Abstract:In this paper, we introduce six axioms for relative Bott-Chern secondary characteristic classes and prove the uniqueness and existence theorem for them. Such a work provides us a natural way to understand and hence to prove the arithmetic Grothendieck-Riemann-Roch theorem.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is about the theoretical construction of relative Bott - Chern secondary characteristic classes and their applications in the arithmetic Grothendieck - Riemann - Roch theorem. Specifically, the author aims to:
1. **Introduce six axioms**: Introduce six axioms for relative Bott - Chern secondary characteristic classes and prove the uniqueness and existence of these classes.
2. **Correct previous mistakes**: Correct the not - entirely - correct assertion regarding the uniqueness of relative Bott - Chern secondary characteristic classes in previous work and fix some other flaws.
3. **Provide a natural way of understanding**: Through the theory of relative Bott - Chern secondary characteristic classes, provide a natural way to understand and prove the arithmetic Grothendieck - Riemann - Roch theorem for locally complete - intersection morphisms.
### Main problem decomposition
#### 1. Introduction of six axioms
In order to define relative Bott - Chern secondary characteristic classes, the author introduces the following six axioms:
- **Downstairs Rule**: Describes the behavior of characteristic classes on the base space.
- **Projection Rule**: Involves characteristic classes of tensor products.
- **Functorial Rule**: Guarantees the functorial properties of characteristic classes between different manifolds.
- **Uniqueness Rule w.r.t. Hermitian Bundles**: Ensures uniqueness under certain specific conditions.
- **Uniqueness Rule w.r.t. Metrized Morphisms**: Deals with the case of composite maps.
- **Deformation to the Normal Cone Rule**: Deals with deformations near singular points.
#### 2. Correction of previous mistakes
The author points out that the uniqueness assertion regarding relative Bott - Chern secondary characteristic classes in previous work is not entirely correct. Therefore, one of the main tasks of this paper is to correct this mistake and fix some other flaws to make the theory more complete.
#### 3. Provide a natural way of understanding
By introducing the theory of relative Bott - Chern secondary characteristic classes, the author hopes to provide a natural way to understand and prove the arithmetic Grothendieck - Riemann - Roch theorem. This theorem has important applications in algebraic geometry and arithmetic geometry, especially for the study of locally complete - intersection morphisms.
### Formula display
For example, the Downstairs Rule can be expressed as:
\[ ddc \, ch_{BC}(E, \rho; f, \tau_f) = f^*(ch(E, \rho) \cdot td(T_f, \tau_f)) - ch(f^*E, L^2(\rho, \tau)) \]
where:
- \( ch_{BC}(E, \rho; f, \tau_f) \) is the relative Bott - Chern secondary characteristic class,
- \( ch(E, \rho) \) is the Chern characteristic class,
- \( td(T_f, \tau_f) \) is the Tod characteristic class,
- \( L^2(\rho, \tau) \) is the L2 metric induced by integration.
Through these axioms and formulas, the author hopes to establish a complete theoretical framework, thereby better understanding the arithmetic Grothendieck - Riemann - Roch theorem.