Up-to-the-boundary differentiability of graph components
DisprovedEthierKurtz.oblique_boundary_differentiability_steppderegularity
This is the up-to-the-boundary differentiability step for graph components.
Let be twice continuously differentiable inside an open set and continuous on its closure. Suppose the interior gradients extend continuously to the closure, i.e. there is a continuous map with at every interior point. Then at every frontier point , is differentiable with derivative .
This packages the Schauder up-to-the-boundary gradient estimate used along reflecting boundaries: it turns the continuous boundary gradient supplied by the graph definition into a genuine derivative, which the Hopf-type boundary step then consumes.
Formalization Note The conclusion is HasFDerivAt u (J x_0) x_0; its proof is the Schauder estimate up to the boundary.
Preamble
import Mathlib open scoped Topology
Formal statement
namespace EthierKurtz
theorem oblique_boundary_differentiability_step {d : ℕ}
{Ω : Set (EuclideanSpace ℝ (Fin d))}
{u : EuclideanSpace ℝ (Fin d) → ℝ} {x₀ : EuclideanSpace ℝ (Fin d)}
{J : (closure Ω) → (EuclideanSpace ℝ (Fin d) →L[ℝ] ℝ)}
(hx : x₀ ∈ frontier Ω) (hxmem : x₀ ∈ closure Ω)
(hC : ContDiffOn ℝ 2 u Ω)
(hcont : ContinuousOn u (closure Ω))
(hJc : Continuous J)
(hJid : ∀ x : (closure Ω), (x : EuclideanSpace ℝ (Fin d)) ∈ Ω →
J x = fderiv ℝ u x) :
HasFDerivAt u (J ⟨x₀, hxmem⟩) x₀ := by sorry
end EthierKurtz
Source
Boundary gradient estimate: continuous extension of interior gradients is the derivative at frontier points (Schauder estimate up to the boundary), cf. Gilbarg-Trudinger, Elliptic Partial Differential Equations of Second Order, Chapter 6; as used for reflecting diffusions in Ethier-Kurtz, Markov Processes, Chapter 4.