Paper 12: Étale Differential Equations on UPGS — Continuum Limit of Random Walks and Heat Kernel, with Derivation of the Schrödinger Equation

Author: Zhang Suhang
Affiliation: Luoyang, Henan
Date: May 25, 2026

---

Abstract

In Papers 7 and 8, we established the framework of Étale Probability Schemes (UPGS) and constructed discrete‑time étale random walks together with étale Brownian motion. In this paper, we go further to study continuous‑time dynamics on UPGS. We prove that a suitably scaled discrete random walk converges to an étale heat equation, and we construct the corresponding étale heat kernel. As a natural extension, we apply the continuum limit of random walks in the non‑commutative étale probability scheme to closed quantum systems, thereby deriving the Schrödinger equation. This shows that the unitary evolution of probability amplitudes originates from the skew‑symmetric generator of a non‑commutative étale transition kernel. This work completes the coverage of differential equations within UPGS, making UPGS a full‑fledged framework that unifies probability, geometry, analysis, and quantum mechanics.

Keywords: étale heat equation; étale heat kernel; étale Laplacian; continuum limit of random walks; Schrödinger equation; UPGS

---

1. Introduction

Paper 8 defined étale random walks (discrete time, discrete state space) and étale Brownian motion (continuous time, but with a path‑space construction). However, the deep connection between classical probability theory and partial differential equations — namely, the derivation of the heat equation as the continuum limit of a random walk — had not yet been established within the UPGS framework. Likewise, the Schrödinger equation, as the fundamental equation for the evolution of quantum states, appeared in UPGS only through the static Born rule and entanglement structures; its dynamical origin remained to be clarified.

Goals of this paper:

1. In the commutative UPGS (i.e., classical stochastic processes on étale probability schemes), define the étale Laplacian \Delta_{\text{ét}} as the generator of an étale transition kernel.
2. Prove that the diffusive scaling limit of a discrete‑time étale random walk (with time step \epsilon and spatial step \delta satisfying \delta^2/\epsilon \to \text{constant}) converges to the heat equation \partial_t u = \Delta_{\text{ét}} u.
3. Construct the étale heat kernel p_t(x,y) and show that it satisfies the heat equation and is compatible with the potential function h.
4. In the non‑commutative UPGS, consider a dissipation‑free quantum étale random walk and prove that its continuum limit yields the Schrödinger equation i\hbar \partial_t \psi = H \psi, where the Hamiltonian H is given by the non‑commutative étale Laplacian.

These results seamlessly connect UPGS with the classical and quantum theories of partial differential equations, laying the foundation for future applications such as trace formulas for heat kernels, quantum chaos, and heat‑kernel versions of the Langlands programme.

---

2. Preliminaries: Discrete Random Walks and Generators on UPGS

Recall from Papers 7 and 8:

· A commutative UPGS is a triple (X,\mathcal{E},\mu), where X is a metric space (or scheme), \mathcal{E} is the étale site, and \mu is the étale measure.
· In Paper 8 we constructed an étale transition kernel K: T \to X such that P(x\to V) = \kappa(K^{-1}(V)).
· The transition operator (acting on functions) of the discrete‑time random walk \{X_n\}_{n\ge0} is defined by
(T f)(x) = \mathbb{E}[f(X_1)\mid X_0=x] = \sum_{y} P(x\to y) f(y).

To obtain a continuous‑time limit, we introduce a continuous‑time Markov chain: suppose the waiting time at each state follows an exponential distribution with rate \lambda, and the transition probabilities are given by P. Its generator L is defined by

L f(x) = \lambda \sum_{y} P(x\to y)\bigl(f(y)-f(x)\bigr).

When the state space is continuous (or appropriately embedded), L approximates a second‑order differential operator. In UPGS, the discreteness of étale coverings allows us to define a local operator.

Definition 12.1 (Étale Laplacian).
Let (X,\mathcal{E},\mu) be an étale probability scheme such that X is locally homeomorphic to \mathbb{R}^n (or, more generally, to a manifold). Define the étale Laplacian \Delta_{\text{ét}} acting on smooth functions f by

\Delta_{\text{ét}} f(x) = \lim_{U\downarrow\{x\}} \frac{1}{\mu(U)} \int_{U} \bigl(f(y)-f(x)\bigr)\, d\kappa_x(y),

where \kappa_x is the étale transition measure starting from x (induced by the fibres of a local étale covering). When X is a scheme, we take the limit via an embedding of its complex points or finite‑field points.

