Covariant tomography of fields

This paper introduces "covariant tomography," a local framework that solves Inverse Boundary Value Problems for parallel transport equations on star-shaped domains by combining geometric decomposition with specific interior extensions and a novel "tower" algorithm to reduce higher-order systems, such as Maxwell equations, into sequentially solvable first-order equations for reconstructing currents and gauge potentials.

The Gist

The Big Picture: Hearing the Shape of a Drum

Imagine you are in a dark room with a drum. You can't see it, but you can hit it and listen to the sound it makes. The famous question is: "Can you hear the shape of the drum?" (i.e., can you figure out the drum's shape just by listening to the sound?)

This paper tackles a similar problem, but instead of sound, it deals with fields (like magnetic or electric fields) and currents (the flow of energy).

• The Problem: We can only measure what happens on the surface (the boundary) of a region. We want to know what is happening inside that region.
• The Goal: To reconstruct the invisible "inner workings" (currents and potentials) of a system based entirely on data collected from its edges.

The author calls this "Covariant Tomography." Think of it like a medical CT scan, but instead of X-rays, we are using math to "scan" invisible fields inside a star-shaped room.

---

The Core Tools: The "Magic Rope" and the "Lego Tower"

To solve this, the author uses two main mathematical tricks.

1. The "Magic Rope" (Geometric Decomposition)

Imagine the room you are trying to scan is shaped like a star (or a starfish). Every point in the room can be connected to the very center by a straight line.

• The Metaphor: Think of the center of the room as a spiderweb hub. The author uses a "homotopy operator," which is like a magic rope that pulls everything from the center out to the edges (and vice versa).
• How it works: This rope allows the author to break down complex, messy equations into simpler pieces. It turns a difficult puzzle into a series of manageable steps, similar to how you might untangle a knot by pulling one specific loop.
• The Catch: This only works if the room is "star-shaped" (no weird caves or donut holes). If the room is too weirdly shaped, the rope gets stuck.

2. The "Lego Tower" (The Tower Algorithm)

Many physics problems (like Maxwell's equations for electromagnetism) are like complex Lego structures built from many layers. They are "second-order" or "higher-order" problems, meaning they are very hard to solve all at once.

• The Metaphor: Imagine a tall, wobbly tower of Legos. Trying to take it apart in one go is impossible.
• The Solution: The author's "Tower Algorithm" breaks this tall tower down into a stack of single Lego bricks (first-order equations).
• The Process: You solve the bottom brick, then use that answer to solve the next one up, and so on.
• The Rule: The paper proves a golden rule: You can only solve the whole tower if you can solve every single brick in the stack, one by one. If one brick is broken (unsolvable), the whole tower falls.

---

The Three Ways to Fill the Room (Extensions)

We know what's happening on the walls (the boundary), but we don't know what's in the middle. To solve the puzzle, we have to guess (or "extend") what the field looks like inside. The paper suggests three ways to do this, like filling a bucket with water:

1. The Radial Extension (The "Flashlight"):

   • Imagine shining a flashlight from the center outward. You just copy the value from the wall straight to the center.
   • Pros: Fast and simple.
   • Cons: It's "jagged." If the wall has a sharp corner, the center might get a "glitch" or a singularity (a mathematical tear). It's like trying to fold a piece of paper perfectly; sometimes it crinkles.

2. The Heat Equation Extension (The "Warm Blanket"):

   • Imagine the boundary values are hot spots. You let "heat" diffuse inward over time.
   • Pros: It smooths everything out. The jagged edges disappear, leaving a nice, soft, continuous field.
   • Cons: You have to decide how long to let the heat sit (a time parameter).

3. The Harmonic Extension (The "Perfect Equilibrium"):

   • Imagine waiting for the heat to settle completely until the temperature is perfectly balanced everywhere.
   • Pros: This gives the smoothest, most "perfect" mathematical solution.
   • Cons: It requires the most complex math to calculate.

The Takeaway: The smoother your "filling" method, the smoother your final answer will be. If you use the "jagged" flashlight method, your reconstructed field might have mathematical spikes.

---

The "Non-Uniqueness" Problem: The Foggy Mirror

Here is the tricky part: There isn't just one answer.
Imagine looking at a reflection in a foggy mirror. You can see the shape, but the details are blurry.

• In physics, there are "gauge modes." Think of these as invisible filters you can put over your solution. You can change the internal math (the "filter") without changing what you see on the boundary.
• The Result: You can reconstruct the field, but you might get a slightly different version of the "truth" depending on which mathematical filter you choose. The paper admits this and suggests using "optimization" (like picking the simplest or most stable version) to pick the best answer.

---

Real-World Example: The Electromagnetic Puzzle

The paper tests this on Maxwell's equations (the rules that govern electricity and magnetism).

• Scenario: You know the magnetic field on the surface of a sphere. You want to know the electric current flowing inside.
• The Method:
  1. Use the "Magic Rope" to break the complex equations into a "Tower" of simple steps.
  2. Use the "Harmonic Extension" to smoothly fill the inside of the sphere with a guess.
  3. Solve the steps one by one.
  4. Result: You successfully reconstruct the hidden currents and potentials inside the sphere, proving the method works.

Summary

This paper is a new mathematical toolkit for seeing the invisible.

1. It turns complex, high-level physics problems into a stack of simple steps (The Tower).
2. It uses a geometric rope to connect the outside world to the inside.
3. It offers different ways to smooth out the data so the answer isn't jagged.
4. It admits that while we can't always find the one perfect answer, we can find a whole class of valid answers that fit the data.

It's like having a new way to solve a 3D puzzle where you can only see the outside, but with the right math, you can figure out exactly how the pieces fit together inside.

Technical Summary

1. Problem Statement

The paper addresses the Inverse Boundary Value Problem (IBVP) for the parallel transport equation (covariant constancy) on star-shaped domains. The core objective is to reconstruct internal physical parameters—specifically currents ($J$) and gauge potentials ($A$)—using only boundary data.

Mathematically, the problem is formulated as finding a solution $\phi$ to the equation:
$$d_\nabla \phi = J$$
subject to the boundary condition:
$$\phi|_B = \alpha$$
where:

• $d_\nabla = d + A \wedge$ is the covariant exterior derivative.
• $A$ is a connection one-form (gauge field).
• $J$ is an inhomogeneity (current).
• $\alpha$ is the measured data on the boundary $B$.

The paper investigates three specific reconstruction scenarios:

1. Current Tomography: Recover $J$ given $A$ and $\alpha$.
2. Gauge Field Tomography: Recover $A$ given $J$ and $\alpha$.
3. Joint Tomography: Recover both $A$ and $J$ given $\alpha$.

The author notes that these problems generally suffer from non-uniqueness, as multiple interior configurations can yield the same boundary values, a characteristic shared with classic inverse problems like the Calderón problem or the "hearing the shape of a drum" question.

2. Methodology

The proposed framework, termed "Covariant Tomography," relies on Geometric Decomposition and the Homotopy Operator. The methodology is local, requiring the domain $U$ to be star-shaped (contractible).

A. Geometric Decomposition

The method utilizes the Poincaré lemma and a linear homotopy operator $H$ defined relative to a center $x_0$. This operator allows for the decomposition of the space of differential forms $\Lambda^k(U)$ into two subspaces:

• Exact forms ($E$): $\text{im}(dH)$
• Antiexact forms ($A$): $\text{im}(Hd)$

This decomposition transforms the partial differential equation (PDE) $d_\nabla \phi = J$ into a structure analogous to ordinary differential equations (ODEs), allowing for iterative solutions via formal series expansions.

B. The Two-Step Solution Process

The reconstruction is achieved through two distinct steps:

1. Extension Problem:
   The boundary data $\alpha$ must be extended from the boundary $B$ into the interior of the domain $U$ to create a field $\Phi$. The paper analyzes three extension strategies, each dictating the regularity of the resulting interior current:

   • Radial Extension: Constant extension along rays from $x_0$. This is simple but can introduce discontinuities at the center, leading to singular (distributional) currents.
   • Heat Equation Extension: Solves the heat equation $\partial_t \Phi = \Delta \Phi$ for a finite time $T$. This yields a smooth ($C^\infty$) interior solution.
   • Harmonic Extension: Solves the Laplace equation $\Delta \Phi = 0$. This provides the smoothest regular solution ($C^\infty$) and is often the preferred choice for physical applications.

