The classical monotonicity result for semilinear elliptic equations in a half-space: if u solves -Delta u = f(u) with u = 0 on the boundary and u > 0 in the interior, and f satisfies a Lipschitz condition, then u is monotone in the direction perpendicular to the boundary. The solution increases as you move away from the boundary.
For the fractional Laplacian (-Delta)^s with 0 < s < 1, the same question becomes harder. The nonlocal operator couples the solution's values at distant points, so the sliding method — the standard technique for proving monotonicity in local PDE — requires global control that the local method doesn't need.
The authors prove monotonicity for fractional semilinear problems in the half-space under optimal conditions. The solution is monotone increasing in the normal direction, provided the nonlinearity f satisfies a Lipschitz condition adapted to the fractional setting. The adaptation accounts for the nonlocal nature of the operator: the Lipschitz constant must be small enough relative to the operator's strength to prevent nonlocal interactions from breaking the monotonicity.
The proof uses a fractional version of the sliding method, where the comparison function is a translated copy of the solution. The nonlocal terms create additional interactions between the original and translated solutions that must be controlled through maximum principle arguments adapted to the fractional setting.
Monotonicity in half-spaces, extended to fractional diffusion. The nonlocal operator tries to break the monotonicity by coupling distant points. The maximum principle, adapted, prevents it.