Abstract:Decoupled fractional Laplacian wave equation can describe the seismic wave propagation in attenuating media. Fourier pseudospectral implementations, which solve the equation in spatial frequency domain, are the only existing methods for solving the equation. For the earth media with curved boundaries, the pseudospectral methods could be less attractive to handle the irregular computational domains. In the paper, we propose a radial basis function collocation method that can easily tackle the irregular domain problems. Unlike the pseudospectral methods, the proposed method solves the equation in physical variable domain. The directional fractional Laplacian is chosen from varied definitions of fractional Laplacian. Particularly, the vector Grünwald-Letnikov formula is employed to approximate fractional directional derivative of radial basis function. The convergence and stability of the method are numerically investigated by using the synthetic solution and the long-time simulations, respectively. The method's flexibility is studied by considering homogeneous and multi-layer media having regular and irregular geometric boundaries.
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is the challenges faced by existing numerical simulation methods in dealing with seismic wave propagation in attenuating media with irregular boundaries. Specifically, the paper focuses on how to effectively solve the fractional Laplacian wave equation in irregular computational domains.
### Problem Background
1. **Limitations of Existing Methods**:
- Currently, the main methods for solving the fractional Laplacian wave equation are Fourier pseudospectral methods. Although these methods are very efficient in solving equations in the frequency domain, they are not effective in dealing with the earth media with curved boundaries, because the fast Fourier transform and its inverse transform are difficult to handle irregular computational domains.
2. **Challenges in Definition Selection**:
- The fractional Laplacian operator has different definition methods, and not all definitions are equivalent on a bounded domain. Selecting the optimal definition is crucial for specific applications.
3. **Matrix Densification Problem**:
- The discretization matrix of the fractional Laplacian operator is usually full - rank, which requires the development of effective pre - processing techniques to solve the resulting linear system.
### Solutions in the Paper
To solve the above problems, the paper proposes a new method based on the radial basis function collocation method (RBF collocation method). This method has the following characteristics:
- **Solution in Physical Space**: Different from the Fourier pseudospectral method, the RBF collocation method directly solves the equation in the physical variable domain, so it can better handle irregular boundaries.
- **Flexibility**: The RBF collocation method can easily deal with high - dimensional and irregular computational domain problems, because its discretization only depends on the distance between nodes, without relying on grids or elements.
- **Mathematical Simplicity**: This method is relatively simple in mathematics and easy to program.
### Specific Applications
The paper verifies the effectiveness and stability of this method through a series of numerical experiments, including but not limited to:
- **Regular and Irregular Geometric Boundaries**: The wave field simulations with regular and irregular geometric boundaries in homogeneous and multi - layer media are studied.
- **Long - term Simulation**: The numerical stability of the method is explored, especially its performance in long - term simulations.
- **Parameter Heterogeneity**: The ability of the RBF collocation method to deal with multi - layer media with spatially - dependent velocity parameters is demonstrated.
### Conclusions
The RBF collocation method proposed in the paper provides a new and effective way to solve the fractional Laplacian wave equation in irregular computational domains, and has potential application value, especially in the field of seismic wave propagation modeling. However, the high computational cost of this method is still a challenge. In the future, it is necessary to develop efficient pre - processing iterative solvers for large - scale simulations. In addition, the treatment of the variable fractional order γ(x,y) and more complex boundary conditions (such as perfectly matched layers) are also the directions of future research.
Hope this information can help you understand the core problems of this paper and their solutions. If you have more questions or need further explanations, please feel free to let us know!