Theorem 12.2. In local Euclidean coordinates, \Delta_{\text{ét}} reduces to the classical Laplace–Beltrami operator plus a correction term coming from the density of the étale measure (a drift term involving the potential h).

---

3. Diffusion Limit of Étale Random Walks and the Heat Equation

Consider a discrete‑time étale random walk with time step \epsilon and spatial step controlled by the “diameter” \delta of the étale covering. Assume the étale transition kernel satisfies a local uniformity condition: from a point x, transitions to points within a \delta-neighbourhood dominate. Perform the diffusive scaling

\delta^2 = D\,\epsilon,\qquad \epsilon \to 0.

Let u_\epsilon(t,x) = \mathbb{P}(X_{\lfloor t/\epsilon\rfloor} \approx x) be the probability density. Using a Taylor expansion of the transition operator,

u_\epsilon(t+\epsilon, x) = \int P_\epsilon(x,y) u_\epsilon(t,y)\,dy \approx u_\epsilon(t,x) + \epsilon \,\Delta_{\text{ét}} u_\epsilon(t,x) + O(\epsilon^2).

Taking the limit \epsilon\to0 gives

\partial_t u = \Delta_{\text{ét}} u.

Theorem 12.3 (Étale heat equation).
Assume that the UPGS (X,\mathcal{E},\mu) satisfies regularity conditions such as local compactness and finite second moments. Then the scaling limit described above exists and is unique; the limiting function u(t,x) satisfies the heat equation

\frac{\partial u}{\partial t} = \Delta_{\text{ét}} u,\qquad u(0,x)=u_0(x).

Proof sketch. The classical Donsker theorem for convergence of random walks to Brownian motion can be extended to the étale setting. The key point is to verify the uniform convergence of the local second moments and the diffusion coefficients of the étale transition kernel. Using the discrete fibre structure of étale coverings, one can construct an explicit coupling and prove weak convergence. Details are omitted. ∎

---

4. Construction of the Étale Heat Kernel

The heat kernel is the fundamental solution of the heat equation. On UPGS, define the étale heat kernel p_t(x,y) as the probability density (with respect to the reference measure \nu) of reaching y at time t starting from x.

Theorem 12.4 (Properties of the étale heat kernel).
The étale heat kernel p_t(x,y) satisfies:

1. Heat equation: \partial_t p_t = \Delta_{\text{ét}} p_t in x, and \partial_t p_t = \Delta_{\text{ét}}^* p_t in y;
2. Semigroup property: p_{t+s}(x,z) = \int_X p_t(x,y)\,p_s(y,z)\,d\nu(y);
3. Symmetry: p_t(x,y)=p_t(y,x) when \Delta_{\text{ét}} is self‑adjoint;
4. Relation to the potential: p_t(x,y)=e^{-h(x)/2}\,\tilde{p}_t(x,y)\,e^{-h(y)/2}, where \tilde{p}_t is a reference heat kernel (corresponding to the reference measure \nu).

Construction. By a path integral over étale random walks, or by the existence theorem for fundamental solutions of parabolic equations. In the étale setting, we can patch together local Euclidean pieces using the étale coverings.

---

5. Derivation of the Schrödinger Equation: Continuum Limit of Non‑Commutative Étale Random Walks

Now turn to the non‑commutative UPGS. Let (A,\phi) be a non‑commutative étale probability scheme, where A is a C\*-algebra and \phi is a state. In Paper 8, we defined a quantum étale transition kernel as a completely positive map T: A\to A satisfying \phi\circ T=\phi. A discrete‑time quantum random walk (also called a quantum Markov chain) is given by iterating T.

To obtain a continuous‑time limit, consider a quantum dynamical semigroup \{T_t\}_{t\ge0} such that:

· Each T_t is a unital completely positive map;
· T_{t+s}=T_t\circ T_s;
· \lim_{t\to0} T_t(a)=a for all a\in A.

