Optimal regularity of solutions to the obstacle problem for the fractional Laplacian with drift (Q477050)

Let \(b(\cdot):\mathbb R^n\to\mathbb R^n,\;c(\cdot):\mathbb R^n\to \mathbb R_+\) be Hölder continuous functions, \(s\in (0,1)\) and \(\varphi(\cdot)\in C^{3s}(\mathbb R^n)\cap C_0(\mathbb R^n)\), and \(Lu(x)=(-\Delta)^su(x)+b(x)\cdot\nabla u(x)+c(x)u(x)\). The authors provide sufficient conditions for the existence and uniqueness of solutions in \(C^{1+s}(\mathbb R^n)\) for \[ \min\left\{Lu(x),u(x)-\varphi(x)\right\}=0\text{ for all }x\in \mathbb R^n.\leqno{(*)} \] Accordingly, for the existence the authors suppose that \(b(\cdot)\in C^s(\mathbb R^n,\mathbb R^n)\), \(c(\cdot)\in C^s(\mathbb R^n,\mathbb R)\) for \(s\in (1/2,1)\), \(L(\varphi(\cdot))_+:=\max(L\varphi(\cdot),0)\) is bounded; for obtaining the uniqueness, the authors also assume that \(b(\cdot)\) is a Lipschitz function and \(c(x)\geq c_0\) for all \(x\in \mathbb R^n\), where \(c_0\) is a fixed positive constant (Theorem 1.1). The proof of the existence is essentially based on a synthesis of the following two results. The first one focuses on the existence and uniqueness of solutions for \(Lu(x)=f\in C^\alpha(\mathbb R^n)\) such that \(\|u\|_{C^{2s+\alpha}\left(\mathbb R^n\right)}\leq C\|f\|_{C^\alpha\left(\mathbb R^n\right)}\) where \(C\) is a constant depending on the norms of \(b(\cdot),c(\cdot)\) and on \(c_0,n,s,\alpha\), see Lemma 2.2, and the proof is based on using comparison principle (Lemma 2.4), i.e., that \(Lu(x)\geq 0\) implies \(u\geq 0\). The second one states that \(Lu(x)=\beta_\varepsilon(\varphi(x)-u(x))\) has a positive solution \(u_\varepsilon\) in some Hölder spaces bounded by \(\|\varphi\|_{C\left(\mathbb R^n\right)}\), which where \(\displaystyle \beta_\varepsilon(t)=\varepsilon^{-1}\max(t,0)\) for all \(t\in \mathbb R\) and \(\varepsilon>0\). Furthermore, \(\|\beta_\varepsilon(\varphi-u_\varepsilon)\|_{C\left(\mathbb R^n\right)}\leq \|(L\varphi)_+\|_{C\left(\mathbb R^n\right)}\) (Lemmas 2.7 and 2.8). Apropos of the proof of the uniqueness, the authors use stochastic representation of a solution to \((*)\), see Proposition 2.11, and the proof is based on employing probabilistic tools.