1: \begin{abstract}

2: In this paper, a backward problem for a time-space fractional diffusion process has been considered.

3: For this problem, we propose to construct the initial function by minimizing data residual error in fourier space domain

4: with variable total variation (TV) regularizing term which can protect the edges as TV regularizing term and reduce

5: staircasing effect. The well-posedness of this optimization problem is obtained under a very general setting.

6: Actually, we rewrite the time-space fractional diffusion equation as an abstract fractional differential

7: equation and deduce our results by using fractional semigroup theory, hence, our theoretical results can be applied to

8: other backward problems for the differential equations with more general fractional operator.

9: Then a modified Bregman iterative algorithm has been proposed to approximate

10: the minimizer. The new features of this algorithm is that the regularizing term altered in each step and

11: we need not to solve the complex Euler-Lagrange equation of variable TV regularizing term (just need to solve

12: a simple Euler-Lagrange equation). The convergence of this algorithm and the strategy of choosing parameters

13: are also obtained. Numerical implementations are provided to support our theoretical analysis to show

14: the flexibility of our minimization model.

15: \end{abstract}

16: