Abstract:In this paper, we consider the atmospheric tomography operator, which describes the effect of turbulent atmospheric layers on incoming planar wavefronts. Given wavefronts from different guide stars, measured at a telescope, the inverse problem consists in the reconstruction of the turbulence above the telescope. We show that the available data is not sufficient to reconstruct the atmosphere uniquely. Additionally, we show that classical regularization methods as Tikhonov regularization or Landweber iteration will always fail to reconstruct a physically meaningful turbulence distribution.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the non - uniqueness and reconstructability issues in Atmospheric Tomography. Specifically, the author explores how to reconstruct the atmospheric turbulence distribution above the telescope from the wave - front data from different guide stars when given such data. However, it has been found that the existing data are insufficient to uniquely reconstruct the atmospheric turbulence, and traditional regularization methods (such as Tikhonov regularization or Landweber iteration) are unable to reconstruct physically meaningful turbulence distributions.
### Specific description of the problem
1. **Non - uniqueness problem**: The paper points out that due to limited data, the Atmospheric Tomography problem usually does not have a unique solution. This means that different atmospheric turbulence distributions may produce the same observational data, resulting in the inability to uniquely determine the real atmospheric state.
2. **Reconstructability problem**: Even when using classical regularization methods (such as Tikhonov regularization, Landweber iteration, etc.), it is not possible to successfully reconstruct a physically meaningful atmospheric turbulence distribution. This is because these methods have inherent limitations when dealing with the Atmospheric Tomography problem.
### Main contributions
- **Analyzing the analytical properties of the Atmospheric Tomography operator**: The author analyzes in detail the mathematical characteristics of the Atmospheric Tomography operator, especially its non - uniqueness and reconstructability problems.
- **Proving non - uniqueness**: Through geometric analysis and mathematical derivation, the author proves that the Atmospheric Tomography problem does not have a unique solution in most cases.
- **Evaluating the limitations of traditional regularization methods**: The author shows the shortcomings of commonly used regularization methods when dealing with the Atmospheric Tomography problem and explains why these methods are unable to effectively reconstruct atmospheric turbulence.
### Numerical verification
To support the above - mentioned theoretical analysis, the author carried out numerical simulations. The results show that although the reconstruction error is large, in areas with more overlaps, the Strehl ratio (an indicator for measuring image quality) is significantly improved. This indicates that although it is not possible to completely and accurately reconstruct the atmospheric turbulence, increasing the number of guide stars can significantly improve the correction effect of the adaptive optics system.
### Conclusion
In general, this paper reveals the non - uniqueness and reconstruction difficulties in the Atmospheric Tomography problem and verifies these problems through theoretical analysis and numerical simulations. Although it is not possible to completely and precisely reconstruct the atmospheric turbulence, increasing the number of guide stars can significantly improve the correction effect of the adaptive optics system, thereby improving the quality of astronomical observations.