Efficiency of random search with space-dependent diffusivity

Abstract: We address the problem of random search for a target in an environment with space-dependent diffusion coefficient $D(x)$. From a general form of the diffusion differential operator that includes It^o, Stratonovich, and H\"anggi-Klimontovich interpretations of the associated stochastic process, we obtain the first-passage time distribution and the search efficiency $\mathcal{E}=\langle 1/t \rangle$. For the paradigmatic power-law diffusion coefficient $D(x) = D_0|x|^{\alpha}$, with $\alpha<2$, which controls whether the mobility increases or decreases with the distance from a target at the origin, we show the impact of the different interpretations. For the Stratonovich framework, we obtain a closed expression of the search efficiency, valid for arbitrary diffusion coefficient $D(x)$. We show that a heterogeneous diffusivity profile leads to lower efficiency than the homogeneous average level, and the efficiency depends only on the distribution of diffusivity values and not on its spatial organization, features that breakdown under other interpretations.