1: \begin{abstract}

2: In this article, we revisit iterative methods for solving the inverse problem of

3: photoacoustic tomography in free space. Recently, there have been interesting developments on

4: explicit formulations of the adjoint operator, demonstrating that iterative methods is an attractive

5: choice for photoacoustic image reconstruction. In this work, we propose several modifications of current formulations

6:  of the adjoint operator which help  speed up the convergence  and yield improved error estimates. We establish

7:   a stability analysis and show that, with our choices of the adjoint operator, the iterative methods  can achieve a linear rate of convergence, in the $L^2$-norm as well as in the $H^1$-norm.

8:   In addition, we analyze the normal operator from the

9:   microlocal analysis point of view. This gives insight into the convergence speed of the iterative methods and choosing proper weights for the mapping spaces. Finally, we present numerical results using various iterative reconstruction methods for full as well as limited view data. Our results demonstrate that Nesterov's fast gradient and the CG  methods converge faster than Landweber's and iterative time reversal methods in the visible as well as the invisible case.

10: 

11:  

12: \bigskip\noindent\textbf{Keywords:}

13: Photoacoustic tomography, variable sound speed, iterative regularization, adjoint operator,  Landweber 's method, Nesterov's method, CG method, visibility  condition, invisibility condition, image reconstruction.

14: 

15: \bigskip\noindent\textbf{AMS Subject Classification:}

16: 35R30;

17: 92C55;  	

18: 65F10;

19: 35A18;

20: 74J05.

21: \end{abstract}

22: