Suppose we have two smooth, real valued functions $\Phi$ and $\hat{\Phi}$ on a manifold $M$. Suppose $\Phi$ and $\Phi$ are close under some function space topology like $L^2$ or $L^\infty$. I am intentionally leaving some details ambiguous.
Let $c$ be a regular value of both functions. What conditions will allow me to conclude $\Phi^{-1}(c)$ and $\hat{\Phi}^{-1}(c)$ are diffeomorphic?
For context, I am studying equipotential surfaces of knotted electric charge distributions. At large distances, the electric potential $\Phi$ can be accurately approximated by the so-called "far-field potential" $\hat{\Phi}$. which treats the knotted charge distribution as a single charged particle. The equipotential surfaces of a single charged particle are spheres, and I want to verify that this approximation is so slight, that it does not affect the topology of a low-voltage (close to $0$) equipotential surface for the actual potential $\Phi$. This result is accepted as true within physics, but I am trying to find a mathematical justification. If it helps, we can assume both functions are harmonic.