2. Projection Problem:
   Once the interior field $\Phi$ is established, the method projects it onto the solution space of the parallel transport equation.

   • If $A$ is fixed, the current $J$ is recovered via $J = d_\nabla \Phi$.
   • If $J$ is fixed, the connection $A$ is determined by solving the algebraic relation $A \wedge \Phi = J - d\Phi$.
   • If both are unknown, a functional minimization (e.g., Tikhonov regularization combined with a Yang-Mills term) is proposed to select an optimal solution.

3. Key Contributions

A. The "Tower" Algorithm (Algorithm 1)

A primary contribution is the development of an algorithm to solve higher-order geometric-based differential equations (such as Maxwell's equations). The author demonstrates that any $k$-th order equation can be decomposed into a sequence of coupled first-order equations, termed a "Tower":
$$\begin{cases} D_k \phi_k = J \\ D_{k-1} \phi_{k-1} = \phi_k \\ \vdots \\ D_1 \phi_1 = \phi_2 \\ j^* \phi_k = \alpha \end{cases}$$
where $D_i$ represents either the covariant derivative ($d_\nabla$) or its dual ($\delta_\nabla$). This reduces complex systems to a sequential chain of solvable first-order IBVPs.

B. Formal Solvability Criterion (Theorem 6)

The paper establishes a rigorous necessary and sufficient condition for solvability:

Theorem 6: An IBVP for a higher-order system is solvable if and only if the associated tower of first-order equations is sequentially solvable.

C. Regularity Analysis

The paper explicitly links the choice of the extension method to the regularity of the recovered fields:

• Radial extension $\to$ Distributional currents (singularities at the homotopy center).
• Heat/Harmonic extension $\to$ Smooth ($C^\infty$) currents.

4. Results and Validation

The framework was validated through low-dimensional examples and electromagnetic applications:

• 1D Examples: Demonstrated the reconstruction of potentials and currents on an interval $(0,1)$, showing how different extension methods (radial vs. harmonic) affect the presence of Dirac delta singularities in the current.
• Maxwell Equations in $\mathbb{R}^3$:
  • The Maxwell system ($dA=F, \delta F = J$) was successfully decomposed into the "Tower" structure.
  • The method successfully reconstructed the electromagnetic potential $A$ from boundary field data $F$ and known current $J$.
  • It highlighted the non-uniqueness of the solution due to gauge modes (elements in the kernel of the connection), which is consistent with physical gauge invariance.

5. Significance and Limitations

Significance:

• Unified Framework: Provides a systematic, geometric approach to solving IBVPs for parallel transport, unifying concepts from exterior calculus, homotopy theory, and inverse problems.
• Algorithmic Clarity: The "Tower" method offers a constructive algorithm for solving high-order PDEs (like Maxwell's or Yang-Mills) by breaking them down into manageable first-order steps.
• Physical Applicability: Directly applicable to electromagnetic tomography, medical imaging, and general relativity where gauge fields and currents must be inferred from boundary measurements.

Limitations:

• Domain Topology: The method is strictly local and requires the domain to be star-shaped (or decomposable into overlapping star-shaped patches). Non-trivial topology (holes) introduces cohomological obstructions.
• Convergence: The series solutions converge only within a radius $r$ determined by the norm of the connection ($||A||$). Strong fields may require domain subdivision.
• Non-Uniqueness: As with most inverse problems, the solution is not unique; it depends on the choice of extension and the specific gauge mode selected.

In conclusion, Kycia's work offers a powerful, mathematically rigorous toolkit for "covariant tomography," bridging the gap between boundary measurements and interior field configurations in gauge theories through the innovative use of geometric decomposition and homotopy operators.