\exp \left(\int_0^t {a({\gamma_{x,v}}(s))ds}\right)dt,\quad (x,v) \in {\partial_ +}SM \tag{1} \] where \(f \in C^\infty(SM,\mathbb{C})\) following the convention of an earlier paper of the last author [\textit{M. Salo} and \textit{G. Uhlmann}, J. Differ. Geom. 88, No. 1, 161--187 (2011; Zbl 1238.53058)]. They provide range characterization for the transform in (1) over pairs of functions and 1-forms when the integrand \(f\) in (1) can be expressed as \(f(x,v) = {f_0}(x) + {\alpha_x}(v)\) in which \({f_0}\) is a function and \(\alpha\) is a 1-form. The authors develop the following main results in this paper: Theorem 2.1. Let \((M, g)\) be a simple surface and let \(a \in C^\infty(M,\mathbb{C})\). Then a function \(u \in C^\infty({\partial_ +}SM,\mathbb{C})\) belongs to the range of \({\mathcal{I}_a}:C^\infty(M,\mathbb{C}) \to C^\infty({\partial_ +}SM,\mathbb{C})\) if and only if \(u = {P_a}w\) for some \(w \in \mathcal{S}_a^\infty ({\partial_ +}SM,\mathbb{C})\). Theorem 2.2. Let \((M, g)\) be a simple surface and let \(a \in C^\infty(M,\mathbb{C})\). The following range characterizations hold: {\parindent=6mm \begin{itemize}\item[1.] A function \(u \in C^\infty({\partial_ +}SM,\mathbb{C})\) belongs to the range of \(I_a^0:C^\infty(M,\mathbb{C}) \to C^\infty({\partial_ +}SM,\mathbb{C})\) if and only if \(u = {P_a}w\) for some \(w \in \mathcal{S}_a^\infty ({\partial_ +}SM,\mathbb{C})\) such that \(w_0^\sharp = 0\). \item[2.] A function \(u \in C^\infty({\partial_ +}SM,\mathbb{C})\) belongs to the range of \(I_a^1:C_{\operatorname{sol}}^\infty ({\Lambda ^1}(M),\mathbb{C}) \to C^\infty({\partial_ +}SM,\mathbb{C})\) if and only if \(u = {P_a}w\) for some \(w \in \mathcal{S}_a^\infty ({\partial_ +}SM,\mathbb{C})\) such that \(w_{ - 1}^\sharp + w_1^\sharp = dp\) some \(p \in C^\infty(M,\mathbb{C})\) (in which \(w^\sharp:SM\to \mathbb{C}\) is the unique solution to the transport equation \(X{w^\sharp} + a{w^\sharp} = 0\;(SM),{w^\sharp}{|_{{\partial_ +}SM}} = w\) for a given \(w \in C^\infty({\partial_ +}SM,\mathbb{C})\) and \({X_{(x,v)}} = \frac{d}{{dt}}{|_{t = 0}}{\varphi_t}(x,v)\) `denotes the generator of the geodesic flow of \(g\)'). \end{itemize}} Theorem 2.3. Let \((M, g)\) a simple surface and \(a \in C^\infty(M,\mathbb{C})\). Define \(\overset{\rightharpoonup}{w}\) and \(\overset{\leftharpoonup}{w}\) smooth holomorphic and antiholomorphic, odd, solutions of \(X\overset{\rightharpoonup}{w}=X\overset{\leftharpoonup}{w}=-a\), and let \(\overset{\rightharpoonup}{\mathcal{B}}\) and \(\overset{\leftharpoonup}{\mathcal{B}}\) as in Theorem 7.5 and Corollary 7.6. Then the functions \(({h_0},f) \in C_0^\infty (M,\mathbb{C}) \times C^\infty(M,\mathbb{C})\) can be reconstructed from data \(\mathcal{I}: = {\mathcal{I}_a}\left[ { \star d{h_0},f} \right]\) (extended by zero to \({\partial_ -}SM\)) via the following formulas: \[ f = - \eta_+ \overset{\rightharpoonup}{\mathcal{D}}_{ -1} - \eta_- \overset{\leftharpoonup}{\mathcal{D}}_1 - \frac{a}{2}\left(\overset{\rightharpoonup}{\mathcal{D}}_0 + \overset{\leftharpoonup}{\mathcal{D}}_0 + i({g_ +} - {g_ -})\right) \] \[ {h_0} = \frac{1}{2}({g_ +} + {g_ -}) - \frac{i}{2}\left(\overset{\rightharpoonup}{\mathcal{D}}_0 - \overset{\leftharpoonup}{\mathcal{D}}_0\right) \] where we have defined \(\overset{\rightharpoonup}{\mathcal{D}} : = {e^{\overset{\rightharpoonup}{w}}}{(\overset{\rightharpoonup}{\mathcal{B}} (\mathcal{I}{e^{ - \overset{\rightharpoonup}{w}}}{|_{\partial SM}}))_\psi},\overset{\leftharpoonup}{\mathcal{D}} : = {e^{\overset{\leftharpoonup}{w}}}{(\overset{\leftharpoonup}{\mathcal{B}} (\mathcal{I}{e^{ - \overset{\leftharpoonup}{w}}}{|_{\partial SM}}))_\psi}\) and where \({g_ \pm} \in {\ker ^0}{\eta_ \pm}\), uniquely characterized by their boundary conditions \({g_ +}{|_{\partial M}} = - i{(\mathcal{I} - \overset{\rightharpoonup}{\mathcal{D}} {|_{\partial SM}})_0},{g_ -}{|_{\partial M}} = i{(\mathcal{I} - \overset{\leftharpoonup}{\mathcal{D}} {|_{\partial SM}})_0}\). Theorem 2.5. Let \((M, g)\) a simple surface and let \(a \in C^\infty(M,\mathbb{C})\). Then the following hold: {\parindent=6mm \begin{itemize}\item[(i)] A function \(u \in C^\infty({\partial_+}SM,\mathbb{C})\) belongs to the range of \(I_a^0:C^\infty(M,\mathbb{C}) \to C^\infty({\partial_ +}SM,\mathbb{C})\) if and only if \(u = {P_a}w\) for some \(w \in \mathcal{S}_a^\infty ({\partial_+}SM,\mathbb{C})\) and \({P_{a,1}}u = {P_{a, - 1}}u = {P_{a, \bot}}u = 0\). \item[(ii)] A function \(u \in C^\infty({\partial_+}SM,\mathbb{C})\) belongs to the range of \(I_a^1:C_{\operatorname{sol}}^\infty ({\Lambda ^1}(M),\mathbb{C}) \to C^\infty({\partial_+}SM,\mathbb{C})\) if and only if \(u = {P_a}w\) for some \(w \in \mathcal{S}_a^\infty ({\partial_+}SM,\mathbb{C})\) and \({P_{a,0}}u = 0\). \end{itemize}} The reviewer is hopeful that these exemplary results should find some practical applications in `some medical imaging modalities such as Computerized Tomography and Ultrasound Doppler Tomography in media with variable refractive index' in the near future.