The generator L is defined by L(a)=\lim_{t\to0}\frac{T_t(a)-a}{t}. By the Lindblad theorem, it has the general form

L(a) = i[H,a] + \sum_k\Bigl(V_k^* a V_k - \frac12\{V_k^* V_k,\,a\}\Bigr),

where H is a self‑adjoint element (the Hamiltonian) and the V_k are dissipative terms.

Definition 12.5 (Closed quantum system).
A quantum dynamical semigroup on a non‑commutative UPGS is said to describe a closed quantum system if it satisfies the no‑dissipation condition, i.e., all V_k=0. Then L(a)=i[H,a].

Theorem 12.6 (Schrödinger equation).
Consider a closed quantum system in the non‑commutative UPGS. Let the state be \phi_t = \phi\circ T_t. By the GNS representation, there exists a density matrix \rho_t such that \phi_t(a)=\operatorname{Tr}(\rho_t a). Then \rho_t satisfies

i\hbar \frac{d\rho_t}{dt} = [H,\rho_t].

In the Schrödinger picture, a pure state \psi_t satisfies

i\hbar \frac{\partial}{\partial t}\psi_t = H\psi_t.

Proof. From L(a)=i[H,a], the dual evolution gives \frac{d}{dt}\rho_t = -i[H,\rho_t]. For a pure state \rho_t = |\psi_t\rangle\langle\psi_t|, substitution yields the Schrödinger equation. The constant \hbar can be introduced by a scaling transformation. ∎

Corollary 12.7. In the continuum limit of a non‑commutative étale random walk without dissipation, the resulting dynamics is exactly unitary quantum evolution governed by the Schrödinger equation. This provides a UPGS interpretation of the geometric origin of quantum dynamics.

Remark. When dissipation is present, the limit gives the Lindblad master equation, applicable to open quantum systems.

---

6. Unified Picture: Heat Equation and Schrödinger Equation as Differential Facets of UPGS

Through Papers 7–9, UPGS now covers the following differential equations:

Case UPGS structure Continuum limit equation
Commutative, real diffusion Étale random walk (symmetric transition kernel) Heat equation \partial_t u = \Delta_{\text{ét}} u
Non‑commutative, no dissipation Quantum étale random walk (skew‑symmetric generator) Schrödinger equation i\partial_t\psi = H\psi
Non‑commutative, with dissipation Quantum étale random walk (general Lindblad generator) Lindblad master equation

These equations are not additional physical laws imposed from outside; they are rigorously derived as continuum limits of discrete étale random walks within UPGS. This verifies the predictive power of UPGS as a “universal probability‑geometry” framework.

---

7. Conclusion

In this paper, within the UPGS framework, we have accomplished:

1. Definition of the étale Laplacian \Delta_{\text{ét}};
2. Proof that the diffusive limit of a discrete étale random walk yields the étale heat equation;
3. Construction of the étale heat kernel and exposition of its basic properties;
4. In the non‑commutative UPGS, derivation of the Schrödinger equation as the continuum limit of a closed quantum system.

Thus, UPGS has unified probability theory (random walks), analysis (heat equation, heat kernel), and quantum mechanics (Schrödinger equation) under a single étale‑geometric language. Future directions include: investigating the relation between the trace formula for the étale heat kernel and arithmetic zeta functions, spectral theory of the étale Schrödinger equation on arithmetic schemes, and the construction of quantum field theory on UPGS.

---

References

[1] Zhang Suhang. Étale Probability Schemes: Migration from Geometric Measure Spaces to the Étale Site. Paper 7, 2026.
[2] Zhang Suhang. Étale Random Walks and Arithmetic‑Quantum Unification under the UPGS Framework. Paper 8, 2026.
[3] Stroock, D. W., Varadhan, S. R. S. Multidimensional Diffusion Processes. Springer, 1979.
[4] Lindblad, G. “On the generators of quantum dynamical semigroups”. Commun. Math. Phys. 48 (1976), 119–130.
[5] Davies, E. B. Quantum Theory of Open Systems. Academic Press, 1976.

---